HEALTHY CHINA DEEPENING HEALTH REFORM IN CHINA Building High-Quality and Value-Based Service Delivery The World Bank and World Health Organization World Health Organization Western Pacific Region Part Title Healthy China Part Title Healthy China: Deepening Health Reform in China Building High-Quality and Value-Based Service Delivery A copublication of the World Bank and the World Health Organization © 2019 International Bank for Reconstruction and Development / The World Bank and World Health Organization 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 22 21 20 19 The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent, or those of WHO. The World Bank and WHO do not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank and/or WHO concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank or the WHO, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the follow- ing conditions: Attribution—Please cite the work as follows: World Bank and World Health Organization. 2019. Healthy China: Deepening Health Reform in China: Building High-Quality and Value-Based Service Delivery. Washington, DC: World Bank. doi:10.1596/978-1-4648-1263-7. License: Creative Com- mons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content con- tained within the work. The World Bank therefore does not warrant that the use of any third-party- owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-1263-7 ISBN (electronic): 978-1-4648-1323-8 DOI: 10.1596/978-1-4648-1263-7 Cover design: Debra Naylor, Naylor Design, Washington, DC Library of Congress Cataloging-in-Publication Data has been requested. Contents Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix Abbreviations Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 China’s Health System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Report Background and Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 The Recommendations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Implementation of Health Care Reforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 China’s Road to Health Care Reform. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Report Objectives and Audience. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Report Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 1  Impressive Gains, Looming Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 China’s Changing Health Care Needs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Health System Reform in China: A Brief Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Meeting China’s Health Care Needs: Key Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Impact of Selected Policy Changes on Health Expenditures. . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 2 Lever 1: Shaping Tiered Health Care Delivery with People-Centered ­Integrated Care. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 The People-Centered Integrated Care (PCIC) Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Benefits of Adopting PCIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 H E A LT H Y C H I N A v vi H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Challenges to PCIC Implementation in China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Core Action Areas and Implementation Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Moving Forward on the PCIC Core Action Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Annex 2A  Summary of PCIC Impact Findings by Case Studies . . . . . . . . . . . . . . . . . . . . . 84 Annex 2B  Methodology and Summaries of 22 PCIC Performance Improvement ­ ­ Initiatives Described in the Case Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Annex 2C  Characteristics of Care Integration Initiatives Involving Hospitals. . . . . . . . . . . 88 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 3 Lever 2: Improving Quality of Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Conceptualizing, Assessing, and Improving the Quality of Care. . . . . . . . . . . . . . . . . . . . . . 99 QoC Challenges in China. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 China’s Health Care Regulatory and Policy Environment . . . . . . . . . . . . . . . . . . . . . . . . . . 104 International Experience in Improving Health System Quality . . . . . . . . . . . . . . . . . . . . . . . 107 Recommendations to China for Improving the Quality of Care. . . . . . . . . . . . . . . . . . . . . . . 114 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 Notes������������������������������������������������������������������������������������������������������������������������������������� 124 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 4 Lever 3: Engaging Citizens in Support of the PCIC Model. . . . . . . . . . . . 131 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Challenges to Engaging Citizens. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Two Routes to Engage People to Improve Health Outcomes and Restore Trust. . . . . . . . . . 134 Recommendations: Strengthening Patient Engagement in the Patient-Provider Relationship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 5 Lever 4: Reforming Public Hospital Governance and Management . . 161 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Overview and Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 Public Hospital Governance: Challenges and Lessons from Reform in China and Internationally. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Hospital Managerial Practices: Challenges and Lessons from China and Internationally. . . . 179 Recommendations for Moving Forward with Public Hospital Reform . . . . . . . . . . . . . . . . . 185 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 6 Lever 5: Realigning Incentives in Purchasing and Provider Payment. . . . . 203 Introduction ������������������������������������������������������������������������������������������������������������������������� 203 Evolution of Health System Incentives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 Challenges and Lessons Learned From Selected Case Studies . . . . . . . . . . . . . . . . . . . . . . . 208 Provider Payment Reforms in China: Lessons From Recent Pilots. . . . . . . . . . . . . . . . . . . . 212 C ontent s vii Core Action Areas and Implementation Strategies to Realign Incentives. . . . . . . . . . . . . . . . 217 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 7 Lever 6: Strengthening the Health Workforce . . . . . . . . . . . . . . . . . . . . . . 235 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 The Chinese Health Labor Market: Trends and Challenges . . . . . . . . . . . . . . . . . . . . . . . . 236 Recommendations for Human Resources Reform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246 An Implementation Road Map. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262 Annex 7A  Supplementary Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276 8 Lever 7: Strengthening Private Sector Engagement in Health Service Delivery��������������������������������������������������������������������������������������������������� 279 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 Evolution of Policies on the Private Health Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280 Scope and Growth of China’s Private Health Sector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 Key Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 Recommendations for Strengthening Private Sector Engagement. . . . . . . . . . . . . . . . . . . . . 295 Annex 8A  Supplementary Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 Notes��������������������������������������������������������������������������������������������������������������������������������������315 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316 9 Lever 8: Modernizing Health Service Planning to Guide Investment. . . . 319 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 Capital Investment Challenges in China’s Health Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . 320 Recommendations to Modernize Service Planning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342 10 Strengthening the Implementation of Health Service Delivery Reform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 Implementation Challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346 A Four-Part Actionable Implementation Framework. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348 Moving Forward: Effective, Sustainable Local Implementation. . . . . . . . . . . . . . . . . . . . . . . 350 Toward a Sequential Plan for Full-Scale Reform Implementation in China. . . . . . . . . . . . . 362 Annex 10A  Case Examples of Transformation Learning Collaboratives (TLCs). . . . . . . . 366 Annex 10B  Plan-Do-Study-Act Cycles of Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369 Annex 10C  Example of a Detailed Implementation Pathway . . . . . . . . . . . . . . . . . . . . . . . 371 Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 A  Supplementary Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 viii H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Boxes I.1 What is value in health care?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 I.2 Communist Party of China’s endorsement of “Healthy China” strategy. . . . . . . . . 19 2.1 People-centered integrated care (PCIC) defined . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 2.2 Potential benefits of people-centered integrated care . . . . . . . . . . . . . . . . . . . . . . . . . 58 3.1  Institutional arrangements for quality improvement in France, Germany, and the Netherlands. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 ­ 3.2 The Ambulatory Care Project of the National Quality Forum. . . . . . . . . . . . . . . . . 109 3.3 U.K. National Health Service outcomes indicator set, 2013/14 . . . . . . . . . . . . . . . . 111 3.4 An innovative U.K. model to improve care for stroke patients. . . . . . . . . . . . . . . . . 112 4.1 Understanding citizen mistrust: Perspectives from patients and providers . . . . . . . 135 4.2 Defining empowerment, engagement, and coproduction of health. . . . . . . . . . . . . 136 4.3 Individual and public routes to patient engagement. . . . . . . . . . . . . . . . . . . . . . . . . 137 4.4 Effects of patient engagement strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 4.5 Social marketing in China: Prevention and control of hepatitis B . . . . . . . . . . . . . . 142 4.6 The Million Hearts Campaign. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 4.7 Self-management of COPD in the U.S. veterans population. . . . . . . . . . . . . . . . . . . 145 4.8  Encouraging self-management of health: Examples from India and the United Kingdom. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 4.9  Improving patient involvement at Beth Israel Deaconess Medical Center, United States. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 ­ 4.10 Decision aid for stable coronary heart disease by the Informed Medical ­Decisions  Foundation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 4.11 Health coaching to coordinate care in Singapore. . . . . . . . . . . . . . . . . . . . . . . . . . . 151 4.12 Changshu: One example of a Chinese Healthy City. . . . . . . . . . . . . . . . . . . . . . . . . 153 5.1 Impacts of public hospital reform in China. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 6.1 A typology of health care provider payment methods . . . . . . . . . . . . . . . . . . . . . . . 213 6.2 Examples of provider payment reforms in China. . . . . . . . . . . . . . . . . . . . . . . . . . . 219 6.3 Bundled payment models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 6.4 Quality-compatible modified global payment systems in the United States. . . . . . 225 7.1 History of the Physician Dual Practice Policy in China . . . . . . . . . . . . . . . . . . . . . . 245 7.2 Sanming’s comprehensive public hospital compensation reform. . . . . . . . . . . . . . . 256 7.3 GP pay-for-performance incentives in the United Kingdom . . . . . . . . . . . . . . . . . 257 8.1 Hospital ownership categories in China: Public, PNFP, and PFP. . . . . . . . . . . . . . 283 8.2 Difficult market conditions for Yunnan Kidney Disease Hospital. . . . . . . . . . . . . 294 8.3 Use of indirect policy tools in Germany’s hospital market . . . . . . . . . . . . . . . . . . . 301 8.4 Use of indirect policy tools in primary care in Denmark . . . . . . . . . . . . . . . . . . . . 303 8.5  Chinese experiences in purchasing community services from private and public providers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304 8.6  Sweden’s “Choice” reforms: From government direct delivery to indirect ­ ­ governance of the primary care market. . . . . . . . . . . . . . . . . . . . . . . . . . . 306 8.7 Health care quality assurance and the role of fair competition . . . . . . . . . . . . . . . 307 8.8 Hospital quality measures at Germany’s Helios group. . . . . . . . . . . . . . . . . . . . . . 308 9.1 Distinguishing features of an effective health service planning approach. . . . . . . . . 321 9.2 Horizon’s three-step CIP model for eldercare in the Netherlands. . . . . . . . . . . . . . 330 9.3 Physical redesign of Northern Ireland’s health system model. . . . . . . . . . . . . . . . . . 339 10.1 Government administrative reforms and international experience. . . . . . . . . . . . . . 355 10.2 Evidence supporting the use of TLC methodology. . . . . . . . . . . . . . . . . . . . . . . . . 357 10.3  Example of a master reform pathway: primary care enhancement for patients with complex needs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364 C ontent s ix Figures ES.1  Life expectancy relative to per capita health expenditure, selected countries, early 2010s. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 ES.2 Main drivers of projected health expenditure in China, 2015–35. . . . . . . . . . . . . . . 4 ES.3 Eight interlinked levers to deepen health care reform in China. . . . . . . . . . . . . . . . . 6 1.1  Aging of the population in China compared with selected countries, 1950–2050. . . . 26 1.2 Prominence of NCDs in the burden of disease and causes of mortality. . . . . . . . . .27 1.3 Smoking and alcohol consumption, international comparisons . . . . . . . . . . . . . . . 28 1.4  Prevalence of hypertension and diabetes in China and selected other countries, early 2010s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 1.5 Coverage of social health insurance in China, 2003–13. . . . . . . . . . . . . . . . . . . . . 31 1.6 Trends in health service use in China, by visit type, 2003–13. . . . . . . . . . . . . . . . . 32 1.7 Hospital beds in China and selected OECD countries, 2000 and 2013 . . . . . . . . . 36 1.8 Trends in the number and use of health care facilities in China, by level . . . . . . . . 37 1.9 Composition of health spending in China, 1997–2013. . . . . . . . . . . . . . . . . . . . . . 39 1.10 Pharmaceutical spending in China and international comparisons. . . . . . . . . . . . . 40 1.11 Composition of health spending in China, by provider type, 1990 and 2013. . . . . 41 1.12 Percentage of outpatient visits in hospitals, China and selected countries, 2013. . . 41 1.13  Trends in out-of-pocket health care payments in China, by insurance type, 2003–13. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 1.14  Life expectancy trends in China relative to total spending on health, 1995–2013. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 1.15  Performance (life expectancy) relative to health spending and income, international comparisons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 1.16 Management of hypertension and diabetes in China, circa 2010 . . . . . . . . . . . . . . 45 1.17 Main drivers of projected health expenditure increases in China, 2015–35 . . . . . . 46 1.18 Hospital discharge rates, selected countries and economies, 2000–14 . . . . . . . . . . 47 1.19  Life expectancy relative to per capita health expenditure, selected countries, early 2010s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 ­ 2.1  Frequency of PCIC interventions in commissioned case studies, by core action area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 2.2 Illness burden scorecard for patient risk stratification. . . . . . . . . . . . . . . . . . . . . . . 65 2.3 Responsibilities of PACT members, U.S. Veterans Health Administration . . . . . . . 67 2.4  Sample health pathway for COPD, Canterbury Health Services Plan, New Zealand. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 2.5 Sample responsibilities for dual referrals in Xi County. . . . . . . . . . . . . . . . . . . . . . 77 2.6 PACE continual feedback loop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 2.7 Types of data collected by PACE centers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 B4.6.1 Million Hearts targets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 5.1 Growth in total health and hospital spending in China, 2005–13 . . . . . . . . . . . . 162 5.2  Trends in hospital beds per 1,000 population in China and selected OECD countries, 2000–13. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 5.3 Growth in number of hospitals in China, by ownership type, 2005–13. . . . . . . . 163 5.4 Growth in number of hospital beds in China, by ownership type, 2005–13. . . . . 164 5.5  Trends in average number of beds in Chinese secondary and tertiary ­hospitals,  2008–14. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 5.6 Growth in admissions to Chinese tertiary and secondary hospitals, 2006–14 . . . 165 5.7 Average managerial practice scores of Chinese hospitals, 2015. . . . . . . . . . . . . . . 182 6.1  Composition of total health expenditure in China, by facility or provider type, 1990–2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 ­ x H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A 7.1  Numbers of medical students and faculty and faculty-student ratios in China, 1998–2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237 7.2 Ratio of nurses to physicians in OECD countries, 2012 (or nearest year). . . . . . . 238 7.3 Number of health professionals per 1,000 population in China, 2003–13. . . . . . 239 7.4  Health professionals per 1,000 population in China, by rural or urban ­location,  2003–13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240 7.5 Proportion of health professionals in the private sector, 2005–13. . . . . . . . . . . . . 240 7.6  Annual average wages of urban employees in China, by sector, 2005 and 2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 7.7 Health worker compensation across levels of care, China 2013 . . . . . . . . . . . . . . 242 7.8  Community health agents’ coverage in Brazil’s Family Health Strategy, 1998–2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 7.9  Ratio of hospital specialist pay to primary GP pay, selected countries and years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 7.10  Ratio of health professionals’ pay to average income per capita of 10th richest decile, China and selected countries . . . . . . . . . . . . . . . . . . 255 8.1 Number of Chinese hospitals, by ownership type, 2005–12. . . . . . . . . . . . . . . . . 285 8.2  Market share of inpatient admissions in China, by hospital ownership type, 2005 and 2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 8.3 Trends in hospital use in China, by ownership type, 2005–12 . . . . . . . . . . . . . . . 286 8.4 Number of beds in Chinese hospitals, by ownership type, 2005–12. . . . . . . . . . . 287 8.5  Composition of Chinese public and private hospitals, by number of beds, 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287 8.6  Correlation between private hospital share of outpatient visits and per capita income across provinces, 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . 288 8.7 Growth in Chinese PHC facilities, by ownership type, 2005–12 . . . . . . . . . . . . . 289 8.8  Composition of hospital beds in OECD countries, by ownership type, 2012 or latest data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298 9.1 Growth rates of health facilities, Sichuan province, 2008–09 to 2011–12 . . . . . . 322 9.2 Distribution of total capital expenditures in health, Sichuan province, 2009–12. . . 323 9.3 Growth rate of health facilities, Hubei province, 2008–09 to 2012–13 . . . . . . . . 324 9.4 Distribution of health capital expenditures, Tianjin municipality, 2000–13. . . . . 324 9.5 Translating health services to costs in OECD countries . . . . . . . . . . . . . . . . . . . . 328 9.6  Delivering a new approach to health sector planning in New South Wales, Australia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332 9.7  Certificate of Need application and approval process in Maine, United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336 9.8 The integrated health services model in Northern Ireland. . . . . . . . . . . . . . . . . . . 340 10.1  Proposed oversight, coordination, and management for implementation and scaling up of health service delivery reform. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352 10.2 The TLC model in three different arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . 358 10.3 Design of a Transformation Learning Collaborative. . . . . . . . . . . . . . . . . . . . . . . 360 10.4 Sequential plan for scaling up reform implementation . . . . . . . . . . . . . . . . . . . . . 362 B10.3.1 Sample “Better Health at Lower Cost” reform pathway. . . . . . . . . . . . . . . . . . . . 364 Tables ­I.1 Deepening Health Reform in China report chapters. . . . . . . . . . . . . . . . . . . . . . . . . 21 1.1  Major reforms to extend financial protection and contain health care costs in China, 2009–15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 1.2 Hypertension diagnosis, treatment, and control in selected countries, 2013 . . . . . . . 45 Contents xi 2.1  Summary list of commissioned case studies on PCIC-based health care reforms in China and other selected countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 2.2 Eight core PCIC action areas and corresponding implementation strategies . . . . . . . 61 2.3 NCQA certification guidelines for patient-centered medical homes. . . . . . . . . . . . . 82 2A.1 Impact frequency of studies on PCIC initiatives, by PCIC model and impact area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 2B.1 Nomenclature and summaries of 22 PCIC performance improvement initiatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 2C.1 Characteristics of six health care integration initiatives at Chinese hospitals. . . . . . 88 3.1 Overuse of prescription drugs and other health interventions in China. . . . . . . . . . 103 3.2 Regulations on health care quality in China. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 3.3 Three core QoC action areas and implementation strategies. . . . . . . . . . . . . . . . . . 115 4.1 Four core citizen engagement action areas and implementation strategies. . . . . . . . 139 4.2 Examples of nudges and regulation to change target behaviors. . . . . . . . . . . . . . . 154 5.1 Hospital governance models in selected countries . . . . . . . . . . . . . . . . . . . . . . . . . . 169 5.2  Characteristics of selected reform models for Chinese public hospital governance, 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 5.3  Five core action areas and implementation strategies for improving ­ public ­hospital governance and management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 5.4 Best practices in hospital management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 6.1  Four core action areas and implementation strategies to realign health system incentives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 6.2 Primary care remuneration systems in OECD countries, 2014. . . . . . . . . . . . . . . . 221 6.3 Hospital remuneration systems in OECD countries with social health insurance. . . . . 222 6.4 Hospital remuneration systems in OECD countries using a tax-based system. . . . 222 7.1 Enrollment of medical students and medical graduates in China, 2013. . . . . . . . . 236 7.2  Share of Chinese doctors who hope their children will attend medical school, 2002–11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 7.3  Four core action areas and implementation strategies to strengthen the health ­workforce. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 7.4 Health system characteristics in selected countries. . . . . . . . . . . . . . . . . . . . . . . . . 249 7.5 Physician payment and predominant service provision mode, selected countries. . . . . 258 7.6  Characteristics of compensation setting for hospital staff, selected OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 7.7 Hospital staff recruitment and remuneration in five countries. . . . . . . . . . . . . . . . 260 7.8 Road map for implementation of health workforce reforms . . . . . . . . . . . . . . . . . 263 7A.1 China’s health workforce: Classifications, numbers, and percentages of total, 2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266 7A.2 Urban and rural distribution of China’s health workforce, 2003 and 2013. . . . . . 267 7A.3 Average salaries in China, by occupational category, 2005–12 . . . . . . . . . . . . . . . 267 7A.4 Governance framework for health human resources management in China . . . . . 269 7A.5 Headcount quota formulation standards for Chinese health institutions. . . . . . . . . 274 8.1 Evolution of policies on the private health sector in China, 1949–present. . . . . . . 281 8.2 Regional distribution of private hospitals in China, 2012 . . . . . . . . . . . . . . . . . . . 288 8.3 Health employees in China, by sector, 2010–12. . . . . . . . . . . . . . . . . . . . . . . . . . . 290 8.4 Health employees in China, by type and sector, 2012 . . . . . . . . . . . . . . . . . . . . . . 290 8.5  Three core action areas and implementation strategies to strengthen private sector engagement in health service delivery. . . . . . . . . . . . . . . . . . . . . . . . 296 8.6  Shares of predominant mode of service provision in OECD countries, by subsector, early 2010s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 8.7 Tools of health care governance or indirect policy . . . . . . . . . . . . . . . . . . . . . . . . . 300 xii H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A 8A.1  Overview and comparison of Chinese national and provincial private sector ­ ­ policies in health care service delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 8A.2  Summary of policies on the social capital sponsoring medical institutions, by region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314 9.1 Distribution of health personnel, Sichuan province, by location type, 2008–12 . . . . 323 9.2  Capital investment planning for health facilities, beds, and health personnel in studied Chinese provinces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325 9.3  Feasibility study results on relocation and expansion of Renshou County People’s Hospital, 2009. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326 9.4  Five core action areas and implementation strategies to strengthen capital investments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 9.5 Services under Northern Ireland’s integrated health services model, by level. . . . . . 341 10.1  Responsibility levels, strategies, and actions to scale up health service delivery reform, by implementation system type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351 10.2  Sample monitoring guidelines for implementation of China’s value-driven future health system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353 10.3 Sample indicators for monitoring health reform implementation, by reform goal. . . . 361 10.4 Scoring system for Transformation Learning Collaboratives . . . . . . . . . . . . . . . . . 362 10.5 TLC sequential implementation plan, by phase, time interval, and jurisdiction. . . 363 10C.1 Pathway 1.1: Increase the use of primary care as first point of contact . . . . . . . . . . 371 10C.2 Pathway 1.2: Stratify panels based on risk for poor outcomes and high utilization, and develop care plans for most at-risk population . . . . . . . . . . . . . . . 372 10C.3 Pathway 1.3: Form multidisciplinary teams and empanel population . . . . . . . . . . 373 10C.4 Pathway 1.4: Review pilot and scale up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374 10C.5 Pathway 1.5: Implement a continuous M&E plan to track reform progress and inform iterative improvements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 A.1  Eight levers and recommended core actions for high-quality, value-based health service delivery in China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 A.2  Government policies in support of the eight levers for health service delivery reform in China. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383 A.3  New policy guidelines on tiered health service delivery and recommended core actions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387 Preface D uring the past three decades, China consumption as well as environmental fac- has achieved a momentous social tors such as air pollution take a huge toll on transformation, pulling 600 million health, and noncommunicable diseases people out of poverty. At the same time, it account for more than 80 percent of the 10.3 has made impressive strides in health. Since million deaths every year. At the same time, the launch of a new round of reforms in with higher economic growth, increased per- 2009, China has invested substantially in sonal incomes, and fast-changing in con- expanding health infrastructure, achieved sumption patterns, people are demanding nearly universal health insurance coverage, more and better health care. As a result of all promoted more equal access to public health these factors, expenditures on health care services, and established a national essential have increased continuously in recent years. medicine system. For China, this rapid growth in health expen- T hese measu res have sig n ificantly diture may be difficult to sustain amid the improved the accessibility of health services, country’s economic slowdown. greatly reduced child and maternal mortality, The Chinese government fully recognizes cut the incidence of infectious disease, and the need to make strategic shifts in the health considerably improved the health outcomes sector to adapt to these new challenges. and life expectancy of the Chinese popula- President Xi Jinping and Premier Li Keqiang tion: average life expectancy reached 76.34 have placed great importance on health care years in 2015, 1.51 years longer than in 2010. reform. As President Xi has pointed out, it And the country’s overall health level has would not be possible to build a well-off soci- reached the average of other middle- and ety without universal health coverage. He high-income countries, achieving better also indicated that China should shift its health outcomes with less input. These focus and resources toward the lower levels achievements have been well recognized of care, aiming to provide its citizens with internationally. public health and basic health services that China has now reached a turning point. It are safe, effective, accessible, and affordable. is starting to face many of the same chal- Premier Li has held several State Council lenges and pressures that high-income coun- executive meetings to set priorities in health tries face. The Chinese population over the care reform and asked for the development of age of 65 is approximately 140 million, and a basic health care system covering all urban that cohort is expected to grow to 230 and rural residents. The State Council has set m illion by 2030. High-risk behaviors like ­ up a Leading Group for Deepening Health smoking, sedentary lifestyles, and alcohol Care Reform to strengthen multisector H E A LT H Y C H I N A xiii xiv H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A coordination, which provides a strong insti- strong primary health care system. Such a tutional guarantee for the reforms. system offers both better health care for its In July 2014 in Beijing, the Chinese gov- citizens and better value for its economy. ernment, the World Bank, and the World To that end, the report offers a compre- Health Organization committed to working hensive set of eight interlinked recommenda- together on a joint health reform study to tions that can prepare the Chinese health sys- further improve the policy formulation and tem for the demog raphic and health to deepen the health reform. This report, challenges it faces. It focuses not only on the Healthy China: Deepening Health Reform in top-level design for reform but also on the China , is the outcome of this joint study. important question of how to make reform Following the successful model of previous work on the ground. It builds on extensive flagship reports such as China 2030 and analysis of literature and case studies from Urban China, this report offers a blueprint high- and middle-income countries as well as for further reforms in China’s health sector. on ongoing innovations in China that offer In July 2016, Minister of Finance Lou lessons and experiences for bringing about Jiwei, Minister of the National Health and desired change. The report draws upon Family Planning Commission Li Bin, and cutting-edge thinking about the science of ­ Deputy Minister of Human Resources delivery that can help in the scaling up of and Social Security You Jun, joined by World health reforms—from prefecture to province Bank Group President Jim Yong Kim and and, ultimately, nationwide. Bernhard Schwartländer, the World Health Our hope is that this report will provide Organization representative to China, jointly the research, analysis, and insight to help launched the Policy Summary of this report at central and local authorities plan and execute the Diaoyutai Guesthouse in Beijing. The major restructuring of the health care deliv- Policy Summary has received wide praise from ery system in China during the 13th Five- the media and academia, has been dissemi- Year Plan period of 2016–20. Getting this nated to the health policy makers in all the reform right is crucial to China’s social and provinces in China, and has served as an economic success in the coming decades. We important instrument for policy making. believe that China’s experience with health The report’s main theme is the need for service delivery reform carries many lessons China to transition its health care delivery for other countries, and we hope this report system toward people-centered, high-quality, can also contribute to a global knowledge integrated care built on the foundation of a base on health reform. Acknowledgments T his study was organized jointly by World Bank vice president, East Asia and China’s Ministry of Finance (MoF), Pacific Region; Timothy Grant Evans, senior National Health and Family Planning director of the World Bank’s Health, Commission (NHFPC), and Ministry of Nutrition & Population (HNP) Global Human Resources and Social Security Practice (GP); Olusoji Adeyi, director of the (MoHRSS); the World Health Organization Bank’s HNP GP; Bert Hofman, World Bank (WHO); and the World Bank. The study was country director for China, the Republic of proposed by Premier Li Keqiang, Vice Premier Korea, and Mongolia; Mara Warwick, World Liu Yandong, Minister Lou Jiwei of MoF, Bank operations manager for China, Korea, Minister Li Bin of NHFPC, Minister Yin and Mongolia; and Toomas Palu, global Weimin of MoHRSS, and World Bank Group practice manager for the Bank’s HNP GP in President Jim Yong Kim. WHO Director- the East Asia and Pacific Region. General Margaret Chan provided valuable Valuable advice was provided by the mem- leadership and guidance at the initiation as bers of the team’s external advisory panel: well as at the critical junctions of the study. In Michael Porter, Bishop William Lawrence particular, Vice Premier Liu hosted two spe- University Professor at the Institute for cial hearings on the progress and main find- Strategy and Competitiveness, Harvard ings for the study in March of 2015 and 2016. Business School; Donald Berwick, president Under the overall leadership of Minister emeritus, senior fellow, and former president Lou (MoF) and World Bank Managing and chief executive officer of the Institute for Director and Chief Operating Officer Sri Healthcare Improvement and former admin- Mulyani Indrawati, the report was overseen istrator of the Centers for Medicare and by a joint team from the five participating Medicaid Services; Winnie Yip, professor of organizations, led by the following: MoF vice health policy and economics at the Blavatnik ministers Yaobin Shi and Weiping Yu; School of Government, University of Oxford; NHFPC vice ministers Zhigang Sun and Ellen Nolte, coordinator of the European Xiaowei Ma; MoHRSS Vice Minister Jun Observatory at the London School of You and former vice minister Xiaoyi Hu; Economics and Political Science and the WHO Regional Director for the Western London School of Hygiene & Tropical Pacific Shin Young-soo; WHO representative Medicine; Yanfeng Ge, director-general, to China Bernhard Schwartländer; Vivian Department of Social Development Research, Lin, director, Division of Health Sector Development Research Center of the State Development , W HO Wester n Pacif ic Council, China; and Shangxi Liu, director- Regional Office; Axel van Trotsenburg, general, Chinese Academy of Fiscal Sciences. H E A LT H Y C H I N A xv xvi H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A A technical working group (TWG), con- La Forgia, Todd Lewis, Hannah Ratcliffe, sisting of technical leads from each of the and Asaf Bitton (chapter 3); Rabia Ali, Todd government agencies as well as the World Lewis, Hannah Ratcliffe, Asaf Bitton, and Bank and WHO, was formed at the begin- Gerard La Forgia (chapter 4); Gerard La ning of the study. The TWG has led technical Forgia, Antonio Duran, Jin Ma, Weiping Li, communications, provided important com- and Stephen Wright (chapter 5); Mukesh ments, and facilitated research data from dif- Chawla and Mingshan Lu (chapter 6); Shuo ferent departments of the Ministries. Its Zhang and Edson Araújo (chapter 7); Karen members included Yao Licheng, Peng Xiang, Eggleston, Barbara O’Hanlon, and Mirja Wang Lei, and Wang Min (MoF); Jiao Yahui, Sjoblom (chapter 8); James Cercone and Zhao Shuli, Zhuang Ning, Liu Yue, Qin Kun, Mukesh Chawla (chapter 9); and Kedar S. and Chen Kai (NHFPC); Wang Guodong, Mate, Derek Feeley, Donald M. Berwick, and Z h ao Z h i hong , a nd S ong C he ng ji n Gerard La Forgia (chapter 10). Mukesh (MoHRSS); Gerard La Forgia, Shuo Zhang, Chawla, Joy De Beyer, Aakanksha Pande, and Rui Liu (World Bank); and Martin Rachel Weaver, and Ramesh Govindaraj did Taylor, Wen Chunmei, and Stephanie Dunn the technical and content editing of the final (WHO). report. Within the World Bank, task team leaders Case studies and background studies Gerard La Forgia and Mukesh Chawla were drawn from 21 provinces, autonomous received significant on-the-ground support regions, and municipalities in China: Beijing, from Shuo Zhang as well as from Elena Shanghai, Tianjin, Chongqing, Sichuan, Glinskaya, Daixin Li, and Rui Liu in the Yunnan, Guizhou, Ningxia, Qinghai, Anhui, World Bank’s Beijing office. Martin Taylor Shandong, Guangdong, Jiangsu, Jiangxi, was the core team member from WHO, with Henan, Zhejiang, Hubei, Hunan, Fujian, support from Clive Tan, Ding Wang, and Tuo Xiamen, and Shenzhen. Other case and back- Hong Zhang of WHO and from Edward Hsu ground studies were drawn internationally and Jiadi Yu of the World Bank Group’s from Brazil, Denmark, Germany, the International Finance Corporation (IFC). Netherlands, New Z ealand, Nor way, Mickey Chopra, Jeremy Veillard, Enis Baris, Portugal, Singapore, Turkey, the United and Patrick Osewe served as the World Kingdom, and the United States. Bank’s internal peer reviewers of the study Many international and China experts con- reports. Valuable inputs were received from tributed through these studies. By chapter, the Simon Andrews (IFC) and Hong Wang (Bill contributors comprised Hui Sin Teo, Rui Liu, and Melinda Gates Foundation). The joint Daixin Li, Yuhui Zhang, Tiemin Zhai, study team also acknowledges media coordi- Jingjing Li, Peipei Chai, Ling Xu, Yaoguang nation work from Li Li; translation and Zhang, David Morgan, Luca Lorenzoni, Yuki proofreading work from Shuo Zhang, Rui Murakami, Chris James, Qin Jiang, Xiemin Liu, and Tianshu Chen; editing work from Ma, Karen Eggleston, and John Goss (chapter Rui Liu and Tao Su; and the tremendous 1); Zlatan Sabic, Rong Li, Rui Liu, Qingyue administrative support from Tao Su, Sabrina Meng, Jin Ma, Fei Yan, Sema Safir Sumer, Terry, Xuan Peng, Lidan Shen, Shunuo Chen, Robert Murray, Ting Shu, Dimitrious and Xin Feng. Kalageropoulous, Helmut Hildebrandt, and The report was prepared and coordinated Hubertus Vrijhoef (chapter 2); Xiaolu Bi, under the technical leadership of Gerard La Agnes Couffinhal, Layla McCay, and Forgia. The chapter authors comprised Tania Ekinadose Uhunmwangho (chapter 3); Rabia Dmytraczenko, Magnus Lindelow, Ye Xu, Ali, Todd Lewis, Hannah Ratcliffe, Asaf and Hui Sin Teo (chapter 1); Asaf Bitton, Bitton, and Gerard La Forgia (chapter 4); Madeline Pesec, Emily Benotti, Hannah Weiyan Jian, Gordon Guoen Liu, and Baorong Ratcliffe, Todd Lewis, Lisa Hirschhorn, and Yu (chapter 5); Christoph Kurowski, Cheryl Gerard La Forgia (chapter 2); Ye Xu, Gerard Cashin, Wen Chen, Soonman Kwon, Min Hu, A c k no w ledgment s xvii Lijie Wang, and Alex Leung (chapter 6); Yujun Jin, Chen Ren, Rui Zhao, Liang Ye, Guangpeng Zhang, Barbara McPake, Xiaoke Chen Meili Zhang, and Ru Yuhong Xiaoyun Liu, Gilles Dussalt, and James (N H F PC); Qinghui Yan, Shuchun Li, Buchan (chapter 7); Jiangnan Cai, Yingyao Chengjin Song, Jun Chang, Yutong Liu, Chen, Qiulin Chen, Ian Jones, and Yi Chen G uo do n g Wa n g , Z h e n g m i n g D u a n , (chapter 8); Dan Liu (chapter 9); and Aviva Yongsheng Fu, Kaihong Xing, Wei Zhang, Chengcheng Liu (chapter 10). Chinese officials Jiayue Liu, and Chao Li (MoHRSS); Yanfeng who provided significant support with the Ge and Sen Gong (Development Research coordination of field studies and mobilization Center of the State Council); Shangxi Liu of research data included Licheng Yao, Xiang (MoF Academy of Fiscal Sciences); Hongwei Peng, and Yan Ren (MoF); Ning Zhuang, Kun Yang, Zhenzhong Zhang, and Weiping Li Qin, Rui Zhao, and Chen Ren (NHFPC); and (N H F PC China National Health officials in provinces. Development Research Center [CNHDRC]); During the study preparation, six techni- Dezhi Yu, Junwen Gao, Lijun Cui, and Beihai cal workshops and several consultative Xia (Anhui Commission for Health and roundtables were organized with active par- Family Planning [CHFP]); Dongbo Zhong ticipation from the MoF, NHFPC, MoHRSS, and Haichao Lei (Beijing CHFP); Xiaochun and select provincial governments. These Chen, Wuqi Zeng, and Xu Lin (Fujian workshops served as platforms for reciprocal CHFP); Xueshan Zhou and Shuangbao Xie policy dialogue and for receiving timely and (Henan CHFP); Patrick Leahy and Henrik constructive feedback from the government Pederson (IFC); Xiaofang Han, Qingyue partners and researchers on the preliminary Meng, Gordon Liu, Jiangnan Cai, Asaf study findings. The following leaders, offi- Bitton, Jin Ma, Wen Chen, James Cercone, cials, and experts made presentations and Ian Forde, Barbara O’Hanlon, Karen important contributions to the discussions: Eggleston, Fei Yan, Guangpeng Zhang, Shaolin Yang, Guifeng Lin, Shixin Chen, Xiaoyun Liu, Qiulin Chen, Min Hu, Lijie Yingming Yang, Qichao Song, Haijun Wu, Wang, Antonio Duran, and Dan Liu (World Aiping Tong, Weihua Liu, Licheng Yao, Bank consultants); and Bang Chen, Junming Yuanjie Yang, Yu Jiang, Wenjun Wang, Lei Xie, Roberta Lipson, Beelan Tan, Sabrinna Wang, Xuhua Sun, Fei Xie, Xiang Peng, Lei Xing, Jane Zhang, Alex Ng, Yuanli Liu, Zhang, Min Wang, Yi Jiang, Shaowen Zhou, Jianmin Gao, Baorong Yu, Mario Dal Poz, Qi Zhang, and Chenchen Ye (MoF); Yan James Buchan, Ducksun Ahn, and Stephen Hou, Wannian Liang, Minghui Ren, Chunlei Duckett. Nie, Yuxun Wang, Wei Fu, Jinguo He, Feng The study team recognizes and appreciates Zhang, Shengguo Jin, Jianfeng Qi, Hongming additional funding support from the Bill & Zhu, Yang Zhang, Ruirong Hu, Ning Melinda Gates Foundation via its Results for Zhuang, Changxing Jiang, Liqun Liu, Yilei Development Institute and from the IFC. The Ding, Yue Liu, Ling Xu, Kun Qin, Ge Gan, joint study team is also grateful for all contri- Zhihong Zhang, Yongfeng Zhu, Kai Chen, butions and efforts from any individuals and Yi Wang, Jianli Han, Yan Chen, Xiaorong Ji, teams not named above. Abbreviations ACC-AHA CVD American College of Cardiology and American Heart Association cardiovascular disease ADHC Ageing, Disability, and Health Care AEHG Aier Eye Hospital Group AHRQ Agency for Healthcare Research and Quality (United States) ARCH Automated Record for Child Health (Boston, United States) BHLC Better Health at Lower Cost BMIs basic medical insurances BoG Board of Governors (foundation trusts, United Kingdom) BoHRSS Bureau of Human Resources and Social Security (China) BRIICS Brazil, Russian Federation, India, Indonesia, China, and South Africa BSC balanced scorecard CAPEX capital expenditure CCGs clinical commissioning groups (United Kingdom) CDC Center for Disease Control and Prevention (United States) CDSS computerized decision support systems CEC Clinical Excellence Commission (Australia) CEO chief executive officer CHA Chaoyang Hospital Alliance (Beijing, China) CHC community health center China CDC Chinese Center for Disease Control and Prevention CHS community health station CHWs community health workers CIF capital investment fund CIP capital investment planning CME continuing medical education CMS Centers for Medicare & Medicaid Services (United States) CON Certificate of Need H E A LT H Y C H I N A xix xx H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A CONU Certificate of Need Unit COPD chronic obstructive pulmonary disease CPC Communist Party of China CQI continuous quality improvement CVA cerebrovascular accident DALY disease-adjusted life year DMC district medical center DRC Development and Reform Commission (provincial) DRGs diagnosis-related groups ECG electrocardiogram EDL Essential Drug List EHR electronic health record FACS Family and Consumer Services FCH Foshan Chancheng Hospital FDS family doctor system FHS Family Health Strategy (Brazil) FTs foundation trusts (United Kingdom) GDP gross domestic product GH Great Health (Zhenjiang, Jiangsu Province) GIS geographic information system GP general practitioner HAS Haute Autorité de Santé (France) HASU hyperacute stroke unit (England, United Kingdom) HCA health care alliance HFPC Health and Family Planning Commission (provincial) HHS Department of Health and Human Services (United States) HIRA  Health Insurance Review and Assessment Service (Republic of Korea) HMC hospital management council/center HSR health services research ICT information and communication technology IFC International Finance Corporation (World Bank) IHI Institute for Healthcare Improvement (United States) IMAI Integrated Management of Adolescent and Adult Illness IOM Institute of Medicine (United States) IPCD Insurance Program for Catastrophic Diseases IQWiG Institute for Quality and Efficiency in Health Care (Germany) IT information technology JCUH James Cook University Hospital (England, United Kingdom) LG leadership group LLG local leading group M&E monitoring and evaluation MBS Medicare Benefits Schedule (Australia) MDT multidisciplinary team A b b re v iation s xxi MFA Medical Financial Assistance MI myocardial infarction MoCA Ministry of Civil Affairs MoF Ministry of Finance MoH Ministry of Health MoHRSS Ministry of Human Resources and Social Security MoLSS Ministry of Labor and Social Security MQCCs medical quality control committees MSA medical savings account MSAC Medical Services Advisory Committee (Australia) MSMGC Medical Service Management and Guidance Center (of NHFPC) NCD noncommunicable disease NCMS New Cooperative Medical Scheme NCQA National Committee for Quality Assurance NDP  National Demonstration Project on Quality Improvement in Health Care (United States) NDRC National Development and Reform Commission NGO nongovernmental organization NHFPC National Health and Family Planning Commission NHIA National Health Insurance Administration (Taiwan, China) NHIS National Health Insurance Service (Republic of Korea) NHS National Health Service (United Kingdom) NICE  National Institute for Health and Care Excellence (United Kingdom) NPDT National Primary Care Development Team (United Kingdom) NPO nonprofit organization NQF National Quality Forum (United States) NRCMS New Rural Cooperative Medical Scheme NSW New South Wales (Australia) OECD Organisation for Economic Co-operation and Development OSS social health organization (Brazil) P4Q pay-for-quality PACE Program of All-Inclusive Care for the Elderly (United States) PACS  Community Health Agents Program (Programa de Agentes Comunitários de Saúde, Brazil) PACS picture archiving and communications system PACT Patient-Aligned Care Team (U.S. Veterans Health Administration) PAD peripheral artery disease PCG primary care group PCIC people-centered integrated care PCMH patient-centered medical home PCT primary care trust PDCA plan-do-check-act (cycle) PDSA plan-do-study-act PFP private-for-profit xxii H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A PHC primary health care PHIFMC Public Health Insurance Fund Management Centre (Sanming, China) PLG provincial leading group PNFP private-not-for-profit PPP public-private partnership PPP purchasing power parity PSA public service announcement PSU public service unit QoC quality of care QOF Quality and Outcomes Framework (United Kingdom) RHS Regional Health System (Singapore) RMB renminbi SCHRO State Council Health Reform Office SES Secretariat of Health, State Government of São Paulo SHI social health insurance SHINe Singapore Healthcare Improvement Network SIKS Integrated Effort for People Living with Chronic Diseases SOE state-owned enterprise SPHCC  Strengthening Primary Health Care Capacity (Feixi County, Anhui Province) SPSP Scottish Patient Safety Programme SRE serious reportable event (NQF, United States) SROS  Regional Strategic Health Plan (Schéma Régional d’Organisation Sanitaire, France) SU stroke unit (England, United Kingdom) TCM traditional Chinese medicine TFY Twelfth Five-Year Plan (Hangzhou, Zhejiang Province) THC township health center THE total health expenditure TLC Transformative Learning Collaborative TQM total quality management UEBMI Urban Employee Basic Medical Insurance UHC universal health coverage ULS unidades de saúde local (Portugal) URBMI Urban Resident Basic Medical Insurance VAT value added tax VC venture capital VHA Veterans Health Administration (United States) WAHH Wuhan Asia Heart Hospital WHO World Health Organization WMS World Management Survey WOFI wholly owned foreign investment Note: All dollar amounts are U.S. dollars unless otherwise indicated. Executive Summary China’s Health System embarking on a high-value path to better health at an affordable cost. Following decades of double-digit growth that lifted more than 600 million people out of poverty, China’s economy has slowed in Reform Initiatives and Benefits recent years. The moderating growth adds a China has already launched major reform ini- new sense of urgency to strengthening human tiatives to improve health sector performance capital and ensuring that the population and meet the expectations of its citizenry. In remains healthy and productive, especially as 2009, the government unveiled an ambitious the economy gradually rebalances toward national health care reform program, com- services and the society experiences shifting mitting to significantly raise health spending demographics and disease burdens. An area to provide affordable, equitable, and effective that demands particular attention in this con- health care for all by 2020. Building on an text is health care, which is critical not only earlier wave of reforms that established a to improving equity but also to ensuring that national health insurance system, the 2009 people live healthier as they live longer. reforms, supported by an initial commitment Furthermore, slower economic growth of RMB 850 billion, reaffirmed the govern- opens the door for much-needed reforms in ment’s role in the financing of health care and the health sector, because continuing on the provision of public goods. present path would be both costly and unaf- After nearly six years of implementation, fordable: government expenditures on health the 2009 reforms have made a number of (including health insurance) would increase noteworthy gains: they have achieved near-­ threefold, to about 10 percent of China’s universal health insurance coverage at a gross domestic product (GDP) by 2060, in speed with few precedents. Benefits have been the absence of cost containment measures, gradually expanded, use of health services but these expenditures would be kept to has increased, and out-of-pocket spending on under 6 percent of GDP if adequate reforms health—a major cause of impoverishment for are undertaken. China now faces an oppor- low-income populations—has fallen. Indeed, tunity to rebalance its health care system by since 2009, the average life expectancy at 1 2 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A birth today has increased by more than care delivery in China. Since 2005, health 30 years; it took rich countries twice that care spending in China has been growing at a long to achieve the same gains. rate of about 5–10 percentage points higher than GDP growth. Affordability of health services remains a concern to both citizens Health Service Delivery Challenges and government. Although out-of-pocket China now faces emerging challenges in expenditures have declined significantly in meeting its citizens’ health care needs associ- recent years, they remain high, at 29.9 percent ated with a rapidly aging society and the of total spending, compared with an average increasing burden of noncommunicable dis- of 21 percent in high-income countries. Social eases (NCDs). The trends of reduced mortal- insurance funds are already under increasing ity and fertility have led to a rapidly aging pressure to not run into debt. society, while social and economic transfor- Although spending growth started from a mation have brought urbanization and life- comparatively low level, the trend is not likely style changes, in turn leading to emerging to reverse in the near future because expendi- risk factors of obesity, sedentary lifestyles, ture pressures will continue to grow. For stress, smoking, abuse of alcohol and other example, addressing the health needs of mil- substances, and exposure to pollution. lions of people with diabetes, hypertension, NCDs are already China’s number one and other chronic diseases who are currently health threat, accounting for more than undiagnosed and not receiving any care will 80 percent of the 10.3 million premature be costly. deaths annually and 77 percent of disability- However, China also needs to address the adjusted life years (DALYs) lost in 2010, not low-value and cost-escalating aspects of its far off the share in Organisation for Economic delivery system. China faces major challenges Co-operation and Development (OECD) in transforming its hospital-centric and countries of 83 percent. Importantly, volume-driven delivery system into one that ­ 39.7 ­percent (males) and 31.9 percent (females) delivers high-quality care at affordable costs at of all NCD deaths in China are “prema- all levels and that meets peoples’ demands and ture”—that is, under the age of 70—­ compared expectations. Motivated by profits and poorly with 27.2 percent (males) and 14.7 ­ percent governed, too many public hospitals are (females) in Japan and 37.2 ­ percent (males) embodiments of both government and market and 25.1 percent (females) in the United States. failures. Health financing is fragmented, and For populations aged 30–70 years, the proba- insurance agencies have remained largely pas- bility of dying from cardiovascular disease, sive purchasers of health services. cancer, diabetes, or chronic respiratory disease As for the quality of care, information is is 19.5 percent in China, compared with 9.3 limited, but available evidence suggests that percent in Japan and 14.3 percent in the there is significant room to improve. A short- United States. age of qualified medical and health workers These trends add to the complexity China at the primary care level compromises the is facing and to which the health system must health system’s ability to carry out the core respond by reducing the major risk factors for functions of prevention, case detection, early chronic disease; addressing those influences treatment of common illnesses, referral, care that drive exposure to these risk factors, integration, and gatekeeping. including the environment; and ensuring the China is transforming its capital invest- provision of services that meet the require- ment planning from input-based parameters ments of those with chronic health problems. (which tend to focus on bed numbers and The 2009 reforms produced substantial facility size) to parameters that are based on positive results in expanded insurance cover- population served. The government has also age and better health infrastructure, but opened the hospital sector to private invest- much still needs to be done to reform health ment, but the private sector’s ability to E x ec u ti v e S u mmary 3 improve access and quality care is con- FIGURE ES.1  Life expectancy relative to per capita health strained by China’s weak regulatory and expenditure, selected countries, early 2010s public purchasing environment. Japan, 74 Average life expectancy in years, 2010 Singapore, The Health Expenditure Outlook 72 Spain Although China is still a comparatively low 70 spender on health care, it needs to avoid the 68 trap observed in several OECD countries of China rapidly increasing health spending combined 66 Netherlands, Luxembourg, with only marginal gains in health outcomes. 64 United States A high-cost path will result in two or three 62 times the per capita spending of the low-cost path and will not necessarily lead to signifi- 60 cantly better outcomes. 0 2,000 4,000 6,000 8,000 Although factors other than health care Health expenditure per capita (PPP), 2012 and health spending contribute to health out- comes, it is instructive that the United States Source: Economist Intelligence Unit 2014; WHO (various years). Note: PPP = purchasing power parity. has a poor-value health care system, spend- ing nearly $9,000 per capita (at purchasing power parity [PPP]). Singapore has a rela- 15,805 billion in 2035—an average increase tively high-value system, spending under of 8.4 percent per year. This will increase $4,000 per capita and achieving better health health expenditure from 5.3 percent of GDP outcomes and higher life expectancy than the in 2015 to 9.1 percent of GDP in 2035. United States (figure ES.1). Under the business-as-usual scenario, more This does not mean that China should than 60 percent of the growth in health expen- emulate one system over the other. The start- diture is expected to be in inpatient services. ing points and contexts are substantially dif- Inpatient expenditure will grow by RMB ferent. As China continues to grow, an incon- 7,915 billion, compared with outpatient venient truth is that health spending will expenditure growth of RMB 3,328 billion, increase. However, the rate of increase can be pharmaceutical expenditure growth of RMB controlled by prudent choices as to the orga- 1,256 billion, and growth of other health nization and production of health services expenditure of RMB 155 billion. and the efficient use of financial and human China could, however, achieve significant resources. savings—equivalent to 3 percent of GDP—if Doing nothing is not an option. A study it could slow down the main cost drivers commissioned by the World Bank and carried (­ f igure ES.2). To realize these savings, the out with researchers from China concluded growth in hospitalization needs to come that business as usual will result in real health down and use of outpatient care needs to go expenditure growth of 9.4 percent a year from up. This implies strengthening the primary 2015 to 2020, during which GDP was pro- care system, raising people’s confidence in the jected to grow by 6.5 percent a year.1 In the health system outside of the hospital setting, period 2030–35, during which GDP is pro- providing high quality people-centered care jected to slow down (this study uses 4.6 per- that is integrated across all levels, and enrich- cent per year as the basis for projection), ing people’s experience with the health care health expenditure will grow by 7.5 percent system. The potential for savings also allows per year. In other words, without deepening for affordable fiscal space for needed invest- the health reform, health expenditure in ments into people-centered integrated care China will increase in real terms (2014 prices) that would be well below the potential sav- from RMB 3,531 billion in 2015 to RMB ings to be achieved. 4 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE ES.2  Main drivers of projected health expenditure in China, 2015–35 10 9.4 8.7 8.4 8 8 7.5 Health expenditure growth, percent 4.9 4.8 4.3 6 4.2 3.8 4 2.5 2.4 2.4 2.4 2.4 2 1.5 1.2 1.3 1.4 1.4 0.6 0.3 0.04 –0.01 0.2 0 –0.01 –0.01 –0.01 –0.1 –0.01 –2 2015–20a 2020–25a 2025–30a 2030–35 2015–35a Disease prevalence rate Unit cost Population aging Population growth Services per caesb Annual growth rate Source: World Bank estimations. a. Growth rates of −0.01 percent pertain to population growth, which has been negative. b. “Services per case” refers to the number of outpatient visits per disease episode—or hospital discharges, in the case of inpatient services— across 19 disease categories. The Health System Reform Outlook These policy directives contain the fun- damental components of service delivery Recognizing these challenges, China’s leaders reform and emphasize strengthening the have adopted far-reaching policies to put in three-tiered system (including primary place a reformed delivery system. On October care and community-based services), insti- 29, 2015, the 18th Session of the Central tuting human resources reform, optimiz- Committee of the Fifth Plenary Session of the ing use of social insurance, and encourag- Communist Party of China (CPC) endorsed a ing private investment in health care. The national strategy known as “Healthy China,” policies also support “people first” princi- which places population health improvement ples such as as the primary strategic goal of the health system. This strategy has guided the planning and implementation of health reforms under • Building harmonious relationships with the 13th Five-Year Development Plan, patients; 2016–20. • Promoting greater care integration The government has also initiated enabling between hospitals and primary care facili- legislative actions. The Basic Health Care ties through tiered service delivery and use Law—which will define the essential ele- of multidisciplinary teams and facility ments of the health care sector, including networks; financing, service delivery, pharmaceuticals, • Shifting resources toward the primary and private investment—has been included in level; the legislative plan of the National People’s • Linking curative and preventive care; Congress of China and is being formulated • Reforming public hospital governance; and by the congress. • Strengthening regional service planning. E x ec u ti v e S u mmary 5 However, although important progress has engagements that are aligned with public pri- been observed, it is mostly limited to pilot orities. However, international experience projects, which suggests the need to strengthen suggests that these tools should be sufficiently implementation and emphasize scaling-up. strong and transparent—and that govern- China already has a mixed health delivery ment should possess adequate enforcement system comprising both public and private and data monitoring capacity—to defend the providers, and this system requires strong gov- public interest and avoid policy and regula- ernment steering to deliver on government tory capture by powerful private (and public) objectives. In this context, the role of the gov- actors. ernment at both the central and provincial lev- els needs to shift from top-down administra- tive management of services and functions Report Background and Structure through mandates and circulars (a remnant of This report was proposed by Chinese Premier the “legacy system”) to indirect governance, Li Keqiang at a July 2014 meeting with the whereby the government guides public and World Bank Group President Jim Yong Kim private providers to deliver health services and and World Health Organization (WHO) results aligned with government objectives. Director-General Margaret Chan. It is a Currently—and despite policy directives product of joint initiatives of five institutions: mandating separation of functions in the China’s Ministry of Finance, Health and health sector—the government is still Family Planning Commission, and Ministry involved in multiple functions, including of Human Resources and Social Security; the oversight, financing, regulation, manage- World Bank; and WHO. It has two objec- ment, and service provision. In contrast, tives: (a) provide advice on core actions and many OECD countries are converging on a implementation strategies in support of health delivery model in which the govern- China’s vision and policies on health reform, ment plays a larger role in financing, over- and (b) contribute technical inputs into the sight, and regulation and a relatively limited implementation of the 13th Five-Year role in direct management and service Development Plan. provision. The report is based on 20 commissioned What matters, however, are the policy background studies; more than 30 case stud- instruments and accountability mechanisms ies from China, middle-income countries, used to align organizational objectives with and OECD countries on various themes; vis- public objectives. Tools include grants, con- its to 21 provinces in China; six technical tracts, regulations, public information and workshops; and inputs from a diversified disclosure rules, independent audits, and tax team of policy makers, practitioners, acade- policies, among others. Some are already in micians, researchers, and interested stake- use in China. Other core government func- holders who came together to dissect, ana- tions in a mixed delivery system include lyze, and discuss the main sectoral reform establishing public purchasing arrangements, areas in this intensive two-year effort. guiding health service and capital investment The report consists of 10 chapters, the first planning, setting and enforcing quality stan- summarizing the major health and health dards and monitoring, regulating public and system challenges facing China and provid- private hospitals, accrediting medical profes- ing a rationale for the recommendations sionals and facilities, and creating a system of detailed in this report. The next eight chap- medical dispute resolution. ters constitute the main body of the report as By using these tools, the government follows: defines public and private roles, creates a level playing field for public and private pro- • “Lever 1: Shaping Tiered Health Care viders, and develops a path for more formal- Delivery with People-Centered Integrated ized and transparent public and private Care” 6 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A • “Lever 2: Improving Quality of Care” representing a comprehensive package of • “Lever 3: Engaging Citizens in Support of interventions to deepen health reform. Each the PCIC Model” lever contains a set of recommended core “ Lever 4: Reforming Public Hospital •  action areas and corresponding implementa- Governance and Management” tion strategies to guide the “what” and • “Lever 5: Realigning Purchasing and “how” of deepening service delivery reform; Provider Incentives” the action areas and strategies are meant to • “Lever 6: Strengthening the Health provide policy guidance at all governmental Workforce” levels. • “Lever 7: Strengthening Private Sector The levers are conceptualized to be inter- Engagement in Health Service Delivery” linked and are not designed to be imple- • “Lever 8: Modernizing Health Service mented as independent actions (figure ES.3). Planning to Guide Investment” For example, actions taken by frontline health care providers will require strong T he f i nal chapter, “St reng t hen i ng institutional support combined with financial Implementation of Health Service Delivery and human resource reforms to achieve the Reform,” focuses on implementation and reform goals. scaling-up. Based on the broader implemen- tation literature, it describes an actionable implementation “system” framework and Service Delivery Levers corresponding strategies relevant to the First, and at the core of the recommenda- Chinese context to promote effective and tions, is the full adoption of a reformed ser- scalable implementation. vice delivery model—referred to as people- c e nte red integ rated c are ( PC IC) — to accelerate progress toward China’s vision of The Recommendations health service delivery reform and improve The report proposes eight sets of strategic value for money. PCIC refers to a care deliv- reform directions, referred to as “levers,” ery model organized around the health needs FIGURE ES.3  Eight interlinked levers to deepen health care reform in China Service Delivery System People-centered integrated care (PCIC) models Rebalanced and value-based health service Reform public Increase citizen Improve quality delivery hospitals and engagement in and personal improve links to health and health experience delivery system care “8-in-1” Reform levers Institutional and Financial Environment Realign incentives Develop a qualified in purchasing and and motivated • Better health provider payment health workforce • Higher quality and patient satisfaction Reorient and better • A ordable costs Strengthen private manage service and sector engagement capital planning E x ec u ti v e S u mmary 7 of individuals and families. The bedrock of a chronic diseases, and initial forms of gate- high-performing PCIC model is a strong pri- keeping. Families who contract with the sys- mary care system that is integrated with sec- tem are assigned to a general practitioner ondary and tertiary care through formal who works with a team to manage care for links, good data, information sharing among 800–1,000 families. Empanelment is also an providers and bet ween providers and integral underlying feature of small-scale but patients, and the active engagement of successful delivery models in Germany (the patients in their care. It uses multidisciplinary Gesundes Kinzigtal integrated care system); teams of providers who track patients with Canterbury, New Zealand (the Health e-health tools, measures outcomes over the Services Plan); and the United States (Patient- continuum of care, and relentlessly focuses Centered Medical Homes). As in Shanghai, on continually improving quality. Curative empanelment is voluntary, and patients can and preventive services are integrated to pro- opt out at any time. vide a comprehensive experience for patients In the PCIC approach, health services are and measurable targets for facilities. Large integrated across provider levels and across secondary and tertiary hospitals have new space, time, and information through alli- roles as providers of complex care and leaders ances or networks. The networks responsible in workforce development. Measurement, for implementing PCIC function on a monitoring, and feedback are based on up- “3-in-1” principle: one system, one popula- to-date, easily available, validated data on tion, one pool of resources. In rural areas, the the care, outcomes, and behaviors of provid- 3-in-1 principle can be applied to the tiered ers and patients. network consisting of village clinics, town- Primary care is a central organizing para- ship health centers (THCs), and county hos- digm for the production of key health system pitals, while in the cities the networks will functions. International experience suggests consist of community health stations, com- that no country can provide high-quality, munity health centers (CHCs), and district person-centered care at lower costs without a hospitals. There can be multiple networks in robust primary care system. In China, front- cities and counties, which would allow for line village, township, and district health patient choice. In geographically dispersed facilities need to continue to be strengthened areas, networks can be established virtually and better staffed to provide an attractive or through contracting arrangements. PCIC model. Improved frontline facilities can In China, the current tiered delivery sys- provide a gatekeeping role for hospital and tem was designed to operate as an integrated specialized services while providing better network. However, separate organization follow-up care for recently discharged and management, loose definition of provider patients. Empanelment can be used to iden- functions across tiers, constrained financial tify reference populations (for example, dia- flows, and fragmented governance arrange- betics) who will receive care by a team of pro- ments have limited the ability to integrate viders who create registries of such patients service provision and provide more continu- to facilitate proactive management and a ous care. Nevertheless, well-organized, inte- population-based approach to care.2 In some grated networks of tiered service delivery— areas, primary care providers are sufficiently such as those emerging in Zhenjiang, Feixi strong to perform these functions. In others, (Anhui Province), and Huangzhong (Qinghai some functions will need the support of Province)—should be mainstreamed. county and district hospitals and can be Within each network, the functions and gradually transferred to primary care once responsibilities of each provider level need to capacity is strengthened. be clearly defined. This will necessarily The “family doctor” system in Shanghai involve shifting low-complexity care out of and other Chinese cities is already piloting hospitals. Initially, at least, networks need to empanelment, registries of patients with avoid incorporating or being operated by 8 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A tertiary hospitals, only because the tertiary many countries (such as the National Institute hospitals might use the network to capture for Health and Care Excellence in the United additional patients rather than to shift low- Kingdom and the National Quality Forum in complexity care to lower levels. Avoiding the United States). China could consider “hospital capture” is also important to pro- establishing a similar agency that, reporting mote the strengthening of primary care and to the central government, would be respon- service integration. sible for coordinating all efforts geared However, secondary hospitals will be toward quality assurance and improvement important network members in terms of and would actively engage all stakeholders in providing technical support and training for implementing quality assurance and improve- the network. Initially, county and district ment strategies. It would develop standard- hospitals will play a strong technical role in ized clinical pathways and oversee their network operations and implementing implementation in clinical practice, set qual- PCIC, in part because these facilities already ity standards, accredit and c­ ertify both public have good working relationships with pri- and private providers, measure and track per- mary care providers in many areas in China. formance, conduct research, and otherwise A networkwide managerial unit will be build capacity in advancing health care qual- responsible for selecting, deploying, and ity. The agency would ideally be co-led by supervising resources in the most efficient representatives of relevant ministries and key way possible to achieve network objectives. professional and scientific bodies and would Ideally, this management unit should be the include other stakeholders such as community executive arm of a governance structure and representatives. be separate from the government adminis- Third, recognizing the key role of patient trative apparatus. trust for the success of the PCIC model, the Second, to improve the quality of care, the report recommends that patients be actively report recommends that a regulatory author- engaged and empowered in the process of ity be established that provides a high level of seeking care through measures that increase technical oversight. PCIC requires strong their knowledge and understanding of the government leadership and stewardship for health system. Optimal use of scarce building capacity to improve the quality of resources requires that patient preferences health care. A regulatory entity would pro- shape decisions about investment and disin- mote scientific, evidence-based medicine by vestment in services, which in turn requires a developing standardized clinical pathways two-way communication between multidisci- and overseeing their implementation in clini- plinary clinical teams and their patients. cal practice. It would also be a key resource Without this exchange, decisions are made for clinical practitioners to access a range of with avoidable ignorance at the front lines of clinical, public health, and social care infor- care delivery, services fall short of meeting mation, including safe practice guidelines, needs while exceeding wants, and efficiency technology appraisals and guidance, quality declines over time. standards, and implementation tools. Quality The report recommends strong patient improvement is recognized as a continuous engagement and self-management practices to effort, which will require continuous moni- help patients manage their conditions. Patient toring and benchmarking of health care ser- self-management refers to patients’ active par- vice delivery and building up the perfor- ticipation in their treatment and providers’ mance information infrastructure to monitor consideration of patient treatment prefer- progress. ences. It offers a more collaborative approach A coordinated institutional architecture in which providers and patients work together committed to helping the nation improve to identify problems, set p­ riorities, establish health care quality and to overseeing related goals, create treatment plans, and solve efforts is increasingly the path followed by issues. Patient self-­ m anagement involves E x ec u ti v e S u mmary 9 systematically educating patients and their Hospitals will continue to play an impor- families about their conditions, how to moni- tant role, but one that becomes less finan- tor them, and how to incorporate healthy cially dominant and more focused on pro- behaviors into their lifestyles. It also involves viding only the specialized services that are training of clinicians to communicate better most needed. As the capacity of primary with patients. By promoting systems for care is strengthened and the PCIC model is patient self-management, health systems can put in place, a wide range of care processes empower individuals to reduce their utiliza- will be shifted out of hospitals to ambula- tion of and make more informed decisions tory units (such as surgical and chemother- r elating to office visits, medication, and ­ apy units) and primary care facilities. procedures. Hospitals will become centers of excellence Several of the case studies commissioned but with adequate volume to deliver high- for this report exemplify such patient engage- quality care. They can perform important ment and self-management approaches, training and workforce development func- including the following: tions. They can also focus more on biomed- ical research and providing clinical support • In Shanghai, the “family doctor” system to lower-level providers. encourages patients and families to jointly Some of these functions are slowly rolling set treatment goals with their providers, out in China. Existing “alliances” in China and monthly patient satisfaction scores already show the potential benefits of these track progress. organizational forms; their use and further • In Germany, Gesundes Kinzigtal (a health development should be considered. In Feixi, care management company whose name county hospitals and THCs share medical translates as “healthy Kinzig valley”) resources and personnel as well as coordinate emphasizes joint treatment goal setting services between local THCs and their asso- and attainment. Shared decision-making ciated village clinics. Similarly, Huangzhong tools augment this process along with case built local alliances to use county hospital managers who support the patient through resources to strengthen and integrate THCs. their conditions and behavior changes. Zhenjiang leveraged key county and ­ academic • In the United States, the Veterans Health hospital resources to set up more integrated Administration encourages self-­ management rehabilitation care. Importantly, payment through disease-specific action planning and schemes need to be adjusted to support these intensive education, especially around medi- functions. cation management. The report suggests comprehensive gover- • In Denmark, the SIKS (Integrated Effort nance arrangements to improve performance for People Living with Chronic Diseases) of public hospitals and promote their integra- project prioritizes patient involvement in tion into the service delivery system. A number developing their own treatment plans, set- of countries where public hospitals historically ting goals through shared care plans, and were directly administered (including Brazil, providing feedback about whether these the Netherlands, Norway, Spain, the United goals were met in partnership with the Kingdom, and others) have taken steps to care team. grant them greater independence. These steps include the following: Fourth, the report suggests deep reforms in the governance and management of • Granting hospitals full autonomy to man- public hospitals to improve their perfor- age all assets and personnel, including ­ civil mance in cost control, quality of care, and service or “quota” staff (for example, to patient satisfaction. Reforming hospitals is hire, dismiss, and determine compensation) part and parcel of reforming service deliv- • Developing independent hospital gover- er y a nd adapt i ng P C IC -l i ke models. nance boards with government and 10 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A nongovernment participation to oversee China has taken many important steps in hospital management and performance recent years to build the role of health pur- • Appoi nt i ng professiona l ma nagers chasing agencies, develop their institutional through a merit-based selection process capacities, and test innovative contracting (although sometimes subject to a consulta- and provider payment approaches. Hence it is tive process with government) well positioned to build on the experiences of • Enacting laws defining the nature of the many successful pilots and experiments— autonomy and specifying board selection, both within and outside the country—to fur- membership, and functions; definition of ther leverage the power of strategic purchas- social function and obligations; separation ing and put in place a set of incentives that of functions between board and manage- motivates providers at all levels. Suggested ment; financial arrangements; and report- core actions include adopting volume-­ ing and other accountability requirements controlled, value-driven approaches to effec- (such as an annual independent audit). tively manage the growth of expenditures; making incentive mechanisms coherent and China may want to consider regulating consistent across the system; rationalizing the public hospitals under a broader legal frame- distribution of services by facility level; and work setting the attributes, accountabilities strengthening the capacity of purchasing (discussed below), and requirements of non- agencies. profit (and for-profit) health care organiza- The report proposes a realignment of tions. Such legislation could also address the incentives within a single, uniform, network- issue of hospital-based “quota” employees wide design in support of population health, and criteria to access social insurance. quality of care, and cost containment. Evaluations have shown that public hospitals Prospective payment is more effective than operating with this full range of decision fee-for-service for improving efficiency and rights frequently perform better than public quality and incentivizing PCIC-based deliv- hospitals that are managed hierarchically by ery. For these mechanisms to work, they must government administration. International (a) be defined and applied consistently across and Chinese experience provide good exam- the full continuum of health care production ples of road maps for improving autonomy. and delivery, from primary care to tertiary For example, China’s Dongyang Hospital interventions; and (b) be aligned so that all manifests many of these features. Other providers, including hospitals, physicians, emerging but less autonomous hospital gover- and health centers, fall within their purview. nance models are also evident in Zhenjiang, Some of the different options for reorient- Shanghai, and other cities. ing incentives are being tested in China. To move public hospitals away from being profit centers to being public-interest entities, the Institutional and Financial report suggests changing how physicians are Environment Levers paid in hospitals and linking their remunera- Fifth, the report makes a strong case for tion to a metric of public interest built around realigning purchasing and provider incen- measures of quality, patient satisfaction, and tives in the health system to motivate the serving vulnerable populations. These mea- establishment of PCIC, strengthen primary sures are consistent with the government’s health care delivery, and integrate services May 2015 policy directive requiring that across the entire spectrum of health care. public hospitals operate for the public good Effectively leveraging the power of strategic instead of seeking lucrative gains and that purchasing, contracting, and paying provid- health services be accessible, equal, and effi- ers could improve the value of the govern- cient for the people. ment’s large investment in the health sector in Sixth, the report recommends strengthen- China and achieve greater value for money. ing the health workforce in China to enable E x ec u ti v e S u mmary 11 the implementation of a PCIC service workers i n r u ra l a nd remote a reas. ­d elivery model. Covering the domains of International experiences suggest that finan- production, recruitment, compensation, cial incentives alone cannot always provide management, regulation, and performance sufficient motivation, and nonfinancial incen- evaluation, the suggested core actions include tives have an important role in meeting spe- raising the status of primary health care cial needs. Commonly used options including workers, paying them well, strengthening rotating housing, job opportunities for their composition and competencies, spouses, and opportunities for further train- and building an effective framework for ing (scholarships for college-level studies, in- governance and regulation of the health ­ service training, and so on). Professional iso- workforce. lation can be avoided by using communication Building a strong enabling environment technologies that facilitate knowledge shar- for the development of the primary health ing with other providers. care (PHC) workforce is key to implementing Seventh, the report recognizes that the PCIC model. To raise the status of pri- although China has formulated several poli- mary care, general practice must be estab- cies to encourage private sector engagement lished as a specialty with equivalent status to in the health system, much remains to be other medical specialties and with the same done to integrate the private entities into the attributes of well-regulated standards of national health system and motivate them to practice. This will require building a consen- deliver good-quality health services that sus and shared understanding among govern- improve the lives and health of China’s pop- ment, health providers, and the public of the ulation. Measures suggested include (a) the centrally important role of primary care enunciation and adoption of a shared vision together with hospitals in providing the full of the private sector’s potential contribution continuum of care to the citizens. to national health system goals; (b) regula- China may like to consider introducing tions that better align private sector health primary-care-specific career development services with social goals; and (c) the estab- prospects to develop and incentivize the pri- lishment of a level playing field for the public mary care workforce. This strategy includes and private sectors to better promote active separate career pathways for general practi- private sector engagement. Through this tioners, nurses, mid-level workers, and com- approach, the incentives and conditions munity health workers that enable career under which the whole health sector operates progression within PHC practice. Current would move China toward a well-integrated pilots of a separate accreditation for rural world-class health system that yields better assistant physicians as well as a separate pro- health outcomes and financial protection for fessional title promotion system for PHC the nation’s investments in health. workers are good examples of this approach. The private sector can play many impor- The report also proposes reforming the tant roles in the production and delivery of compensation system to provide strong incen- health services, and it is important that China tives for good performance. The compensa- articulate a clear vision to steer the course of tion system needs to be revised to reduce reli- private engagement. If properly harnessed, ance on service revenue-based bonuses and to the private sector can deliver value through increase base salaries and hardship allow- business model innovation and a commit- ances. Although a combination of fixed ment to quality and transparency. The pri- payment with variable performance-based ­ vate sector can contribute most effectively in payments is desirable, the latter should focus ­ areas where the public sector is currently on quality improvements (for example, pay- weakest and where market forces can play an for-quality schemes). important role—that is, where patients can In addition, nonfinancial incentives should make informed choices, as in the case of be introduced to attract and retain health long-term care, home care, and so on. 12 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A In areas where patients are typically not able Provincial governments should receive clearer to make choices, the expansion of the private guidance on private sector planning, entry sector should be gradual and cautious as well requirements, surplus use, and other commu- as predicated upon the establishment of a nity service requirements, and enforcement strong purchasing function. In other words, should be strictly monitored. Likewise, the China should leverage the potential gains private sector should be assured that it will that involvement of the private sector in enjoy treatment similar to public institutions health would bring but be careful not to get in such aspects as access to health profession- into a situation that would make it difficult als, land use, equipment purchasing, desig- to reverse course. nated medical insurance, and professional China should also adopt policies and regu- title appraisal. A critical factor to leveling the latory measures to guide private sector playing field is ensuring that social insurance engagement and minimize the risks associ- payments follow the patients to their chosen ated with growth of poor-quality private pro- providers. As social insurers continue to viders. The private sector in China and strengthen their purchasing functions, the abroad contains examples of business models government could consider introducing equal that deliver high quality at low cost, as well contracting standards and payment principles as poor models that rely on overprescribing (“pay for quality, not quantity”) for both services, false advertising, and cherry-picking public and private providers for health ser- patients and thus fail to serve social interests. vices. This will encourage a virtuous circle, As China moves from a wholly public system where both public and private providers spur to one of mixed delivery, it needs to have in each other toward achieving better value. place the right regulation and payment incen- Eighth, the report recommends a funda- tives to motivate all health providers to oper- mental change in how capital investment ate in the best medical and social interest, decisions are made in China’s health sector irrespective of whether they are publicly or by modernizing health service planning. privately owned. Indeed, China is at a critical More specifically, the report suggests moving point in private sector development and must away from traditional input-based planning avoid many of the pitfalls encountered in toward capital investments based upon other countries as they opened up their health region-specific epidemiological and demo- sectors. China will need to consider the full graphic profiles. Shifting from a strategy that range of regulatory instruments—including is driven by macro standards to one that is legal prohibition, disclosure rules, industry determined by service planning based on real self-regulation, and audits—to foster private population needs will help China better align engagement in the health sector in areas its huge capital investments—projected to where it can best serve the social interest and reach $50 billion annually by 2020—with to deter companies with vested interests from the demands of an affordable, equitable influencing hospital (and physician) behav- health care system and achieve value for iors, whether for-profit, nonprofit, or public. money for its massive investments in the Through appropriate regulation and over- health sector. sight, China can accelerate the shift in the Moving from capital investment planning private sector from low-quality to high-­ to a people-centered service planning model quality private providers. will require prioritization of public invest- Private and public providers of health ser- ments according to burden of disease, where vices should be subject to the same set of people live, and the kind of care people need rules and regulations. Licensing a private on a daily basis. Service planning offers the facility remains cumbersome, unpredictable, opportunity to remake the health provider and costly compared with public facilities network—its design, culture, and practices— and to a large extent depends on the whims to better meet the needs of patients and and will of local government officials. families and the aspirations of those who ­ E x ec u ti v e S u mmary 13 provide their care. Within this service plan- guidelines and reduce excess capacity and ning approach, capital investment planning duplication in the network. (which is necessary to optimally use funding Another practice to consider is periodic opportunities such as insurance and public issuance of specific guidance on implement- reimbursements) can guide the development ing standards and investment appraisals—a of facilities of the future, change the status drill that OECD countries with advanced quo of today, and ensure that excess capacity capital planning processes (such as Australia is not created to further exacerbate ineffi- and the United Kingdom) routinely carry out ciency and capital misallocation Allowing by issuing “green papers” on various policies population needs to drive service and capital to support local authorities in interpreting investment planning will make an important those policies. correction in the current system and will Finally, China should consider setting pro- direct delivery of health services toward a vincial caps on capital spending or “ear- people-centered model. marked” allocations by level of care to pro- Countries that have strong planning mote new development of ambulatory t raditions, such as France and the United ­ solutions for surgery, chemotherapy, dialysis, Kingdom, follow a needs-based planning imaging, and so on that would reduce the approach linked to specific health challenges. need for hospital beds and expensive infra- These countries incorporate demographic structure and bring services closer to the and epidemiological considerations in devel- people. oping their service plans, and they factor in private sector capacity in planning for a bal- ance between market demand and supply. Implementation of Health Care This approach allows them to focus on inte- grated networks delivering services for Reforms defined catchment populations, allocate capi- The report’s final chapter addresses the cen- tal funds to provinces to acquire and upgrade tral challenge of how to implement the physical assets such as property and equip- important changes suggested in the eight ment, and correct for equity and the level of levers and recommends tools to operational- population vulnerability. ize and sustain the core actions and imple- Ensuring that available assets deliver the mentation strategies suggested. It presents an most cost-effective delivery solution requires operational framework that focuses on four the development of a regulatory framework “implementation” systems: macro implemen- that directs capital investment away from tation and influence, coordination and sup- expansion and toward deepening of the exist- port, service delivery and learning, and moni- ing infrastructure’s capacity to better meet toring and evaluation. Recognizing the strong the population’s health needs. This regula- association between high-quality implemen- tory framework should encourage integrated tation and the probability of obtaining better capital planning and allocation across sectors program performance, it recommends estab- of care to capture the potential cost and qual- lishing an enabling organizational environ- ity advantages of integration. In addition, ment as a precondition for effective imple- capital planning needs to be integrated into a mentation. Without it, progress may be medium-term expenditure framework to elusive. bring together planning and budgeting, Transforming the commitment of central- strengthen capital spending by facilitating level leadership to deepening health care multiyear funding programs, and incorporate reforms by operationalizing a value-based the operation and maintenance costs of delivery system will require (a) defining cen- investments into expenditure projections. At tral and local governmental roles within a the same time, planning standards should be policy implementation framework and tightened to close loopholes in the existing (b) putting in place the right governance, 14 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A organizational, and shared learning plat- political support and be fully empowered forms. Despite a consensus that China’s (and accountable) to support reform imple- reform policies are sufficiently robust, most mentation within their jurisdictions. observers acknowledge that the country has The proposed councils should consist of had difficulty translating these policies into representatives from the various government scalable and sustained actions. Current insti- agencies involved in the health sector as well tutional fragmentation and vested interests as representatives from providers, the private make it difficult to maintain or scale up even sector, and community leaders. The councils effective pilots. Appropriate governance, should be held accountable to the central gov- organizational, and shared learning plat- ernment through central-local intergovern- forms are key preconditions to effective mental performance or “task” agreements implementation and represent the critical first that specify implementation benchmarks, steps in the prioritization and sequencing of anticipated results of the reforms, and, ulti- interventions necessary to build a modern mately, population health indicators. These 21st-century health system. These platforms implementation performance measures will need strong and persistent central gov- should also be incorporated into the career ernment support to make them work. promotion system for provincial and local The central government must take the lead leaders. Importantly, the councils will direct in guiding and overseeing implementation of government agencies involved in human the reforms, including the eight levers. China resource management, planning, and financ- may like to consider assigning this mandate ing to enact the changes required to create an to the State Council, which would prepare a enabling environment for the reforms. uniform policy implementation framework to C h i na m ay a lso w ish to create orient reform planning and execution by Transformation Learning Collaboratives local governments. (TLCs) at the network and facility levels as This framework would not be a one-size- the fundamental building blocks to imple- fits-all blueprint but would need to be opera- ment, sustain, and scale up reforms on the tional in nature, specifying categorically front line. The shift in organizational goals what to do as well as what not to do. In turn, from a treatment orientation to an outcome local governments would need to have full orientation will require f u ndamental authority to decide on how to do what needs changes in organizational culture. Health to be done, including developing, executing, care organizations—whether networks, and sequencing implementation plans based hospitals, CHCs, or THCs—must adopt on local conditions but according to the pol- continuous learning and problem-solving icy implementation framework specified by approaches to encourage innovation and a the State Council. new culture of care. At the same time, pol- Strengthening accountability arrange- icy guidance from national and provincial ments, particularly at the provincial and local officials will need to be customized and levels, is another essential ingredient to facili- adapted at the front lines of service deliv- tate effective implementation. Any gover- ery. The service delivery model envisioned nance arrangement should be sufficiently for China includes several impor tant powerful to align institutional standpoints changes at care sites, as specified under and to leverage government interests when Lever 1 (shaping service delivery with dealing with providers and vested interests. PCIC). Although these changes can and One solution is to form empowered leader- should be driven by national, provincial, ship groups or councils at the provincial or and local leadership, implementing them at prefecture levels led by government leaders local sites will require assistance for local (governors and mayors). A few such councils learning, problem solving, and adaptation. already exist in China. At any level, the coun- The driving vision behind the TLC con- cils will require strong leadership and cept is to assist and guide local care sites E x ec u ti v e S u mmary 15 (such as village clinics, THCs, CHCs, and The most successful clinics and centers county and district hospitals) to implement would become mentors and coaches to those and scale up the reformed service delivery that are struggling. Using both meetings and model and close the gap between knowing ongoing virtual exchanges, participating sites and doing. A TLC is a structure for rapidly will help each other to overcome barriers disseminating better practices for change to and accelerate their progress. Such learning all facilities in a network, whether in a county alliances have been successfully applied to or city. Each TLC can be organized as a support service delivery reform in England, short-term (12–15-month) learning system, Scotland, Singapore, Sweden, and the United which brings together teams from each par- States. ticipating facility, ideally within a specific Implementing suitable reform pathways in network. an ordered manner has the potential to begin Before launching a TLC, for example, rebalancing China’s health system toward county network officials agree to the slate people-centered integrated care. Done rigor- of interventions that will be implemented as ously and with effective learning, measure- well as a set of measures to track the imple- ment, and feedback loops, these reform mentation progress of all participating facili- pathways have the potential to improve the ­ ties (and institutions). The facility-level teams quality and efficiency of key services deliv- meet face-to-face in “learning sessions” every ered across the entire system. four to six months to discuss implementa- tion successes, barriers, and challenges; share better practices; and describe lessons learned. ­ Notes In between these face-to-face meetings are “action periods” when facility teams test 1. According to the latest World Bank growth estimates (issued June 2018), China grew by and implement interventions in their local 6.9 percent in 2017 and is projected to grow s ettings—and collect and report data to ­ by 6.5 percent in 2018, 6.3 percent in 2019, measure the impact of these reforms. Teams and 6.2 percent in 2020. submit regular progress reports for the entire 2. Empanelment is the process by which all TLC to review and are supported by site visits, patients in a given facility or geographic area conference calls, and other web-based discus- are assigned to a primary care provider or sions facilitated by implementation experts. care team. Introduction China’s Road to Health conditions no longer represents a poverty- Care Reform inducing shock for rural ­ residents. Massive investments in health infrastruc- Deepening health sector reform is arguably ture and human resource formation at the one of the major social undertakings facing grassroots level and significant expansion of C hina. In 2009, China unveiled an ambi- ­ access to basic public health services have tious national health care reform program, fueled impressive ­ g ains. For example, committing to significantly raise health health insurance coverage has stayed above spending to provide affordable, equitable, 95 percent. Service capacity has increased, use ­ and effective health care for all by ­ 2 020. of health services has risen, and out-of-pocket Building on an earlier wave of reforms that spending as share of total health expenditures established a national health insurance sys- has fallen, leading to more equitable access to tem, the 2009 reforms, supported by an care and greater ­ affordability. For example, initial financial commitment of R M B by 2014 the reimbursement rates were raised 1,380 billion, reaffirmed the government’s for inpatient services under the three main role in the financing of health care and the social insurance schemes (Urban Employee provision of public ­ goods. Basic Medical Insurance [UEBMI] scheme, After nearly six years of implementation, Urban Resident Basic Medical Insurance the 2009 reforms have made many notewor- [URBMI], and New Cooperative Medical gains. Among them, China has achieved thy ­ Scheme [NCMS]); consequently, the differ- near universal health coverage (UHC) at a ences between them significantly nar- speed with few precedents globally or rowed, reaching 80 percent, 70 percent, and historically. Benefits have also been gradually ­ 75 percent, r ­ espectively. Twelve categories of e xpanded. For example, the New Rural ­ basic public services, including care for sev- Cooperative Medical Scheme (NRCMS), eral chronic conditions, are now covered free which targets rural populations, has become charge. The essential drug program is of ­ more comprehensive, incrementally adding helping to reduce irrational drug use and outpatient benefits while including coverage improving access to effective d ­ rugs. The for specific ­ d iseases. Treatment for many reforms, including subsequent regulations, 17 18 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A have encouraged greater private sector partici- exposure to pollution and traffic a ­ ccidents. pation, in part to reduce overcrowding in pub- NCDs are already China’s number one facilities. The governments have spent lic ­ health ­threat. huge sums on the construction of primary These trends add to the complexity health care f ­acilities. The capacity of pri- China is facing and to which the health sys- mary health care services has been greatly tem will have to respond by reducing the ­strengthened. major risk factors for chronic disease; Finally, the 2009 reforms also spear- addressing those influences that drive expo- headed hundreds of innovative pilots in sure to these risk factors (such as the envi- health financing, public hospitals, and ronment); and ensuring the provision of grassroots service delivery— several of services that meet the requirements of those which are examined in this report—and with chronic health p ­ roblems. R ising provided a strong foundation for the next incomes and levels of education contribute stage of ­ r eform. China is progressing to population demands for more and better quickly to achieving UHC, and some of the health s ­ ervices. China’s health system will reforms’ achievements have attracted world- be judged on how well it handles these new wide ­attention. ­challenges. China needs to avoid the risks of devel- oping a high-cost, low-value health system Challenges Ahead (box ­ I .1). The health system is hospital-­ China now faces emerging challenges in meet- centric, fragmented, and v ­ olume-driven. ing the health care needs of its citizens, associ- Cost-inducing provider incentives and lack ated with a rapidly aging society, the increas- of attention to quality are major system ing burden of noncommunicable diseases ­ s hortcomings. The delivery system has a (NCDs), and the rising prevalence of risk bias toward doing more treatment rather ­ factors. The trends of reduced mortality and than improving population health out- fertility have led to a rapidly aging society, comes and toward admitting patients to while social and economic transformation hospitals rather than treating them at the has brought urbanization and changed life- primary care ­ l evel. Services are uninte- styles, leading to the emerging risk factors of grated (or uncoordinated) across provider obesity, sedentary lifestyles, stress, smoking, tiers (tertiary, secondary, and primary) and abuse of alcohol and other substances, and between preventive and curative ­ s ervices. BOX ­I.1  What is value in health care? Value is defined as health outcomes for the money making effective links between health care and spent (Porter ­2010). Others offer a more expanded health ­outcomes. definition involving a combination of better out- “Low-value care” refers to services that have little comes, quality and patient safety, and lower costs or no benefit in terms of health outcomes, are clinically (Yong, Olsen, and McGinnis ­ 2 010). In terms of ineffective or even harmful, and are cost-ineffective reform or change strategies to improve health ser- (compared with ­ alternatives). The term encompasses vices, value involves “shift[ing] the focus from the multiple concepts (and terms) that contribute to excess volume and profitability of services provided— costs, low-quality care, and poor health outcomes, physicians visits, hospitalizations, procedures, including inappropriate care, unsafe care, unnecessary and [diagnostic] tests—to the patient outcomes care, overutilization, misuse, overtreatment, overdiag- achieved” (Porter 2010, ­ 3). The concept involves nosis, missed prevention opportunities, and ­waste. I ntrod u ction 19 The high prevalence of NCDs suggests that The government has also initiated enabling care is ­suboptimal. legislative ­actions. The Basic Health Care In addition, health financing is institution- Law—which will define the essential ele- ally fragmented, and insurance agencies have ments of the health care sector including remained passive purchasers of health financing, service delivery, pharmaceuticals, services. Effective engagement with the pri- ­ private investment, and so on—has been vate sector is in its infancy, and service plan- included in the legislative plan of the National ning has not been ­ modernized. There is a People’s Congress of China and is being shortage of qualified medical and health formulated by the ­ congress. workers at the primary care level, which fur- The CPC Central Committee Suggestions ther compromises the system’s ability to carry for the 13th Five-Year Development Plan and out the core functions of prevention, other recent policy directives (Guo Wei Ji case detection, early treatment, and care Ceng Fa N ­ o. 93, 2015) contain the funda- ­integration. mental components of service delivery r eform. For example, policies emphasize ­ strengthening the three-tiered system The Next Stage: “Healthy China” (including primary care and community- Recognizing these challenges, China’s leaders based services), instituting human resources have adopted far-reaching policies to put in reform, optimizing use of social insurance, place a reformed delivery s­ ystem. On October and encouraging private investment (“social 29, 2015, the 18th Session of the Central capital”) to support health ­ care. Committee of the Fifth Plenary Session of the The government policies also support Communist Party of China (CPC) endorsed a “people first” principles such as national strategy known as “Healthy China,” which places population health improvement • Building harmonious relationships with as the main system g­ oal. This strategy guided patients; the plan n ing and i mplementation of • Promoting greater care integration between health reforms under the 13th Five-Year hospitals and primary care facilities through Development Plan, 2016–20 (State Council tiered service delivery and use of multidisci- 2016), as noted in box ­I.2. plinary teams and facility networks; BOX ­I.2  Communist Party of China’s endorsement of “Healthy China” strategy “China will deepen the reform of the medical and of medical resources to the grassroots level and health systems, promote the interaction of medical rural areas, and promote work concerning general services, health insurance, and pharmaceutical sup- practitioners, family doctors, and the medical ser- ply, implement the tiered delivery system, and estab- vice capacity of highly needed areas, and electronic lish primary care and modern health care systems medical ­records. that cover both urban rural ­ areas. “Efforts should be made to encourage social “Efforts should be made to optimize the layout forces to develop the health service industry, pro- of medical institutions, improve the medical service mote the equal treatment of non-profit private hospi- system featuring the interaction and complementar- tals and public hospitals, strengthen supervision and ity of higher and lower levels of institutions, improve control of medical quality, improve mechanisms for the model of medical service at the grassroots level, dispute resolution, and build harmonious relations develop distance medical service, promote the flow between doctors and ­ patients.” Source: “Suggestions of the CPC Central Committee on the 13th Five-Year Plan for National Economic and Social Development on the Promotion of 42–43. a ‘Healthy China’” (English translation), ­ 20 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A • Shifting resources toward the primary than more ­ t reatment. This would suggest level; shifting the focus from rewarding volume • Linking curative and preventive care; and sales to rewarding health outcomes as • Reforming public hospital governance; well as achieving more value for the money and ­ spent. It would also suggest paying particular planning. • Strengthening regional service ­ attention to providing affordable and equita- ble health care for all population groups, so These are some of the essential features that the poor and disadvantaged people do and supporting elements of a value-driven not face the risks of catastrophic medical delivery system that incorporates a new ser- spending and forgo medical care because of vice delivery model, the full adoption of ­unaffordability. which will facilitate achievement of China’s Shifting from a health care delivery system vision of service delivery ­reform. However, focused on production of treatments to one although important progress has been focused on value and producing health sug- observed, it is mostly limited to pilot proj- gests a strategic agenda that aligns all ects, which suggests the need to strengthen stakeholders and works toward three goals: implementation and emphasize ­ scaling-up. attaining better health for the population, Acknowledging the difficulty of implement- providing better quality and care experience ing these reforms and the time required for individuals and families, and achieving to achieve scale, they are collectively affordable ­costs. referred to as reforms of the emerging “deepwater” ­phase. China also faces an unenviable conun- Report Objectives and Audience drum: as its economy slows down, health The objective of this report is to provide spending is not likely to follow ­ suit. Indeed, advice on core actions and implementation as the population ages and new technologies strategies in support of China’s vision and are further integrated into preferred treat- policies on health reform, particularly in rela- ment options, the upward pressures on health tion to service ­delivery. A more immediate spending will become even more ­ pronounced. objective is to contribute technical inputs for In the face of these opposing trends, China the implementation of the 13th Five-Year will soon need to come up with a new model Development ­Plan. of health production, financing, and delivery, There is much to learn from national and which responds to the needs and expecta- international innovations and experiences to tions of its population but at the same time is successfully reform service d­ elivery. In China, grounded in the economic reality of today, for example, many successful pilot initiatives based on the economic new ­ normal. have not yet been scaled u ­ p. These initiatives China has already decided that doing represent opportunities that China can build nothing is not an option: continuing to pro- upon to shape a world-class service delivery vide quality health services under the current ­ system. At the same time, China can draw on arrangement will result in increasing health the experience of Organisation for Economic costs and a heavier burden on the state exche- Co-operation and Development (OECD) quer or households or b ­ oth. In fact, because countries that are reshaping their health reforms take time to work their way through delivery systems to address similar challenges the complex health care system, the time to posed by chronic diseases, aging populations, implement and scale up transformative mea- and cost ­ pressures. Drawing on commis- sures is now, before it gets too late and even sioned case work and analysis as well as the more ­expensive. broader literature, the report summarizes les- In moving forward with the delivery sons learned from Chinese and international reforms, China should consider maintaining experiences and recommends actions to its focus on achieving more health rather support policy ­implementation. I ntrod u ction 21 The report is intended for central- and Report Structure provincial-level policy makers and regulators as well as planners and implementers at the Chapter 1 summarizes the major health and local level, including insurers and ­ providers. health system challenges facing China and Policy makers may want to focus on the rec- provides a rationale for the recommendations ommended levers and corresponding core detailed in this r­ eport. More-specific chal- actions. The strategies for central and pro- ­ lenges are highlighted in each of the subse- vincial government proposed in the imple- quent chapters according to ­ theme. mentation model (described in chapter 10) The next eight chapters (chapters 2–9) would also be of interest to this ­ g roup. constitute the main body of the report Meanwhile, planners and implementers can (table I.1): each chapter concentrates on a ­ center their attention on the core actions and single “lever” or strategic direction to sup- corresponding specific implementation port the planning and implementation of the s trategies. They would also benefit from ­ government’s vision of ­ reform. The levers aim the frontline elements of the proposed imple- to provide policy implementation guidance to mentation ­model. all governmental ­ levels. Each lever contains a Before proceeding, a couple of caveats are set of recommended core action areas and in order: This study centers on reforms to corresponding implementation strategies to improve health service delivery and the sup- guide the what and how of deepening service porting financial and institutional environ- delivery ­reform. ment in ­China. Resource and time constraints These levers are interlinked and should did not allow for analysis of other important not be considered or implemented as inde- reform themes that can be the subjects of pendent sets of a ­ ctions. To be sure, actions future ­research. These include China’s phar- taken by frontline providers will require maceutical industry, its tobacco industry, the strong institutional support combined with education and licensing of medical profes- financial and human resource reforms to sionals, traditional Chinese medicine (and its achieve the triple goals (attaining better integration with Western medicine), and dis- popu lat ion hea lt h , b et ter qu a l it y of semination and use of medical ­ technologies. care, and affordable c ­ osts). In short, the Some of the links between aged care, health eight levers represent a comprehensive care, and social services in China will be package of interventions to deepen health taken up in a forthcoming World Bank s ­ tudy. ­reform. TABLE ­I.1  Deepening Health Reform in China report chapters Chapter ­no. Chapter title 1 Impressive Gains, Looming Challenges 2 Lever 1: Shaping Tiered Health Care Delivery with People-Centered Integrated Care 3 Lever 2: Improving Quality of Care 4 Lever 3: Engaging Citizens in Support of the PCIC Model 5 Lever 4: Reforming Public Hospital Governance and Management 6 Lever 5: Realigning Incentives in Purchasing and Provider Payment 7 Lever 6: Strengthening the Health Workforce 8 Lever 7: Strengthening Private Sector Engagement in Health Service Delivery 9 Lever 8: Modernizing Health Service Planning to Guide Investment 10 Strengthening the Implementation of Health Service Delivery Reform 22 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Service Delivery Levers integrated to provide a comprehensive expe- rience for patients and measurable targets for How health services are organized and ­ facilities. Hospitals have new roles as provid- delivered—and how providers relate to each ers of complex care and leaders in workforce other and to ­ p atients—matters. “People- development. They also adopt more robust ­ centered integrated care” (PCIC) is the term governance arrangements and management used to refer to a health care delivery model practices. Measurement, monitoring, and ­ organized around the health needs of indi- feedback are based on up-to-date, easily viduals and f ­amilies. PCIC—also discussed available, and validated data on the care, in the World Health Organization’s (WHO) outcomes, and behaviors of providers recently proposed “WHO Global Strategy of and ­patients. People-Centered and Integrated Health Internationally, many countries are imple- Services” (WHO 2015a, 2015b)—comprises menting PCIC-like models to address chal- characteristics that seek to achieve better lenges similar to those facing China: cost health and better quality at affordable costs, escalation, questionable quality of care, and or in other words, more value for the money stagnant gains in health ­ outcomes. Australia, ­ odel. How ­spent. It is not a one-size-fits-all m Brazil, Canada, Denmark, Germany, New PCIC is implemented in practice depends on Zealand, Singapore, the United Kingdom, local ­conditions. and the United States are some of the coun- Based on the WHO strategy and the tries testing reformed service delivery models broader literature, PCIC involves several stra- that incorporate features of P ­ CIC. Though tegic directions, referred to as “levers,” at the expanding rapidly, PCIC-like approaches service delivery level (covered in chapters 2–5): remain local or regional in most of these 1. Reorienting the model of care, particu- countries. Preliminary results suggest that ­ larly in terms of strengthening primary gains can be made in outcomes, quality, health care, changing the roles of hospi- and cost containment, but results vary tals, and integrating providers across care considerably within and across c ­ ountries. levels and among types of services Implementing these reforms at scale would 2. Continuously improving the quality of care make China a world leader in reform service 3. Engaging people to make better decisions delivery and place it at the vanguard in health about their health and health-seeking system innovation and development with behaviors insightful lessons for many ­ countries. 4. Improving the governance and manage- ment of ­hospitals. Financial and Institutional Environment Levers Broadly, the bedrock of a high-performing PCIC model is a strong primary care system Establishing an enabling institutional envi- that is integrated with secondary and ­ tertiary ronment together with strengthening incen- care through formal links, good data, and tives and accountabilities are underlying but information sharing among providers and recognized drivers of successful PCIC imple- between providers and patients, and active mentation and improved service delivery engagement of patients in their ­care. It uses globally (WHO ­ 2015b). China is no ­different. multidisciplinary teams of providers who Implementation and sustained development track patients with e-health tools, measures of service delivery reform in China will outcomes over the continuum of care, and require fundamental shifts in incentives, relentlessly focuses on improving q ­ uality. capabilities, and accountabilities, especially Feedback and audit mechanisms ensure in the ways that services are purchased, pro- continuous learning and quality improve- ­ viders are paid, people are reimbursed, and ment. Curative and preventive services are providers report on performance and are held I ntrod u ction 23 accountable for better care and alignment contracts, regulations, public information with public ­priorities. and disclosure rules, independent audits, and These shifts will require strong gover- tax policies, among ­ others. Some are already nance arrangements and sustained high-level in use in ­China. Other core government func- government ­ support. The success of PCIC, tions in a mixed delivery system include for example, will depend on improving the establishment of public purchasing arrange- primary care workforce, raising compensa- ments, guidance of health service and capital tion and competencies of primary care investment planning, setting and enforcement clinicians, and reforming human resource ­ of quality standards and monitoring, regula- management ­ practices. The implementation tion of public and private hospitals, accredi- of service delivery reform will also be tation of medical professionals and facilities, enhanced through developing more effective and creation of a system of medical dispute forms of public-private ­ engagement. Finally, ­resolution. new approaches to service and capital invest- By using these tools, the government ment planning will be required to align defines public and private roles, creates a investment planning with the new service level playing field for public and private pro- delivery ­model. Realigning incentives, devel- viders, and develops a path for more formal- oping a qualified and motivated workforce, ized and transparent public and private strengthening private sector engagement, and engagements that are aligned with public improving capital and service planning are priorities. However, international experience ­ the levers taken up in chapters ­ 6 –9. suggests that these tools should be sufficiently China already has a mixed health delivery strong and transparent—and that govern- system consisting of both public and private ment should possess adequate enforcement providers, and this system requires strong and data monitoring capacity—to defend the government steering to deliver on public public interest and avoid policy and regula- objectives. In this context, the role of the gov- ­ tory capture by powerful private (and public) ernment, at both the central and provincial ­actors. levels, needs to shift from top-down adminis- trative management of services and functions Moving Forward with Implementation through mandates and circulars (a remnant of the “legacy system”) to indirect governance The final chapter concludes with recom- whereby government guides public and pri- mended strategies, coordination arrangements, vate providers to deliver health services and and organizational platforms to facilitate sus- results aligned with government ­ objectives. tained implementation and full scaling-up of Currently—and despite policy directives reforms. Based on the broader implementation ­ mandating separation of functions in the literature, it describes an actionable implemen- health sector—the government is still involved tation “system” framework and correspond- in multiple functions, including oversight, ing strategies relevant to the Chinese con- financing, regulation, management, and ser- text to promote effective and scalable provision. By comparison, many OECD vice ­ i mplementation. Recommendations on the ­ countries are converging on a health delivery sequencing and timing of rollout to reach full model in which the government plays a large scale are also ­provided. role in financing, oversight, and regulation and a relatively limited role in direct manage- Supporting Case Studies and ment and service ­ provision. Appendixes What matters, however, are the policy instruments and accountability mechanisms Finally, case studies commissioned for this used to align organizational objectives with report. study are referenced throughout the ­ public ­ o bjectives. Tools include grants, For details about each of the case studies, 24 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A 2B. In addition, at the see chapter 2, annex ­ State ­ Council. ­ 2016. “The 13th Five-Year Plan for end of the volume, appendix A presents the Economic and Social Development of the following supplementary tables: People’s Republic of China ­ (2016–2020).” Translated by the Compilation and Translation Bureau, Central Committee of the Communist • Table ­A.1: Eight Levers and Recommended Party of ­ China. Beijing: Central Compilation Core Actions & Translation ­ http://en.ndrc.gov.cn​ P ress. ­ • Table ­A .2: Government Policies in Support / ­p olicyrelease /201612 / P0201612076457​ of the Eight Levers 66966662.pdf. • Table A.3: New Policy Guidelines on WHO (World Health O ­ rganization). ­ 2 015a. Tiered Service Delivery and Recommended “People-Centered and Integrated Health Core Actions Services: An Overview of the ­ Evidence. Interim ­R eport.” Report ­W HO/ HIS/SDS/2015.7, WHO, ­Geneva. References ———. 2 ­ 015b. “WHO Global Strategy on People- Guo Wei Ji Ceng ­ 2015. “NHFPC Opinions F a. ­ Centered and Integrated Health Services: on Further Standardizing the Management of Interim R ­ eport.” Report W ­ HO/ HIS/SDS​ Community Health Services and Improving /2015.6, WHO, ­ Geneva. ­ ervice.” Guo Wei Ji Ceng Fa, the Quality of S Yong, ­ P. L ­ ., ­ L. A ­ . Olsen, and ­ M. McGinnis, ­ J. ­ eds. ­No. ­93. ­2010. Value in Health Care: Accounting for Porter, Michael ­E. ­ 2 010. “What Is Value in C ost, Q u alit y, S afet y, O utcomes, and Health Care?” New England Journal of ­I nnovation . Washington, DC: National Medicine 363 (26): ­2477–81. Academies ­Press. 1 Impressive Gains, Looming Challenges Introduction same time, the increase in government subsi- dies to social insurance schemes has increased China has been a pioneer in primary care and the utilization of health services and reduced public health, and, more recently, in universal the share of out-of-pocket spending in total insurance coverage. The introduction of bare- health expenditures. foot doctors,1 community or workplace health Notwithstanding recent accomplishments, insurance, and ambitious public health cam- however, the health care system in China is paigns drove improvements—combined with experiencing major challenges that are con- higher incomes, lower poverty, and better liv- tributing to cost escalation, increasing citizen ing standards (sanitation and water quality, discontent, and threatening future health sys- education, nutrition, and housing)—that sig- tem gains: nificantly reduced mortality rates and brought an unprecedented increase in life expectancy • Emerging demographic and epidemio- (Caldwell 1986; Yang and others 2008). logical trends, including a rapidly aging The health sector in China has undergone population and the onslaught of noncom- ­ a series of wide-ranging reforms aimed at pro- municable diseases (NCDs) and corre- viding affordable, equitable, and effective sponding risk factors health care. Social health insurance was intro- • Quality of care issues, including insuffi- duced in phases beginning in 1998, first cov- cient attention to measuring and improv- ering formal sector workers, then expanding ing the quality of health care service to the rural population in 2003, and finally delivery extending to informal sector workers, chil- • Internal systemic factors, such as the top- dren, and the elderly in urban areas in 2007. heavy structure of the delivery system and Health spending rose significantly in 2009, unreasonable provider incentives, that all when t he gover n ment i njec ted R M B contribute to rising costs 850 ­billion into the financing of health care. As a result, China has achieved near-­universal This chapter first reviews these three health insurance coverage at a speed that has ­ hallenges and then examines the resulting c few precedents, reaching over 95 percent in inefficiencies and their potential spending both urban and rural areas since 2011. At the implications. 25 26 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A China’s Changing Health Care changes are creating a rapidly aging popula- Needs tion. The “graying” of the population has profound implications for China’s economic Driven by improvements in disease prevention and social policies and places new demands and access to medical care—as well as by on the health system to deliver care that higher incomes, lower poverty, and better liv- ensures that people live healthy longer lives. ing standards (stemming from progress in In 2013, China had 202 million people aged sanitation and water quality, education, nutri- 60 years or older, who made up 15 percent of tion, and housing), China has achieved a rapid its total population. 2 Their number is decline in mortality and an unprecedented expected to double by 2030 and grow to increase in life expectancy (Blumenthal and more than a third of the population by 2050 Hsiao 2005; Caldwell 1986; Yang and others (UN DESA 2013, 2015). 2008). A child born in China today can China will have far less time than the expect to live more than 30 years longer than Organisation for Economic Co-operation his forebears half a century ago; it took rich and Development (OECD) countries to adjust countries twice that long to achieve the same to the challenges imposed by an aging popu- gains (Deaton 2013). lation (figure 1.1). For example, at the current In addition, fertility has declined sharply, rate of demographic change, it will experi- declining from 6.0 children per woman of ence in 26 years the degree of population reproductive age in 1950 to 2.9 children per aging that took 115 years to occur in France woman in 1979 (Hesketh, Lu, and Xing 2005; (Kinsella and Phillips 2005). Smith, Strauss, and Zhao 2014). Fertility con- The share of working-age people in tinued to fall, albeit more gradually, after China’s population peaked in 2010 at China’s introduction in 1979 of the one-child 74 ­percent and has now started to decline. policy, and it has stabilized at approximately In 2015, at the current retirement age of 60, 1.7, well below the replacement rate of 2.1. there were 4.1 workers for each retiree; this ratio will fall to 1.4 workers per retiree by An Aging Population 2050. According to some estimates, if China is to avoid postponing the retirement age, Although reductions in mortality and fertility this shift in the dependency ratio will require represent progress, these demographic a doubling of the tax rate to finance elderly people’s income benefits (Smith, Strauss, and FIGURE 1.1  Aging of the population in China compared with selected countries, 1950–2050 Zhao 2014). The graying of China’s population also has profound implications for the country’s Share of population age ≥ 60 years, percent 40 mortality and morbidity profile. A mere 35 36.5 quarter century ago, 41 percent of the burden 30 of disease in China came from injuries; com- 32.8 25 municable diseases; and newborn, nutri- 20 tional, and maternal conditions—a profile 29.3 little different from that of the average low- 15 19.4 to middle-income country today (figure 1.2). 10 Now NCDs are responsible for 77 percent of 5 the loss of healthy life and for 85 percent of 0 all deaths, giving China a profile similar to that of most OECD countries (IHME 2010). 19 0 19 0 19 0 19 0 19 0 20 0 20 0 20 0 20 0 20 0 50 75 20 5 19 5 20 5 20 5 20 5 20 5 20 5 55 65 5 6 7 8 9 0 1 2 3 4 1 8 9 0 3 4 2 19 19 19 19 Cardiovascular diseases and cancers alone More-developed regions China India Brazil account for more than two-thirds of China’s total mortality (WHO 2014). Strokes, ische­ Source: UN DESA 2015. mic heart disease, chronic obstructive pulmo- Note: “More-developed” countries are defined according to World Health Organization (WHO) criteria. nary disease, and lung cancer top the list I mpre s s i v e G ain s , L ooming C hallenge s 27 FIGURE 1.2  Prominence of NCDs in the burden of disease and causes of mortality a. Share of NCDs in total disease burden b. Causes of mortality in China 90 8% 5% Share of NCDs in total burden of disease, percent 83 81 80 6% 77 70 2% 59 45% 60 11% 50 49 40 36 23% 30 Cardiovascular diseases Other NCDs Cancers Communicable, maternal, 20 perinatal, and nutritional Chronic respiratory 1990 1995 2000 2005 2010 diseases conditions Developed China Developing Diabetes Injuries Sources: IHME 2010; WHO 2014. Note: NCDs = noncommunicable diseases. Groups of “developed” and “developing” countries are defined according to the World Health Organization (WHO) criteria. of causes of premature mortality, while dia- the stronger increase taking place among men. betes, along with musculoskeletal disorders An alarming 49 percent of Chinese men are and major depressive disorders, has emerged daily smokers—a proportion more than twice as a principal cause of years lived with dis- the OECD average (figure 1.3, panel a). And ability (IHME 2010; Yang and others 2013). alcohol consumption per capita nearly dou- bled between 2000 and 2010, to 5.8 liters of pure alcohol per capita per year—a steeper NCDs and Changing Risk Factors increase than in Brazil and India and quickly The NCD epidemic will continue to grow. By catching up to the OECD average of 9 liters some estimates, the number of NCD cases (figure 1.3, panel b). among Chinese people older than 40 years of These risk factors are strongly associated age is predicted to double or even triple over with the major causes of morbidity and mor- the next two decades; diabetes will be the tality in China. Causality has been established most prevalent disease, and lung cancer between smoking and lung, liver, stomach, cases are likely to increase fivefold (Wang, esophageal, and colorectal cancers, which Marquez, and Langenbrunner 2011). together contribute close to a fifth of all pre- Even more than the aging of the popula- mature mortality in the Chinese population tion, other powerful forces behind the growth (HHS 2014; IHME 2012). Cardiovascular of chronic illnesses in China are high-risk diseases, China’s leading cause of mortality, behaviors such as smoking, poor diets, seden- are also attributable to smoking as well as to tary lifestyles, and alcohol consumption, as air pollution, poor diet, and high blood pres- well as environmental factors such as air pol- sure. In 2005, 2.33 million cardiovascular lution (Batis and others 2014; Gordon-Larsen, deaths among Chinese adults aged 40 years Wang, and Popkin 2014; Ng and others 2014; and older were due to high blood pressure Yang and others 2008). Adult overweight (He and others 2009). prevalence nearly tripled between 1991 This hypertension problem is worsening. (11.8 percent) and 2009 (29.2 percent), with According to the China Nutrition and Health 28 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 1.3  Smoking and alcohol consumption, international comparisons a. Daily smoking among malesa 60 Daily smokers among males aged ≥ 15 years, percent 50 40 30 20 10 0 xe nm a bo rk Tu ep. Ne Fin ico w Sta y ite orw a Ko Esto le Sp ia CD erl a d Is m a Po nce Ice den Ire age Ch n Ch ce ala s re nia Sw lov m Fr ain rtu c M rg y ep lic er d Gr ey Ja ry ec ep ly Be nds av and Au gal Lu De ad Ca nd Hu and Au and rm d ng el Ne d a Ze te Un N trali OE itz eni in Po ubli an pa m a i str th lan iu S do Cz k R Ita Ge lan ee a u h R ub ex Ki ra rk R n ng a er e lg la l a, l s Sw va ite Slo Un b. Alcohol consumption per capita, 2000–10 Liters of pure alcohol per capita, population aged ≥ 15 years 12 10 9.1 8 7.4 5.8 6 4 2.5 2 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 OECD Brazil China India Source: OECD 2015. Note: OECD = Organisation for Economic Co-operation and Development. a. Data on smoking are for 2013 or nearest year. Survey, the crude prevalence of hypertension people living with hypertension and a preva- in the population aged 15 years and older lence higher than in many middle- and high- increased from 5.1 ­ p ercent in 1959 to income countries (figure 1.4, panel a). 18.8 percent in 2002. By 2012, the estimated Likewise, the prevalence of diabetes—a prevalence of hypertension among middle- cause of premature mortality and a contrib- aged adults reached 40 ­ percent (Chow and uting factor to cardiovascular disease— others 2013; Feng, Pang, and Beard 2014)— is high and rising (figure 1.4, panel b). which translates into more than 200 million The diabetes prevalence rate rose from I mpre s s i v e G ain s , L ooming C hallenge s 29 2.5 percent in the early 1990s to between faced financial constraints (Wang, Marquez, 9.7 percent and 11.6 percent of the adult and Langenbrunner 2011). ­ p opulation two decades later (depending The rising burden of NCDs and their asso- on the estimates) and is projected to reach ciated costs poses a significant challenge for 13 percent by 2035 (Guariguata and others China’s health system. There is no quick, 2014; Xu and others 2013; Yang and others simple way to meet the challenge of chronic 2010). Using data from the nationally repre- illnesses, but there is much that the health sentative China National Diabetes and ­ s ystem can do to improve outcomes and Metabolic Disorders Study (2007–08), Yang reduce these costs—by preventing the onset and others (2011) found that 30 percent of of disease; providing early diagnosis and the sample studied had three or more cardio- effective management through regular moni- vascular risk factors; applied to the Chinese toring; supporting changes in behavior; and population, this translates into more than encouraging appropriate use of pharmaceuti- 300 million people—equivalent to roughly cals and other therapies. Dealing effectively the entire population of the United States— being at high risk for cardiovascular disease. FIGURE 1.4  Prevalence of hypertension and diabetes in China and Chronic disease can have disastrous out- selected other countries, early 2010s comes for individuals and society. If not a. Hypertensiona effectively managed, diabetes, hypertension, Prevalence of hypertension, percent 50 and other chronic conditions tend to result 45 40 in complications, which in turn may lead to 35 disability, suffering, or premature death. 30 The economic costs of chronic disease—­ 25 20 associated with health care, lost productivity, 15 caregiving, and loss of healthy life—can be 10 enormous. At the system level, the direct 5 0 medical costs of NCDs in China were m n n a ico es ca y $210 billion in 2005 and are estimated to ey d in an tio pa do an at fri rk ex Ch rm Ja ra St hA ail ng Tu M de reach more than $500 billion by 2015 (Bloom Ge d Th Ki ut ite Fe d So Un ite ian and others 2013). Considering the impact of Un ss Ru NCDs on labor supply and capital accumula- tion, the total economic impact of the five b. Diabetesb major NCDs on China is projected to be 16 $27.8 trillion for the period 2012–30. Prevalence of diabetes, percent 14 NCDs also pose a threat to the financial 12 health of households because they are expen- 10 sive to treat and often require care over an 8 extended period. In 2009, the average out-of- pocket spending per hospital admission due 6 to NCDs had already mounted to 50 percent 4 of the disposable annual income of an urban 2 resident ($750 per capita per year) and 1.3 0 times that of a rural resident ($291 per capita Ge co ca n na es n y ey d d pa Un ratio an an fri an i at i rk ex Ch Ja per year), while a coronary artery bypass rm St hA gl Tu ail M de En d Th ut ite Fe So operation cost 1.2 times and 6.4 times the ian ss annual disposable income of an urban Ru and rural resident, respectively (Chen and Zhao 2012). A 2011 study found that Sources: Chow and others 2013; Guariguata and others 2014; Ikeda and others 2014. Note: The figures show the percentage of the population diagnosed with hypertension or 37.6 percent of low-income patients reported diabetes in specific age groups. not being hospitalized despite being advised a. The hypertension age group is 35–84 years, except for China and Germany, for which the upper age limit is 70 and 74, respectively. to do so, because most of them (89.1 percent) b. The diabetes age group is 20–79 years. 30 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A with chronic medical conditions requires in the form of markups for drugs and some both continuity and coordination of care services. This created incentives for drug because, unlike episodic acute medical condi- ­ prescribing and overservicing that remain tions, NCDs often call for interventions from a feature of China’s health system today. multiple providers over long periods. Over Disparities between urban and rural areas the past decade, many OECD countries have and among regions widened drastically, and introduced reforms to strengthen public health care expenditure grew rapidly as a health programs and primary health care and result of pervasive overservicing and waste improve the coordination of care. A growing (Chen 2009; Ma, Lu, and Quan 2008). body of evidence shows that such reforms can have important benefits not only in Expansion of Health Care Coverage improving outcomes but also in controlling and Utilization cost escalation. In the late 1990s, concerns about the afford- Health System Reform in China: ability of health care led the state to begin a major expansion of health insurance A Brief Overview coverage. Initially, the expansion focused on ­ The collapse of the planned economy during reestablishing insurance for formal sector the 1980s, and the massive unemployment workers with the introduction of the Urban and layoffs associated with the reform of Employee Basic Medical Insurance (UEBMI) state-owned enterprises during the 1990s, scheme in 1998. This was followed by the resulted in a drastic reduction in health pro- introduction of the New Cooperative tection in both urban and rural areas, and by Medical Scheme (NCMS) in 2003, offering 2001 households were paying 60 percent of subsidized health insurance for China’s all health care costs out-of-pocket (Feng, rural population, and the Urban Resident Lou, and Yu 2015). Alongside the economic Basic Medical Insurance (URBMI) for the transition from the planned economy to a urban poor and informal sector workers market economy since the 1970s, the Chinese in 2007. government took a series of policy measures Insurance coverage expanded at a remark- to reform the health care facilities. The focus able pace, reaching more than 90 percent of of these measures was to enhance hospitals’ the population in urban areas and 97 percent financial and operational autonomy as well of those in rural areas by 2013 (figure 1.5).3 as to improve their efficiency and effective- By design, the depth of insurance coverage ness. Because of these policy measures, the expanded more gradually. When they were hospitals became more and more financially first launched, the NCMS and URBMI cov- independent, though they still remained ered only inpatient services. Reimbursements u nder t he ju risd ic tion of t he healt h were later extended to outpatient services, departments. and reduced copayments were permitted for Over the past three decades, hospitals in public health priorities such as hospital deliv- Chinese cities have improved dramatically, eries and the identification and treatment of and difficulties in seeking health services, chronic diseases (Yip and others 2012). receiving surgeries, and getting admitted for Although these changes have helped to reduce inpatient care have been greatly reduced. disparities across the fragmented insurance However, the regulatory role of governments schemes over time, the UEBMI continues to was not strong enough during the autonomy be better funded and to offer more generous enhancement process. As a result, chaotic benefits than the other two schemes (Liang competition occurred in the market, and the and Langenbrunner 2013; Meng and Tang hospital sector became more profit-driven. 2010; Yip and others 2012). Health facilities were given full financial China’s achievement of almost universal autonomy and allowed to generate revenues health insurance coverage has coincided with I mpre s s i v e G ain s , L ooming C hallenge s 31 FIGURE 1.5  Coverage of social health insurance in China, 2003–13 a. Urbana b. Ruralb 100 100 Insurance coverage, percentage of population Insurance coverage, percentage of population 90 90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 2003 2008 2011 2013 2003 2008 2011 2013 URBMI UEBMI Other insurance NCMS Other insurance Sources: Center for Statistics and Information data, National Health and Family Planning Commission; Meng and others 2012. a. UEBMI = Urban Employee Basic Medical Insurance. URBMI = Urban Resident Basic Medical Insurance. b. NCMS = New Cooperative Medical Scheme (2003 data include the predecessor to NCMS, the Cooperative Medical Scheme). “Other” insurance comprises the government insurance scheme, the labor insurance scheme, and other commercial and noncommercial schemes. an equally remarkable increase in the use of Together these features create incentives for some health services (figure 1.6). Hospital increasing patients’ average length of stay admissions, for instance, increased more than and service volume, particularly for items twofold between 2003 and 2013 (Meng and with high price–cost margins, such as drugs others 2012). Insurance has significantly and high-technology services. They inflate raised the use of inpatient services (Babiarz the charges to patients (including, in some and others 2012; Liu and Zhao 2014; Yu and cases, postreimbursement charges) and drive others 2010; Zhang, Yi, and Rozelle 2010). up health care costs overall (Li and others Insurance uptake has also been associated 2012; Liu, Wu, and Liu 2014; Meng and oth- with a rise in specific services such as annual ers 2012; Yang and Wu 2014). medical checkups and services for other pub- The government recognizes these short- lic health priorities that are either free or comings and has attempted to address them almost free of charge to patients (Lei and through national policies that reduce outpa- Lin 2009; Meng and others 2012; Yip and tient copayments for priority chronic condi- others 2012). tions, establish zero markup on sales of drugs in primary care facilities and (more recently) in some hospitals, and separate revenues Reforms to Address Coverage from expenditures in lower-level facilities Shortcomings (table 1.1).4 Additionally, in July 2015 the Several studies point to shortcomings in the government launched insurance to cover cat- design of the insurance schemes that make it astrophic illnesses, extending financial pro- difficult to achieve better results. Providers tection for families at risk of being impover- face pressure to generate revenue to cover ished by health expenditures. operational costs and a bonus system that ties These policies may be having the desired staff remuneration to facility revenues. effects, 5 but they are stopgap measures that 32 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 1.6  Trends in health service use in China, by visit type, 2003–13 a. Most-used services b. Less-used services 20 100 18 Utilization rate, percentage of population 90 Utilization rate, percentage of population 16 80 70 14 60 12 50 10 40 8 30 6 20 4 10 2 0 0 2003 2008 2011 2013 2003 2008 2011 2013 2003 2008 2011 2013 2003 2008 2011 2013 Five or more ANC visits Hospital delivery Outpatient visits Hospital admissions Rural Urban Sources: Center for Statistics and Information data, National Health and Family Planning Commission (NHFPC); Meng and others 2012. Note: ANC = antenatal care. address the symptoms rather than the struc- State Council on Deepening the Health Care tural deficiencies in financing and service System Reform” further emphasized the delivery that prevent China from having a importance of primary care and promoted world-class health system. regional health planning and coordinated management of the health system across lev- els of government. Since 2009, more-specific Reforms to Strengthen Prevention, reforms have been introduced to strengthen Primary Care, and Integration of primary care and promote integration. As Services discussed further in chapter 2, these have Chinese policy makers have increasingly seen included proposed increases in salary levels, that the health system suffers from hospital- and the introduction of performance-based centrism and fragmentation. As part of a pay, in community-level institutions. 8 The reform of the rural health system that was concept of “general practice” was subse- launched in the early 2000s, policies envisaged quently introduced, identifying primary care that township health centers and village clinics as the first point of contact and emphasizing would play important roles, with effective continuity in the relationship between gen- coordination and collaboration across the tiers eral practitioners and the population. 9 of the system.6 The importance of primary Policies also call for large-scale investments care was reinforced in subsequent years, to construct and equip community-level including a growing emphasis on developing health institutions and to train health work- community health services and their associ- ers at the village and township levels.10 ated workforces as well as on adjusting the The government made further commit- coverage and reimbursement arrangements of ments in 2013 and 2014 to enhance integra- the NCMS and URBMI to promote the use of tion between hospitals and primary care services at the community level.7 facilities11 and in 2015 to shift resources The 2009 “Opinions of the [Communist toward the primary level. In 2013, the central Party of China] Central Committee and the government invested in a program to improve I mpre s s i v e G ain s , L ooming C hallenge s 33 TABLE 1.1  Major reforms to extend financial protection and contain health care costs in China, 2009–15 Year Reform document Issuing agencies (document number) Key content 2009 Opinion on Establishing the Ministry of Health (Weiyaozheng fa Establishes an essential medicines list and suggests that National Essential Medicines 2009, No. 78) community health centers and country hospitals pilot zero System markup for medicines on the list 2009 Opinion on Improving Ministry of Finance, National Suggests that local government pilot the retention of all Government Health Subsidy Development and Reform revenue from health facilities, including from the sale of Commission, Ministry of Civil medicines (as opposed to allowing facilities to retain such Affairs, Ministry of Human revenues) Resources and Social Security, and Ministry of Health (Caishe 2009, No. 66) 2011 Opinions on Further Improving Ministry of Human Resources and Implements payment reforms in the UEBMI scheme (for the Reform of Health Insurance Social Security (Renshenbu fa 2011, example, DRGs, capitation, and global budget) to control Payments No. 63) the increase of medical expenditures 2012 Notification on Health Sector State Council (Guo fa 2012, No. 57) For NCMS and URBMI: Development in the 12th Five- ·   Gradually increases government subsidy to RMB 360 per Year Plana capita by 2015 ·   Increases reimbursement ceiling ·   Increases reimbursement rates for inpatient care to 75 percent; outpatient care to 50 percent; specific treatments for select chronic conditions (such as diabetes and mental health) and NCDs (such as cancer and Parkinson’s disease) to 100 percent ·   Stipulates that basic outpatient care should be covered broadly by 2015 2012 Opinions on Promoting Ministry of Health (Weizhengfa fa Supplements insurance coverage for 20 high-expense Rural Residents’ Catastrophic 2012, No. 74) priority conditions (such as childhood leukemia, congenital Insurance heart disease, uremia, and lung cancer) 2012 Notification on Implementing National Development and Stipulates that a portion of the URBMI and NCMS funds be the Catastrophic Insurance for Reform Commission, Ministry used to finance catastrophic expenditures of beneficiaries Urban and Rural Residents of Health, Ministry of Finance, of the respective schemes (the exact portion to be Ministry of Civil Affairs, and China determined by local government) Insurance Regulatory Commission (Guoyigaiban fa 2012, No. 2605) 2012 Opinion on Implementing Ministry of Human Resources and Suggests implementation of payment reforms across the Control of Total Medical Social Security, Ministry of Finance, all schemes (for example, DRGs, capitation, and global Insurance Payment and Ministry of Health (Renshebu fa budget) to control the increase of medical expenditures 2012, No. 70) 2012 Guidance on Promoting Ministry of Health, Ministry Implements payment reforms (for example, for DRGs, Reform of the NCMS Payment of Finance, and National capitation, and global budget) to control the increase of System Development and Reform medical expenditures Commission (Weinongwei fa 2012, No. 28) 2013 Opinion on Consolidating General Office of the State Council Establishes zero markup policy for medicines on the and Improving the Essential (Guoban fa 2013, No. 14) essential list at the grassroots level Medicine System and New Operating Mechanism 2013 Opinion on Establishing an General Office of the State Council Reimburses medical expenses for emergency room Emergency Medical Assistance (Guoban fa 2013, No. 15) admission for patients without capacity to pay System (Table continued next page) 34 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 1.1  Major reforms to extend financial protection and contain health care costs in China, 2009–15  Continued Year Reform document Issuing agencies (document number) Key content 2014 Guidance on Further Ministry of Human Resources and Suggests that local government pilot the portability of Implementation of Social Security, Ministry of Finance, insurance across pooled funds Basic Health Insurance and National Health and Family Reimbursement Across Planning Commission (Renshebu fa Pooling Regions 2014, No. 93) 2015 Opinion on Comprehensively General Office of the State Council Extends the zero markup policy to all county hospitals Scaling Up Reform of County- (Guoban fa 2015, No. 33) Level Public Hospitals 2015 Main Tasks of Deepening General Office of the State Council ·   Increases government subsidy to NCMS and URBMI to Medical Reforma (Guoban fa 2015, No. 34) RMB 380 per capita ·   Increases reimbursement rate for outpatients to 50 percent and inpatient services to 75 percent 2015 Opinion on Pilot General Office of the State Council Extends the zero markup policy to all pilot cities Comprehensive Reform of (Guoban fa 2015, No. 38) Urban Public Hospitals 2015 Notification of the Opinion General Office of the State Council ·   E xtends supplemental coverage for catastrophic medical to Comprehensively Scale Up (Guoban fa 2015, No. 57) expenditures to 700 million patients (with reinsurance Urban and Rural Residents’ with commercial companies) Catastrophic Medical ·   Increases reimbursement for eligible patients by Insurance 10–15 percent 2015 Notification of Measures for National Health and Family Creates a database of essential drug prices for all schemes Administration of National Planning Commission Essential Medicine Directory (Guoweiyaozheng fa 2015, No. 52) 2015 Guidance on Promoting Tiered General Office of the State Council Promotes health insurance payment reform and improves Care Service Delivery System (Guoban fa 2015, No. 70) health service pricing mechanism to support the tiered care system 2015 Several Opinions on the National Health and Family Eight cost-containment measures and a list of 21 Control of Unreasonable Planning Commission, National monitoring indicators, including the following: Growth of Medical Expenses in Development and Reform ·   Reduces the percentage of drug revenue to 30 percent Public Hospitals Commission, Ministry of Finance, of overall revenue (excl. TCM) Ministry of Human Resources ·   Reduces the revenue of consumables below RMB 20 per and Social Security, and State RMB 100 of revenue generated (excl. revenue from Administration of Traditional medicines) in pilot public hospitals by 2017 Chinese Medicine (Guoweitigai fa ·   Promotes provider payment reform, ensures application 2015, No. 89) of clinical pathway management to 30 percent of hospitalization patients, and implements DRG system on at least 100 disease types Note: DRG = diagnosis-related group; NCDs = noncommunicable diseases; NCMS = New Cooperative Medical Scheme; TCM = traditional Chinese medicine; UEBMI = Urban Employee Basic Medical Insurance; URBMI = Urban Resident Basic Medical Insurance. a. Umbrella reforms that are implemented through subsequent more specific policy documents. the capacity and service provision of tradi- government invested $19 billion in building, tional Chinese medicine as a way to further renovating, and equipping thousands of vil- augment frontline service delivery capacity. lage clinics, community health centers, and In 2015, the central government enacted a township health centers. New training pro- series of policies and supported investments grams for primary health care providers have to promote a more patient-focused delivery spread across the country, and thousands of system with multiple tiers.12 new workers have been trained to provide Policies in support of community-level frontline primary health care. A number of health care have also been backed by invest- technical cooperation relationships between ments. Between 2009 and 2012, the central hospitals and village clinics have begun to I mpre s s i v e G ain s , L ooming C hallenge s 35 improve the skills of frontline health workers Prevention and Treatment of Chronic and to encourage coordination across levels Diseases” sets ambitious goals, including cov- of the health system. China has also made erage of key interventions and reduction of large investments in information systems in chronic-disease prevalence and related community-level health institutions, as well mortality.14 as in piloting the general-practice model, ver- In parallel, the central government has tical integration, and gatekeeping with developed protocols for chronic-disease man- patient-referral systems. agement and has promoted the introduction Notwithstanding the impressive expan- and expansion of disease management pro- sion and improvement of community-level grams by the local governments. Some parts health facilities, the overall impact of the of China have achieved significant progress reforms to promote primary care and inte- in implementing these policy commitments. gration has been limited for several reasons: • Difficulty in attracting and retaining qual- Meeting China’s Health Care ified health professionals at the commu- Needs: Key Challenges nity level. Despite a policy push to improve Recalibrating the health system has become conditions, salaries and incentives for more urgent as China’s changing demo- work at this level have not been adequate graphic and epidemiological profiles—as well considering the professional and financial as its rising health care costs and slowing opportunities that are available in higher- economy—exert growing pressures. The rap- level institutions. idly evolving needs pose important challenges • C o n t ra d i c t i o n s i n po l i c y. C ro s s - for policy makers in the drive to create an government policies on personnel and equitable system that uses resources effi- budgeting restrict the scope for raising ciently to produce good health outcomes. compensation for work in the health sec- These challenges, discussed below, include tor a nd for ex pa nd i ng t he hea lt h excessive use of hospitals for care that could workforce. be provided effectively and much more • Lack of incentives for local governments cheaply in primary care facilities; uneven and and providers to restructure the health inadequate quality of care; strong incentives system and pursue integration. Indeed, to provide medically unnecessary services; rather than promoting coordination and rising costs and poor value for money; and cooperation, current financing arrange- disappointing health outcomes. ments have stimulated expansion of the volume and complexity of care as well as Hospital-Centrism and Weakness competition among providers. in Primary Care In addition to broader financing and service China’s health system remains hospital-­ centric delivery reforms, the government has intro- and fragmented. The number of hospital beds duced various initiatives to improve the care of doubled between 1980 and 2000 (from 1.2 chronic diseases. Some local governments million to 2.17 million) and doubled again in started experimenting with community-­ based 13 years (to 4.58 million in 2013). The number disease management programs in the early of hospital beds per 1,000 population has more 2000s. Building on their experiences, the cen- than doubled from 2000 to 2015 (figure 1.7). tral government defined the management of Although starting from a lower base, the chronic diseases as a priority public health ser- expansion of hospital capacity in China runs vice area in 2009, highlighting the important counter to international trends. Most OECD role of com mu n it y-level providers. 13 countries, with the notable exception of the The National Health and Family Planning Republic of Korea, significantly reduced the Commission’s 2012–15 “Work Plan for number of hospital beds per 1,000 population 36 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 1.7  Hospital beds in China and selected OECD countries, 2000 and 2013 16 –9% Hospital beds per 1,000 population 14 12 136% 10 –9% 8 –36% –29% –21% –26% 95% 6 –27% –28% –29% –33% –16% –20% –7% 4 –20% 33% 2 0 p. ce ain a n en ile da a a ey m es ly y d d in ali ni pa an Re Ita an lan lan do at Ch ed na rk Sp Ch to str Ja rm St a, Tu Fr ng Sw Ca Es er Fin Au re Ge d itz Ki Ko ite Sw d Un ite Un 2000 2013 Source: OECD 2015. Note: OECD = Organisation for Economic Co-operation and Development. Percentages shown above the bars for each country represent the change between 2000 and 2013. over the past decade, in many cases by as much medicine sales profits)—has disproportion- as 30 percent (OECD 2015), as shown in figure ately favored the hospital sector (figure 1.8, 1.7. China today has more hospital beds per panel b). The gap between the value of high- 1,000 population than Canada, Spain, the priced medical devices in hospitals versus United Kingdom, or the United States. those in township health centers and below China’s hospitalization rate reached 14.1 widened by a factor of three, from RMB 157 percent in 2013, up from 4.7 percent in 2003, billion to RMB 473 billion (Xu and Meng an annual growth rate of 11.5 percent. The 2015). volume of hospitalization in both secondary This trend is significant in view of the and tertiary hospitals tripled in roughly the experience in high-income economies, where same period (Xu and Meng 2015). technology has been a major driver of the The expansion and use of China’s hospital increase in health care expenditures (de la capacity has been shifting toward higher- Maisonneuve and Oliveira Martins 2013; level facilities. Between 2002 and 2013, the Smith, Newhouse, and Freeland 2009). numbers of tertiary and secondary hospitals Appropriate use of medical technologies can rose by 82 percent and 29 percent, respec- improve the quality of care, but if providers’ tively, while the number of primary care pro- incentives are not well aligned, there is a dan- viders declined by 6 percent (figure 1.8, ger that the mere availability of equipment panel a). Health workers, especially those can induce its overuse. with formal medical education (a measure of Although secondary hospitals still provide health service quality), have been moving to the largest volume of inpatient services, hos- high-level facilities and have become particu- pitalizations are growing by 18.3 percent per larly concentrated in hospitals (Meng and year at the tertiary level compared with others 2009; Xu and Meng 2015). annual growth of 14.1 percent at the second- Moreover, China’s technology boom— ary level (Xu and Meng 2015). Further, f ueled by a n i ncreasi ng rel ia nce on county hospitals are replacing township patient examination (to generate revenue to health centers as the principal providers of compensate for the tighter regulation of inpatient services in rural areas. I mpre s s i v e G ain s , L ooming C hallenge s 37 FIGURE 1.8  Trends in the number and use of health care facilities in China, by level a. Growth in number of facilities, 2000–14a b. Share of inpatient services, 1981–2011 200 90 Share of total inpatient services, percent 180 80 Number of facilities (2002 = 100) 160 70 140 60 120 50 100 40 80 30 60 40 20 20 10 0 0 2000 2002 2004 2006 2008 2010 2012 2014 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 Tertiary hospitals Secondary hospitals Hospitals at county level or above Primary care facilities Township health centers Source: Xu and Meng 2015. a. “Primary care facilities” refers to village clinics and township health centers in rural areas and to community health centers and stations in urban areas. As for outpatient visits, all types of pro- pilot show indications of success: the propor- viders have experienced rapid growth since tion of inpatients in hospitals who had been 2004. However, hospitals are playing a referred there by CHCs increased rapidly, greater role in such services. From 2010 from 2.7 percent in 2007 to 57 percent in to 2014, the share of all health care facil- 2010, and average medical expenditure per ity outpatient services that occurred in hospital admission was 6 percent lower for hospitals increased from 34.9 percent to patients referred by CHCs than for those 39.1 ­percent, while the proportion in primary who had sought care directly at hospitals. care facilities dropped from 61.9 percent to Similarly, the Shenzhen pilot achieved a 57.4 percent. 40 percent increase in the share of insured As described earlier, China’s initiatives patients who used CHCs as their first point aimed at strengthening tiered service provi- of contact. In Chongqing, by contrast, studies sion include (a) capital investments and of the two-way referral policy found no evi- capacity building at the primary care level; dence of improvement in two-way referral (b) national policies that promote gatekeep- after the reforms; fewer than a quarter of all ing, referral systems, and vertically integrated acute outpatient visits were the result of refer- networks (such as medical groups); and (c) ral services, and the vast majority of those payment reforms to improve financing and were upward referrals. In the vertically inte- provider incentives at lower levels of care. grated networks in Beijing, upward referrals Little systematic information is available were also more frequent than downward by which to gauge the success of these initia- referrals. tives. On the coordination of care across dif- Hospitals are full-service facilities and have ferent providers, the results from localized little financial incentive to turn away patients experiments have thus far been limited and and the associated revenues. With their uneven (Meng 2015). In Fuzhou, McCollum enhanced infrastructure and human resource and others (2014) found that coordination capacity, which draw patients to them, hospi- across levels of health providers was unsatis- tals are in the driver’s seat. Any attempt to factory. In Qingdao, case studies of the com- create an effective integrated tiered delivery ­ munity health center (CHC) gatekeeping system will require fundamental changes—at 38 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A both the hospital and the primary care to make a proper diagnosis and only slightly levels—in a system that has significant room ­ more than a third of the questions that were for marked improvement in quality, whether deemed essential (Sylvia and others 2015). ­ perceived or real. When presented with an unstable-angina case, village doctors performed only 15 ­p ercent of the recommended examina- Quality of Care: Disparities and tions, and only 26 percent of their clinical Inappropriate Incentives diagnoses were correct. Overall, treatment The quality of health care services is often was considered correct or partially correct in understood as the degree to which the ser- only about half of the interactions. In addi- vices increase the likelihood of desired tion, in 75 percent of the interactions, village health outcomes and are consistent with doctors dispensed medication, 64 percent of current professional knowledge. Quality is which was determined to be unnecessary or affected by the availability of basic inputs harmful by an auditing physician. Efforts (adequate supplies of equipment, drugs, and have been made to address these problems by personnel) as well as by the process of care expanding training and the use of clinical delivery. protocols and guidelines, but the impact In assessing quality, fundamental ques- needs to be further improved. tions concern whether the nurse or physician Arguably, the most salient quality issue in asks the right questions, performs the appro- China is that of overservicing. Excessive pre- priate tests and exams, reaches the correct scription of drugs and procedures not only diagnosis, communicates effectively with the increases the risk of medical harm to patients patient, prescribes the appropriate treatment, and undermines trust in the system, but also and provides comfort to the patient. These wastes scarce resources that, if used appro- questions are fundamental both from an priately, could improve population health individual patient’s perspective and at the outcomes and reduce health inequalities. system level: systematic underuse, overuse, China’s health care facilities derive signifi- and inappropriate use of drugs and proce- cant incomes from the sale of medicines and dures results in wasted resources and subpar certain services. Over time, this feature has health outcomes. translated into financial incentives for indi- Although systematic evidence is hard to vidual providers to prescribe drugs and per- come by, quality is known to be a significant form diagnostic and other procedures, while issue in China’s health system. Concerns at the same time shaping patients’ expecta- about quality are the main reason why it is tions of what constitutes “good” health care. difficult to redirect patients to primary care Numerous studies have shown that over- facilities: patients perceive huge disparities in prescription is now pervasive in China. the quality of care among different levels of A ­prescription audit study on village clinics in providers (Bhattacharyya and others 2011; Shandong found that the average number of Jing and others 2015; Yang and others 2014). drugs per prescription (three) and the rates of Available evidence shows that many health use of antibiotics (60 percent), intravenous professionals lack the basic skills needed to injection drugs (53 percent), and steroids diagnose and treat common conditions effec- (20 percent) all exceeded the reference levels in tively. A doctor’s qualifications are a strong the World Health Organization (WHO) index correlate of technical quality, yet large varia- system of rational drug prescription (Yin and tions persist between rural and urban areas others 2015). A series of experiments to test in doctor training and qualification stan- the underlying motives for overprescription in dards. In one study using simulated patients, China concluded that financial incentive is the village doctors asked on average just 18 per- major driver (Currie, Lin, and Meng 2014; cent of the questions that were recommended Currie, Lin, and Zhang 2011). I mpre s s i v e G ain s , L ooming C hallenge s 39 Rising Costs Amid Room for Efficiency rise in China raises concerns about afford- Improvement ability, particularly considering the country’s relatively small government revenue base Health expenditures in China have been ris- (11.3 percent of GDP), which is below the ing steadily at a faster rate than income. Over average for upper-middle-income, and even the past two decades, total spending on low-income, countries (14.4 percent and 13.4 health increased fourteenfold from about percent, respectively. Even during China’s RMB 220 billion (1985) to RMB 3,170 ­ billion period of record-breaking growth in GDP (2013) in real terms.15 In 2015, China spent and increases in central government subsidies 5.98 percent on health as a share of gross resulting from reform policies, local govern- domestic product (GDP), close to the average ments in the country’s less-developed regions of 6.1 percent in middle- to high-income were feeling the fiscal constraints of meeting countries. This gap has been narrowing, centrally mandated levels of funding to sup- spurred mostly by growth in public spend- port social programs, including health care ing—including for social health insurance, (Long and others 2013).17 Concerns about wh ich i s he av i ly sub sid i z e d by t he affordability have prompted an increased government. focus on cost containment, with measures China’s tax-financed government budget ranging from direct price controls or budget for health nearly doubled its share of national caps (through supply-side measures to con- health spending between 2001 and 2013 trol the volume and unit prices of services) to (from 16 percent to 30 percent). Most of the increased cost-sharing and demand-side mea- budget increase was used to increase public sures designed to affect usage patterns. subsidies for social health insurance.16 Since Experience from other countries has 2001, the share of China’s total health spend- shown that cost-control measures differ in ing paid by social health insurance has grown how well they contain costs and that they can from a quarter to more than one-third, while sometimes adversely affect broader health sys- that of households’ out-of-pocket payments tem goals such as quality and r ­ esponsiveness. has dropped from 60 percent to less than Hence, in many OECD countries, instead of one-third (figure 1.9). simply attempting to control costs, the focus There is no single correct or best level of of health system reform has increasingly government spending on health, but the rapid shifted to promoting value for money. FIGURE 1.9  Composition of health spending in China, 1997–2013 3,000 2,500 Constant renminbi, trillions 2,000 1,500 1,000 500 0 99 00 01 03 11 13 97 98 02 04 05 06 07 08 09 10 12 19 20 20 20 20 20 19 19 20 20 20 20 20 20 20 20 20 Household out-of-pocket Social health insurance Government tax-funded (excl. insurance subsidies) Source: Global Health Expenditure Database, World Health Organization, http://apps.who.int/nha/database. 40 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A There are ample opportunities to improve OECD average of 16 percent.18 As noted value for money in China’s health care sys- above, concerns have been rising about irra- tem. The most obvious is to curb waste and tional use of medicines in China, and in par- inefficiency. Chisholm and Evans (2010) esti- ticular about excessive use of antibiotics mated that, globally, 20–40 percent of total (Currie, Lin, and Zhang 2011; Yin and health spending was wasted, mainly because others 2013; Yin and others 2015). A key ­ of technical inefficiencies related to human driver of this practice is the financial incen- resource management, inappropriate use of tive to prescribe antibiotics (Currie, Lin, medicines, medical errors and other types of and Meng 2014), which results in antibiotic suboptimal quality, and corruption and abuse, unnecessary health expenditure, fraud. In China, despite scant systematic data and the public health threat of antimicrobial on these issues, there is plenty of scope both resistance. Yip and others (2014) show that to control costs and to improve outcomes by reforms of provider payment ­ arrangements— reducing unnecessary use of medicines and in particular, capitation with pay for procedures. performance—have been effective in curtail- ­ Pharmaceutical spending per capita, for ing overprescribing and inappropriate pre- example, has increased more than threefold scribing as well as in reducing ­per-visit costs over the past decade (figure 1.10, panel a). in parts of China. A lthough spending on medicines has Another indicator of the efficiency of declined as a share of total health expendi- health expenditure is the share of health ture, it still accounts for more than 40 per- spending that is incurred in hospital settings. cent (figure 1.10, panel b). This share is large Hospital-based care often entails intensive compared with those of other countries in use of resources, including advanced medical East Asia and the Pacific and with the technology and procedures, resulting in a FIGURE 1.10  Pharmaceutical spending in China and international comparisons a. Spending and share of THE in China, 1990–2012 b. Share of THE in Asian countries and OECD average, 2010 120 60 60 Spending per capita, constant renminbi, trillions 100 50 50 44% Share of THE, percent Share of THE, percent 80 40 40 60 30 30 40 20 20 16% 20 10 10 0 0 0 Ca akis ep. ei A OE a i L an Sin ng ka Ko gap lia a, e Ph mb tan pi ia n a lad m Chesh ew sia ss lia Sr Jap m N nes Ba Vie mar a ail ia Da us CD M one R 90 00 20 2 12 Gu Fiji 92 94 96 98 04 06 20 8 20 0 pu Ma and In o P l d e ya si re or La epa in d D ilip od 0 Th Ind 0 1 an ng tna ala M an o in ru tra P R a N lay 19 20 20 19 19 19 19 20 20 al Ze o w Pharmaceutical expenditure per capita Ne un Pharmaceutical expenditure as a share of THE Pa Br Sources: 2014 China National Health Accounts Report data, China National Health Development Research Center (CNHDRC); OECD and WHO 2014. Note: THE = total health expenditure. OECD = Organisation for Economic Co-operation and Development. The data vary slightly depending on the source: according to the CNHDRC, pharmaceutical expenditure as a share of GDP in China in 2010 was 42 percent, whereas the OECD and WHO estimate is 44 percent. I mpre s s i v e G ain s , L ooming C hallenge s 41 FIGURE 1.11  Composition of health spending in FIGURE 1.12  Percentage of outpatient visits in hospitals, China China, by provider type, 1990 and 2013 and selected countries, 2013 100 70 Share of total health expenditure, percent 90 60 occurring in hospitals, percent 80 Share of all outpatient visits 70 50 21% 10% 60 40 50 40 30 30 54% 20 44% 20 10 10 0 1990 2013 0 p. Other Retail sale of medical goods n ain Sw ia en Fr a Cz No e a re c Fin y h R ay d pa d in c Ko ubli Re e n lan an ed rk na rw Sp to Ch Ja a, Tu Ambulatory facilities Township health centers Ca Es ep Hospitals ec Source: 2014 China National Health Accounts Report data, China Sources: 2014 China National Health Accounts Report data, China National Health National Health Development Research Center (CNHDRC). Development Research Center (CNHDRC); OECD 2015. Note: Data for Spain are for 2011. high cost per episode of treatment. Hospitals copayment, distorted price schedules that account for 54 percent of China’s total health favor drugs and high-­ technology procedures expenditure (figure 1.11), compared with the over other health care services, concentration OECD average of 38 percent (OECD 2015). of health workers and other resources in urban And the average length of stay for inpatient areas, and medical staff remuneration tied to services—a key driver of higher costs—is ­ volume- and ­ revenue-based bonus payments longer in China (9.8 days) than in OECD (Li and others 2012; Liu, Wu, and Liu 2014). countries (7.3 days). China will need to adopt a more compre- Good tertiary care is essential in a health hensive approach to reform that corrects dys- system, but in many instances, care can be functions across levels of care if it is to more appropriately and less expensively pro- achieve appropriate care-seeking behavior by vided in a community facility or in an outpa- patients and to increase the use of affordable, tient setting. A large share of outpatient con- quality primary care services (He and Meng sultations in China (nearly 40 percent) take 2015). Current inefficiencies are costly not place in hospitals (figure 1.12). only for the health system but also for patients Many studies point to the inefficiencies that who face congestion in high-level hospitals arise when patients bypass lower-level facili- and incur expenditures associated with some- ties to seek care in hospitals, particularly the times unnecessary procedures. better-equipped and better-staffed tertiary Out-of-pocket payments have been rising in hospitals, where provider-induced overuse of real terms as a whole, although the trend may medical technologies and of procedures with vary for specific health insurance schemes high profit margins is well documented ­ (figure 1.13). This is to be expected: as incomes (Eggleston and others 2008; He and Meng rise, households are better able to afford goods 2015; Sun, Wang, and Barnes 2015). These and services, and health care is no exception. inefficiencies have been attributed to specific But the important question is whether house- features of the financing and delivery system holds are having to pay catastrophically high such as reliance on the fee-for-service payment costs for health services or are being pushed method, lack of effective referral or tiered into poverty by health care costs. 42 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 1.13  Trends in out-of-pocket health care payments in continue to spend a considerable share of their China, by insurance type, 2003–13 income on health: 8.4–10 percent of their annual household income, depending on the 14,000 measure used (Liang and Langenbrunner 12,000 2013; Long and others 2013). 10,000 Cost pressures in China’s health sector are likely to worsen in coming decades. As in Renminbi 8,000 6,000 many other countries, the aging of the popu- lation; the growing prevalence of chronic dis- 4,000 ease; and the introduction and expanded use 2,000 of new drugs, procedures, and other medical 0 technology are all putting upward pressure 2003 2008 2013 2003 2008 2013 2003 2008 2013 on spending. Expenditure pressures will also UEBMI URBMIa NCMS come from addressing coverage gaps and dis- Reimbursement Out-of-pocket payment parities in the health system. China’s health insurance coverage is now Source: Tang 2014. nearly universal, but it is still shallow and Note: NCMS = New Cooperative Medical Scheme; UEBMI = Urban Employee Basic Medical there are important exclusions. An analysis Insurance; URBMI = New Urban Resident Basic Medical Insurance. a. No 2003 data are available for URBMI, which was not launched until 2007. of reimbursement data shows that the effec- tive reimbursement rate under both URBMI and the New Rural Cooperative Medical As noted earlier in this chapter, evidence Scheme (NRCMS) is about 50 percent (con- of the effects of China’s reforms in extending sidering deductibles, exclusions, and so on), financial protection to patients is mixed. By even though the reimbursement rates of the some estimates, the ratio of out-of-pocket schemes are 70 –75 percent (Liang and expenditures to disposable personal income Langenbrunner 2013). The shortfall is largely may be rising (Zhang and Liu 2014). At the due to insufficient funds, especially in the household level, there is some evidence that NCMS. health insurance has had positive impacts. In practice, insurance beneficiaries con- For instance, the rate of self-discharge from tinue to pay significant health costs out of hospital for financial reasons has declined pocket and to incur catastrophic expendi- steadily since 2003 (Meng and others 2012). tures despite having insurance coverage; this But the incidence of catastrophic levels of is particularly true for rural residents, the health care spending has remained stable: poor, and households with members suffering Liu, Wu, and Liu (2014) found that reim- from chronic disease, and it is an important bursements through insurance mechanisms cause of citizen discontent (Yang 2015; were more than offset by increases in spend- Zhang, Yi, and Rozelle 2010). Further, sev- ing associated with patients’ use of higher- eral studies find that insurance coverage is level facilities, longer lengths of stay, and use positively associated with prereimbursement of more expensive treatment items. charges because insured patients are more Though the share of out-of-pocket pay- likely than the uninsured to receive more ments in total health expenditures in China types of treatment, seek care in higher-level has declined impressively—from 60 percent to facilities, and stay longer in the hospital, all 30 percent in little more than a decade—it of which contribute to higher costs (Liu, Wu, remains high relative to the OECD average of and Liu 2014; Wang, Liu, and Liu 2014; 21 percent. Moreover, the decline has not Yang and Wu 2014). The resulting trend is an benefited urban and rural populations evenly. increase in prereimbursement costs with no For the rural population, out-of-pocket pay- significant difference in out-of-pocket or ments still account for 50 percent of total per postreimbursement payments, except for capita health spending, and rural households patients insured under UEBMI. I mpre s s i v e G ain s , L ooming C hallenge s 43 Extending financial protection and reduc- of healthy life, large gaps in health care ing rural–urban disparities are important ­ c overage remain despite steady increases policy objectives, but they entail significant in diagnosis, awareness, and treatment. fiscal costs. Additionally, as detailed in the Hypertension statistics provide an example. following section of this chapter, millions of Between 1991 and 2002, China’s hyperten- people with diabetes, hypertension, and other sion awa rene s s rat e i nc re a s e d f rom chronic diseases are currently undiagnosed 26.3 ­percent to 30.2 percent, the treatment and not receiving the care they need. Any rate i ncreased f rom 12 .1 p ercent to effort to deepen coverage, improve the qual- 24.7 percent, and the control rate increased ity of care, and reduce the large gaps in enti- from 2.8 percent to 6.1 percent. This trans- tlements under China’s different health insur- lates into about 30 million hypertensive ance schemes will require substantial patients who received treatment and about increases in government spending. 6 million whose symptoms were under control (Liu 2011). However, 130 million hypertensive patients Poor Outcomes (65 percent) are still unaware of their condi- Life expectancy: After a period of rapid tion. Most live in rural areas, where mortality improvements in health, during which China from the major complication of hyperten- recorded impressive progress in reducing sion—stroke—exceeds that in urban areas. maternal and infant mortality, recent invest- Among people with hypertension who are ments are not translating into greater longev- aware of their condition, 30 million ity for the population (figure 1.14). Although percent) have not received treatment, and (43 ­ China still performs well relative to other among those who are receiving treatment, countries—having a higher life expectancy at 75 percent do not have their blood pressure birth than would be expected at its level of under control. income and health spending—its global com- In their analysis of the 2011–12 China parative position of advantage has deterio- Health and Retirement Longitudinal Study rated (figure 1.15). of people aged 45 years or older, Feng, Hypertension: For chronic conditions Pang, and B ea rd (2014) fi nd f u r t her associated with the principal causes of loss improvements in diagnosis, treatment, and FIGURE 1.14  Life expectancy trends in China relative to total spending on health, 1995–2013 6 80 Health spending as a share of GDP, percent 5 Life expectancy, years 4 75 3 2 70 1 0 65 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 Total health expenditure (% of GDP) Life expectancy at birth (total years) Source: World Bank 2015; China Health Statistics. 44 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 1.15  Performance (life expectancy) relative to health spending and income, international comparisons a. 1995 b. 2013 Better than average Vietnam Better than average Sri Lanka China Vietnam Lao PDR Sri Lanka Indonesia Cambodia Indonesia Thailand Performance relative to health spending Performance relative to health spending India India China Philippines Solomon Islands Philippines Thailand Lao PDR Malaysia Malaysia Cambodia Ghana Papua New Guinea Brazil Solomon Islands Russian Federation Russian Federation Papua New Guinea Brazil Ghana Worse than average South Africa Worse than average Nigeria Nigeria South Africa Worse than average Bettter than average Worse than average Bettter than average Performance relative to income Performance relative to income Sources: Global Health Expenditure Database, World Health Organization, http://apps.who.int/nha/database; World Bank 2015. Note: Performance is measured by life expectancy at birth. Both axes are log scale. control (to 56.2 percent, 48.5 percent, and Diabetes: Better prevention and manage- 19.2 percent of the sample, respectively). ment of diabetes is another enormous oppor- Nonetheless, 33 percent of the randomly tunity and challenge for public health policy selected sample had hypertension that was in the next few decades. Low awareness of not well controlled (­fi gure 1.16). In short, diabetes is a major obstacle for treatment and there is still much room for improvement control of blood glucose, and 60–70 percent in hyper tension prevention and of China’s diabetic population are unaware management. of their condition (Xu and others 2013; Yang The proportions of hypertensive people and others 2010). Between 1979 and 2012, who are aware, treated, and controlling their no obvious improvement took place in aware- high blood pressure are lower in China than ness of diabetes. the averages in middle-income countries— Treatment and control of diabetes increased whose overall management of hypertension among those who were aware of their condi- is, in turn, worse than that of high-income tion. But only 25–40 percent of patients with countries (table 1.2). In the United States, diabetes received treatment, and only 20–40 85.3 percent of people aged 35–84 years percent of those treated achieved adequate gly- who have hypertension are aware of their cemic control (defined as HbA1c<7 percent) (Li health condition, 80.5 ­ percent are on medi- and others 2013; Xu and others 2013). By com- cation, and 59.1 percent have their blood parison, 50 percent of diabetic patients in the pressure controlled (Chow and ­ others 2013; United States were adequately controlled Ikeda and others 2014). (Selvin and others 2014). I mpre s s i v e G ain s , L ooming C hallenge s 45 TABLE 1.2  Hypertension diagnosis, treatment, and control in selected countries, 2013 Adults aged 35–84 years, percent Country Diagnosed Treated Controlled China 41.6 34.4 8.2 Thailand 46.0 38.4 17.7 Turkey 49.7 29.0 6.5 South Africa 52.8 37.6 21.0 Germany 53.1 39.2 7.4 Mexico 55.8 49.5 28.0 United Kingdom 62.5 53.5 32.3 Bangladesh 62.7 54.6 30.2 Jordan 73.9 71.0 38.2 Russian Federation 74.9 59.9 14.2 United States 85.3 80.5 59.1 Japan — 48.9 22.9 Sources: Chow and others 2013; Ikeda and others 2014. Note: — = not available. FIGURE 1.16  Management of hypertension and diabetes in China, circa 2010 a. Hypertensiona b. Diabetesb 45 45 3.1 40 19.8 40 Surveyed population ages 45 years Surveyed population ages 45 years 35 35 30 30 or older, percent or older, percent 25 25 33.0 20 17.9 20 15 15 0.5 3.0 10 10 8.1 7.0 5 5 7.9 4.6 0 0 Hypertensive, Diagnosed Diagnosed, Controlled Diabetic, Diagnosed Diagnosed, Controlled not aware and treated not treated not diagnosed and treated not treated Sources: Feng, Pang, and Beard 2014; Xu and others 2013; Yang and others 2010. a. Hypertension figures are for 2011–12. b. Diabetes figures are for 2007 and 2010, using a mid-range of estimates from Xu and others (2013) and Yang and others (2010). Impact of Selected Policy 2060 to approximately 8 percent of GDP Changes on Health Expenditures (de la Maisonneuve and Oliveira Martins 2013). There is no doubt that with further As noted earlier, China spends less on health declines in poverty and increases in prosper- as a share of GDP than most OECD ity, the demand for health care will increase. countries—but its rate of growth of spending ­ But it does not have to increase threefold. on health, especially public spending, is Indeed, the same OECD study also concluded higher than that of all OECD countries. An that China could limit its public expenditures OECD study projects that, at this rate, public on health to reasonable levels under 5 percent spending on health in China will triple by of GDP if adequate reforms are undertaken. 46 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A The salience and nature of these reforms is inpatient and outpatient service provision to the focus of this book. bring it to world best practice. To answer this A study commissioned by the World question, consider the rate of hospital dis- Bank, conducted together with researchers charge in high-income economies from China (the China National Health figure 1.18). France had a discharge rate of (­ D e velopm e nt R e s e a rc h C e nt e r) a nd 166 per 1,000 persons in 2013, compared Australia (the University of Canberra), proj- with 179 per 1,000 in Australia. The median ects that if the health spending between hospital discharge rate in OECD countries in 1993 and 2012 is any indication (as ana- 2013 was 162 per 1,000. In East Asia, dis- lyzed and decomposed by Zhai and others charge rates vary, from 111 per 1,000 in 2015),19 current health spending in China Japan in 2011; 138 per 1,000 in Taiwan, will increase in real terms (2014 prices) from China; and 147 per 1,000 in the Republic of RMB 3,591 billion in 2015 to RMB 18,039 Korea in 2011. (Zhai and others 2015). billion in 2035, equivalent to an average In OECD countries, the median discharge increase of 8.4 percent per year. This would rate rose from 159 discharges per 1,000 per- increase China’s health expenditure from sons in 2003 to 162 per 1,000 in 2013. Over 6 percent of GDP in 2016 to 9.1 percent of the same period—10 short years—the dis- GDP in 2035. 20 charge rates in China jumped from 47 per The most important driver of the increase 1,000 (30 percent of the OECD median) to in health expenditures (accounting for about 134 per 1,000 (83 percent). At this rate, dis- 55 percent) is the rising number of services charge rates will increase to 201 per 1,000 by per case of disease. 21 Excess health price 2020 and to 241 per 1,000 by 2030. In com- inflation (the increase in prices of medical parison, discharge rates in Taiwan, China, goods and services relative to prices of other will increase to 152 per 1,000 in 2020 and to goods and services) accounts for a further 182 per 1,000 in 2030. 26 percent, while demographic factors, such Likewise, outpatient services per case of as population aging and population growth, disease in China are currently growing at an make up the balance (19 percent), as shown annual rate of 5.2 percent. By 2035, outpa- in figure 1.17. tient services in China will be 62 percent of An important question to consider here is the outpatient service levels in Taiwan, how much further China needs to increase its China. The World Bank study shows that if China FIGURE 1.17  Main drivers of projected health expenditure increases in China, 2015–35 were to reduce inpatient use growth and keep it growing at levels comparable to Taiwan, 10 9.45 China, hospital spending would be reduced Health expenditure growth, percent 8.73 8.40 to RMB 3,566 billion in 2035, or 1.8 percent 7.95 7.50 8 of GDP, compared with RMB 9,562 billion, 6 or 4.8 percent of GDP, under the business-as- 4 usual scenario of no spending reduction. In other words, health system reforms that 2 lower the utilization of inpatient services 0 will result in significant savings, equivalent to 3 percent of GDP, and bring health spending –2 down to 6 percent of GDP by 2035. This 2015–20 2020–25 2025–30 2030–35 2015–35 finding is supported by the considerable evi- Services per casea Excess price in ation Population aging dence from pilot health reforms in China, Population growth Disease prevalence rate Annual growth rate which have succeeded in reducing the use of health inputs without compromising the Source: World Bank estimations. a. “Services per case” refers to the number of outpatient visits per disease episode—or quality of care and in curbing expenditures hospital discharges, in the case of inpatient services—across 19 disease categories. and out-of-pocket outlays (Cheng 2013; Gao, I mpre s s i v e G ain s , L ooming C hallenge s 47 FIGURE 1.18  Hospital discharge rates, selected countries and economies, 2000–14 200 180 160 Hospital discharges per 1,000 population 140 120 100 80 60 40 20 0 2000 2002 2004 2006 2008 2010 2012 2014 Australia Canada China France Japan Korea, Rep. OECD median Taiwan, China Source: Zhai and others 2015. Note: OECD = Organisation for Economic Co-operation and Development. Xu, and Liu 2014; Yip and others 2014). reflected in the use of health services, reduced Most importantly, this finding highlights the labor supply and productivity, and possible significant impact that health system reforms impacts on savings and investment associated can have on health spending in the medium with illness and premature death (Bloom and to long term. others 2013). Of course, the extent to which We note that though health policy deci- savings are realized depends on how well the sions can noticeably affect trends in health health system performs in improving health spending, rising expenditures on health also outcomes.22 reflect improvements in medical technologies, rising incomes, and demographic and epide- miological factors. Societies invest in health Conclusion because these investments have the potential Weaknesses in primary care, hospital-­ to generate significant value. This value centrism, lack of integration, and uneven comes from longer life and absence of disabil- quality impede China’s achievement of better ity, which, although not reflected in GDP, health outcomes and higher returns to invest- increase individual well-being and tend to be ments in health. Shifting spending from highly valued by society. ­ h ospitals to primary care facilities and Value also comes from reducing the direct improving the quality and coordination of economic costs of poor health—which are care would improve outcomes and enhance 48 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A the efficiency of spending. These gains would has a poor-value health care system, spend- come from strengthened primary and second- ing $9,146 per capita (2013, at purchasing ary health care, reductions in medical com- power parity [PPP]), whereas Japan and plications and hospital admissions, and fewer Singapore get much higher value, spending years of disability and poor health in old age. only a little over $3,500 per capita (2013) Savings in the short term are likely to be and achieving much better health outcomes ­ outweighed by the cost of expanding the cov- and higher life expectancy than the United erage of interventions and improving quality, States (figure 1.19). A large part of Japan’s but the returns over the longer term—from success in ensuring value for money, for slower growth in health spending and instance, is its management of expenditures reduced economic costs of ill health—are by encouraging the use of more cost-effective likely to be significant. services, such as emphasizing primary care in As China’s economy continues to grow, the insurance benefits package and promot- health spending will inevitably rise. But how ing investment in facilities that provide high- fast it rises can be controlled by prudent priority services (Maeda and others 2014). choices about the organization and produc- China will need to implement a new model tion of health services and the more efficient of health care production, financing, and use of resources. Continuing along the cur- delivery that responds to the needs and rent high-cost path to improved health out- expectations of the population but at the comes would result in two or three times the same time is grounded in today’s economic per capita spending associated with a low- reality. Doing nothing is not an option: cost path and would not necessarily lead to continuing to provide health services in the ­ significantly better outcomes. current environment using the current deliv- Although factors other than health care ery model will raise health costs and impose and health spending contribute to health out- heavier burdens on the government or house- comes, it is instructive that the United States holds, or both. FIGURE 1.19  Life expectancy relative to per capita health expenditure, selected countries, early 2010s 78 Singapore 76 Japan 74 Average healthy life expectancy, years 72 Luxembourg Norway 70 United States 68 China 66 64 62 60 0 2,000 4,000 6,000 8,000 Health expenditure per capita (PPP) Source: Global Health Observatory data repository, World Health Organization, http://apps.who.int/gho/data. Note: PPP = purchasing power parity. I mpre s s i v e G ain s , L ooming C hallenge s 49 Notes  9. “Gu idance on t he E stablish ment of a General Practitioner System,” State Council 1. Barefoot doctors are “farmers who received a (2011). short medical and paramedical training, to 10. “Planning and Implementation Plan of offer primary medical services in their rural Deepening Health System Reform during the villages” (Yang and Wang 2017). Twelfth Five-Year Plan,” State Council (2012). 2. Demographic data, by age group, from 11. “Announcement on Publishing 2013 National National Bureau of Statistics of China, 2014 Health Work Conference Docu ment,” China Statistical Yearbook. Beijing: China Ministry of Health (2013); “Announcement Statistics Press. on carrying out some key work in the New 3. The urban/rural classification is according to Rural Cooperative Medical Scheme,” Ministry hukou (residence registration). Migrant work- of Health (2014). ers are captured as rural residents and typi- 12. The 2015 State Council documents include cally have NCMS coverage. Not being insured the following: “Guidance on Comprehensive in the cities where they spend most of their Pilot Reform of Urban Public Hospitals,” time hinders their access to health care N H F P C ( G u oba n fa 2 015, N o. 38); because NCMS copayments are higher for “Guidance on Comprehensively Scaling-Up providers outside rural areas (Peng and others Reform of County-Level Public Hospitals” 2010; Yip and others 2012). (Guoban fa 2015, No. 33); “Guidance of the 4. In addition, a number of subnational govern- General Office of the State Council on ments have experimented with provider pay- Promoting Multi-L evel Diag nosis and ment reforms, including capitation and Treatment System” (No. 70); and “Guidance prospective payments, to lessen the incentives on Promoting Tiered Care Service Delivery to drive up service volume. System” (No. 93). 5. Analysis of the 2013 China Health and 13. “Opinions on Promoting the Gradual Nutrition Survey, which would shed light on Equalization of Basic Public Health Services,” the impact on these policies on health service Ministry of Health and Ministry of Finance use and out-of-pocket payments, was not pub- (2009). licly available at the time of this publication. 14. Specific goals include ensuring that 40 percent 6. Sources of the rural health system reform poli- of the hypertensive and diabetic patients cies include the 2002 “Decision on Further receive standardized disease control interven- Strengthening Rural Health Work” issued by tions, limiting the increase in stroke incidence the Central Committee of Communist Party to 5 percent, reducing stroke mortality by of China and State Council and the Ministry 5 percent, and reducing prevalence of chronic of Health’s 2002 “Opinion on Reform and obstructive pulmonary disease (COPD) to Management of Rural Health Facilities.” below 8 percent. The policy calls for further 7. Government guidance on development of pri- specific targets to be set at the local level. mary care and community health services 15. Health spending data from the 2014 China included the State Council’s 2006 “Guidance National Health Accounts Report, China on Development of Community Health National Health Development Research Services in the Cities”; the 2007 “Guiding Center (CNHDRC), Beijing. Opinions of the State Council about the Pilot 16. According to China’s system of health Urban Resident Basic Medical Insurance”; accounts (2011 edition), social health insur- and the 2007 “Guiding Opinions Regarding ance is classified as part of general govern- Per fe c t i ng M e a su re s of Po ol i ng a nd ment health expenditures, and this category Reimbursement of the New Rural Cooperative includes tax-financed subsidies to social Medical Scheme” issued by the Ministry of health insurance. For international compara- Health, Ministry of Finance, and State bility, we present data according to this clas- Administration of Traditional Chinese sification in the charts of this report. Medicine. Disaggregated data on tax-financed insurance 8. “Guiding Opinions Regarding Implementing subsidies were provided by the China National Pe r for m a n c e - B a s e d S a l a r y i n P ubl i c Health Development Resea rch C enter Institutions on Public Health and Primary (CNHDRC). Care,” Ministry of Human Resources and 17. Public subsidies mainly benefited the least- Social Security (2009). developed Western provinces. 50 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A 18. The data vary slightly depending on the Improved Finances of Township Health source. According to the China National Centers but Not the Number of Patients Health Development Research Center’s Served.” Health Affairs 31 (5): 1065–74. (CNHDRC) 2014 National Health Accounts doi:10.1377/hlthaff.2010.1311. Report, pharmaceutical expenditure as a Batis, Carolina, Daniela Sotres-Alvarez, Penny share of GDP in China in 2010 was 42 per- Gordon-Larsen, Michelle A. Mendez, Linda cent, whereas the OECD and WHO 2010 esti- Adair, and Barry Popkin. 2014. “Longitudinal mate is 44 percent. The CNHDRC estimates Analysis of Dietary Patterns in Chinese Adults that the share was 40 percent in 2013. from 1991 to 2009.” British Journal of 19. Zhai and others (2015) found that the growth Nutrition 111 (8): 1441–51. doi:10.1017​ in total health expenditure from 1993 to 2012 /S0007114513003917. was driven mainly by rapid increases in Bhattacharyya, Onil, Yin Delu, Sabrina T. Wong, real expenditure per case, which contributed and Chen Bowen. 2011. “Evolution of Primary 8.2 percentage points to the 11.6 percent total Care in China 1997–2009.” Health Policy 100 growth in health expenditures. Excess health (2–3): 174–80. doi:10.1016/j.healthpol.2010​ price inflation and population aging contrib- .11.005. uted 1.3 and 1.2 percentage points, respectively. Bloom, David E., Elizabeth T. Cafiero, Mark Reduction in disease prevalence led to modest E. McGovern, Klaus Prettner, Anderson savings, and population growth was a small Stanciole, Jonathan Weiss, Samuel Bakkila, contributor to the growth in expenditures. and Larry Rosenberg. 2013. “The Economic 20. Figures differ slightly from the China National Impact of Noncommunicable Disease in Health Accounts data presented earlier because China and India: Estimates, Projections, projections were done for current expenditures and Comparisons.” NBER Working Paper only, excluding capital formation, which is No. 19335, National Bureau of Economic lumpy and difficult to predict. Projections also Research, Cambridge, MA. differ from those produced for OECD coun- Blumenthal, David, and William Hsiao. 2005. tries and the BRIICS countries (Brazil, Russian “Privatization and Its Discontents—The Federation, India, Indonesia, China, and South Evolving Chinese Health Care System.” New Africa) by de la Maisionneuve and Oliveira England Journal of Medicine 353 (11): Martins (2013) and for China by Lorenzoni 1165–70. and others (2015), which used a component- Caldwell, John C . 1986. “Routes to Low based model that considers the age and sex Mortality in Poor Countries.” Population and structure of the population; population growth Development Review 12 (2): 171–220. doi: and aging (adjusted for longevity to account for 10.2307/1973108. savings in health costs due to additional years Chen, Junshi, and Wenhua Zhao. 2012. “Diet, of good health); income; health-relative prices; Nutrition, and Chronic Disease in Mainland and other unexplained factors. China.” Journal of Food and Drug Analysis 21. The number of “services per case” is the num- 20: 222–25. ber of outpatient visits (or hospital discharges, Chen, Zhu. 2009. “Launch of the Health-Care in the case of inpatient services) per disease Reform Plan in China.” The Lancet 373 episode by age group across 19 disease catego- (9672): 1322–24. ries and 17 age groups. Cheng, Tsung-Mei. 2013. “A Pilot Project Using 22. In the United Kingdom, this consideration led Evidence-Based Clinical Pathways and the Treasury-appointed Wanless Commission Payment Reform in China’s Rural Hospitals to recommend increases in government Shows Early Success.” Health Affairs 32 (5): spending on health both to improve health 963–73. doi: 10.1377/hlthaff.2012.0640. and to control health care costs (Wanless and Chisholm, Dan, and David B. Evans. 2010. others 2007). “Improving Health System Efficiency as a M e a n s of Mov i ng towa rd s Un iver s a l Coverage.” World Health Report (2010) References B ackg rou nd Pap er 28 , World H e a lt h Babiarz, Kimberly S., Grant Miller, Hongmei Yi, Organization, Geneva. Linxiu Zhang, and Scott Rozelle. 2012. Chow, Clara K., Koon K. Teo, Su mathy “China’s New Cooperative Medical Scheme Rangarajan, Shofiqul Islam, Rajeev Gupta, I mpre s s i v e G ain s , L ooming C hallenge s 51 Alvaro Avezum, Ahmad Bahonar, Jephat Diabetes Research and Clinical Practice 103 Chifamba, Gilles Dagenais, and Rafael Diaz. (2): 137–49. 2013. “Prevalence, Awareness, Treatment, and He, Alex Jingwei, and Qingyue Meng. 2015. “An Control of Hypertension in Rural and Urban Interim Interdisciplinary Evaluation of China’s Communities in High-, Middle-, and Low- National Health Care Reform: Emerging Income Countries.” Journal of the American Evidence and New Perspectives.” Journal of Medical Association 310 (9): 959–68. Asian Public Policy 8 (1): 1–18. doi:10.1080/17 Currie, Janet, Wanchuan Lin, and Juanjuan 516234.2015.1014299. Meng. 2014. “Addressing Antibiotic Abuse in He, Jiang, Dongfeng Gu, Jing Chen, Xigui Wu, China: An Experimental Audit Study.” Journal Tanika N. Kelly, Jian-feng Huang, Ji-chun of Development Economics 110: 39–51. Chen, Chung-Shiuan Chen, Lydia A. Bazzano, doi:10.1016/j.jdeveco.2014.05.006. and Kristi Reynolds. 2009. “Premature Deaths Currie, Janet, Wanchuan Lin, and Wei Zhang. Attributable to Blood Pressure in China: 2011. “Patient Knowledge and Antibiotic A Prospective Cohort Study.” The Lancet 374 Abuse: Evidence from an Audit Study in (9703): 1765–72. China.” Journal of Health Economics 30 (5): Hesketh, Therese, Li Lu, and Zhu Wei Xing. 933–49. doi:10.1016/j.jhealeco.2011.05.009. 2005. “The Effect of China’s One-Child de la Maisonneuve, Christine, and Joaquim Family Policy after 25 Years.” New England Oliveira Martins. 2013. “A Projection Method Journal of Medicine 353 (11): 1171–76. doi: for Public Health and Long-Term Care doi:10.1056/NEJMhpr051833. Expenditures.” OECD Economics Depart­ HHS (U.S. Department of Health and Human ment Working Paper No. 1048, Organisa­ Services). 2014. The Health Consequences of t ion for E conom ic C o - op erat ion a nd Smoking—50 Years of Progress: A Report of Development, Paris. the Surgeon General. Atlanta: Centers for Deaton, Angus. 2013. The Great Escape: Health, Disease Control and Prevention. Wealth, and the Origins of Inequalit y. I H M E (I nstitute for Health Metrics and Princeton, NJ: Princeton University Press. Evaluation). 2010. “Global Burden of Diseases Eggleston, Karen, Li Ling, Meng Qingyue, Study: China Profile.” IHME, University of Magnus Lindelow, and Adam Wagstaff. 2008. Washington, Seattle. “Health S er vice Deliver y in China: A ———. 2012. “Global Burden of Diseases Literature Review.” Health Economics 17 (2): Study: China Profile.” IHME, University of 149–65. doi: 10.1002/hec.1306. Washington, Seattle. Feng, Jin, Pingyi Lou, and Yangyang Yu. 2015. Ikeda, Nayu, David Sapienza, Ramiro Guerrero, “Health Care Expenditure over Life Cycle in Wichai Aekplakorn, Mohsen Naghavi, Ali H. the People’s Republic of China.” Asian Mokdad, Rafael Lozano, Christopher J. L. Development Review 32 (1): 167–95. Murray, and Stephen S. Lim. 2014. “Control of Feng, Xing Lin, Mingfan Pang, and John Beard. Hypertension with Medication: A Comparative 2014. “Health System Strengthening and Analysis of National Surveys in 20 Countries.” Hypertension Awareness, Treatment and Control: Bulletin of the World Health Organization Data from the China Health and Retirement 92 (1): 10–19C. Longitudinal Study.” Bulletin of the World Health Jing, Limei, Zhiqun Shu, Xiaoming Sun, John Organization 92 (1): 29–41. F. Chiu, Jiquan Lou, and Chunyan Xie. Gao, Chen, Fei Xu, and Gordon G. Liu. 2014. 2 015. “ Fa c t o r s I n f lu e n c i n g Pat i e nt s’ “Payment Reform and Changes in Health Care Contract Choice with General Practitioners in China.” Social Science and Medicine 111: in Shanghai: A Preliminary Study.” Asia- 10–16. doi:10.1016/j.socscimed.2014.03.035. Pa c i f i c Jo u r n a l o f P u b l i c H e a l t h 27 Gordon-Larsen, P., H. Wang, and B. M. Popkin. (2 Suppl): 77S –85S. doi:10.1177/1010539​ 2014. “Overweight Dynamics in Chinese 514561654. Children and Adults.” Obesity Reviews 15: Kinsella, Kevin, and David R. Phillips. 2005. 37–48. doi: 10.1111/obr.12121. “Global Aging: The Challenge of Success.” Guariguata, L., D. R. Whiting, I. Hambleton, Population Bulletin 60 (1): 5–42. J. Beagley, U. Linnenkamp, and J. E. Shaw. Lei, Xiaoyan, and Wanchuan Lin. 2009. “The 2 014. “G loba l E st i m ate s of Di ab e te s New Cooperative Medical Scheme in Rural Prevalence for 2013 and Projections for 2035.” China: Does More Coverage Mean More 52 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Service and Better Health?” Health Economics Maeda, Akiko, Edson Araujo, Cheryl Cashin, 18 (S2): S25–S46. doi: 10.1002/hec.1501. Joseph Harris, Naoki Ikegami, and Michael R. Li, Min-zhi, Li Su, Bao-yun Liang, Jin-jing Tan, Reich. 2014. Universal Health Coverage for Qing Chen, Jian-xiong Long, Juan-juan Xie, Inclusive and Sustainable Development: and others. 2013. “Trends in Prevalence, A Synthesis of 11 Country Case Studies. Awareness, Treatment, and Control of Directions in Development Series. Washington, Diabetes Mellitus in Mainland China from DC: The World Bank. 1979 to 2012.” International Journal of McCollum, Rosalind, Lieping Chen, Tang Chen Endocrinology 2013: Article ID 753150. Xiang, Xiaoyun Liu, Barbara Starfield, Zheng doi:10.1155/2013/753150 Ji n hu a n , a nd R achel Tol hu r s t . 2 014. Li, Ye, Qunhong Wu, Ling Xu, David Legge, “Experiences with Primary Health Care in Yanhua Hao, Lijun Gao, Ning Ning, and Gang Fuzhou, Urban China, in the Context of Wan. 2012. “Factors Affecting Catastrophic Health Sector Reform: A Mixed Methods Health Expenditure and Impoverishment from Study.” The International Journal of Health Medical Expenses in China: Policy Implications Planning and Management 29 (2): e107–e126. of Universal Health Insurance.” Bulletin of the doi:10.1002/hpm.2165. World Health Organization 90: 664–71. Meng, Qingyue. 2015. “Overview of Efforts to Liang, Lilin, and John C. Langenbrunner. 2013. Improve Primary Health Care and Enhance “The Long March to Universal Coverage: I nt e g r at io n i n C h i n a .” World B a n k , Lessons from China .” Universal Health Washington, DC. Coverage (UNICO) Study Series 9, World Meng, Qingyue, and Shenglan Tang. 2010. Bank, Washington, DC. “Universal Coverage of Health Care in China: Liu, Hong, and Zhong Zhao. 2014. “Does Health Challenges and Opportunities.” World Health Insurance Matter? Evidence from China’s Report (2010) Background Paper, 7, World Urban Resident Basic Medical Insurance.” Health Organization, Geneva. Journal of Comparative Economic s 42 (4): Meng, Qingyue, Jing Yuan, Limei Jing, and 1007–20. doi:10.1016/j.jce.2014.02.003. Junhua Zhang. 2009. “Mobility of Primary Liu, Kai, Qiaobing Wu, and Junqiang Liu. 2014. Health Care Workers in China.” Human “Examining the Association between Social Resources for Health 7 (24): 5. Health Insurance Participation and Patients’ Meng, Qun, Ling Xu, Yaoguang Zhang, Juncheng Out-of-Pocket Payments in China: The Role of Qian, Min Cai, Ying Xin, Jun Gao, Ke Xu, Institutional Arrangements.” Social Science J. Ties Boerma, and Sarah L. Barber. 2012. and Medicine 113: 95–103. doi:10.1016/j. “Trends in Access to Health Services and socscimed.2014.05.011. Financial Protection in China between 2003 Liu, L. S. 2011. “2010 Chinese Guidelines for the and 2011: A Cross-Sectional Study.” The Management of Hypertension.” Zhonghua xin Lancet 379 (9818): 805–14. doi: http://dx.doi. xue guan bing za zhi 39 (7): 579–615. org/10.1016/S0140-6736(12)60278-5. Long, Qian, Ling Xu, Henk Bekedam, and Ng, S. W., A. G. Howard, H. J. Wang, C. Su, and Shenglan Tang. 2013. “Changes in Health B. Zhang. 2014. “The Physical Activity Expenditures in China in the 2000s: Has the Transition among Adults in China: 1991– Health System Reform I mproved 2011.” Obesity Reviews 15: 27–36. doi:10.1111​ Affordability?” International Journal for /obr.12127. Equity in Health 12: 8. OECD (Organisation for Economic Co-operation L or e n z o n i , L u c a , D av id M or g a n , Yu k i and Development). 2015. “OECD Health Murakami, and Chris James. 2015. “The Statistics 2015.” Database, OECD, Paris. O E C D H e a l t h a n d L o n g -Te r m C a r e http://www.oecd.org/els/health-systems/oecd​ E xpendit u re Projec tion Model and Its -health-statistics-2015-country-notes.htm. Application to China.” Background report, OECD and WHO (Organisation for Economic Organisation for Economic Co-operation and Co-operation and Development and World Development (OECD), Paris. Health Organization). 2014. Health at a Ma, Jin, Mingshan Lu, and Hude Quan. 2008. Glance: Asia/Pacific 2014. Paris: OECD. “From a National, Centrally Planned Health Peng, Yingchun, Wenhu Chang, Haiqing Zhou, System to a System Based on the Market: Lessons Hongpu Hu, and Wannian Liang. 2010. from China.” Health Affairs 27 (4): 937–48. “Factors Associated with Health-Seeking I mpre s s i v e G ain s , L ooming C hallenge s 53 Behavior among Migrant Workers in Beijing, Wan less , Derek, Joh n Appleby, A nthony China.” BMC Health Services Research 10: 10. H a r r i s o n , a nd D a r s h a n Pat el . 2 0 0 7. Selvin, Elizabeth, Christina M. Parrinello, David Our Future Health Secured? A Review of B. Sacks, and Josef Coresh. 2014. “Trends in NHS Funding and Performance. London: Prevalence and Control of Diabetes in the King’s Fund. United States, 1988–1994 and 1999–2010.” W HO (World Health Organization). 2014. A n n al s of Int e r n al M edicin e 160 (8): “Noncommunicable Diseases (NCD) Country 517–25. Profiles: China.” Factsheet, WHO, Geneva. Smith, James P., John Strauss, and Yaohui Zhao. h t t p : / / w w w.w h o . i n t / n m h / c o u n t r i e s​ 2014. “Healthy Aging in China.” Journal of /­chn_en.pdf. t he E c o n o m i c s o f A ge i n g 4 : 37– 43 . World B a n k. 2015. Wo rl d De velop m e nt doi:10.1016/j.jeoa.2014.08.006. Indicators 2015. Washington, DC: World Smith, Sheila, Joseph P. Newhouse, and Mark S. Bank. Freeland. 2009. “Income, Insurance, and Xu, Jin, and Qingyue Meng. 2015. “People- Technology: Why Does Health Spending Centered Health Care: Towards a New Outpace Economic Growth?” Health Affairs Structure of Health Service Delivery in China.” 28 (5): 1276–84. doi:10.1377/hlthaff.28.5.1276. World Bank, Washington, DC. Sun, Zesheng, Shuhong Wang, and Stephen R. Xu, Y., L. Wang, J. He, Y. Bi, M. Li, T. Wang, Barnes. 2015. “Understanding Congestion in L. Wang, Y. Jiang, M. Dai, and J. Lu. 2013. China’s Medical Market: A n I ncentive “Prevalence and Control of Diabetes in Structure Perspective.” Health Policy and Chinese Adults.” Journal of the American Planning. doi:10.1093/heapol/czv062. Medical Association 310 (9): 948. Sylvia, Sean, Yaojiang Shi, Hao Xue, Xin Tian, Yang, Gonghuan, Lingzhi Kong, Wenhua Zhao, Huan Wang, Qingmei Liu, Alexis Medina, and Xia Wan, Yi Zhai, Lincoln C. Chen, and Scott Rozelle. 2015. “Survey Using Incognito Jeffrey P. Koplan. 2008. “Emergence of Standardized Patients Shows Poor Quality Chronic Noncommunicable Diseases in Care in China’s Rural Clinics.” Health Policy China.” The Lancet 372 (9650): 1697–1705. and Planning 30 (3): 322–33. doi:10.1093​ doi:10.1016/S0140-6736(08)61366-5. /heapol/czu014. Yang, Gonghuan, Yu Wang, Yixin Zeng, George Tang, Shenglan. 2014. “Developing More F. Gao, Xiaofeng Liang, Maigeng Zhou, Xia Equitable and Efficient Health Insurance in Wan, Shicheng Yu, Yuhong Jiang, and Mohsen China.” Paulson Policy Memorandum, Paulson Naghavi. 2013. “Rapid Health Transition in Institute, Chicago. China, 1990–2010: Findings from the Global U N DESA (United Nations Department of Burden of Disease Study 2010.” The Lancet Economic and Social Affairs). 2013. “World 381 (9882): 1987–2015. Population Aging 2013.” Report ST/ESA/ Yang, Huajie, Xiang Huang, Zhiheng Zhou, Harry SER.A/348, UN DESA Population Division, H. X. Wang, Xinyue Tong, Zhihong Wang, Jiaji United Nations, New York. Wang, and Zuxun Lu. 2014. “Determinants of ———. 2015. “World Population Prospects: The Initial Utilization of Community Health Care 2015 Revision, Key Findings and Advance S er v ic e s a mong Pat ie nt s w it h M ajor Tables.” Working Paper No. ESA/P/WP.241, Noncommunicable Chronic Diseases in South U N DE SA Population Division, United Ch i na.” PL o S ON E 9 (12): e116051. Nations, New York. doi:10.1371/journal.pone.0116051. Wang, Shan, Lihua Liu, and Jianchao Liu. 2014. Yang, Wei. 2015. “Catastrophic Outpatient “Comparison of Chinese Inpatients with Health Payments and Health Payment-Induced Different Types of Medical Insurance before Poverty under China’s New Rural Cooperative and after the 2009 Health Reform.” BMC Me d ic a l S cheme.” A p pl i e d E c o n o m i c Health Services Research 14 (443): 8. Perspectives and Policy 37 (1): 64 – 85. Wang, Shiyong, Patricio Marquez, and John doi:10.1093/aepp/ppu017. Langenbrunner. 2011. “Toward a Healthy and Yang, Wei, and Xun Wu. 2014. “Paying for Harmonious Life in China: Stem m ing Outpatient Care in Rural China: Cost the Rising Tide of Non- Communicable Escalation under China’s New Co-operative Diseases.” Report 62318-CN, World Bank, Medical Scheme.” Health Policy and Planning. Washington, DC. doi:10.1093/heapol/czt111. 54 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Yang, Wenying, Juming Lu, Jianping Weng, Pay-for-Performance Improves Antibiotic Weiping Jia, Linong Ji, Jianzhong Xiao, Prescribing Practices in Rural China.” Health Zhongyan Shan, Jie Liu, Haoming Tian, and A f fa i r s 33 (3): 502 –10. doi :10.137 7/ Qiuhe Ji. 2010. “Prevalence of Diabetes among hlthaff.2013.0702. Men and Women in China.” New England Yip, Winnie, Chi-Man Yip, William C. Hsiao, Journal of Medicine 362 (12): 1090–1101. Wen Chen, Shanlian Hu, Jin Ma, and Alan Yang, Zhao-Jun, Jie Liu, Jia-Pu Ge, Li Chen, Zhi- Maynard. 2012. “Early Appraisal of China’s Gang Zhao, and Wen-Ying Yang. 2011. Huge and Complex Health-Care Reforms.” “Prevalence of Cardiovascular Disease Risk The Lancet 379: 10. doi:10.1016/S0140​ Factors in the Chinese Population: The 2007– -6736(11)61880-1. 08 China National Diabetes and Metabolic Yu, Baorong, Qingyue Meng, Charles Collins, Disorders Study.” European Heart Journal 33: Rachel Tolhurst, Shenglan Tang, Fei Yan, 213–20. doi: 10.1093/eurheartj/ehr205. Lennart Bogg, and Xiaoyun Liu. 2010. “How Yin, Wen-qiang, Zhong-ming Chen, Hui Guan, Does the New Cooperative Medical Scheme Xue-dan Cui, Qian-qian Yu, Hai-ping Fan, Influence Health Service Utilization? A Study Xin Ma, and Yan Wei. 2015. “Using Entropy in Two Provinces in Rural China.” BMC Weight RSR to Evaluate Village Doctors’ Health Services Research 10: 9. Prescription in Shandong Province under the Zhai, Tiemin, John Goss, Jinjing Li, Rachel E s s e nt i a l M e d ic i ne Sy s t e m .” M o d e r n Davey, and Yohannes Kinfu. 2015. “Main Preventive Medicine 42 (3): 2. Drivers of Recent Health Expenditure Growth Yin, Xiaoxv, Fujian Song, Yanhong Gong, in China: A Decomposition Analysis.” The Xiaochen Tu, Yunxia Wang, Shiyi Cao, Lancet 386, Suppl 1: S46. doi:10.1016/S0140​ Ju n a n L iu , a nd Zu x u n Lu . 2 013. “A -6736(15)00627-3 Systematic Review of Antibiotic Utilization Zhang, Linxiu, Hongmei Yi, and Scott Rozelle. i n C h i n a .” Jo u r n a l of A n t i m i c ro b i a l 2010. “Good and Bad News from China’s Chemotherapy 68 (11): 2445–52. doi:10.1093​ New Cooperative Medical Scheme.” IDS /jac/dkt223. Bulletin 41 (4): 10. Yip, Winnie, Timothy Powell-Jackson, Wen Zhang, Lufa, and Nan Liu. 2014. “Health Reform Chen, Min Hu, Eduardo Fe, Mu Hu, Weiyan and Out-of-Pocket Payments: Lessons from Jian, Ming Lu, Wei Han, and William C. China.” Health Policy and Planning 29 (2): Hsiao. 2014. “Capitation Combined with 217–26. 2 Lever 1: Shaping Tiered Health Care Delivery with People- Centered Integrated Care The People-Centered Integrated effective PCIC while also keeping costs low. Care (PCIC) Model Effective PCIC promotes primary care as the patients’ first point of contact for most of A country’s health care service delivery sys- their health care needs, coordinating care tem should ensure that patients receive appro- with other providers, such as hospitals, at dif- priate, high-quality care in the best setting ferent levels of the health care system and for their needs in a timely, equitable, and across the spectrum of health needs. affordable fashion. This report uses “people- Ultimately, adopting PCIC implies a process centered integrated care” (PCIC) to refer to a of rebalancing and structuring China’s deliv- flexible model that is organized around the ery system into functional and accountable health needs of individuals and their families. networks of tiered and interconnected The term is shortened, for easier translation, providers. from the World Health Organization’s In China, a paradigm shift toward a (WHO) global strategy on “people-centered PCIC-like model is already under way from a and integrated health services” (W HO policy perspective, as chapter 1 discussed. Of 2015a). Box 2.1 defines PCIC, drawing on particular relevance are recent State Council the WHO strategy. guidelines that outline the roles and responsi- The goal of PCIC is to provide the right bilities of different levels of a tiered delivery service at the right place and the right time. system.1 These guidelines establish the essen- In addition to responding to patient needs tial tenets and features of the PCIC delivery and perspectives, this approach prioritizes model in China and set the stage for the core integration and coordination of services actions outlined later in this chapter. across the spectrum of care, from promotion The new guidelines include the following and prevention to curative and palliative important features: needs to reduce fragmentation and wasteful use of resources across a health system. • Strengthening grassroots providers The foundation of PCIC is primary health • Promoting first contact at the grassroots care. Without a robust primary health care level system, no country can provide high-quality, • Establishing two-way referrals 55 56 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A BOX 2.1  People-centered integrated care (PCIC) defined People-centered care is “an approach to care that receive a continuum of health promotion, ­ d isease consciously adopts the perspectives of individuals, prevention, diagnosis, treatment, disease manage- families and communities, and sees them as partici- ment, rehabilitation and palliative care services, at pants as well as beneficiaries of trusted health sys- the different levels and sites of care within the health tems that respond to their needs and preferences in system and according to their needs throughout their humane and holistic ways.” life course.” Integrated care consists of “health services that are managed and delivered in a way that ensures ­ people Source: WHO 2015a, 10–11. • Defining provider roles while fostering the initiatives for establishing PCIC—this chap- integration of providers across a tiered ter assesses the benefits of a shift to PCIC as delivery system well as the challenges China is likely to face • Emphasizing special care arrangements to in broadening the reliance of its health care treat and manage chronic diseases system on PCIC. It then discusses actions in • Expanding the supply of general practice eight core areas that will help guide the deci- physicians (“general practitioners”) to sions of Chinese authorities as they seek to staff primary care facilities broaden the implementation of PCIC. The • Organizing provider networks chapter concludes by summarizing the out- • Advancing the use of information and look in China on the PCIC core action areas. communications technology for electronic The focus of the discussion is on organizing health systems the delivery system around the needs of patients and achieving integration between Across the globe, PCIC initiatives that pri- the different components of the system. oritize primary health care are gaining trac- tion as central parts of health care reform. Their core features—strengthened primary Benefits of Adopting PCIC care, a focus on patient needs, and care inte- Although results are often context-specific gration across provider levels—are ubiqui- and most of the available evidence comes tous. In the United States, the patient-­ from PCIC initiatives in high-income coun- centered medical home (PCMH) model has tries, preliminary findings suggest that become an important form of primary care adopting PCIC can improve health out- improvement. Across highly functioning comes, quality, and patient experience. health systems such as those in Australia, Annex 2A reviews the evidence on effects of Canada, Denmark, the Netherlands, and the PCIC-like models on health outcomes, qual- United Kingdom, PCIC-like reforms are tak- ity, and costs, drawing on findings from ing shape. Even middle-income countries more than 300 studies, including 10 Chinese such as Brazil, Costa Rica, Singapore, and case studies that were commissioned for this Turkey show a marked orientation toward report. In summary, PCIC adoption has had reshaping service delivery toward PCIC. the following results: Though they are expanding rapidly, PCIC- like approaches remain local or regional in • Reduced hospitalizations and use of emer- many countries. gency care. Reviews of a wide variety of Based on a review of international PCIC approaches—including those of the ­ experience—including a range of Chinese U.S. Program of All-Inclusive Care for the L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 57 Elderly (PACE) and the U.S. Veterans investments and services, unnecessary use Health Administration’s Patient Aligned of health care facilities, and waiting times Care Teams (PACT)—highlight reduc- for care. PCIC can also improve patient tions in patients’ visits to emergency safety by reducing medical errors, increas- departments, unscheduled readmissions, ing the uptake of screening and prevention and lengths of hospital stays (Ali 2015; Ali programs, improving diagnostic accuracy, and Li 2015b). Hospital admission rates increasing the appropriateness and timeli- for conditions that can be treated with ness of referrals, and improving equity ambulatory care have often declined. across the care system. • Improved clinical care processes. PCIC interventions improved pain assessment An exhaustive review of the literature on and treatment, adequacy of medicine PCIC initiatives globally (WHO 2015a, dosages, patients’ adherence to prescrip- ­ 2015b) identified an array of potential benefits tions, use of care plans, and patient educa- to individuals, communities, health ­ workers, tion. For example, of the 48 clinical and health systems (as listed in box 2.2). processes studied in the PACT case study, International experience shows that better 41 improved (Ali and Li 2015b). outcomes at potentially lower costs are pro- • Improved outcomes and patient satisfac- duced by systems that prioritize the critical tion. PCIC interventions decreased primary health care functions: accessibility, patients’ pain, improved the quality of life, comprehensive capacities for most general and reduced the severity of depression. nonemergent clinical needs, continuity of care Other benefits included better glycemic and information, continual quality improve- control and lipid profiles and improve- ment, and integration of care (Friedberg, ments in physical function, nutritional Hussey, and Schneider 2010; Macinko, status, and physical balance. When mea- ­ Starfield, and Erinosho 2009). It also suggests sured, patient satisfaction almost always that, though there is no one model for provid- increased. ing PCIC, at the service delivery level PCIC • Mixed impacts on costs. Nearly all the should encompass at least the following four studies examined short-term impacts on strategic goals:2 costs, and their findings varied consider- ably within and across countries. Some • Organizing the model of care around the PCIC interventions in Europe and the health needs of patients, which in China is United States have generated savings, but likely to entail strengthening primary health the clear majority of studies show limited care and changing the roles of hospitals or inconclusive evidence on cost stabiliza- • Integrating providers across care levels tion or curtailment, and a handful even and among types of services report cost increases (Hebert et al. 2014). • Continuously improving the quality of Further research is needed to determine care whether better quality and outcomes will • Engaging people to make better decisions bring about cost savings in the long term. about their health and health-seeking In China, with its current bias toward behaviors. specialty services over primary care ser- ­ vices, there is probably greater potential Challenges to PCIC for cost savings. • Improved balance of health system Implementation in China resources and needs. At the health system The 22 case studies commissioned by the level, adopting PCIC enables a shift in the World Bank for this report yield insights into balance of care so that resources are the challenges China is likely to face in imple- allocated to better respond to needs. This ­ menting PCIC-based reforms in service shift can reduce the duplication of health delivery. Ten of the studies were of PCIC ­ 58 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A BOX 2.2  Potential benefits of people-centered integrated care Benefits to individuals and their families • Greater influence and better relationships with • Increased satisfaction with care and better care providers that build community awareness ­ relationships with care providers of, and trust in, care services • Improved access to and timeliness of care • Greater engagement and participatory representa- • Improved health literacy and decision-making tion in decision making about the use of health skills that promote patient independence resources • Shared decision making with professionals along • Clarification of citizens’ health care rights and with increased patient and family involvement in responsibilities care planning • Care that is more responsive to community needs. • Increased ability to self-manage and control long- term health conditions Benefits to health systems • Better coordination of care across different care • Better balance of care so that resources are allo- settings. cated closer to needs • Improved equity and enhanced access to care Benefits to health professionals and community for all health workers • Improved patient safety through reduced medical • Improved job satisfaction errors and adverse events • Decreased workloads and reduced burnout • Increased uptake of screening and preventive • Role enhancement that expands workforce skills, so programs workers can assume a wider range of responsibilities • Improved diagnostic accuracy and appropriate- • Education and training opportunities to learn new ness and timeliness of referrals skills, such as working in team-based health care • Reduced hospitalizations and lengths of stay environments. through stronger primary and community care services and better management and coordination Benefits to communities of care • Improved access to care, particularly for marginal- • Reduced unnecessary use of health care facilities ized groups and waiting times for care • Improved health outcomes and healthier commu- • Reduced duplication of health investments and nities, including greater levels of health-seeking services behaviors • Reduced overall costs of care per capita • Better ability for communities to manage and con- • Reduced mortality and morbidity from both infec- trol infectious diseases and respond to crises tious and noncommunicable diseases. Source: WHO 2015a, 12. improvement initiatives in China, and 12 highlighted in this and other chapters, inno- were of initiatives in other middle- and high- vative initiatives are under way in China to income countries (table 2.1; for more details, address these challenges. see annex 2B). First among these challenges, the case Through interviews with planners and studies found that, at the primary care level, ­ personnel in hospitals and primary care cen- systems need to be established for registering ters in China and other countries, these case or empaneling patients3 and stratifying them studies highlighted the typical issues that by their conditions and risks. Experience with arise when improving primary care and gatekeeping in China is limited, and referral ­ b etter integrating care across the tiers of systems need improvement to support the goal the health care system. Importantly, as of having patients’ first contact be at the L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 59 TABLE 2.1  Summary list of commissioned case studies on PCIC-based health care reforms in China and other selected countries In-text reference Case study Location Chinese case studies Beijing, CHA Beijing Chaoyang Hospital Alliance (CHA), four cases Beijing Beijing, PKU IDS Peking University Renmin Hospital Integrated Delivery Beijing System (PKU IDS), four cases Feixi, SPHCC Strengthening Primary Health Care Capacity (SPHCC) Feixi (Anhui province) Hangzhou, TFY Twelfth Five-Year Plan (TFY) Hangzhou (Zhejiang province) Huangzhong, HCA Health Care Alliance (HCA) Huangzhong (Qinghai province) Shanghai, FDS Family Doctor System (FDS) Shanghai Shanghai, RLG Shanghai Ruijin-Luwan Hospital Group (RLG), four cases Shanghai Xi, IC Integrated Care (IC) Xi (Henan province) Zhenjiang, GH Great Health (GH) Zhenjiang (Jiangsu Province) Zhenjiang, ZKG Jiangsu Zhenjiang Kangfu Hospital Groups (ZKG), Zhenjiang (Jiangsu Province) four cases International case studies Canterbury, HSP Health Services Plan (HSP) Canterbury, New Zealand Denmark, SIKS Integrated Effort for People Living with Chronic Denmark Diseases (SIKS) Fosen, DMC District Medical Center (DMC) Fosen, Norway JCUH, AEC James Cook University Hospital (JCUH), Ambulatory England Emergency Care (AEC) Kinzigtal, GK Gesundes Kinzigtal (GK) Kinzigtal, Germany Maryland, CareFirst CareFirst Patient-Centered Medical Home Maryland, United States Netherlands, DTC Maastricht Diabetes Care (DTC) Netherlands Portugal, ULS Local Health Unit (ULS) Portugal Singapore, RHS Regional Health Systems (RHS) Singapore Turkey, HTP Health Transition Plan (HTP) Turkey United States, PACE Program of All-Inclusive Care for the Elderly (PACE) United States VHA, PACT Veterans Health Administration (VHA), Patient Aligned United States Care Teams (PACT) Note: For more detailed descriptions of each of the case studies, see annex 2B (table 2B.1). primary care level rather than at a hospital. remains insufficient throughout the country. Downward referral systems—that is, from Unattractive compensation levels discourage hospital to primary care—function irregu- qualified professionals and health workers larly. Although there is a clear movement from seeking and retaining positions at the toward forming multidisciplinary teams, in grassroots level. much of China the health care workforce Second, hospitals in China have strong lacks the knowledge, skills, and culture needed fin anci al inc e ntives to c apture both to work collaboratively. Despite the govern- ­ inpatients and outpatients and not to shift or ment’s calls to integrate preventive and cura- integrate care provision to lower levels, tive care at the primary level, integration because of their high dependency on 60 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A fee-for-service revenues. Hospitals’ willing- overlapping services, financing arrangements, ness to shift care to facilities at lower levels of human resources, and logistics. The formation the health system will usually favor those of health maintenance organizations has not patients for whom the revenue receipts do not overcome this difficulty. Two cases with argu- cover costs (such as geriatric patients with ably the best examples of care coordination, long stays). Each facility is paid separately Xi County and Zhenjiang-Kangfu, operate for the care it provides, and except in the within a single administrative jurisdiction Xi County (Henan province) experiment (Ali and Li 2015a; Li and Jiang 2015). (Ali and Li 2015a), few facilities have Finally, though China has adopted many attempted to share earnings or savings from innovations in health information and com- improved coordination. Though China is munication systems, these initiatives often experimenting with the formation of inte- lack interoperability. Many of them center grated facility networks, known as hospital on supporting hospitals rather than grass- alliances, some of these alliances are domi- roots providers. nated by large hospitals and become channels The case studies suggest that China should to capture patients at higher levels of care. consider unified, standardized local and Third, providers and facilities are not national systems to measure and improve the compensated—or not fully compensated— quality of primary health care service deliv- for the provision of integrated care. For ery, chronic disease management, and patient example, Beijing Chaoyang Hospital provides satisfaction. Such measurement systems a small yearly stipend to physicians who should be linked to improvement efforts. rotate to community health centers on a part- China is not alone in facing challenges time basis, but at a rate too low to cover the to the reform of health services. Many “lost income” they could have made by Organisation for Economic Co-operation providing hospital-based care (Jian and Yip ­ and Development (OECD) countries have 2015). Primary care physicians rarely receive begun adopting PCIC as part of broader additional income for activities related to reform efforts to invest more in health care coordination, and primary care facilities s ystems (rather than in individual health ­ cannot retain savings they may earn if they programs and facilities)—reforms that are coordinate care and increase efficiency. designed to address the needs of aging pop- Fourth, patients in China have few incen- ulations, the high and increasing incidence tives to use primary care as a first point of of noncommunicable diseases (NCDs), and entry. Government policy allows patients escalating costs. The challenges they have direct access to hospitals for all care. Relative faced in doing so include transforming the to hospitals, primary care facilities have acute care and “illness treatment” orienta- l imited drug formularies and fewer well-­ ­ tion of traditional care models; reversing qualified professionals—which limits the the often top-heavy, pyramidal structures demand from patients with complex condi- of their service delivery systems; and reduc- tions. Because the copayments charged at ing fragmentation of management and ser- hospital outpatient departments are little vices between facilities and service levels higher than those at primary care facilities, (Cercone and O’Brien 2010; Porter 2010; out-of-pocket cost differences do not deter WHO 2008). patients from preferring hospitals. Some of the OECD reforms have also Fifth, integration requires administrative aimed to alter traditional relationships coordination beyond what currently exists in between providers and patients away from China. Networks that consist of facilities in the current model, which has tended to be different political-administrative jurisdictions limited to the moment of the consultation, (such as municipality, district, and county) centered on discrete interventions, limited to have difficulty coordinating decision making a single medical professional’s advice to an and staff behavior regarding patient flows, individual patient during the consultation, L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 61 and conceiving of the patient’s responsibility continuity to ensure the right care is given at as merely following the provider’s advice. the right place. Almost all of them sought to strengthen horizontal integration as well as Core Action Areas and performance measurement and feedback (core areas 4 and 7, respectively). Implementation Strategies The least frequent area of focus was certi- Based on the commissioned case studies and, fication and accreditation (core area 8), which where appropriate, the broader case litera- was seen in about one-quarter of the cases; ture, this section outlines eight core action none of the initiatives in China used this as areas identified as fundamental to the estab- an improvement strategy. Other approaches lishment of effective PCIC systems (table 2.2). used for improving accountability and the For each, we outline strategies that China’s standardization of care (core areas 1 and 2) authorities may wish to adopt to guide include linking payment to quality, imple- implementation. menting clinical pathways and other stan- All the improvement initiatives featured in dards of care (core area 6), and measuring the case studies have used multiple strategies patient satisfaction. in pursuit of PCIC (figure 2.1). All initiatives Within each action area, the PCIC initia- included work to strengthen vertical integra- tives used a range of strategies, chosen to tion and information and communication match their identified priorities, the scope of systems (core areas 3 and 5, respectively)— the targeted areas for improvement, existing reflecting a need to improve coordination and challenges, the health care system’s strengths TABLE 2.2  Eight core PCIC action areas and corresponding implementation strategies Core action area Implementation strategies 1: Primary health care as the first • Use empanelment to facilitate population health management point of contact • Stratify empaneled population based on risk • Strengthen gatekeeping • Ensure accessibility 2: Multidisciplinary teams • Define team goals, composition, roles, and leadership • Form individualized care plans between care teams and patients 3: Vertical integration, including new • Redefine the roles of facilities within a vertically integrated network roles for hospitals • Establish provider-to-provider relationships • Develop formalized facility networks 4: Horizontal integration • Promote the integration of different types of health care facilities • Provide integrated care around the individual user to promote more patient- centered care 5: Advanced information and • Establish standardized electronic health records systems accessible to providers communication technology and patients (e-health) • Establish communication and care management functions • Ensure interconnectivity and interoperability 6: Integrated clinical pathways and • Craft integrated pathways to facilitate care integration and decision support for functional dual referral systems providers • Promote dual referrals within integrated facility networks 7: Measurement, standards, and • Use standard performance measurement indicators feedback • Create continuous feedback loops linked to action plans to drive quality improvement 8: Accreditation and certification • Develop nationally and locally relevant accreditation criteria • Set indicator targets for certification Note: PCIC = people-centered integrated care. 62 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 2.1  Frequency of PCIC interventions in commissioned case studies, by core action area 100 100 100 100 100 100 95 95 91 90 90 90 86 80 82 80 70 Percentage of studies 60 60 50 40 30 23 20 10 0 0 First contact MDT Vertical Horizontal E-health Clinical Measurement Certi cation access integration integration pathways and and feedback dual referrals All China Note: PCIC = patient-centered integrated care; MDT = multidisciplinary team. For descriptions of all 22 case studies commissioned for this report, see annex 2B (table 2B.1). and structure, the system’s history of reform, ensuring that primary health care is the first and local traditions and culture. point of contact for patients for most of their health care needs: empanelment, risk stratifi- cation, gatekeeping, and accessibility. Core Action Area 1: Primary Health Care as the First Point of Contact Strategy 1: Use empanelment to facilitate Primary health care is the focal point of population health management PCIC, addressing both individual and com- Empanelment is the process by which all munity health. One of the basic characteris- patients in a given facility or geographic area tics of a strong primary health care system is are assigned to a primary care provider or that it establishes primary health care as the care team. It was considered a fundamental first point of contact for most patients’ needs. component in 10 of the 22 initiatives stud- When patients consistently use trusted and ied, including three of those in China. competent primary health care providers as Empanelment is the mainstay of service their entryway into a tiered health system, delivery systems in a number of European they can receive care that is continuous and countries, including Denmark; England, coordinated across the range of health care U.K.; Finland; the Netherlands; Scotland, delivery levels (hospital, primary care pro- U.K.; and Turkey. In China, adopting vider, and specialist). They thus receive the empanelment is likely to be an important needed care at the right place and avoid step in improving patient-provider relation- unnecessary hospital admissions, procedures, ships and trust, in ensuring responsibility is risks, and medical expenses. taken at the primary care level for a popula- Based on the findings from the case stud- tion’s health, and in shifting health-seeking ies, four strategies were identified for behavior away from hospitals. L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 63 There are two main ways to approach government initially decided to assign empanelment: allowing patients an element patients geographically to family medicine of choice or assigning patients by geographic doctors, creating “patients’ registries,” while region—which is typically done using preex- allowing patients to request to leave the origi- isting community demarcations. For exam- nal registry and join the panel of a different ple, the success of the Shanghai “family doc- family physician even if the physician was tor” system (FDS) largely hinged on contracts outside their geographic area. This freedom between residents and primary health care of choice proved to be a challenge for conti- providers (Ma 2015a). The FDS empaneled nuity of care, particularly when patients populations by neighborhood in all of its dis- moved between panels without effective tricts. It focused on building strong relation- communication between physicians. Further, ­ ships between patients and primary care pro- transferring patient records could take sig- viders, which furthered community trust in nificant time. If China were to implement a the family doctors as the first point of contact similar choice-based empanelment system, it in the health system. would need to be supported by an effective, China also used empanelment as a tool in real-time information management system, shifting from fee-for-service care to providing to ensure that patient information is trans- more accessible, high-quality care in lower- ferred efficiently whenever patients change level health facilities through integrated providers. chronic disease management programs and In the United States, the private health improved e-health tools. Empanelment in insurance company CareFirst prioritized Hangzhou (through implementation of its patient preference in the state of Maryland Twelfth Five-Year Plan [TFY]) was conducted but also worked to ensure that all patients by geographically restricted patient choice: were assigned a regular source of primary residents were able to choose any primary health care as part of its PCMH payment health care provider in a community health model (Murray 2015). The company devised center (CHC) within a specific area desig- a complex process of “aligning” or attribut- nated by their district health insurance. Each ing patients who did not have a chosen pri- resident was contracted for a full year and mary health care provider to a provider. It could only choose one primary care provider developed an algorithm to assign a primary at a time (Yan 2015a). Similarly, in Kinzigtal, care provider based on which physician the Germany, health care management company patient had last seen and how many times Gesundes Kinzigtal (whose name translates they had seen him or her. Importantly, as “healthy Kinzig Valley”) empaneled its CareFirst prioritized patient choice and population by region to make sure that all reached out to the patients to confirm and the covered patients were linked to a primary update their empanelment assignment. At health care team, but it allowed the patients any point, a patient could contact CareFirst to choose any primary care team within their to switch to another primary health care pro- region (Hildebrandt and others 2015; Nolte vider. While labor-intensive, this alignment and others 2015). process allowed for patient empanelment in a Though simple, geographic empanelment country where there are no established can limit patients’ choice of physician and empanelment practices. thereby decrease their acceptance of the A core component of empanelment is system. Empanelment by patient choice is ­ maintaining workable panel sizes (the num- an alternative approach that was used ber of patients assigned to a specific provider in other initiatives. In Turkey, the Health or team). Limiting this panel size is critical Transformation Plan of 2003 sought to estab- to achieve the goals of empanelment, allow- lish family medicine centers in every district ing the primary care provider and team to of the country, each with a defined reference deliver effective PCIC by focusing on the population (Sumer 2015). The Turkish patient’s needs. Overly large panel sizes can 64 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A discourage the formation and maintenance care and improve health outcomes (Ali and of strong patient-provider relationships and Li 2015a). inhibit the provision of coordinated quality Other means to identify high-risk patients care by limiting the time available for indi- include a patient’s past use of care facilities or vidual patients. Large panel sizes are often current illness burden. In the United States, the result of fee-for-service based payment CareFirst’s PCMH model in Maryland found systems, which can lead providers and prac- that risk stratification based on history of use tices to prioritize the number of visits was highly effective without being overly bur- and patients seen over the quality of their densome to the provider (Murray 2015). The interactions, care provided, and health program uses an illness burden score to outcomes. quantify patients’ risk. Illness burden scores are calculated using the past 12 months of Strategy 2: Stratify the risks of the empan- health insurance claims data and diagnoses eled population (figure 2.2). One of the first tasks that PCIC planners In Germany, physicians in Gesundes must consider is defining the health needs of Kinzigtal complete a risk status question- the target population. Risk stratification is naire for each new patient as a part of the the proactive identification of individuals enrollment procedure. They use this informa- within an empaneled reference population tion to compute risk for poor outcomes, who are at higher risk of developing poor which helps tailor a specialized care plan outcomes or who have, or risk having, high and goal-setting process for each patient rates of service utilization, particularly hos- (Hildebrandt and others 2015; Nolte and pitalization. The individuals thus identified others 2015). This process ensures that can be proactively targeted for interventions higher-intensity services target those patients designed to provide the needed higher- who are at higher risk for poor outcomes. intensity, coordinated care within the The Netherlands’ Maastricht Diabetes primary health care setting. At the same ­ Care and Denmark’s Integrated Effort for time, these high utilizers can be engaged to People Living with Chronic Diseases (SIKS) understand and address their needs, as well initiatives both applied risk stratification by as to reduce their preventable use of higher- identifying specific diseases that were associ- cost, higher-intensity services. Risk stratifi- ated with high costs, required complicated cation was featured in 10 of the 22 initia- management, or were associated with high tives studied, including only one of the risk for poor outcomes (Nolte and others Chinese initiatives. 2015; Runz-Jørgensen and Frølich 2015; Risk stratification can be done at an Vrijhoef and Schulpen 2015). SIKS created individual patient level or based on disease rehabilitation centers for people with four burden. At the individual level, risk can be diseases: chronic obstructive pulmonary dis- assessed based on clinical guidelines, on the ease, diabetes, heart failure, and hip fracture, presence of particular target conditions, or specifying clearly defined clinical entry crite- on a recent history of high utilization. ria for each. Disease-based risk stratification Clinical staff can also use a summative pro- allowed SIKS to reduce hospital admissions cess of their clinical intuition to create lists by 18 percent and outpatient visits by of patients whom they believe will need a 24 ­ p ercent (Runz-Jørgensen and Frølich higher level of attention from the team. The 2015). Based on the success of the SIKS cen- Xi Integrated Care initiative used the staff’s ters in Copenhagen, Denmark has scaled the summary clinical judgment to stratify approach into a national disease manage- patients by risk and to target the higher- ment program, which provides integrated risk patients for integrated clinical path- comprehensive care for people with chronic ways to increase the delivery of appropriate diseases. L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 65 FIGURE 2.2  Illness burden scorecard for patient risk stratification Percent of Percent Cost population of cost PMPM Illness burden (5.00 and above) Advanced critical illness Extremely heavy health care users with 3% 29% $4,436 band 1 signi cant advanced / critical illness. Multiple Illness burden (2.00–4.99) chronic illnesses Heavy users of health care systems, 8% 23% $1,160 band 2 mostly for more than one chronic disease. Illness burden (1.00–1.99) At risk Fairly heavy users of health care system 12% 21% $578 band 3 who are at risk of becoming more ill. Illness burden (0.25–0.99) Stable Generally healthy, with light use of 27% 20% $218 band 4 health care services. Illness burden (0–0.24) Healthy Generally healthy, often not using 50% 7% $49 band 5 health system. Source: O’Brien 2014. Note: Graphic depicts the risk stratification used by insurance provider CareFirst in the state of Maryland, United States. PMPM = per member per month. Strategy 3: Strengthen and target gatekeeping penalties on noncompliant patients or their Gatekeeping is an important mechanism for providers. Hangzhou’s TFY initiative used ensuring that patients receive the right care at explicit gatekeeping for patients with hyper- the right place at the right time. Because tension or diabetes (Yan 2015a). These patients may perceive gatekeeping as limiting patients had to access the health care system their choice and imposing undue restrictions, through their primary care providers, who gatekeeping systems must be designed with could then refer them to more-advanced care both patient autonomy and overall utilization at the CHC. controls in mind. Having primary health care In implicit gatekeeping systems, patients providers as the gatekeepers is a way to man- are strongly encouraged to see their primary age patients’ access to specialty care and can health care provider before they visit a spe- help to reduce overuse of inappropriate care. cialist, but they are not formally required to Gatekeeping arrangements must include a do so. This arrangement may be preferable strong referral system so that those patients to explicit gatekeeping because it allows who need a higher level of care have access to it. greater patient choice. Turkey’s Health Gatekeeping can be done explicitly or Transformation Plan chose not to adopt a implicitly. In explicit gatekeeping, patients formal gatekeeping program. Instead, it cannot receive secondary or tertiary care gave patients a financial incentive to use without first seeing and getting approval family medicine practices as their first con- from their primary health care provider, the tact for problems by waiving the hospital “gatekeeper.” This mechanism is often copayment for patients who come to the enforced by imposing financial or regulatory hospital with a referral from their family 66 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A physicians (Sumer 2015). This initiative Other ways to increase access to primary decreased the number of patients coming to health care include mobile clinics and home- hospitals, but it also made family medicine based care. In a rural part of northwest physicians feel that they are sometimes used China, Huangzhong in Qinghai Province only for referrals to hospitals. implemented a health care alliance system in 2013 with the goal of fully integrating Strategy 4: Expand accessibility county, township, and village health centers Providing options for patients to see or speak (Meng, Luyu, and others 2015). For eldercare to their providers when they perceive the need follow-up visits, primary health care facilities is a vital function of primary health care. were supplemented with mobile clinics that Primary health care must be even more acces- were extended out from the county hospitals. sible and convenient than hospitals. ­ After-hours The mobile clinics were capable of screening care options and same-day visit opportunities patients, transporting the critically ill to strengthen the ability of primary health care to larger hospitals, and moving specimens to avoid unnecessary upstream utilization of more and from the primary health care facilities expensive care options. Increasing accessibility and the laboratory. In Zhenjiang, in China’s for patients was addressed in 14 (64 percent) of eastern Jiangsu Province—Great Health’s the 22 PCIC initiatives. “3+X” teams were required to spend three Financial incentives for providers can be days per week providing home visits to com- used to improve patient access. In Maryland, munity members (Yan 2015b). Additional the CareFirst insurance company paid each services they provided included appointment of its primary health care providers a monthly booking and online communication. These non-visit-based payment with quality bonuses services were most often used by the elderly. that are partly determined by the providers’ Under Turkey’s Health Transformation accessibility to patients (Murray 2015). Plan, family medicine physicians were Obtaining high scores on measures such as required to conduct home visits and mobile accessibility (on weekends, evening hours, clinics for patients who could not attend tra- and by telephone) and by giving patients ditional clinics (Sumer 2015). These options access to their own medical records made were especially important in rural areas. The providers eligible to receive bonuses. The provision of home-based care, mobile health financial incentives gave physicians the impe- services, and mobile pharmacies are written tus to increase patients’ access, including into the contracts that family medicine prac- after-hours and electronic access, to their tices make with the Ministry of Health. own electronic health records (EHRs). However, there have been some complaints Access standards can also be legislated. that the time spent away from the office may Before the district medical center (DMC) ini- create a gap in the continuity of care. Seven tiative began in Fosen, Norway, many dis- years after the start of the initiative, in 2010, tricts did not have their own emergency care Turkey also began to require that family beds. Patients had to travel long distances to medicine physicians serve nursing homes, reach emergency care in tertiary medical cen- prisons, and childcare centers by conducting ters. After the success of the DMC model, the community visits. national government mandated that every district establish emergency care beds, Core Action Area 2: Functioning although it allowed the districts to choose Multidisciplinary Teams where to do so, based on their infrastructure and their populations’ needs (Forde, Multidisciplinary teams are a building block Auraasen, and Moreira 2015). By dictating for most successful PCIC initiatives. In prin- the goal but allowing the districts choice in ciple, these teams are nonhierarchical groups implementation, Norway ensured increased of clinical and nonclinical staff whose goal is access to 24/7 emergency care for its citizens. to provide comprehensive and integrated care L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 67 for patients. Teams composed of clinical physician, and the teams consisted of a nurse, and nonclinical members with a variety of medical ­assistant, pharmacist, care coordina- training backgrounds can provide a fuller tor, and community social worker. The pro- range of services than individual health care gram required each care team to clearly providers. define the role of each team member Multidisciplinary teams were implemented figure 2.3) but allowed each team to adapt (­ by 17 (77 percent) of the 22 initiatives stud- the roles of its members to its individual ied, and most were viewed as facilitators to needs and context. PCIC. The initiatives used various approaches Multidisciplinary teams can designate a to make the multidisciplinary teams success- care coordinator to relieve stress on other ful, including ensuring appropriate team team members, counsel patients on improv- composition and leadership and providing ing their health, and help patients navigate comprehensive, coordinated patient care. the delivery system. Because a large propor- tion of V HA patients had complicated Strategy 1: Define team goals, composition, chronic conditions that required well-­ roles, culture, and leadership coordinated care, each PACT team included a The starting point for defining an effective designated care coordinator who managed team is to define the health needs of the popu- patients’ appointments, follow-ups, referrals, lation and the role that primary care teams test data, and discharge from the hospital. will play in responding to those needs. The care coordinator was a critical position Aligning the goals of population health within on the care team, explicitly responsible for a community to the composition and tasks of coordinating the clinical staff and the range the team is a key initial planning step. of provided services (Ali and Li 2015b). The personnel on a multidisciplinary The multidisciplinary family doctor teams team can vary, but clearly defining their roles within Shanghai’s FDS included a primary and responsibilities is critical for success. care physician as the core leader, along with a Typically, the team leader is an experienced nurse, a public health physician, an assistant, primary health clinician. For example, in the and sometimes other professionals such as U.S. Veterans Health Administration’s nutritionists and pharmacists (Ma 2015a). (VHA) PACT, the leader of each team was a Similarly in Zhenjiang, Great Health’s 3+X FIGURE 2.3  Responsibilities of PACT members, U.S. Veterans Health Administration Physician Medical Care Community (team leader) Nurse assistant Pharmacist coordinator social worker Performs Conducts Performs Conducts Performs Conducts Leads in Patient Previsit Makes Manages Works developing education, preparation, medication patient data, closely with team priorities, goal setting, documentation, adjustments tracks results, patients patient goals, self-management follow-up based on participates and and care plans; teaching and after visit, medical records in follow-up, care team approves test coaching, care team and patient facilitates to facilitate orders, medication outreach health status; referral and community medication, reconciliation, assignments, educates patients discharge outreach and and and and maintaining about process health fairs referrals education room stocking medication use Source: Stout and others 2015. Note: PACT = Patient Aligned Care Team. 68 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A teams comprised three members from a com- the plan clearly defined the targets and used munity health center and a varying number emotional appeals to ensure that everyone of volunteers from hospitals; generally, the understood the importance of working in three members were a primary care physi- multidisciplinary teams—rather than just cian, a nurse, and one other preventive care mandating that they do so, as in the staff member, while the hospital representa- Portuguese case. To foster a collaborative cul- tives (the “X”) included other physicians, ture from the start, the Canterbury leader- nurses, and administrative staff (Yan 2015b). ship committees signed charters outlining the Multidisciplinary teams function best in a culture they agreed to uphold (Love 2015). collaborative culture where providers com- municate openly, trust one another, and Strategy 2: Form individualized care plans are treated as equals. Teams must meet regu- between care teams and patients larly and engage in frequent training and A care plan provides a road map for all the improvement efforts as a team. A comparison providers who care for a patient. Care plans between the Norwegian and Portuguese are generally used for high-risk patients but PCIC improvement initiatives is illustrative. can be applied to all patients. They can also In the Norwegian case, a long history of be used by patients themselves to manage cooperation and team mentality facilitated their conditions at home. Successful care the development of multidisciplinary teams. plans act as a contract of mutual commit- Though they faced few requirements or man- ments and contingency plans between the dates to work in teams, the DMC staff physician or nurse practitioner and the quickly developed a team spirit as the model patient. was implemented (Forde, Auraasen, and For example, in the Maryland case, Moreira 2015). Even the public health depart- CareFirst developed care plans for particu- ment and community psychiatric care service larly high-risk and high-utilizing patients. used the team approach, though their coun- Although CareFirst’s local care coordinator terparts elsewhere in the world use hierarchi- played a role in initiating and maintaining cal management models. the care plan, the key to a successful plan was The Portuguese PCIC initiative also recog- the involvement of both the primary health nized the importance of multidisciplinary care provider or nurse practitioner and the teams, and the reform explicitly called for the patients themselves. When care plans are sub- newly developed local health units (unidades ject to patient consent and can be accessed de saúde local, or ULS) to work in multidisci- online by both the patient and the provider, plinary teams to provide primary, inpatient, they can also promote patient empowerment. and public health services. Unfortunately, the To encourage providers to contribute the providers did not fully accept the concept of necessary time and effort to develop and ­ teamwork, and they reported that the process maintain a care plan, CareFirst offered pay- of creating functional multidisciplinary ments to primary care providers for estab- teams was a major barrier to care integration. lished and maintained plans (Murray 2015). Although on paper their multidisciplinary teams appeared similar to those elsewhere, a Core Action Area 3: Vertical Integration culture of teamwork and collaboration was Including New Roles for Hospitals generally missing. Fifteen years after the beginning of the ULS initiative, the teams Vertical integration is a key element of tiered still struggle to function (Forde, Auraasen, service delivery and involves communication and Moreira 2015). and coordination among the primary, second- Importantly, collaborative cultures can be ary, and tertiary health facilities ­ delivering cultivated, as shown by the Health Services care. For China, achieving vertical integration Plan in Canterbury, New Zealand. To foster will involve redefining the roles of, and inter- a culture of collaboration, the architects of actions among, the facilities in all three tiers, L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 69 especially hospitals, so that they work technology and expertise, focusing highly together toward the “3-in-1 principle: one sys- complex care, and providing valuable rescue tem, one population, one pot of resources.” services for life-threatening conditions. They China’s State Council guidance on tiered can also share their personnel to provide health care delivery recently outlined roles technical assistance and training to facilities and responsibilities of facilities at different at lower levels of the system. levels of the system,4 and it is a good basis to In China, the recent policy reform plans build upon. As seen in some of the Chinese and policy directives of the central govern- initiatives featured in the case studies, verti- ment emphasize the integration of health cal integration can also link providers across care across a tiered delivery system, with the tiers of the system to provide support and close collaboration between hospitals and technical assistance and to strengthen the primary care providers (State Council 2012, quality of care. 2015). 5 Integrating county hospitals, town- Of the 22 PCIC initiatives studied for this ship health centers (THCs), CHCs, and vil- report, 15 sought to strengthen vertical inte- lage clinics is not a particularly new concept gration. Their strategies were of three broad in China but is one that continues to be dif- types: (a) redefining facility roles within a ficult to implement. As in other reforms, the vertically integrated network; (b) strengthen- government aims to draw lessons from ing relationships among providers through pilots. Several mostly hospital-led, small- technical assistance and skill building; and scale initiatives seek to coordinate care or at (c) developing formal networks of facilities least link hospitals with other providers. based on the 3-in-1 principle. Most of these initiatives are linked to other reforms, with the integration of care being Strategy 1: Redefine the role of facilities, one of multiple goals for improving service especially hospitals, within a vertically delivery. integrated network The experience of six of the Chinese initia- To ensure coordination and continuity, tives highlights some of the issues associated vertical integration requires cooperation ­ with greater hospital integration into the among health facilities at different levels of delivery system. (For details of these initia- the health care system, many of which may tives, see annex 2C.) Except for Xi County’s not have traditionally collaborated. It is Integrated Care delivery model in Henan necessary to define the roles of facilities so ­ Province, the initiatives took place in large that they function within a robust vertically cities and involved tertiary hospitals, usually integrated network; to determine what as the lead facility. Four were managed by range of services specific health facilities hospital management groups or councils will provide; and to decide how higher-level (HMCs), 6 one by a county health bureau, facilities will support lower-level facilities and another by a lead hospital. through supervision, technical assistance, Though their objectives were varied, all and partnership. six initiatives sought to improve the capacity Internationally, the role of hospitals is of affiliated primary care providers: that is, changing. They are no longer stand-alone CHCs in urban areas and THCs in rural facilities at the center of the delivery system, areas. They used rotating specialists who nor are they the point of entry to care or provided training and technical support; they “one-stop shops” for all services. Rather, improved referral systems; and they estab- they are becoming a part of networks of lished a “green channel” to facilitate upward facilities that include other providers such referrals from affiliated lower-level facilities. as primary care facilities, diagnostic units, Only a few of the initiatives also designed and social services (Porignon and others payment systems to incentivize the coordina- 2011). Within these networks, hospitals can tion of care. Two initiatives—the Shanghai become centers of excellence—concentrating Rujin-Luwan Group and the Zhenjiang 70 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Kangfu Hospital Group—each horizontally At the same time, challenges remain. Key integrated a subset of diagnostic services in a features of integrated systems—gatekeeping, single location. Both involved some degree of use of multidisciplinary teams, use of indi- e-health innovation, such as introducing elec- vidual care plans coordinated by primary tronic consultations among providers for care facilities, patient tracking, and postdis- diagnostics and teleconferencing for training charge care—and the associated policy and clinical guidance (Jian and Yip 2015; ­ m easures still need to be improved and Li and Jiang 2015; Ma 2015b). Programs in ­ i mplemented. In Xi County, the flow of Xi County and Hangzhou established upward referrals continues to dwarf that of e-­consultations so that patients attending pri- downward referrals, and the “green channel” mary care facilities could interact with hospi- (upward referral) admissions represent only a tal specialists (Ali and Li 2015a; Yan 2015a), fraction of total admissions to hospitals.7 while their provinces (Henan and Zhenjiang, Care shifting from hospitals to lower levels respectively) implemented electronic record has also not been fully achieved. systems that their affiliated providers could access. Strategy 2: Establish provider-to-provider Often, integration can force health facili- relationships through technical assistance ties into new roles that they find unfamiliar and skill building and uncomfortable. In these circumstances, Links between providers across the vertical clarifying roles from the outside can provide levels of care can be established and strength- needed direction and guidance. A prime ened through hospitals helping to improve example of this is the Xi County case. quality and competency at the lower levels of In June 2014, four of the county’s hospitals care facilities. Most of the Chinese initiatives and 19 of its THCs were contracting with used technical assistance provided by hospi- each other for inpatient care. The county tals to primary health care facilities to estab- established these contracts and clearly laid lish the interfacility relationships and com- out roles and responsibilities for each level of munication required for effective vertical facility. By linking payment and reimburse- integration. ment to performance, the authorities Two such examples are Feixi County’s i ncentivized the facilities to fulfill their ­ Strengthening Primary Health Care Capacity responsibilities. (SPHCC) and Huangzhong County’s health Findings from these case studies show care alliance (HCA) initiatives, both of which some good results. Zhenjiang reports established technical assistance programs improvements in the number of patients between village clinics, THCs, and county under management for NCDs, while Xi hospitals (Meng, Luyu, and others 2015; County reports a significant increase in fol- Meng, Yinzi, and others 2015). The upper- low-up care of hypertensive patients, as well level facilities were responsible for providing as in flows of two-way referrals. Further, the clinical technical assistance through training, Zhenjiang Kangfu Hospital Group estab- education, and joint consultations to physi- lished multidisciplinary family health teams cians in lower-level facilities. This interaction consisting of personnel from hospitals, strengthened coordination between the levels CHCs, and public health teams to support and was further supported by an e-health integrated management of NCDs and mater- system that allowed health facilities to com- nal and child care. The same program estab- municate with one another. lished rehabilitation wards in four commu- In Hangzhou, the TFY health care reform nity health centers for patients who were created 46 “joint centers” in communities discharged from geriatric and neurology (Yan 2015a). The joint centers were staffed by departments and also implemented a pay- hospital specialists (from four municipal hos- ment system to incentivize chronic-care man- pitals) and primary care physicians to manage agement in CHCs. patients with diabetes and hypertension. L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 71 Features included tracking and coordinating Strategy 3: Develop formalized facility care across the delivery chain, using integrated networks care pathways, crafting individual care plans In many health systems, vertical integration for patients, and fostering the active involve- has been driven by the creation of provider ment of hospital specialists in the postdis- networks. At their most developed, these net- charge care delivered in CHCs. The joint works offer a broad continuum of care across ­ centers also established a peer-to-peer mentor- all possible service lines, connected seam- ing program that paired CHC physicians with lessly through e-health tools, and often take hospital endocrinologists and cardiologists to on financial risk for the health outcomes of improve the CHCs’ management of chronic the populations they serve. Looser, including diseases. “virtual,” networks also exist for vertical Zhenjiang’s Great Health initiative required integration. These virtual networks often hospital specialists to work at least five half- form out of proximity or with the goal to days per week at community health centers to negotiate favorable contracts with payers. provide outpatient services, inpatient rounds, Because they often lack strong governance case discussion, and lectures (Yan 2015b). The structures and shared e-health tools, such as CHCs also provided ­ training opportunities unified patient records, looser networks are for their employees at the core hospital. For often less successful at curbing costs while example, the Liming Community Health integrating care. Center in the Rehabilitation Health Care There are many ways in which China can Group sent one doctor per year for training at create networks that achieve PCIC goals the neurology department in the First People’s without fostering control by hospitals. In Xi Hospital. Further, three or four specialists in County, the integration of care between the traditional Chinese medicine, internal medi- county hospital, THCs, and village clinics cine, and pediatrics were placed in CHCs for a was one goal of an externally financed proj- year to provide intensive training for staff ect that broadly aimed to improve the acces- there. Unique to Zhenjiang’s Great Health ini- sibility, affordability, and quality of rural tiative, the integrated health care groups health care. The Xi Integrated Care initia- addressed multisite licensed physicians and tive created a more formalized network of could ­ effectively use them and their skills health facilities that jointly cared for when it came to providing technical assis- patients (Ali and Li 2015a). It also used a tance. Zhenjiang also created a special fund financial incentive scheme to reinforce the for downward mobilization of resources, such integration across facilities and encouraged as for technical assistance; from this fund the providers to recognize the connectedness of government reimburses a tertiary hospital their system. RMB 80,000 for each specialist that it sends The initiative greatly emphasized the to support a CHC. importance of following guidelines for clini- In Shanghai, the Ruijin-Luwan Hospital cal and integrated care. These guidelines Group implemented a “specialist–general explicitly advised how and at what facility practitioner joint outpatient” service in level to care for a patient with a given condi- CHCs and then developed a training plan for tion, and also specified the criteria for refer- primary health care providers in these centers rals and postdischarge care for more than (Jian and Yip 2015; Ma 2015b). The training 100 conditions at each provider level. Liaison covered basic theories of general practice; officers were hired at the THCs to manage basic knowledge of internal medicine, sur- care coordination and referrals and to over- gery, and diagnosis; treatment of frequently see the use of customized care plans for occurring diseases; provision of people-­ ­ f ollow-up by village clinics. The project centered community health services; com- developed metrics to assess the application of puter skills; and community health service the integrated pathways and introduced management. EHRs that were accessible countywide. 72 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Service agreements among county hospitals, Strategy 1: Promote horizontal integration THCs, and village clinics reflected the above through service colocation features. Although no penalties were imposed At the systems level, horizontal integration for noncompliance with the service agree- may take the form of colocation of services ments, Xi County introduced a payment within a single facility. For example, the mechanism in which insurance payments for DMC initiative in Fosen, Norway, integrated inpatient care were shared between the public health, primary health care, and emer- county hospital and the THCs, encouraging gency care into one facility, thus allowing the providers to shift care out of the hospital to population to access public health and pri- ensure postdischarge care and hospital–THC mary care services—from vaccinations to coordination. emergency medical care—within one loca- However, networks should not be solely tion (Forde, Auraasen, and Moreira 2015). operated by hospitals. In Singapore, the Hangzhou’s TFY created joint centers for movement to integrate public health services, NCDs in CHCs. The centers integrated pub- secondary hospital care, and contract with lic health, specialty care, and primary care primary health care providers through for NCDs, successfully transforming previ- regional health systems (RHSs) aimed to ously fragmented care delivery and making it move away from the concept of the hospital easier for patients to receive a broader array as the anchor of the system. Instead, the of services within a single visit to a frontline Singapore RHSs aimed to center the system facility (Yan 2015a). on the patient’s needs (Teo 2015). Horizontal integration can also help to Hospital capture can occur when hospitals achieve greater economies of scale. The “capture” patients who could be treated in pri- Zhenjiang Great Health initiative consoli- mary care and pull them up into the hospital dated clinical diagnosis facilities and labora- system. To avoid hospital capture, the manage- tories across hospitals and CHCs into single ment of the RHS was separate from hospital units, reducing service overlap and allowing management, and the chairperson of the pri- for the more efficient use of resources vate corporation that oversees all RHSs was a (Yan 2015b). government-appointed employee. These actions signaled an important shift away from the hos- Strategy 2: Provide integrated care around pital-centric model and toward a PCIC system. the individual user to promote more patient-centered care Horizontal integration enables the provision Core Action Area 4: Horizontal of holistic and comprehensive care for the Integration individual patient, bringing together preven- Horizontal integration aims to provide more tive and curative treatments and looking complete and comprehensive services— beyond specific diseases to the person as a including promotional, preventive, curative, whole. At the patient level, a holistic approach reh abi l it at ive , a nd pa l l iat ive c a re —­ to care goes beyond the traditionally defined coordinated by the providers at the frontline medical components and addresses psychoso- facility. Such service integration allows for cial and contextual contributors to disease. more effective management of health care In the United States, for example, the delivery and better-coordinated care within a Program of All-Inclusive Care for the Elderly cohesive health system centered on the needs (PACE) case study highlighted how the pro- of the patient rather than the convenience of gram sought to provide both comprehensive the delivery system. Horizontal integration medical care and holistic care by integrating can also contribute to more efficient use of services across disciplines (Ali 2015). PACE resources by reducing wasteful service dupli- achieved holistic care through home-based cation. Half of the 22 cases reported horizon- care services, meal delivery, intensive social tally integrating care. work, and a nuanced understanding of frail L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 73 elders’ challenges to maintaining their home- This coordination has been shown to result based independence. in more effective care and to decrease the As part of the primary health care initia- unnecessary costs related to duplication of tive in Hangzhou’s TFY implementation, testing, inappropriate medication, and avoid- contracts between physicians and area resi- able complications due to gaps in follow-up. dents covered integrated home-based ser- Use of ICT also helps enable PCIC by facili- vices, living support, and community day tating new forms of interaction beyond short care or nursing centers for the elderly (Yan in-person visits—including, for instance, 2015a). The services were based on residents’ multifaceted shared EHRs with registries, needs and included medical care and social telephone or web consultations, and online activities. appointment scheduling systems. Feixi County’s SPHCC initiative empha- The time, effort, and resources needed to sized the importance of integrating holistic achieve the putative savings from adopting care into modern medical services and cre- e-health systems are substantial. E-health ated a partnership between a traditional strategies were used by 21 (95 percent) of the medicine center and a THC in Zipeng (Meng, 22 PCIC improvement initiatives, which Yinzi, and others 2015). As a result, the clearly shows the importance and core func- Zipeng branch of the Feixi Hospital of tion of the Core Action Area 5 strategies in Traditional Chinese Medicine fully inte- strengthening the health service system. grated health care organizations of varying Three main e-health strategies emerged from levels and types, including both traditional the cases: applying EHRs, establishing elec- and modern Chinese medicine. Because tra- tronic communication and management ditional Chinese medicine is important to functions, and ensuring interoperability. many of the country’s citizens, combining it with modern care better accounted for Strategy 1: Establish EHR systems that are patients’ desires and belief systems and accessible to providers and patients encouraged their overall engagement in their At the center of an effective e-health system is treatment plans. the EHR, which has been shown to improve clinical decision support, registries, team care, care transitions, personal health Core Action Area 5: Advanced ICT records, telehealth technologies, and mea- (e-Health) surement (Bates and Bitton 2010). When Within an advancing technological environ- these key factors function smoothly in a ment, a robust e-health platform is the back- health care setting, both providers and bone of an interconnected health care system patients experience a more coordinated care that puts patients at the center of their care pathway. Providers across different levels can (Bates and Bitton 2010). Use of information communicate in real time and easily access and communication technology (ICT) lays patients’ current and updated health infor- the foundation for successful communication mation in one place. between facilities and also provides health For example, Xi County’s integrated care workers and patients with the opportunity to initiative implemented a new clinical man- participate in the improved service process, agement system using EHRs, including key care management, and decision making. information such as how the dual referral With ICT, patients have the tools to more system operates (Ali and Li 2015a). This sys- fully engage with their care. Such an e-health tem links inpatient and outpatient care, platform can greatly enhance the functional- enabling THCs to monitor the clinical ser- ity and effectiveness of a primary health care vices provided by village clinics. Physicians at system by connecting providers to achieve the THCs could also view the outcome of horizontal and vertical integration, coordina- follow-up appointments as well as the clinical tion, and continuity of information over time. pathways and the individualized care plans 74 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A developed by the upper-level facility doctors. reach a younger generation by using WeChat, Xi’s new system also captured patient a Chinese messaging app. This proved to be a referrals. quick and easy way to get health information Hangzhou’s TFY reform provided area to patients, who could use it to check physi- residents with a citizen card and a database cian information, make appointments, and of “intelligent services” that held all patients’ update their registration and payment forms. lifetime health information and allowed it to Telemedicine and video conferencing be shared between providers (Yan 2015a). played a particularly important role in The database housed information such as Norway at the rural Fosen DMC. Video con- records of antenatal care and delivery, child ferencing expanded access in two ways: care, health screenings, NCD management, (a) primary health care providers could diagnosis and treatment, hospitalization, and ­ c onsult with secondary and tertiary care laboratory testing. providers, and (b) patients could see second- EHR systems that allow patients access to ary care providers (Forde, Auraasen, and their own health records can also increase Moreira 2015). When the center was first patients’ empowerment and engagement in established, six videoconference units were their care. For example, the U.S. VHA devel- purchased and connected to the secondary oped a patient portal called My HealtheVet hospital, St. Olav’s (a three-hour drive from to support patients’ self-management of their Fosen). The primary health care providers at health needs (Ali and Li 2015b). A patient the DMC all participated in a daily morning portal is a website where patients can view video meeting with St. Olav’s. In these meet- their personal health record, refill prescrip- ings, physicians discussed current DMC tions, view lab results, send secure messages patients and could seek consultations from to their physician, and review their ­ physician’s specialists as needed. As providers became notes. The portal was reported to have been comfortable with the video conferencing helpful in coordinating and integrating care software and developed relationships with for veterans. Recent studies show it has each other, they expanded the videoconfer- improved patients’ experiences of care and ences to include patient consultations. This connections with their personal health teams. capability was used especially during night visits to the acute care center. Strategy 2: Establish communication and The primary care providers’ workstations care management functions in Huangzhong’s health care alliance provide ICT can help expand access through online another example of how to expand access to appointment scheduling, video conferencing, quality care through e-health. These mobile and mobile workstations; it can also help workstations, which were piloted in 2013 in improve patient safety. Online appointment 30 districts, could be carried by village doc- scheduling is one method to improve patient tors to more-remote locations (Meng, Luyu, access to health services. Turkey’s Health and others 2015). Doctors could use them to Transformation Plan created a central physi- conduct a number of medical exams (includ- cian appointment system that schedules ing elderly checkups and health assessments, appointments for primary, secondary, and hypertension management, diabetes manage- tertiary facilities over the telephone and ment, electrocardiogram monitoring, and online (Sumer 2015). This system allowed pulse oximetry testing) and to upload data patients to request an appointment with a and results to the system’s health information specific physician, at a specific office loca- services. The workstations gave clinicians tion, or in a specialty area and reduced the better mobility when it came to follow-up long waiting times at clinics. care, and local residents trusted the results of In China, both Shanghai’s FDS and Xi the high-tech device more than the unaided County’s Integrated Care initiative aimed to diagnoses made by the village doctors. L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 75 The use of e-health tools has effectively simultaneously with the providers at St. Olav’s improved patient safety. For example, Feixi (Forde, Auraasen, and Moreira 2015). County’s SPHCC initiative used ICT to reduce overprescription and to promote Core Action Area 6: Integrated more accurate and careful administration Clinical Pathways and Functional Dual of medications to patients in THCs and vil- Referral Systems lage clinics (Meng, Yinzi, and others 2015). Medical information technology systems, Integrated clinical pathways attempt to stan- using clinical decision-support tools, stored dardize the treatment and referral pathways recommended lists of drugs for 50 outpa- that providers use in at least two levels of a tient diseases that were cared for in pri- health system to address particular condi- mary health care facilities. Once a diagno- tions. They also clarify relationships and sis was put into the system, a set list of responsibilities between the different provid- possible medications appeared, and the ers in the system. Because these pathways physician could only choose a drug from may finally lead to referrals to another level that list. of care, they are most effective in the context Similarly, in Xi County, the EHR system of strong vertical integration. Dual referrals improved patient safety by linking the tiers of include referrals from primary to secondary the health system, allowing clinical pathways care and also referrals back to primary from to be put in place with guidelines about how secondary care. Integrated pathways and to care for a patient with a particular illness strong dual referral systems are important to (Ali and Li 2015a). If a physician deviated facilitate providing the right care at the right from the pathway, the change was recorded; time. approved deviations were monitored by a Of the 22 case studies, 15 (68 percent) quality control officer, and any changes in applied integrated care pathways, and 13 care had to be made in consultation with the (59 percent) used dual referrals. Two main hospital unit director. strategies were applied: (a) crafting integrated pathways to facilitate care integration and Strategy 3: Ensure interoperability of decision support for providers, and (b) pro- e-health tools across facilities and services moting dual referrals within integrated facil- E-health tools carry great potential to ity networks. improve the quality and safety of care, but they must be interoperable between facilities; Strategy 1: Craft integrated care pathways that is, they must be capable of being accessed to facilitate care integration and decision by different providers in different facilities. support for providers Where multiple e-health systems function, a Clinical pathways help integrate care across major challenge exists in getting them to providers and provide valuable decision “talk to each other.” support. As a part of the Canterbury Health ­ Interoperability (when the records can be Services Plan in New Zealand, clinicians lawfully used across institutions) needs to be developed a program called Health Pathways, built into an e-health system from the start. which created 570 clinical pathways for In Norway, Fosen’s DMC achieved interoper- referral (figure 2.4). The pathways made ­ ability between its records and those of the secondary-care referral decisions explicit to ­ secondary hospital with which it partnered ­ reduce variation in referral patterns and (St. Olav’s). Because the DMC was developed avoid unnecessary or duplicate referrals. with this partnership in mind, it adopted the Measures used to maximize their effective- same EHR system as St. Olav’s rather than ness included biannual reviews for quality by create its own information system. Thus, the a group of clinicians; periodic clinical audits; physicians at the DMC could view any patient and updating through a formal modification record they needed at any time of day, process. 76 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 2.4  Sample health pathway for COPD, Canterbury Health Services Plan, New Zealand Assessment Management Request • Make a diagnosis of COPD • Record confirmed case in Disease Register • Request respiratory physician • Consider COPD in current and • Nonpharmacological interventions are assessment if ex-smokers who are symptomatic as important as drug treatments for all • Diagnostic uncertainty • Consider COPD for any current or patients with COPD, particularly • Age > 40 years ex-smoker who had an emergency • Smoking cessation: This is the single • Severe disease group (group D) and treatment for a respiratory condition most important intervention. Provide age > 55 years • In patients who have never smoked, advice at every opportunity. • Frequent exacerbations consider diagnosis other than COPD • Exercise: Encourage exercise and • Difficult to control symptoms • Differentiate between asthma and consider pulmonary rehabilitation for all, • Uncommon sputum pathogens COPD especially groups B and D, as it improves • Sleep-disordered breathing • Arrange spirometry with bronchodilator quality of life, reduces dyspnea, and • Home oxygen required • Mandatory for accurate diagnosis improves exercise tolerance. • If appropriate, consider community • Peak flow testing is inadequate • Identification and management of respiratory services • Other investigations commonly associated comorbidities: • Your patient may also wish to consider • Chest x-ray: only arrange if suspicious Patients in groups B and D have high private respiratory sociality assessment. of other conditions rates of cardiovascular comorbidity. • Blood tests: CBC and BMP if possible • Important interventions for all patients coexisting heart failure include • Pulse oximetry: if < 92% on two • Annual influenza immunization occasions when resting, refer for • Advice on occupational factors arterial gas testing • Acute exacerbation management • Look for comorbidities • COPD action plan • Determine COPD classification • Adequate home heating or subsidy for (A, B, C, or D) heating • Consider other interventions, especially for group D patients • Specific management depends on the severity group Source: Adapted from Love 2015. Note: COPD = chronic obstructive pulmonary disease; CBC = complete blood count; BMP = basic metabolic panel. Based on international best practices, the (Ali and Li 2015a). The pathways defined pathways were tailored to the needs and the scope of responsibility for hospitals and interests of the local population. Because of THCs, clarified when patients should be the high degree of clinical rigor used to transferred to a THC for continued ­ inpatient develop and evaluate these pathways, physi- care, and provided guidelines for discharge cians placed high trust in them and used and follow-up care at village clinics them intensely: more than 80 percent of phy- (figure 2.5). sicians viewed the Health Pathways website All the county hospitals and THCs had at least six times per week. Physicians full-time liaison officers who were responsi- reported feeling that their referral decisions ble for soliciting feedback from patients and were more rational and that care was more staff, coordinating dual referrals, making tailored to patient needs (Love 2015). appointments, and transferring patient The health care initiative in Xi County records when a referral was necessary. At the also emphasized the importance of adhering village clinic level, 21 disease-specific path- to clinical pathways, which were established ways were created and implemented to sup- for 188 diseases in county hospitals and 104 port decision making and strengthen a dual diseases within THCs at an inpatient level referral system between the clinics and THCs. L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 77 FIGURE 2.5  Sample responsibilities for dual referrals in Xi County Village clinic Township health center County hospital • Must develop and use patient referral • Must develop and use patient referral • Must develop and use patient referral form form form • Acquire an understanding of the • For referrals with village clinics: set up • For downward referral: transfer patient specialties of THC doctors and the services green channel referral system; acquire records to THC; arrange for transfer back they provide patient consent for referral; ensure THC to hospital if necessary; complete • Assist patients with selecting a physician physicians mentor village clinic physicians patient transfer form and specify at the THC • For referrals with county hospitals: acquire existing conditions of treatment, • Identify patients in need of care an understanding of the specialties of treatment plan, and recovery pathway • Acquire patient consent for referral to county hospital physicians and the services • For receiving referrals: register patients THC they provide; assist patients with transferred from THC; designate person • Complete patient information before selecting a physician at the county responsible for directing referred referral to THC hospital patients to appropriate care Source: Ali and Li 2015a. ©World Bank. Permission required for reuse. Note: THC = township health center. Strategy 2: Promote dual referrals within Core Action Area 7: Measurement integrated facility networks Standards and Feedback All the Chinese PCIC initiatives that were studied employed upward referrals using the Establishing a performance measurement “green channel.” As described in the case system is critical to ensuring the quality and studies, patients who were referred upward performance of a PCIC-based system. The through the green channel from participating performance measurement indicators need to facilities in their system were expected to reflect the national standards, which in turn receive expedited care at hospitals. In prac- should reflect the core functions and goals of tice, however, green channels functioned the PCIC-based service delivery system, irregularly. Downward referrals from hospi- including coordination, comprehensiveness, tals to CHCs were rare, and some patients integration, and technical and experiential resisted them. quality. However, collecting only the indica- Notably, the dual referral system in Xi tor information would not improve the sys- County’s Integrated Care initiative was tem’s performance; formation of a feedback incentivized by cost sharing and reimburse- cycle is required to ensure that the perfor- ment. Under this scheme, upper-level facili- mance results are communicated to stake- ties were reimbursed for the entire cost of a holders at all levels, ranging from the com- referred case and shared the payment they munity and providers through to management received with the lower-level facility, depend- and policy makers ing on a previously determined price and the Performance measurement can also iden- care workload. The reimbursement depended tify early positive outliers that can teach oth- on whether the patient care pathway had ers and point to effective intervention compo- been satisfactorily fulfilled in both health nents for broader implementation. Of the facilities (Ali and Li 2015a). This system 22 initiatives studied, 20 (91 percent) used encouraged hospitals to refer patients to measurement and evaluation to strengthen lower-level health facilities to save costs and their performance.8 Two common strategies bed space. for promoting measurement and feedback 78 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A emerged: (a) development and use of stan- Strategy 2: Create continuous feedback dardized performance metrics; and (b) cre- loops linked to action plans to drive quality ation of feedback loops to drive continuous improvement quality improvement. To drive improvement, standardized perfor- mance measures must be fed back to the Strategy 1: Use standardized indicators for appropriate stakeholders so that results can performance measurement be used to improve quality. A commitment Performance measurement should be stan- to identifying, learning from, celebrating, dardized through the use of common, verifi- and spreading identified effective practices is able, and meaningful performance indicators. important at all levels of the system. These For example, the Gesundes Kinzigtal initia- processes will accelerate change, motivate tive in Germany used standardized reports providers and managers, and fur ther based on a core set of measurement standards increase the value and use of performance-­ for care providers, the management team, measurement data. and other stakeholders (Hildebrandt and Measurement information needs to be others 2015; Nolte and others 2015). The ­ communicated to all levels of the PCIC measurement covered the system, technical system—starting with patients and commu- ­ quality, and service-experience quality (the nities, who can become active partners in patients’ experiences), and it included the strengthening primary health care by pub- dimensions of structure, process, outcomes, licly sharing data on facility performance quality, integration, patient experience, and and community health and by building efficiency. The use of performance measure- their capacity to understand results and ment facilitated communication related to engage in improvement and advocacy. Such progress and enabled comparisons across engagement allows communities and pro- facilities. viders to hold health systems accountable The outcomes and processes chosen for for the quality, responsiveness, and out- measurement should reflect the priorities of comes of the care provided. Care teams and the system. The leaders and managers of the managers at the facility and subnational Huangzhong health care alliance developed a level are other vital users of measurement quality and safety evaluation system that information. They may need support and could measure a set of core components encouragement to use feedback to under- broader than medical services, including man- stand and improve their performance agement (financial management and village within the context of the facility, network, clinic management), disease and prevention and broader care system. control, patient satisfaction, and Chinese tra- To transform performance data into ditional medicine services (Meng, Luyu, and action and improvement requires regular others 2015). The metrics used for primary feedback loops that enable the identification health care reflected the priority goal of of gaps in services and that drive and improvement within the primary health care support continuous learning and correction. ­ workforce—focusing on provider availability, A strong focus on feedback, linked to action motivation, and performance; patient satisfac- l evels of the system, is critical. To at all ­ tion; the number and quality (enthusiasm, ensure ongoing learning requires a resilient skills, and ability) of the workforce; and some system with the following main elements: system measures, including indicators of coor- performance measurement, feedback and dination within the health service network. review of the data, identification of gaps, Many OECD countries have established and design and implementation of interven- and implemented patient-reported outcome tions. Each of these elements needs to be measures and patient-reported experience underpinned by support and training of measures as part of their health system perfor- staff in improvement methods. The cycle mance assessment systems (Klazinga 2014). continues with remeasurement to assess L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 79 whether gaps have been closed and whether In Feixi’s SPHCC initiative, the county’s new gaps have been identified. THCs used a nine-dimensional evaluation For example, the PACE case outlines how tool to measure and improve their perfor- the program’s charter provided for continual mance (Meng, Yinzi, and others 2015). This feedback (figure 2.6) through a Quality tool was used twice per year and became a Assessment and Performance Improvement basis for budget allocation and performance- program that is data-driven, community-led, based financing. THC directors’ salaries were and iterative (Ali 2015). Providers received tied to evaluation results; negative reviews the performance measurement results regu- resulted in penalties for the directors and also larly so they could review their personal per- affected governmental funding. Some health formance and identify problem areas across facilities found evaluation to be helpful the practice. Many PACE centers instituted because it pushed them to provide high-­ monthly, biweekly, or even daily feedback quality care, and top-performing facilities processes, combining required, optional, and received praise and recognition from the facility-initiated performance measurement. government. All PACE centers followed the Health Plan Management System of the Centers for Core Action Area 8: Accreditation and Medicare and Medicaid Services (CMS) and Certification were required to report basic information on core performance measures every quarter to Accreditation is a formal process by which CMS (figure 2.7). Following a recommenda- an independent body conducts an external tion by the National PACE Association, indi- assessment of a health care organization’s vidual institutions within PACE have recog- performance in relation to previously nized the value in continual qualit y defined metrics and published standards. At improvement efforts and have begun collect- its core, accreditation is a defined mecha- ing additional measures focused on local nism for externally assuring accountability priorities. for minimal standards to be met across the FIGURE 2.6  PACE continual feedback loop Design and implement intervention Collect data • Incorporate improvements into • Establish and maintain a health information standard practice for the delivery of care; system that collects, integrates, and reports data track performance to ensure that • Train staff in data integrity concepts and improvements are sustained practices Identify Gaps Obtain Feedback and Review Data • Use data collected to identify areas of good or poor performance and prioritize • Document and disseminate QAPI activities performance improvement • Immediately correct problems that threaten the activities health or safety of participants Source: Ali 2015. ©World Bank. Permission required for reuse. Note: PACE = Program of All-Inclusive Care for the Elderly; QAPI = Quality Assessment and Performance Improvement. 80 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 2.7  Types of data collected by PACE centers • Locally developed performance measurements, focusing on individual facility priorities for improvement • Examples: Facility- Cost per participant for primary care, meals, and transport initiated Information on falls: where the fall occurred, cause of fall, who else was present, resulting injury, and location of treatment • Additional data the National PACE Association suggests centers collect, focusing on PACE priority areas • Allows facilities to measure individual performance for improvement Optional and compare performance between institutions • Examples: Days spent in nursing homes, meals given • Required by the government’s Centers for Medicare and Medicaid Services • Allows for monitoring at a national level Required • Examples: Deaths, hospitalizations, and grievances Falls, burns, and medication errors Source: Ali 2015. ©World Bank. Permission required for reuse. Note: PACE = Program of All-Inclusive Care for the Elderly. health care delivery system. Accreditation Although the changes are in the right direc- differs from licensure, which is generally tion, the guidelines leave considerable dis- considered a government regulatory respon- cretion to local government in operational- sibility and is designed to set minimum stan- i z i n g a nd i mpl e m e nt i n g t h e s y s t e m dards to protect public health and safety. (Wagstaff and others 2009). As a result, the Accreditation sets standards that are consid- accreditation process does not cover many ered optimal and achievable but are more key health care delivery organizations, rigorous than the minimum standards used including THCs, village clinics, and private for licensure, and it has the stated intent to sector health delivery organizations. This foster a culture of improvement (Mate and limits comparability of performance across others 2014). The terms accreditation and localities. certification are often used interchangeably, A recent review of 44 health care accredi- but accreditation usually applies only to tation programs in 38 countries reported six organizations, while certification may apply key features (Braithwaite and others 2012): to individuals as well (Salmon and others 2003). • Development or adoption of a set of health I n 20 05, t he C h i ne se gover n ment care standards reformed its hospital accreditation system, • Enrollment of members who assess moving away from a focus on infrastruc- their own performance against those ture and equipment and toward a broader standards range of criteria including scientific man- • Recruitment, education, and management agement, patient safety, and service quality. of a workforce of industry surveyors L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 81 • Deployment of teams of surveyors to by the NCQA and other organizations. health care organizations to assess prog- When setting criteria, it will be important to ress against the standards consider the reporting burden and to stream- • Evaluation of survey teams’ reports line the number of measures adopted. For • Award of accreditation status (if eligible) example, the Aarogyasri Health Care Trust, for a period of three to five years. which insures more than 65 million people in the Indian state of Andhra Pradesh, External organizational and clinical cer- recently began identifying standards that tification (or accreditation) standards are will be required of all hospitals that are critical to ensuring high-quality, reliable, empaneled to provide care to the Trust’s ben- and safe care organizations (Greenfield and eficiaries. The Trust is working closely with others 2012). However, accreditation fea- India’s National Accreditation Board of tured in only 5 of the 22 PCIC performance Hospitals, as well as with insurers from sev- improvement initiatives studied for this eral states, to develop a shared set of stan- report. Their strategies to launch accredita- dards that will encourage other hospitals to tion include developing criteria and setting also achieve accreditation and thus improve targets. their quality of care (Smits, Supachutikul, and Mate 2014). Strategy 1: Develop certification criteria that are nationally and locally Strategy 2: Set targets and use them to relevant accredit facilities Criteria for accrediting primary care facilities Once the accreditation criteria have been need to reflect the priorities and structure of developed, the next steps are to set targets a PCIC-based delivery system. They should and apply the criteria through a transparent define model standards in areas ranging from and reliable mechanism. Under the VHA’s infrastructure (resources, information tech- PACT model (Ali and Li 2015a), a primary nology, and human resources); systems orga- health care facility must meet certain crite- nization (integration and hospital and pri- ria to be recognized as a PCMH by the mary health care roles); and care delivery NCQA. The NCQA uses a point-based sys- characteristics (people-centeredness, compre- tem with three levels of classification, and it hensiveness, continuity, and coordination) to also requires a facility to score higher than facilitate desired health outcomes. 50 percent on each of six “must-pass” ele- For example, for a facility to be recog- ments to receive accreditation (table 2.3). nized as a PCMH—a form of PCIC facility Many primary care facilities publicize their recently launched in the United States—it attainment of these recognition levels must meet the following criteria established to attract patients (Bitton, Martin, and by the National Committee for Quality Landon 2010). Assurance (NCQA): team-based care, care New accreditation programs benefit from coordination, patient self-management, setting relatively achievable standards based enhanced access and continuity, care man- on the current status of local health care agement , a nd qu a l it y i mprovement . facilities and making a commitment to Performance metrics based on these stan- upgrade their standards over time. Malaysia dards can be used to assess the quality of and Thailand have each adopted this care being provided and to compare perfor- approach. Malaysia has issued four versions mance across providers and facilities in a of its hospital standards since starting its standardized way. accreditation program in 1999. Thailand’s Looking ahead, China may wish to draw progressive changes include introducing a on a wide array of easily available and scien- stepwise recognition program in 2004, tifically proven protocols and guidelines for followed by patient safety goals in 2006 ­ care that are available on websites sponsored (Smits, Supachutikul, and Mate 2014). 82 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 2.3  NCQA certification guidelines for patient-centered medical homes Standards Elements Possible points Patient-centered access Patient-centered appointment access 4.5 24/7 access to clinical advice 3.5 Electronic access 2.0 Total 10.0 Team-based care Continuity 3.0 Medical home responsibilities 2.5 Culturally and linguistically appropriate services 2.5 Practice team 4.0 Total 12.0 Population health management Patient information 3.0 Clinical data 4.0 Comprehensive health assessment 4.0 Use data for population management 5.0 Implement evidence-based decision support 4.0 Total 20.0 Care management and support Identify patients for care management 4.0 Care planning and self-care support 4.0 Medication management 4.0 Use electronic prescribing 3.0 Support self-care and shared decision making 5.0 Total 20.0 Care coordination and care transitions Test tracking and follow-up 6.0 Referral tracking and follow-up 6.0 Coordinate care transitions 6.0 Total 18.0 Performance measurement and Measure clinical quality performance 3.0 quality improvement Measure resource use and care coordination 3.0 Measure patient and family experience 4.0 Implement continuous quality improvement 4.0 Demonstrate continuous quality improvement 3.0 Report performance 3.0 Use certified e-health record system technology Not scored Total 20.0   Overall total 100.0 Source: NCQA 2014. Note: NCQA = National Committee for Quality Assurance. Patient-centered medical homes are a primary care model for providing comprehensive, coordinated care across all elements of the broader health care system in ways centered on individual patients’ needs, accessibility of services, high quality of care, safety, and accountability (“Defining the PCMH,” Patient Centered Medical Home Resource Center, U.S. Department of Health and Human Services [accessed June 13, 2018], https://pcmh.ahrq.gov/page/defining-pcmh). L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 83 Moving Forward on the PCIC Core ways, but at their core is an EHR that is Action Areas accessible across providers, and by patients, in a given region or locality. China needs China has already begun reforms in some of to implement e-health systems that not the eight core action areas for achieving only facilitate and document interactions PCIC, including introducing nearly univer- between providers and patients, but also sal health insurance coverage, reducing serve as coordination mechanisms for dual markups on sales of medication, and referrals and promote comprehensive, team- strengthening inputs, including health facil- based care. Importantly, e-health systems ity infrastructure and training of health care must allow providers and facilities at differ- workers. Although recent reforms have led ent levels of the network to communicate to improvements, developing a high-quality with each other so that patients’ care can be PCIC system that can rein in costs, increase transferred seamlessly across them. In addi- value for money, and improve population tion, an effective e-health system that can health outcomes will require f ur ther detect patients who need more-intensive ser- reforms. vices (for example, patients with recurrent These reforms must be shaped by current hospitalizations, poor control of NCDs, or circumstances and by lessons from improve- frequent primary health clinic visits) will be ment initiatives in China and in other coun- needed, while to respond effectively to the tries. It is not feasible to address all eight needs of these patients will require individu- action areas at the same time, and within alized care plans and multidisciplinary each action area, strategies will need to be teams. prioritized. Thus, it will be crucial for The case studies highlight the role played Chinese authorities to develop clear imple- by stable and consistent management. mentation guides or toolkits on what to Attention to building the necessary skills and implement and how, with indications on capacities within a cadre able to support the where and how local adaptations are goals of the improvement is essential. needed. These guidelines will need to be Important management functions will be supported with measurement and feedback to support performance measurement to systems to monitor the implementation of identify gaps, drive improvement through improvement efforts and troubleshoot prob- supportive supervision, and ensure account- ­ lems as needed.9 ability. One-third of the case studies noted China will also need to address provider that flexibility and “opportunistic” imple- payment mechanisms and ensure they are mentation were major factors in success. aligned with and actively support the chosen Allowing the managers of facilities some improvement interventions for PC IC . degree of autonomy to address unexpected Achieving integration will require joint challenges and local opportunities for inno- accountability for costs and quality through vation—and encouraging them to share the networks of care, strong dual referrals, and a lessons they have learned—will be impor- health information system that helps patients tant, regardless of the pathway taken. receive coordinated care across providers and Finally, effective measurement and feed- facilities. back loops need to be designed and imple- Effective use of ICT to create e-health mented to monitor and strengthen initiatives systems has been a key to success in almost as they are started and scaled up. And mean- all the cases studied. These e-health systems ingful patient engagement will be crucial are structured and implemented in varying throughout the change process. 84 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Annex 2A  Summary of PCIC Impact Findings by Case Studies TABLE 2A.1  Impact frequency of studies on PCIC initiatives, by PCIC model and impact area Number of studies Intermediate health Hospitalizations Processes outcomes and Patient Model Impact and ED use of care mortality experience Costs Citations General PCIC Improvement 17 7 21 9 22 Guanais and Macinko (2009); (52 studies) Hildebrandt and others (2015); No change or 5 4 10 1 2 Macinko and others (2011); worsened Nolte and Pitchforth (2014); Insufficient or 30 41 21 42 28 RAND (2012); Schulte and others inconclusive evidence, (2014); World Bank (2015) or not measured PCMH (14 Improvement 12 7 4 2 6 Bitton (2015); DeVries and studies) others (2012); Fifield and others No change or 2 1 1 0 2 (2013); Friedberg and others worsened (2014); Friedberg and others Insufficient or 0 6 9 12   (2015); Gilfillan, Tomcavage, inconclusive evidence, and Rosenthal (2010); Hebert or not measured and others (2014); Nelson and others (2014); Reid and others (2010); Reid and others (2013); Rosenthal and others (2013); van Hasselt and others (2015); Wang and others (2014); Werner and others (2014); World Bank (2015) PACE Improvement 9 0 7 1 0 Beauchamp and others (2008); (16 studies) Chatterji and others (1998); No change or 1 0 1 3 0 DHCFP (2005); Kane, Homyak, worsened and Bershadsky (2002); Kane and Insufficient or 6 16 8 12 16 others (2006a, 2006b); Mancuso, inconclusive evidence, Yamashiro, and Felver (2005); or not measured Meret-Hanke (2011); Mukamel, Bajorska, and Temkin-Greener (2002); Mukamel and others (2006); Mukamel and others (2007); Mukamel, Temkin- Greener, and Clark (1998); Temkin-Greener, Bajorska, and Mukamel (2008); Weaver and others (2008); Wieland and others (2000); Wieland and others (2010) Disease or Improvement 82 22 64 28 34 Elissen and others (2012); Elissen case and others (2015); Frølich, No change or 29 6 25 8 37 management Jacobsen, and Knai (2015); Nolte worsened (257 studies) and Pitchforth (2014); Runz- Insufficient or 17 0 14 10 9 Jørgensen and Frølich (2015); inconclusive evidence Struijs, de Jong-van Til, and others Not measured 129 229 154 211 178 (2012); Struijs, Mohnen, and others 2012); Vadstrup and others (2011) China (6 Improvement 1 6 1 1 2 World Bank (2015) studies) No change or 0 0 0 0 0 worsened Insufficient or 5 0 5 5 4 inconclusive evidence Note: ED = emergency department; PACE = Program of All-Inclusive Care for the Elderly; PCIC = patient-centered integrated care; PCMH = patient centered medical home. L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 85 Annex 2B  Methodology and Summaries of 22 PCIC Performance Improvement Initiatives Described in the Case Studies Case study research consisted of three stages: research and qualitative work, including focus g roup d iscussions and key-­ 1. In the first stage, each author completed a informant interviews. rapid mapping of the county or country’s performance improvement initiative, the Because no applicable tool was identified progress to date, and important contex- from the existing literature, the authors tual factors (policy, institutional, and developed a tool designed to capture the financial environments). follow ing in formation from the case 2. In the second stage, the details of the ini- studies: tiative were further analyzed, focusing on key components of PCIC, implementation • Challenges prompting the design and readiness, implementation strategies, and implementation of the PCIC performance initial and intermediate outcomes of the improvement initiative initiative. • Core initiative elements 3. The third stage examined critical contex- • Strategies for initiative implementation tual factors influencing the initiative • Facilitators and barriers to implementation design, implementation approach, and • Lessons learned successes and challenges. Researchers • Adaptability of the initiative to the Chinese used a mixed-methods approach that health care system consisted of both quantitative secondary ­ • Potential sustainability. TABLE 2B.1  Nomenclature and summaries of 22 PCIC performance improvement initiatives PCIC performance improvement initiative Description Citations Chinese case studies Zhenjiang, Jiangsu province: Zhenjiang city, on the Yangtze River in eastern China, implemented the Great Health Yan (2015b) Great Health initiative in 2011 to service its two main districts. Through this initiative, two health care groups, Rehabilitation Health Care Group and Jiangbin Health Care Group, were created that focused on vertical and horizontal integration with new “3+X” family health teams managing the care of all contracted residents. Shanghai: Family Doctor Huangpu and Pudong, two neighboring districts within coastal Shanghai, China, Ma (2015a) System implemented the family doctor system (FDS) in April 2011. This case study focused on five community health centers within these districts. The FDS centered on strengthening the relationship between the general practitioner and contracted resident by using empanelment and improved frontline service delivery to establish a continuous health care relationship with a particular focus on the management of chronic diseases. Huangzhong, Qinghai In northwest China, Huangzhong County implemented a health care alliance system in Meng, Luyu, and province: Health Care Alliance 2013 to vertically integrate county, township, and village health centers. By focusing on others (2015) the creation of a unified administration, the integration of human resources, a tight dual referral arrangement, the interconnection of health information systems, and shared medical resources, an integrated “county-township-village” health system emerged. Hangzhou, Zhejiang province: Hangzhou, capital of China’s Zhejiang province, is home to more than 8 million Yan (2015a) Twelfth Five-Year Plan individuals and has traditionally struggled to provide equal and sufficient health care to its citizens. To curb such obstacles, the Twelfth Five-Year Plan was implemented in 2011, and key aspects included integrated e-consultation services, noncommunicable disease joint centers, and collaborative services for medical and living support and nursing care. (Table continued next page) 86 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 2B.1  Nomenclature and summaries of 22 PCIC performance improvement initiatives  Continued PCIC performance improvement initiative Description Citations Feixi, Anhui province: Feixi County, of Anhui province in eastern China, has a population of roughly 850,000 Meng, Yinzi, and Strengthening Primary Health citizens. In 2009, Feixi became the pilot site for the government’s Strengthening Primary others (2015) Care Capacity Health Care Capacity initiative, which focused on strengthening four sectors: human resources, network building, organization and management, and working conditions. Xi, Henan province: Integrated In Xi County, China, the integrated care (IC) reform in 2012 addressed the low quality Ali and Li (2015a) Care Reform of care for noncommunicable diseases and disjointed health systems by implementing contracts between county hospitals, township health centers, and village clinics. The initiative focused on building a strong referral mechanism, providing technical assistance to lower-level facilities, and altering the payment system to support cost sharing, all of which have had considerable success, even in their early stages. Beijing: Beijing Chaoyang The Beijing Chaoyang Hospital Alliance (CHA), started in late 2012, aimed to attract Jian and Yip Hospital Alliance (four cases) patients to use community health centers more frequently for minor ailments and (2015) to strengthen the collaboration between upper- and lower-level facilities. The CHA comprised a core hospital, a second tertiary hospital, a secondary hospital, and a number of community health centers, which coordinated care for patients. As a result of this structure, the growth rate of participating facilities rose from 2012 to 2013. Beijing: Peking University- Started in 2007, the Peking University (PKU)–Renmin Hospital Integrated Delivery System Jian and Yip Renmin Hospital Integrated (IDS) in Beijing increased technical assistance between health facilities and improved (2015) Delivery System (four cases) communication between providers through an information technology system. Through this system, providers were able to engage in telediscussions and specialist education and training, thus supplementing available continuing education for all providers in the IDS. Shanghai: Shanghai Ruijin- In 2011, the Shanghai Ruijin-Luwan Group was established, consisting of Shanghai Jiaotong Jian and Yip Luwan Hospital Group (RLG) University as the core hospital, two secondary hospitals, and four community health (2015); Ma (2015b) (four cases) centers, which serviced people in the immediate area. This health care group created a shared imaging and testing center that increased access for residents, provided “specialist– general practitioner joint outpatient” visits for patients in community health centers, and strengthened its previously existing primary care provider training base. Zhenjiang, Jiangsu province: The Jiangsu Zhenjiang Kangfu Hospital Group began in late 2009 in Zhenjiang, China. Li and Jiang Jiangsu Zhenjiang Kangfu This initiative integrated imaging, chemical laboratory, and pathology test departments (2015) Hospital Group (ZKG) and required primary health care facilities to take more responsibility for chronic-disease (four cases) outpatient services. Additionally, the hospital group established “3+X” health teams and supported more frequent information exchange. International case studies Denmark: Integrated Effort for Denmark piloted its chronic disease rehabilitation programs in Copenhagen with Nolte and others People Living with Chronic four centers, called SIKS rehabilitation centers. Owing to the success of the centers, (2015); Runz- Diseases (SIKS) Denmark embarked on a national disease management program, which provides Jørgensen and integrated comprehensive chronic disease care. Frølich (2015) England, U.K.: James The James Cook University Hospital (JCUH) is in northern England, where hospitals Forde, Auraasen, Cook University Hospital are public but semiautonomous. In the early 2000s, JCUH developed an Ambulatory and Moreira Ambulatory Emergency Care Emergency Care Center where patients could receive same-day care using (2015) predetermined clinical guidelines for certain conditions instead of being hospitalized. Simultaneously, it developed patient care pathways and explicitly strengthened the interface between primary care physicians and the hospital. Kinzigtal, Germany: Gesundes Gesundes Kinzigtal, a health care management company in the Black Forest area of Hildebrandt and Kinzigtal (GK) Germany, launched in 2005 a unification of a nonprofit, physician-run organization others (2015); (MQNK) and a health-science management and investment company (OptiMedis). Nolte and others The integrated organizational model focused on improving the health of the (2015) population as well as patient experience while considering a fair business plan that appropriately incentivized patients and providers to join. (Table continued next page) L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 87 TABLE 2B.1  Nomenclature and summaries of 22 PCIC performance improvement initiatives  Continued PCIC performance improvement initiative Description Citations Netherlands: Maastricht The Maastricht region in the south of the Netherlands developed an integrated Vrijhoef and Diabetes Care (DTC) framework for diabetes care whereby the insurers negotiate with the primary care Schulpen (2015); (In-text reference: physicians a price for a complete package of care for a specific disease. Based on the Nolte and others Netherlands, DTC) care package’s success, the Netherlands expanded this program nationwide in 2010. (2015) Canterbury, New Zealand: Canterbury, a district in central New Zealand, developed its Health Services Plan in Love (2015) Health Services Plan 2007. The plan included initiatives like the Acute Demand Management Services, Health Pathways standardization of care for hundreds of conditions, and the Community Rehabilitation and Enablement Support Team. Concurrent enabling initiatives, including an electronic medical record system, an electronic referral system, clinical continuing education programs, and a formal alliance between health care facilities, supported the mission of developing people-centered, coordinated, and integrated health care. Fosen, Norway: District Fosen, Norway, a municipality in the fjords of northern Norway, developed a Forde, Auraasen, Medical Center comprehensive district medical center (DMC) model. The Fosen DMC provides and Moreira integrated, coordinated acute medical care to people in the community to help (2015) them avoid hospital stays. In 2012, Norway modeled its national health care initiative on Fosen’s successful DMC model. Portugal: Local Health Unit In 1999, a small province in northwest Portugal created a local health unit (unidad de Forde, Auraasen, saúde local, or ULS) that provides integrated primary and secondary care to a defined and Moreira geographic area (Matosinhos) with centralized management and coordinated services. (2015) Starting in 2007, seven more ULSs have been established and now serve 10 percent of the Portuguese population. Singapore: Regional Health Singapore reorganized its health care system by developing six regional health Teo (2015) Systems systems (RHSs), which aim to provide horizontally and vertically integrated health care ecosystems. Each RHS developed interventions to provide integrated coordinated care. Examples of these interventions include Aged Care Transition, Aging in Place, Post-Acute Care at Home, Community Health Assist Schemes, Family Medicine Centers, and Integrated Care Pathways. Turkey: Health Transformation Turkey’s 2003 national Health Transformation Plan focused on the establishment of Sumer (2015) Plan high-quality family medicine centers accountable for individual and population health in every district of the country. Restructuring of hospitals, physician payment, data management, and national health insurance facilitated this transformation. Maryland, United States: This case study describes the patient centered medical home payment model created Murray (2015) CareFirst Patient-Centered by the health insurance company CareFirst of Maryland. Support from the insurance Medical Home company and a new financial incentive structure supported improvement of frontline delivery services across the state, resulting in improved quality and lower utilization of hospital and specialty care services. United States: Program of All- PACE centers across the country provide coordinated, integrated, holistic care for frail, Ali (2015) Inclusive Care for the Elderly nursing-home-eligible patients in their own homes. Funded by capitation payments (PACE) from Medicare and Medicaid, each PACE center cares for around 300 patients. The PACE model originated in California and has now spread to 30 states in the United States. Veterans Health Across the United States, the patient-centered medical home (PCMH) model has been Ali and Li (2015b) Administration, United States: used to integrate and improve primary care. The Veterans Health Administration (VHA) Patient Aligned Care Teams drew on the PCMH model and created the Patient Aligned Care Team (PACT) model to reorganize the way it provides primary care and to be integrated into the rest of the system. VHA primary care is now based entirely on the PACT model, with early evidence of success. Note: PCIC = patient-centered integrated care. 88 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Annex 2C  Characteristics of Care Integration Initiatives Involving Hospitals TABLE 2C.1  Characteristics of six health care integration initiatives at Chinese hospitals Beijing: Renmin Shanghai: Ruijin- Beijing: Chaoyang Hospital Integrated Luwan Hospital Zhenjiang: Kangfu Hangzhou: NCD Henan: Xi County Characteristic type Hospital Alliance Care Delivery System Group Network joint centers Integrated Care Basic features Administrative Municipal and Municipal but Municipal and Mainly district Municipal and County, township, level district with significant district district and village national reach Initiation date 2012 2007 2011 2009 2013 2012 Participating Tertiary and Mainly tertiary Tertiary and Tertiary and Tertiary hospitals County hospitals, facilities secondary and secondary secondary secondary and CHCs THCs, and VCs hospitals and hospitals and hospitals and hospitals and CHCs CHCs CHCs CHCs Lead facility Tertiary hospital Tertiary hospital Tertiary hospital Tertiary hospital Tertiary hospital County hospital Lead organization Hospital alliance Lead hospital Hospital Group Hospital Informal County health to oversee “council” Council management municipal bureau and manage council “leading team” coordination Payment system Subsidy to No No Payment to No Yes to support care hospital CHCs for NCD integration physicians to management consult in CHCs Main innovation Formation of a Academic Horizontal Rehabilitation and Joint outpatient Integrated “Hospital Alliance” seminars, integration of recovery wards in centers in CHCs care pathways, to improve care at clinical research lab, imaging, and CHCs; horizontal for hypertension “service CHCs exchanges, and radiotherapy integration of and diabetes agreements,” and technical training services lab, imaging, liaison officers to (teleconferencing) and pathology coordinate care services Elements of integrated care Tracking patient Referrals only No Referrals and Referrals and tests Yes Yes contacts with subset of delivery system diagnostic tests Gatekeeping No No No For NCMS and No No URBMI PHC-based care No No No No For diabetes and Partial: THC coordination hypertension physician across providers management at monitors; VC CHCs provides care Use of integrated No No No No Care plans for Yes care pathways hypertension and/or individual and diabetes care plans Use of metrics No No Limited to Limited to referrals Limited to Compliance to measure care referrals and and diagnostic referrals and with integrated coordination diagnostic tests tests diagnostic rates pathways; referral tracking (Table continued next page) L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 89 TABLE 2C.1  Characteristics of six health care integration initiatives at Chinese hospitals  Continued Beijing: Renmin Shanghai: Ruijin- Beijing: Chaoyang Hospital Integrated Luwan Hospital Zhenjiang: Kangfu Hangzhou: NCD Henan: Xi County Characteristic type Hospital Alliance Care Delivery System Group Network joint centers Integrated Care Hospital-related activities and roles Use of No No No Yes Hospital No multidisciplinary specialists only care teams with participation of hospital professionals Hospital No No No Information Inpatients with Yes involvement in provided to lower diabetes and postdischarge levels hypertension follow-up care Two-way referral Limited Limited Yes Yes Yes Yes system Hospital “green” Yes Very limited Yes, but limited Yes, but limited Yes Yes channela Care shifted out Limited: subset of No Some shifting of Limited to No No of hospitals to patients requiring tests from tertiary one hospital appropriate levels rehabilitation to secondary department hospitals Technical Yes Yes Yes Yes Yes Yes support, training, or supervision provided by hospitals to lower levels E-health E-consultations Remote Teleconferencing EHRs accessed E-consultations EHRs accessed for imaging laboratory testing for training and by providers at with hospital countywide; services and imaging; technical support different levels specialists e-consultations teleconferencing with hospital for training and specialists technical support Note: CHC = community health center; EHR = electronic health record; NCD = noncommunicable disease; NCMS = New Cooperative Medical Scheme; PHC = primary health care; THC = township health center; URBMI = Urban Resident Basic Medical Insurance; VC = village center. a. A hospital “green” channel refers to facilitation of upward referrals from affiliated lower-level facilities, with the intent of expediting the care of the patients so referred. Notes (2013); Hofmarcher, Oxley, and Rusticelli (2007); Øvretveit (2011); Shortell and others 1. For example, “Guidance of the General Office (2014); Wenzel and Rohrer (1994); and WHO of the State Council on Promoting Multi-level (2007, 2015a, 2015b). Diagnosis and Treatment System,” State 3. Empanelment is the process by which all Council (2015, No. 70). patients in a given facility or geographic area 2. The strategic service delivery goals for PCIC are assigned to a primary care provider or are developed from the following sources: care team. Barr and others (2003); Berwick, Nolan, 4. “Guidance of the General Office of the State and Whittington (2008); Craig, Eby, and Council on Promoting Multi-level Diagnosis Whittington (2011); Curry and Ham (2010); and Treatment System,” State Council Curtis and Hodin (2009); Ham and Walsh (2015, No. 70). 90 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A 5. In addition, see “Supervision of Comprehensive Bitton, A. 2015. “Taking the Pulse of PCMH Reform of County-Level Public Hospitals,” Transformation Nationwide.” Lecture presented State Council (2015). at World Bank, Washington, DC, May 29. 6. Hospital management groups or councils are Bitton, A., C. Martin, and B. E. Landon. 2010. described further in chapter 5. “A Nationwide Survey of Patient-Centered 7. For example, “green channel” admissions Medical Home Demonstration Projects.” accounted for less than 1 percent of total Journal of General Internal Medicine 25 (6): admissions in the Beijing Renmin Hospital 584–92. system. Braithwaite, J., C. D. Shaw, M. Moldovan, 8. Implementation of improvement initiatives D. Greenfield, R. Hinchcliff, V. Mumford, with feedback loops is further examined in M . B . K r i s ten s en , a nd ot her s . 2012 . chapter 10. “Comparison of Health Service Accreditation 9. Chapter 10 outlines an implementation plan Programs in Low- and M iddle-I ncome that starts with the replication of effective Countries with Those in Higher-Income practices at a provincial level in a targeted C ou nt r ie s: A C ross - S e c t ion a l St udy.” selection of counties or municipalities within International Journal for Quality in Health the province. As learning is integrated into the Care 24 (6): 568–77. process, change will be accelerated as the Cercone, James, and Lisa O’Brien. 2010. reforms spread. “Benchmarking Hospital Performance in Health.” White paper, Sanigest International San Jose, Costa Rica. References Chatterji, Pinka, Nancy R. Burstein, David Ali, Rabia. 2015. “Programs of All-Inclusive Care Kidder, and Alan White. 1998. “Evaluation of for the Elderly (PACE) in the U.S.: Background the Program of All-Inclusive Care for the and Rationale.” Case study commissioned by Elderly (PACE) Demonstration: The Impact of the World Bank, Washington, DC. PACE on Participant Outcomes.” Report sub- Ali, Rabia, and Rong Li. 2015a. “Integrated Care m it ted to t he He a lt h C a re Fi na nci ng Services in Xi County, Henan Province.” Case Administration by Abt Associates, Cambridge, study commissioned by the World Bank, MA. Washington, DC. C ra ig , C at her i ne , D ou g E by, a nd Joh n ———. 2015b. “Patient-Centered Medical Home Whittington. 2011. “Care Coordination (PCMH) Model of U.S. Veterans Health Model: Better Care at Lower Cost for People Administration.” Case study commissioned by with Multiple Health and Social Needs.” IHI the World Bank, Washington, DC. Innovation Series white paper, Institute for Barr, Victoria J., Sylvia Robinson, Brenda Marin- Healthcare Improvement, Cambridge, MA. Link, and Lisa Underhill. 20 03. “T he Curry, Natasha, and Chris Ham. 2010. Clinical Expanded Chronic Care Model: An Integration and Ser vice Integration: The Route to of Concepts and Strategies from Population Improved Outcomes. London: King’s Fund. Health Promotion and the Chronic Care Curtis, Jessica, and Renée Markus Hodin. 2009. Model.” Hospital Quarterly 7 (1): 73–82. “Special Delivery: How Coordinated Care Bates, D. W., and A. Bitton. 2010. “The Future of Programs Can Improve Quality and Save Health Information Technology in the Patient- Costs.” Paper, Community Catalyst, Boston. Centered Medical Home.” Health Affairs DeVries, Andrea, C. H. Li, Gayathri Sridhar, Jill (Millwood) 29 (4): 614–21. Rubin Hummel, Scott Breidbart, and John J. Beauchamp, Jody, Valerie Cheh, Robert Schmitz, Barron. 2012. “Impact of Medical Homes on Peter Kemper, and John Hall. 2008. “The Quality, Health Care Utilization, and Costs.” Effect of the Program of All-Inclusive Care for American Journal of Managed Care 18 (9): the Elderly (PACE) on Quality.” Mathematica 534–44. Policy Research report for the Centers for DHCFP (Division of Health Care Finance and Medicare & Medicaid Services, Princeton, NJ. Policy). 2005. “PACE Evaluation Summary.” Berwick, Donald M., Thomas W. Nolan, and U n p u b l i s h e d m a n u s c r i p t , D H C F P, John Whittington. 2008. “The Triple Aim: Commonwealth of Massachusetts, Boston. Care, Health, and Cost.” Health Affairs 27 (3): Elissen, A . M., I . G. Duimel-Peeters, C . 759–69. Spreeuwenberg, M. Spreeuwenberg, and H. J. L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 91 Vrijhoef. 2012. “Toward Tailored Disease Gilfillan, R. J., J. Tomcavage, and M. B. Management for Type 2 Diabetes.” American Rosenthal. 2010. “Geisinger Medical Home Journal of Managed Care 18: 619–30. Pilot Demonstrates Success.” Managed Care Elissen A. M. J., I. G. P. Duimel-Peeters, 16: 607–14. C. Spreeuwenberg, H. J. M. Vrijhoef, and Greenfield, D., M. Pawsey, R. Hinchcliff, M. E . Nolte. 2015. “The Netherlands.” In Moldovan, and J. Braithwaite. 2012. “The Assessing Chronic Disease Management in Standard of Health Care Accreditation European Health Systems. Country reports,” Standards: A Review of Empirical Research edited by Ellen Nolte, Cécile Knai, and Underpinning Their Development and Impact.” Richard B. Saltman, 99–110. Copenhagen: BMC Health Services Research 12: 329. World Health Organization Regional Office Guanais, Frederico C., and James Macinko. 2009. for Europe (acting as the host organization for, “The Health Effects of Decentralizing Primary and secretariat of, the European Observatory Care in Brazil.” Health Affairs 28 (4): on Health Systems and Policies). 1127–35. Fifield, Judith, Deborah Dauser Forrest, Melanie Ham, Chris, and Nicola Walsh. 2013. “Making Martin-Peele, Joseph A. Burleson, Jeanette Integrated Care Happen at Scale and Pace.” Goyzueta, Marco Fujimoto, and William Briefing paper, Lessons from Experience Gillespie. 2013. “A Randomized, Controlled Series, The King’s Fund, London. Trial of Implementing the Patient-Centered Hebert, Paul L., Chuan-Fen Liu, Edwin S. Wong, Medical Home Model in Solo and Small Susan E. Hernandez, Adam Batten, Sophie Lo, Practices.” Journal of General Internal and Jaclyn M. Lemon. 2014. “Patient-Centered Medicine 28 (6): 770–77. Medical Home Initiative Produced Modest Forde, Ian, Ane Auraasen, and Liliane Moreira. E conomic Results for Veterans Health 2015. “Delivering Person-Centered Integrated Administration, 2010 –12.” Health Affairs Care: Three Case Studies from OECD Health 33 (6): 980–87. Systems.” Case studies commissioned by the Hildebrandt, Helmut, Alexander Pimperl, Oliver World Bank, Organisation for Economic Gröne, Monika Roth, Christian Melle, Timo Co-operation and Development, Paris. Schulte, Martin Wetzel, and Alf Trojan. 2015. Friedberg, M. W., P. S. Hussey, and E . C. “Gesundes Kinzigtal: A Case Study on People- Schneider. 2010. “Primary Care: A Critical Centered/Integrated Health Care in Germany.” Review of the Evidence on Quality and Costs Case study commissioned by the World Bank, of Health Care.” Health Affairs (Millwood) Washington, DC. 29 (5): 766–72. Hofmarcher, Maria M., Howard Oxley, and Friedberg, Mark W., Meredith B. Rosenthal, Elena Rusticelli. 2007. “Improved Health Rachel M. Werner, Kevin G. Volpp, and Eric System Performance through Better Care C. Schneider. 2015. “Effects of a Medical Coordination.” OECD Working Paper No. 3. Home and Shared Savings Intervention on Organisation for Economic Co-operation and Quality and Utilization of Care.” JAM A Development, Paris. Internal Medicine 175 (8): 1362–68. Jian, Weiyan, and Winnie Yip. 2015. “Integrated Friedberg, Mark W., Eric C. Schneider, Meredith B. Care in China: Case Study: PKU-Renmin Rosenthal, Kevin G. Volpp, and Rachel M. Hospital IDS, Beijing Chaoyang Hospital Wer ner. 2 014. “A s s o c i at ion b e t we e n Alliance, Shanghai Ruijin-Luwan Hospital Participation in a Multipayer Medical Home Groups, Zhenjiang Kangfu Hospital Groups.” Intervention and Changes in Quality, Utilization, School of Public Heath, Peking University. and Costs of Care.” Journal of the American Case study commissioned by the World Bank, Medical Association 311 (8): 815–25. Washington, DC. Frølich, A , R . Jacobsen, C . K nai. 2015. Kane, Robert L., Patty Homyak, and Boris “Denmark.” In Assessing Chronic Disease Bershadsky. 2002. “Consumer Reactions to Management in European Health Systems. the Wisconsin Partnership Program and Its Country Reports,” edited by E. Nolte and C. Parent, the Program for All-Inclusive Care of Knai, 17–26. Copenhagen: World Health the Elderly (PACE).” The Gerontologist 42 (3): Organization (acting as host for, and secretar- 314–320. iat of, the European Observatory on Health Kane, Robert L ., Patricia Homyak, Boris Systems and Policies). Bershadsky, and Shannon Flood. 2006a. 92 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A “The Effects of a Variant of the Program for Meng, Qingyue, Zhang Luyu, Zhu Weiming, and All-Inclusive Care of the Elderly on Hospital Ma Huifen. 2015. “People-Centered Health Utilization and Outcomes.” Journal of the Care: A Case Study from Huangzhong County, American Geriatrics Society 54: 276–83. Qinghai Province.” China Center for Health ———. 2006b. “Variations on a Theme Called Development Studies, Peking University. Case PACE.” Journal of Gerontology 61a (7): study commissioned by the World Bank, 689–93. Washington, DC. Klazinga, Niek. 2014. “Update on OECD’s Meng, Qingyue, Jin Yinzi, Yue Dahai, and He Li. Health Care Quality Indicator Project: 2015. “People-Centered Health Care: A Case Measurement, Information Systems, Policies Study from Feixi County, Anhui Province.” and Governance in OEC D Countries.” China Center for Health Development Studies, Presentation to the European Commission’s Peking University. Case study commissioned Patient Safety and Quality of Care Working by the World Bank, Washington, DC. Group, Brussels, March 13. Meret-Hanke, Louise. 2011. “Effects of the Li, Weiping and Mengxi Jiang. 2015. “China’s Program of All-Inclusive Care for the Elderly Public Hospital Governance Reform: Case on Hospital Use.” The Gerontologist 51 (6): Study of Zhenjiang.” China National Health 774–85. Development Research Center, Beijing. Case Mukamel, Dana B., Alina Bajorska, and Helena study commissioned by the World Bank, Temkin-Greener. 2002. “Health Care Services Washington, DC. Utilization at the End of Life in a Managed Love, Tom. 2015. “People-Centered Health Care Program Integrating Acute and Long- Care in Canterbury, New Zealand.” Sapere Term Care.” Medical Care 40 (12): 1136–48. Research Group. Case study commissioned by Mukamel, Dana B., Derick R. Peterson, Helena the World Bank, Washington, DC. Temkin-Greener, Rachel Delavan, Diane Ma, Jin. 2015a. “Primary Care–Centered Care in Gross, Stephen J. Kunitz, and T. Franklin Two Districts in Shanghai.” Shanghai Jiao Williams. 2007. “Program Characteristics and Tong University. Case study commissioned by Enrollees’ Outcomes in the Program of All- the World Bank, Washington, DC. Inclusive Care for the Elderly (PACE).” The ———. 2015b. “C h i na’s P ubl ic Hospit a l Milbank Quarterly 85 (3): 499–531. Governance Reform: Case Study of Shanghai.” Mukamel, Dana B., Helena Temkin-Greener, and Shanghai Jiao Tong University. Case study Marleen L . Clark. 1998. “Stabilit y of c o m m i s s i o n e d b y t h e Wo r l d B a n k , Disability among PACE Enrollees: Financial Washington, DC. and Programmatic Implications.” Health Care Macinko, J., V. B. D. Oliveira, M. A. Turci, F. C. Financing Review 19 (3). Guanais, P. D. F. Bonolo, and M. F. L. Costa. Mukamel, Dana B., Helena Temkin-Greener, 2011. “The Influence of Primary Care and Rachel Delavan, Derick R. Peterson, Diane Hospital Supply on A mbulator y Care – Gross, Stephen Kunitz, and T. Franklin Sensitive Hospitalizations among Adults in Williams. 2006. “Team Performance and Risk- Brazil, 1999–2007.” American Journal of Adjusted Health Outcomes in the Program of Public Health 101 (10): 1963–70. All-Inclusive Care for the Elderly (PACE).” The Macinko, J., B. Starfield, and T. Erinosho. 2009. Gerontologist 46 (2): 227–37. “The Impact of Primary Health Care on Murray, Robert. 2015. “The CareFirst Blue Cross– Population Health in Low- and Middle-Income Blue Shield Patient-Centered Medical Home Countries.” Journal of Ambulatory Care Model, Maryland, U.S.A.” Case study commis- Management 32 (2):150–71. sioned by the World Bank, Washington, DC. Mancuso, David, Greg Yamashiro, and Barbara NCQA (National Com mit tee for Qualit y Felver. 2005. “PACE: An Evaluation. 2005.” Assurance). 2014. “PCMH 2014 Content and R e p o r t N o. 8 . 2 6 , Wa s h i n g to n S t at e Scoring Summary.” NCQA, Washington, DC. Department of Social and Health Services Nelson, Karin M., Christian Helfrich, Haili Sun, Research and Data Analysis Division, Olympia. Paul L. Hebert, Chuan-Fen Liu, Emily Dolan, Mate, K. S., A. L. Rooney, A. Supachutikul, and and Leslie Taylor. 2014. “Implementation of G. Gyani. 2014. “Accreditation as a Path to the Patient-Centered Medical Home in the Achieving Universal Quality Health Coverage.” Veteran’s Health Administration: Associations Globalization and Health 10: 68. with Patient Satisfaction, Quality of Care, L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 93 Staff Burnout, and Hospital and Emergency Admissions.” Annals of Family Medicine Department Use.” JAMA Internal Medicine 11 (1): S1–S26. 174 (8): 1350–58. Rosenthal, Meredith B., Mark W. Friedberg, Sara Nolte, Ellen, Anne Frølich, Helmut Hildebrandt, J. Singer, Diana Eastman, Zhonghe Li, and Alexander Pimperl, and Hubertus J. Vrijhoef. E ric C . S ch neider. 2013. “Ef fec t of a 2015. “Integrating Care: A Synthesis of Multipayer Patient-Centered Medical Home Experiences in Three European Countries.” on Health Care Utilization and Quality: The European Observatory. Case studies commis- Rhode Island Chronic Care Sustainability sioned by the World Bank, Washington, DC. Initiative Pilot Program.” JAMA Internal Nolte, Ellen, and Emma Pitchforth. 2014. Medicine 173 (20): 1907–13. “What Is the Evidence on the Economic Runz-Jørgensen, Sidsel Marie, and Anne Frølich. Impacts of Integrated Care?” Policy Summary 2015. “SIKS—The Integrated Effort for People 11, World Health Organization Regional Living with Chronic Diseases: A Case Study on Office for Europe, Copenhagen (acting as the People Centered/Integrated Health Care in host organization for, and secretariat of, the Denmark.” The Research Unit for Chronic European Observatory on Health Systems Conditions, Bispebjerg Hospital, Denmark. and Policies). Case study commissioned by the World Bank, O’Brien, John Michael. 2014. “PCMH, TCCI & Washington, DC. the CareFirst Model: ‘Focus on Medication Salmon, J. W., J. Heavens, C. Lombard, and Paula Management.’” PowerPoint presentation to Tavrow, 2003. “The Impact of Accreditation on event, “A Prescription for Savings: Medication the Quality of Hospital Care: KwaZulu-Natal Management and Improved Adherence,” Province, Republic of South Africa.” Operations N at io n a l C o a l it io n o n H e a lt h C a r e , Research Results Series report, published for Washington, DC, March 6. the U.S. Agency for International Development Ø v r e t v e i t , J o h n . 2 011. D o e s C l i n i c a l (USAID) by the Quality Assurance Project, Coordination Improve Quality and Save University Research Co., Bethesda, MD. Money? Volume 1: A Summary Review of the Schulte, T., A. Pimperl, A. Fischer, B. Dittmann, Evidence. London: The Health Foundation. P. Wendel, and H . H ildebrandt. 2014. Porignon, Denis, Rey naldo Holder, Olga “Ergebnisqualität Gesundes Kinzigtal— M aslovska ia , Teph a ny G r i f f it h , Av r i l ­ quantifiziert durch Mortalitätskennzahlen. Ogrodnick, and Wim Van Lerberghe. 2011. Eine quasi-experimentelle Kohortenstudie: “The Role of Hospitals within the Framework Propensity-Score-Matching von of t he Renewed Pri ma r y Healt h Ca re Eingeschriebenen vs. Nicht-Eingeschriebenen Strategy.” World Hospital Health Service des Integrierten Versorgungsmodells auf Basis 47 (3): 6–9. von Sekundärdaten der Kinzigtal-Population.” Porter, Michael E. 2010. “What Is Value in Health Data Analytics study for Gesundes Health Care?” New England Journal of Kinzigtal GmbH, Hausach, Germany. Medicine 363: 2477–81. Shortell, Stephen, Rachael Addicott, Nicola R AND. 2012. “National Evaluation of the Walsh, and Chris Haim. 2014. “Accountable Department of Health’s Integrated Care Pilots: Care Organisations in the United States and Appendices.” Technical report, R A N D England: Testing, Evaluation, and Learning Corporation, Santa Monica, CA. What Works.” Briefing paper, The King’s Reid, Robert J., Katie Coleman, Eric A. Johnson, Fund, London. Paul A. Fishman, Clarissa Hsu, Michael P. Smits, H., A. Supachutikul, and K. S. Mate. 2014. Soman, Claire E. Trescott, Michael Erikson, “Hospital Accreditation: Lessons from Low- and Eric B. Larson. 2010. “The Group Health and Middle-Income Countries.” Globalization Medical Home at Year Two: Cost Savings, and Health 10: 65. Higher Patient Satisfaction, and Less Burnout State Council. 2012. “The Twelfth Five-Year Plan for Providers.” Health Affairs 29 (5): 835–43. for Health Sector Development.” Guo Fa 2012, Reid, Robert J., Eric A. Johnson, Clarissa Hsu, No. 57. http://www.wpro.who.int​ /­h ealth​ Kelly Ehrlich, Katie Coleman, Claire Trescott, _­services/china_nationalhealthplan​.pdf. and Michael Erikson. 2013. “Spreading ———. 2015. “Evaluation Report on Urban a Medical Home Redesig n: Effects on Public Hospital Reform Pilots.” State Council Emergency Department Use and Hospital of the People’s Republic of China, Beijing. 94 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Stout, S., C. Klucznik, A. Chevalier, R. Wheeler, Wagstaff, A., M. Lindelow, S. Wang, and S. J. Azzara, L. Gray, D. Scannell, L. Sweeney, Zhang. 2009. Reforming China’s Rural Health M. Saginario, and I. Lopes. 2015. “Cambridge System . Directions in Development Series. Health Alliance Model of Team-Based Care: Washington, DC: World Bank. I mple m e nt at ion G u id e a nd To ol k it .” Wang, Qiuyan Cindy, Ravi Chawla, Christine M. Cambridge Health Alliance, Cambridge MA. Colombo, Richard L. Snyder, and Somesh Struijs, J., J. de Jong-van Til, L. Lemmens, H. W. Nigam. 2014. “Patient-Centered Medical Drewes, S. de Bruin, and C. Baan. 2012. Home Impact on Health Plan Members “Three Years Bundled Payment System for with Diabetes.” Journal of Public Health Diabetes Care: Effects on Care Process and Management and Practice 20 (5): E12–E20. Quality of Care.” National Institute for Public Weaver, Frances M., Elaine C. Hickey, Susan L. Health and the Environment, Bilthoven, Hughes, Vicky Parker, Dawn Fortunato, Julia Netherlands. Rose, Steven Cohen, Laurence Robbins, Struijs, J. N., S. M. Mohnen, C. C. M. Molema, J. Willie Orr, Beverly Priefer, Darryl Wieland, T. de Jong-van Til, and C. Baan. 2012. “Effects and Judith Baskins. 2008. “Providing All- of Bundled Payment on Curative Health Care Inclusive Care for Frail Elderly Veterans: Costs in the Netherlands: An Analysis for Evaluation of Three Models of Care.” Journal Diabetes Care and Vascular Risk Management of the American Geriatrics Societ y 56: based on Nationwide Claim Data, 2007–10.” 345–53. National Institute for Public Health and the Wenzel, Richard P., and James E. Rohrer. 1994. Environment, Bilthoven, Netherlands. “The Iron Triangle of Health Care Reform.” Sumer, Safir. 2015. “Case Study for People- Clinical Performance and Quality Health Centered Health Care in Turkey.” Case study Care 2 (1): 7–9. c o m m i s s i o n e d b y t h e Wo r l d B a n k , Werner, Rachel M., Anne Canamucio, Judy A. Washington, DC. Shea, and Gala True. 2014. “The Medical Temkin-Greener, Helena, Alina Bajorska, and Home Transformation in the Veterans Health Dana B. Mukamel. 2008. “Variations in Administration: An Evaluation of Early Service Use in the Program of All-Inclusive Changes in Primary Care Delivery.” Health Care for the Elderly (PACE): Is More Better?” Services Research 49 (4): 1329–47. Journal of Gerontology 63A (7): 731–38. W HO (World Health Organization). 2007. Teo, Hui Sin. 2015. “Singapore’s Experience with People- Centered Health Care: A Policy Care Integration.” Case study commissioned Framework. Geneva: WHO. by the World Bank, Washington, DC. ———. 2008. The World Health Report 2008: Vadstrup, E., A. Frølich, H. Perrild, E. Borg, and Primary Health Care—Now More than Ever. M. Røder. 2011. “Effect of a Group-Based Geneva: WHO. Rehabilitation Programme on Glycaemic ———. 2015a. “WHO Global Strategy on People- Control and Cardiovascular Risk Factors in Centered and Integrated Health Services.” Type 2 Diabetes Patients: The Copenhagen WHO/HIS/SDS/2015.6, WHO, Geneva. Type 2 Diabetes Rehabilitation Project.” Patient ———. 2015b. “People-Centered and Integrated Education and Counseling 84: 185–90. Hea lt h S er v ice s: A n O ver v iew of t he Van Hasselt, M., N. McCall, V. Keyes, S. G. Evidence .” WHO/HIS/SDS/2015.7, WHO, Wensky, and K. W. Smith. 2015. “Total Cost Geneva. of Care Lower among Medicare Fee-for- Wieland, Darryl, Rebecca Boland, Judith Baskins, Service Beneficiaries Receiving Care from and Bruce Kinosian. 2010. “Five-Year Survival Patient-Centered Medical Homes.” Health in a Program of All-Inclusive Care for Elderly Services Research 50 (1): 253–72. Compared with Alternative Institutional and Vrijhoef, J. M. Hubertus, and Guy Schulpen. Home- and Community-Based Care.” Journal 2015. “Care Group Zio: Integrated Health of Gerontology 65 (7): 721–26. Care for People with Type 2 Diabetes: A Case Wieland, Darryl, Vicki L. Lamb, Shae R. Sutton, Study on People Centered/Integrated Health Rebecca Boland, Marleen Clark, Susan C a r e i n t h e M a a s t r i c ht R e g io n , t h e Friedman, Kenneth Brummel-Smith, and Netherlands.” Maastricht University Medical G. Paul Eleazer. 2000. “Hospitalization in the Center. Case study commissioned by the Program of All-Inclusive Care for the Elderly World Bank, Washington, DC. (PACE): Rates, Concomitants, and Predictors.” L E V E R 1: S H A P I N G T I E R E D H E A LT H C A R E D E L I V E RY 95 Journal of the American Geriatrics Society 48: University. Case study commissioned by the 1373–80. World Bank, Washington, DC. World Bank. 2015. World Development Indicators ———. 2015b. “The Integrated Health Care 2015. Washington, DC: World Bank. System Based on Health Care Groups in Yan, Fei. 2015a. “Integrated Health Services Zhenjiang, Jiangsu Province, China.” School Reform between Community Health Service of Public Health, Fudan University. Case study Centers and Hospitals in Hangzhou, Zhejiang c o m m i s s i o n e d b y t h e Wo r l d B a n k , Province.” School of Public Health, Fudan Washington, DC. 3 Lever 2: Improving Quality of Care Introduction uniformly delivered, with the appropriate uti- lization of resources and services.” China’s success in rebalancing service delivery In the context of health systems, the term toward people-centered integrated care (PCIC) “quality” incorporates a range of positive will depend on the health system’s ability to features that contribute to the overall per- produce and deliver high-quality services. This formance of health care systems, a view that means providing clinically safe, effective, and underscores the “systems property” of qual- timely care—at all levels of peoples’ interac- ity rather than simply the duty of a physician, tion with the health system, but especially in department or facility (IOM 2001). Indeed, primary care, which is the patient’s first point evidence-based, high-quality, clinically of contact in an effective PCIC model. Indeed, appropriate care—delivered with high tech- efforts to improve the quality of care (QoC) nical skills—is critical if China is to improve to support PCIC will serve as a key lever to population health, patient experience, and achieve China’s reform aims of better popu- efficiency of health care. lation health, better patient experience, and QoC matters because it is a determi- more efficient health care. nant of health outcomes. Approximately “Quality” in health care—an abstract and 10–30 percent of the reduction in premature complex concept (Dayal and Hort 2015; La mortality over the past decade in Organisa- Forgia and Couttolenc 2008)—was described tion for Economic Co-operation and Devel- in an Institute of Medicine (IOM) report as opment (OECD) countries can be attributed “the degree to which health services for to QoC improvements (Nolte and McKee individuals and populations increase the like- ­ 2011, 2012). Low QoC can lead to medical lihood of desired health outcomes and are errors that harm rather than help patients: consistent with current professional knowl- for example, the IOM documented up to edge” (Kohn, Corrigan, and Donaldson 98,000 deaths per year because of medical 1990). Drawing in part on the IOM report, errors in U.S. hospitals (Kohn, Corrigan, Dlugacz, Restifo, and Greenwood (2004) and Donaldson 1999). Potentially prevent- define QoC as “care that is measurably safe, able hospitalization due to poor primary of the highest standard, evidence-based, care accounted for 1 out of every 10 hospital 97 98 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A stays in 2008 in the United States (Stranges evidence of variations in clinical quality sug- and Stocks 2010). Low-quality care, as indi- gests room for quality improvement in hospi- cated by medical errors and adverse events, tals as well (Xu and others 2015). also drives up health spending. Medical Quality shortcomings have been associ- errors alone cost the United States an esti- ated with low utilization of primary care ser- mated $19.5 billion in 2008 (Andel 2012; vices (Bhattacharyya and others 2011; Zhang Stranges and Stocks 2010). and others 2014) and an increasing number of Better QoC is not only associated with bet- patient-doctor disputes over medical practice, ter patient outcomes and experience, but is resulting in litigations and violence (Hesketh also vital to the efficiency and sustainability and others 2012). A well-documented quality of the delivery system because of the close link problem is the overprescription of unneces- between quality and costs. Research shows sary services and drugs (Li and others 2012; that high-quality care is not necessarily more Yin and others 2015; Yin and others 2013). expensive, but that low-quality care is associ- Patients have expressed dissatisfaction about ated with more hospitalizations, more inten- overprescription as well as poor attitude, lack sive treatments and use of medicine, longer of effort, and short consultation time with hospital stays, and unnecessary readmissions, doctors and nurses (NCHS 2010). resulting in wasted resources and poor out- In China, the National Health and Fam- comes (Baicker and Chandra 2004; ­ Berwick, ily Planning Commission (NHFPC) has Nolan, and Wittington 2008). launched initiatives to improve QoC through Estimates of health care costs stemming promoting clinical pathways, regulating from improper and unnecessary use of medi- market access, and developing clinical stan- cines exceeded $200 billion in 2012 in the dards (NHFPC 2014). Implementation is just United States (IMS Institute for Health Care under way. However, some important quality Informatics 2013). Similarly, the United improvement functions are not yet addressed, Kingdom’s National Health Service was including developing, validating, and man- found to be wasting up to £2.3 billion a dating the use of national standardized qual- year on a range of unnecessary procedures ity measures; managing the monitoring and and processes (Campbell 2014). Preventing evaluation of quality at the facility level; and medical errors could have saved US$3 billion coordinating efforts for quality improvement annually in the Australian health system dur- across various stakeholders. ing 1995–96 (AHMAC 1996). In short, low- In the past decade, most OECD countries quality care can harm patients’ health and have recognized continual quality improve- compromise the efficiency of health systems. ment as a central goal of health sector devel- In China, there is a need for information opment and have implemented systematic on QoC and the implications for spending. reforms to improve QoC. Governments It is safe to assume that the quality-cost increasingly act as stewards of the public and links observed elsewhere also exist in China, payers for health care, leading the changes in though more research would be needed to health care delivery to improve QoC. Draw- confirm this hypothesis. A dearth of reliable ing on their experience combined with rel- data and a weak institutional infrastructure evant experience from China, this chapter is for monitoring and evaluation of quality organized as follows: makes it difficult to assess how deficiencies in quality are affecting patient outcomes or • “Conceptualizing, Assessing, and Improv- driving up health spending (for example, by ing the Quality of Care” describes a con- causing medical errors, adverse events, and ceptual framework for analyzing QoC. unnecessary readmissions to hospitals). There • “Challenges in the Quality of Care in is nonetheless a consensus that quality needs China” examines opportunities and chal- improvement, especially at primary health lenges for quality improvement in the facilities (Yip and Hsiao 2015). Emerging ­Chinese context. L e v er 2 : I mpro v ing Q u ality of C are 99 • “International Experience in Improving • I mproved individual experience of Health System Features for High Quality” care; improved health of population; presents the experience of OECD coun- and reduced per capita costs of care for tries in establishing the support infrastruc- ­ p opulation (Institute of Health Care ture for quality improvement. Improvement, United States) • “Recommendations to China for Improv- • The degree of care given to the patient, ing the Quality of Care” proposes a set of w it h ex ist i ng med ic a l tech nolog y, core actions and implementation strategies resources, and capacity, and following China may wish to adopt for scaling up professional ethics and clinical standards systematic quality improvement efforts. (National Health and Family Planning Commission, China) (NHFPC 2014). Conceptualizing, Different stakeholders are likely to empha- Assessing, and Improving size different aspects of quality. Medical pro- the Quality of Care fessionals tend to emphasize the consistency This section first examines what QoC of practice with current professional stan- means through a review of definitions and dards and evidence (IOM 2001). Patients concepts used internationally. It then pres- value certain features such as infrastructure, ents frameworks for assessing quality chal- provider communication, and waiting time lenges and for improving quality in the (de Silva and Bamber 2014). Policy makers broader policy and institutional environ- and health managers tend to focus on strik- ment. These frameworks are applied to the ing a balance between quality, cost, and Chinese and OECD contexts in subsequent equity. Moreover, quality is often a moving sections. target, because new medical knowledge and technology tend to alter understanding of what is considered high-quality (or higher- Conceptualizing the Quality of Care quality) care, requiring a constant revisiting Defining QoC, clarifying the goals of qual- and updating of processes, standards, and ity improvement, and measuring progress metrics. are not straightforward tasks. Most coun- Quality can be considered to have two tries define it in terms of desirable features dimensions: technical and personal. This of care processes and outcomes that com- chapter centers on the technical dimension bine technical, patient experience, and of quality, which refers to the correctness of affordability dimensions, as in the following diagnosis, the appropriateness of prescribed examples: interventions based on best evidence, and the competency of the clinical team in delivering • Safe, effective, efficient, timely, patient- those interventions, resulting in an increased centered, equitable (Institute of Medicine, likelihood of an improved health outcome at United States) an affordable cost. • Clinically effective, safe, and good patient The personal dimension of quality, experience (National Health Service, including patient satisfaction, is discussed United Kingdom) in chapter 4. Briefly, it refers to the respon- • Doing the right thing for the right patient, siveness of care to patients’ preferences: the at the right time, in the right way to ability to see a preferred clinician, continu- achieve the best possible results (Agency ity of care, good communication, demon- for Health Care Research and Quality, stration of empathy, and respect for p ­ rivacy U.S. Department of Health and Human contribute to perceived higher QoC. Ensur- Services; and a similar definition by the ing the highest standard of quality means National Committee for Quality Assur- that all patients receive the right care, at the ance, United States) right time, in the right ­setting, every time. 100 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Assessing the Quality of Care recommended preventive care, acute care, care for chronic conditions, screenings, or The most salient definition of quality care, by follow-up care; these findings indicated a Donabedian (2005), provides a useful frame- massive underuse of effective interventions, work of structure, process, and outcomes for especially cost-effective preventive care critically examining problems related to the (McGlynn and others 2003). In low- and QoC in China—all three of which provide middle-income countries, though evidence valuable information for measuring quality: needs to be strengthened, inappropriate overuse of antibiotics has been widely docu- • Structural quality evaluates the relatively mented (Laxminarayan and Heymann 2012). stable characteristics of the environ- Low-quality care processes are often asso- ment in which care takes place, including ciated with poor patient outcomes. A retro- infrastructure, equipment, and human spective review of medical records associated resources. with hospital admissions in eight countries • Process quality assesses interactions (the Arab Republic of Egypt, Jordan, Kenya, between clinicians and patients, and Morocco, South Africa, Sudan, Tunisia, and whether the clinician follows recom- the Republic of Yemen) documented an aver- mended care or clinical guidelines to age adverse event rate of 8.2 percent (rang- reach a correct diagnosis and an appropri- ing from 2.5 percent to 18.4 percent). Of the ate treatment plan, and skillfully delivers adverse events, 83 percent were judged to be treatments. preventable and 30 percent were associated • Outcomes offer evidence about changes in with death of the patient. One-third were patients’ health status as a result of health caused by therapeutic errors in relatively non- care. complex clinical situations (Wilson and oth- ers 2012). A systematic review of published Failure to adhere to best evidence or pro- studies on health care–associated infections fessional standards leads to three types of in low- and middle-income countries also problems in process quality: overuse, unde- found much higher rates than in high-income ruse, and misuse (for example, of lab tests, countries, particularly of nosocomial infec- procedures, or prescription of drugs). A tions in adult intensive care units; surgical health intervention is considered appropriate site infections; and methicillin resistance and worthwhile if its expected health benefits (Allegranzi and others 2011). exceed its expected health risks by a wide Systematic information collection to mea- enough margin (Brook 1995). Overuse of sure processes and outcomes in a health ser- services occurs when an intervention is given vice plays an important role in generating without medical justification—for example, insights into quality shortfalls, contributing by using antibiotics to treat patients with to the development of actionable improve- viral colds or to treat children with simple ment plans and monitoring arrangements. infections. Underuse of services occurs when The World Health Organization and oth- recommended or necessary care, including ers have done considerable work to develop effective preventive care, is not delivered to quality indicators for use at a national level.1 patients. Misuse can consist of incorrect diag- Because of the complexity and uncertainty nosis or inappropriate interventions delivered of health care, these measures may not be to patients that may harm their health. In perfect; nevertheless, once endorsed and turn, a poor process of care results in poor adopted by health care stakeholders, they patient outcomes. serve the purposes of signaling priority areas Inappropriate or harmful care occurs for quality improvement and respective across the globe. For example, a landmark quality standards and of holding providers RAND study found that adults in the United accountable to the agreed-upon performance States received only about 55 percent of the framework. L e v er 2 : I mpro v ing Q u ality of C are 101 Improving the Quality of Care: The Need unified vision for QoC and confirmation for System Support of QoC goals. • It provides technical leadership in harmo- International experience suggests that nizing and disseminating quality measures addressing any quality gap will require strong and standards, identifying QoC deficien- stewardship by government, usually in part- cies, and developing quality improvement nership with nongovernmental stakeholders, strategies. to develop national support for sustained • It offers high-level oversight of the imple- quality improvement. Improving QoC is not mentation of quality improvement efforts merely the responsibility of one doctor or one and allows monitoring of their progress. health facility. The capacity of any country to measure and raise QoC is critical to system- Monitoring and evaluation (M&E) is a wide improvement of health care delivery and critical component of the architecture of a patient outcomes. health care system, linking health policy and Health workers and facilities acting alone health care practice and providing the essen- are often ill prepared or lack incentives to tial information to guide quality improvement take on such complex tasks. They require and enforce accountability. Diligent M&E of support from within their organizations, as quality improvement is indispensable because well as from the broader health system, to investments in quality improvement at the foster a culture of quality improvement. system or facility level can only be justified Appropriate policies, regulations, incentives, in terms of the positive changes they achieve and monitoring systems are needed to guide in support of predefined goals. If an improve- continuous quality improvement. Of equal ment falls short of expectations, evaluation importance, institutions are needed to mea- of the existing strategy can be used to shape sure and monitor quality, provide guidance modifications early on, thus maximizing the to health care organizations while strength- likelihood of success. Recognition of prog- ening their capacity, and support systematic ress and celebration of achievements also help research and evaluation of clinical practices. to maintain the motivation and commitment System support and institutional leader- of stakeholders in the change process. Public ship are key to creating a high-level vision for reporting, based on continuous monitoring of quality improvement and a conducive policy indicators, benchmarking, and comparison, environment. In many countries, the lack of has become an increasingly common mecha- an overall vision of QoC has led to wide vari- nism to complement monitoring systems. ations in quality across health care facilities. Leaving quality improvement to the discre- tion of individual facilities will not produce QoC Challenges in China higher QoC at scale; instead, it may favor China faces increasing public pressure to facilities that are well endowed with physi- raise quality. Large disparities exist between cal, financial, and knowledge resources and the country’s elite tertiary hospitals and the thereby exacerbate inequities. vast majority of lower-level facilities in physi- Strong, unified leadership from a national- cal and human resources, quality culture, level professional institution is critical. Insti- and clinical practices. The government aims tutional leadership in quality assurance and to raise the bar of quality for all providers improvement serves at least three essential and has invested in expanding and upgrading functions (WHO 2006): the health care infrastructure, particularly at the grassroots level, but it has only recently • It leads analysis of quality in health service directed attention to improving the processes delivery, including stakeholder involve- and outcomes of care. Following the struc- ment and situational analysis, which ulti- ture-process-outcome framework outlined in mately contributes to the creation of a the previous section, this section reviews key 102 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A aspects of China’s health care delivery sys- of five pieces of medical equipment valued tem, briefly examines recent policy initiatives at RMB 10,000 or above at THCs, includ- related to quality, and outlines the challenges ing X-ray machines, ultrasound machines, to achieving higher-quality care. biochemical analyzers, electrocardiogram Evidence on China’s QoC is thin. Most machines, ventilators, blood-cell-count people (and stakeholders) rate health care machines, and urine analyzers. Many THCs facilities often using subjective assessments have multifunction life monitors, electric suc- based on prestige, possession of high-­ tion devices, electric lavage machines, and technology equipment, or the presence of anesthesia machines (Wang and Xue 2011). distinguished senior specialists. As a result, In addition, the 2009 health sector reform evidence on the QoC in most facilities is lack- (as discussed in chapter 1) prioritized the ing, and efforts to measure and assess quality establishment of a network of urban CHCs are few. that are geographically accessible to any The scope of QoC research also varies. resident within a 15- to 20-minute walk. Most assessments of quality are descriptive These measures are laudable, as they aim to studies of a single hospital or a handful of improve timely access to health services. tertiary hospitals (for example, Nie, Wei, and QoC is also a product of the knowledge, Cui 2014; Wei and others 2010). Studies of the attitudes, and practices of health profes- quality of inpatient care at tertiary hospitals sionals. Like many countries, China faces are more common than studies of secondary shortages of well-trained health profession- ­ hospitals and primary care facilities. Given als, especially in grassroots facilities and the conflicting definitions of quality, there rural areas. Among all registered (assistant) is no agreed-upon framework for undertak- physicians, only 45 percent have bachelor’s ing a comprehensive quality assessment. The degrees or higher; most work at large hospi- wide range of quality measures and analyti- tals. At the grassroots level, about 75 percent cal methods in use makes the results hard to of the doctors at THCs and CHCs have voca- compare. Prominent themes of QoC-related tional high school or two-year college medi- studies are the prevalence of inappropriate cal education, and only 10–15 percent have care and the relation between quality and effi- a bachelor’s degree. Most village doctors are ciency (for example, Li and others 2012; Yin former barefoot doctors who only have a and others 2015; Yin and others 2013). vocational high-school medical education. 2 Continuing medical education (CME) is lim- ited for doctors in grassroots facilities: THC Structure of the Delivery System and CHC doctors register 15 days of CME Most QoC studies in China focus on struc- training each year. In addition, most nurses tural aspects of the delivery system. After in China are not college educated. the massive expansion of facilities over the past two decades, China is well endowed Process of Care with modern infrastructure and state-of-the- art equipment even at the lower levels of its Restricted by shortcomings in the avail- health care system. able data, few studies in China have directly The government has done a remarkable examined the mechanisms and procedures job in regulating building standards (for used to deliver health care. But existing evi- example, setting minimum requirements for dence suggests shortfalls in care processes. total facility areas and floor plans) and estab- Overuse of prescription drugs and health lishing an inventory of essential equipment services—particularly high-technology, high- for grassroots facilities such as village clinics, cost interventions—prevails in most facilities. township health centers (THCs), and com- Li and others (2012) studied 230,900 outpa- munity health centers (CHCs). In Guizhou, tient prescriptions written between 2007 and for example, facility surveys show an average 2009 in 28 Chinese cities and found that half L e v er 2 : I mpro v ing Q u ality of C are 103 of the prescriptions were for antibiotics and insufficient among grassroots-level doctors that 10 percent were for two or more types (Liu, Hou, and Zhou 2013; Wu and others of antibiotics. 2009; Xu 2010). Wu and others (2009) tested Overprescription of antibiotics is particu- 651 village doctors’ knowledge of hyperten- larly problematic in lower-level facilities and sion treatment. They found that less than poorer regions, because these facilities rely 40 percent of village doctors performed the more heavily on drug revenue. According to a correct procedure to take a patient’s blood senior health official, the overuse of CT scan- pressure and that only 7 percent regularly ning, cesarean section, coronary artery stent measured patients’ blood pressure in their implantation, and coronary artery bypass work. Less than 50 percent knew the recom- graft are conspicuous problems in China’s mended treatment for hypertension or that clinical practice (Liao 2015) (table 3.1). hypertension is the single most important Underuse of effective care is another con- risk factor for cardiovascular diseases. cern in China. Underuse of effective pro- Evidence on appropriate processes of care phylaxis was found to have contributed to is mixed for secondary and tertiary hospitals; the occurrence of preventable surgical-site for example, Wei and others (2010) found a infections (Fan and others 2014). In addition, high uptake of secondary prevention of ische­ cost-effective behavioral counseling such as mic stroke by doctors in a nationwide sample smoking-cessation counseling may be unde- of urban hospitals. However, Qian and oth- rused; even among physicians, the smoking- ers (2001) showed that obstetric care did cessation rate was found to be low, and the not follow best practice in four hospitals in provision of advice to patients on smoking Shanghai and Jiangsu, with three out of six cessation was not common (Abdullah and procedures that should be avoided being rou- others 2013). Other researchers found unde- tinely performed more than 70 percent of the ruse of early beta-blocker therapy among time. Similar results were found regarding patients with acute myocardial infarction medication for patients with acute coronary who could benefit from it as well as poten- syndromes (Bi and others 2009). tial overuse among patients who might be harmed by it (Zhang and others 2015). Both Outcomes overuse and underuse represent failures to follow best practice. As in many countries, limited access to medi- Chronic diseases have become the major cal records, uneven record keeping, and unre- disease burden in China, and studies have liable clinical service data present challenges found that knowledge and experience in to outcomes research. The available evidence managing common chronic diseases are in China suggests large variations in patient TABLE 3.1  Overuse of prescription drugs and other health interventions in China Type of health intervention Findings Overprescription of drugs Average number of drugs per prescription (three) exceeds WHO rational drug use reference level (Yin and others 2015); 50 percent of prescriptions were for antibiotics, and 10–25 percent were for two or more types of antibiotics (Li and others 2012; Yin and others 2013). Overuse of intravenous Intravenous injection rate (53 percent) exceeds WHO rational drug use reference level injection drugs (Yin and others 2015). Overuse of surgical Cesarean section rate is 46 percent of all deliveries; 50 percent of cesarean sections were procedures unnecessary (Liao 2015). Overuse of CT True positive rate of CT scans is only 10 percent, compared with global average of scans 50 percent (Liao 2015). Note: WHO = World Health Organization. 104 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A outcomes in tertiary public hospitals (Xu and to the NHFPC) created a new Department others 2015). Reduction in hospital mortality of Medical Service Management that year. has lagged over time (Li and Wang 2015), and The department’s mandate is to develop and preventable adverse events remain common. implement strategies for improving pub- A meta-analysis showed that the surgical-site lic hospital management and performance, infection rate averaged 4.5 percent between including quality assurance and improve- 2001 and 2012 (Fan and others 2014), com- ment. Under the department’s leadership, pared with 1.9 percent in the United States quality standards have been revisited and between 2006 and 2008 (CDC 2018). specific quality issues have been addressed by updating standards and guidelines, pro- moting investigations, and strengthening China’s Health Care Regulatory regulation.3 The “Guidelines on Antimicro- and Policy Environment bial Drug Use” have been updated several times, with the latest version published in The health care sector requires a strong regula- 2015.4 The department has also drafted a set tory regime, in part because of the complexity of condition-specific quality control guide- of health services and the information asym- lines, measures, and clinical pathways5 to be metry between providers and consumers. piloted in some public hospitals, with sub- Some health services can be considered “cre- stantial inputs from other countries’ experi- dence goods,” whose utility and impact the ence. These efforts have contributed to a consumer finds difficult or impossible to ascer- more comprehensive, up-to-date set of qual- tain even after the services have been com- ity-related regulations. pleted. The government can play an important In addition, the Ministry of Health estab- role in safeguarding the QoC by developing lished national and local committees for med- quality standards and mandating providers’ ical quality control (MQCCs) to be respon- compliance. In addition, government can also sible for developing standards and enforcing provide political and financial incentives to quality control within medical specialties at promote continuous quality improvement. the facility level. Local MQCCs are situated in tertiary or teaching hospitals that are con- sidered technical leaders—which empowers Institutional and Policy Mechanisms for them to undertake quality evaluation. Zhe­ Quality Improvement jiang province, for example, currently has 47 The government of China has long consid- MQCCs, and its provincial Medical Qual- ered quality assurance a core responsibility of ity Control and Evaluation Office serves as health administration and has been a leader a coordinator of quality improvement work in this regard. Table 3.2 summarizes the main and as an external evaluator of medical qual- regulatory strategies and policy initiatives in ity (ZJOL 2011). The office’s Continuous force. Basic regulatory mechanisms to ensure Quality Improvement Work Plan (2009–12) safety and quality have been in place since the has provided a model for other facilities. late 1970s, specifying minimum equipment Based on the plan, many hospitals have and staffing standards by type of facility; established a quality management committee governing the qualification of health profes- for clinical care, which is typically chaired sionals; and setting out general management by the hospital president or a vice president principles. Between 1989 and 1998, China in charge of quality and consists of medical implemented a first round of hospital accredi- directors from various specialty departments. tation that was intended to categorize and Since 2009, the NHFPC has initiated rank hospitals based on their relative quality. various activities on hospital management, In part because the 2009 reforms posi- with safety and quality assurance as inte- tioned public hospital reform as a priority, gral components. These include the ­ Hospital the then-Ministry of Health (predecessor Management Year initiative (2005– 09), L e v er 2 : I mpro v ing Q u ality of C are 105 TABLE 3.2  Regulations on health care quality in China Regulatory strategy Major policy documents or initiatives Purpose and implementation Regulatory Medical Service Management and Guidance Center •  • MSMGC carries out health technology institutions (MSMGC) established in March 2015 assessment and provide technical National and local medical quality control •  support to local quality improvement committees (MQCCs) established since 2009 efforts. Department of Medical Service Management •  • MQCCs develop standards and (DMSM) within Ministry of Health (MoH) established enforce quality control within medical in 2008 specialties. • DMSM provides high-level political authority and leadership in public hospital management and quality regulation. General Clinical Quality Management Regulations •  • Traditionally, the MoH only develops regulations (opinion-seeking draft) issued in 2014 and publishes general hospital Work plan for comprehensively improving the •  management regulations that include a capacity of county hospitals developed in 2014 quality component. Accreditation Standards for Tertiary Hospitals •  established in 2011; Accreditation Standards for Secondary Hospitals established in 2012 General Hospital Evaluation Standards revised in •  2009 Guiding Principles on Clinical Pathway •  Management issued in 2009 Clinical practice Chinese Medical Association has developed clinical •  Forty-seven such specialty guidelines •  guidelines and diagnosis and treatment guidelines and clinical have been developed and distributed. standards technology and operations standards by specialty However, there is limited monitoring •  since 2006. of clinical practice against these Provincial and local medical associations •  guidelines. Only certain secondary and established various local clinical guidelines, tertiary hospitals are required to report specifications, expert opinions, expert consensus, clinical data. guiding opinions and other advice on specific Guidelines cannot be used as a legal •  clinical issues. basis in malpractice litigation. NHFPC publishes guidelines on specific issues: for •  example, “Nosocomial Infection Management” (2011); “Guidelines on Clinical Use of Antibiotics” (2015); and “Evaluation System of Single-Disease Quality Management” (2008). Health facility Two rounds of hospital accreditation (2011–12 and •  Although new accreditation standards •  evaluation 1989–98) were developed, the NHFPC withdrew and quality Further Improving Medical Services Initiative •  accreditation for more than 240 improvement (2015–17) newly accredited tertiary hospitals initiatives The 10,000 Miles Medical Quality Inspection Tour •  between 2011 and 2012. Fraud and annually since 2009 corruption were suspected during the •  The Hospital Management Year initiative (2005–09) accreditation process.a “100 Best Hospitals” recognition •  On-site survey is the most commonly •  used method for hospital evaluation. Source: National Health and Family Planning Commission (NHFPC) website (accessed June 27, 2018), http://en.nhfpc.gov.cn/. a. Dajiang Net 2012. the 10,000 Miles Medical Quality Inspec- initiative (2015–17), which aims to improve tion Tour (an annual national tour to per- convenience of access and patients’ service form on-site surveys of hospitals), and rec- experience (for example, by enabling online ognition of 100 Best Hospitals. The Further medical appointments and weekend outpa- Improving Quality of Health Care Services tient consultations, reducing waiting times, 106 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A and providing access to medical records) has MQCCs are conceived of as technical work- been implemented since 2015 with some suc- ing groups rather than as permanent agencies cess. Two recent policy documents on urban with implementation capacity or administra- public hospital reforms and county hospital tive authority. The MSMGC’s current focus reforms have reiterated the call for quality is on health technology assessment, and it assurance and improvement.6 appears to give less attention to quality assur- In March 2015, the NHFPC took another ance, improvement, and M&E. Important important step in developing national insti- functions that the MQCCs and the MSMGC tutional infrastructure and leadership in sup- are not set up to perform include harmo- port of quality improvement. It established a nizing and mandating the use of national national Medical Service Management and standardized measures of quality, manag- Guidance Center (MSMGC) with a range ing quality M&E at the facility level, and of mandates, including health technology coordinating efforts for quality improvement assessment and technical support to local across various stakeholders. Moreover, as quality improvement efforts. part of the NHFPC system, the focus is more on public hospitals rather than on providing institutional support to all providers. Within Challenges for Quality the NHFPC, the Department of Medical Ser- Improvement Policy vice Management has the authority and tech- The rapid expansion and upgrading of health nical capacity but insufficient resources to care infrastructure in China has laid the monitor and evaluate quality as a foundation foundation for delivering higher-quality care, for continuous quality improvement. and the NHFPC has made earnest efforts to Third, to better support quality improve- create the institutional and policy architec- ment across the health care system requires ture essential to improve quality. The speed of greater efforts to strengthen information sys- policy change is impressive, but the depth of tems, quality evaluation, and facility-level policy implementation and its impact remain improvement initiatives, as follows: to be seen. In addition, some shortcomings in the existing approach may limit its potential • Information systems, e-health, and for quality improvement, as follows. reporting. Regular reporting of clinical First, QoC regulation and evaluation data has room for improvement in China. emphasize entry qualifications and structural Facility-based information systems are readiness (as indicated, for example, by setting not set up to collect such data. An unde- up internal quality control committees). These termined number of tertiary hospitals tasks need to be broadened to encompass (and, to a lesser extent, secondary hospi- audit, control, and clinical processes and out- tals) are required to report clinical data, comes. For example, the current standards and albeit limited in scope. Even the avail- assessment criteria cover more items related to able data are not regularly analyzed to structure (and give more weight to structural provide insights on quality practices and factors) than to processes and outcomes— outcomes. Researchers must sample often that is, to how services are provided and the difficult-to-access medical records for clin- resulting benefits for patients. Although many ical data, which is a costly process. Data health facilities in less-developed areas may are generally lacking on the QoC provided still require investment in basic physical and by secondary hospitals and primary care human resources, China as a whole needs to facilities. advance its quality agenda with a stronger • Quality evaluation. The government focus on monitoring and improving clinical assesses and enforces regulatory compli- processes and patient outcomes. ance through periodic inspections. On Second, institutional leadership for qual- their own, these inspections are unlikely ity improvement remains elusive. The local to be sufficient to drive continuous L e v er 2 : I mpro v ing Q u ality of C are 107 quality improvement on the front lines. International Experience in The scope of annual on-site surveys is Improving Health System Quality limited in part by the absence of verifi- able data on processes and outcomes. As In the past 15 years, many OECD countries such, the assessments can be subjective, have recognized continuous quality improve- and the results are only shared between ment as a central goal of health sector devel- the evaluated hospital and the health opment and have implemented systematic administration. They are not dissemi- reforms to improve QoC. This new wave of nated to the public or other stakehold- quality improvement is occurring in the con- ers or used to generate financial rewards text of aging populations, an increasing bur- and penalties. den of chronic diseases, and surging expen- • Support for facility-level improvement ini- ditures on health care. Governments in these tiatives. Incentives and accountabilities for countries are increasingly acting as stewards continuous quality improvement can be of public health and payers for health care, improved in China. Frontline providers are and they are driving the changes in health not incentivized to improve quality pro- care delivery to improve QoC. They have cesses and outcomes. Health facilities are evaluated the cost-effectiveness of new health oriented toward increasing the quantity of interventions and technologies, developed services so they can generate revenue and quality measures and proposed new quality enlarge market share, and their income is standards, and popularized modern manage- not directly affected by quality. The gov- ment tools and practices to improve QoC. ernment does not provide funding sup- Their lessons are relevant for China’s fast- port or incentives for quality improvement evolving health system. activities, nor (when evidence is available) As discussed earlier, certain features does it penalize hospitals for having low and institutional arrangements in a coun- quality or poor patient outcomes. The lack try’s health system form critical parts of of an institutional focus on quality means the supporting infrastructure that is needed that few hospitals are willing to invest in for improving quality and sustaining qual- quality improvement initiatives. ity improvement. Box 3.1 summarizes BOX 3.1  Institutional arrangements for quality improvement in France, Germany, and the Netherlands The Haute Autorité de Santé (HAS) is the French In Germany, the Federal Joint Committee is independent public authority that certifies health legally authorized to evaluate all new and existing facilities and practitioners; draws up best-practice technologies for their appropriateness and cost effec- recommendations; and evaluates health products tiveness, and the Institute for Quality and Efficiency and services, professional practice, and the organi- in Health Care provides technical capacity to scien- zation of care and public health (Chevreul and oth- tifically evaluate health services. Their recommenda- ers 2010). The Ministry of Health and HAS together tions can lead to restriction of services under social establish a set of quality indicators, and both public health insurance. and private health facilities are required to publish The Netherlands in 2012 set up a Quality their quality performance yearly. In France, phar- Institute that imposes a mandatory framework for maceuticals are evaluated for effectiveness and for the development of care standards, clinical guide- their added benefits over other pharmaceuticals for lines, and performance measures, including using the same indicators. Low-value pharmaceuticals are evidence-based medicine principles to assess estab- delisted from social health insurance. lished medical science and medical practices. 108 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A institutional arrangements in support of The IOM subsequently published two quality improvement in three European seminal reports that called for an overhaul countries. The experience of Australia, the of U.S. health care, proposed a strategic United Kingdom, and the United States in but action-oriented framework for quality putting in place these building blocks is then improvement, and set a goal of a 50 ­ percent examined in more detail below. The cases reduction in medical errors in five years examine the countries’ overall approaches to (IOM 2001; Kohn, Corrigan, and Donald- quality assurance and improvement, institu- son 1999). The findings and the proposed tional leadership, accountability and control improvement strategy in these reports were mechanisms, and management innovations. well received by policy makers, payers, and the medical professionals. The IOM placed quality improvement on the health care United States: Public and Private agenda; created a vision for quality improve- Collaboration for Quality Improvement ment; and galvanized nationwide campaigns, The United States has been a leader in mod- further research, and quality improvement ern quality improvement in health care. activities. In the 1980s, a double-digit increase in In 1999, the Agency for Healthcare health spending drove a sense of urgency to Research and Quality (AHRQ) was estab- strengthen regulations to prevent abusive use lished within HHS as the U.S. government’s of medicine. In 1989, the U.S. Department lead agency for research in health care qual- of Health and Human Services (HHS) estab- ity. The AHRQ finances and acts as a clear- lished the Agency for Health Care Policy and inghouse for evidence to make health care Research to enhance the quality, appropriate- safer; of higher quality; and more accessible, ness, and effectiveness of health care services. equitable, and affordable. It also works with The HHS used updated clinical guidelines stakeholders to ensure that such evidence and stringent peer review, combined with is understood and used. For example, the financial incentives, but it achieved little buy- AHRQ develops and updates clinical guide- in from health professionals and had little lines for a large and growing number of impact on QoC. ­ conditions. It also has produced quality indi- The IOM has played a critical role in driv- cators for inpatient and outpatient services, ing the quality improvement agenda in the including for preventive measures, inpatient United States. Established in 1970, the IOM quality, patient safety, and pediatric care. is an independent organization of eminent The AHRQ plays a significant role in the professionals from diverse fields including national effort to improve patient safety health and medicine; the natural, social, and through national programs to reduce health behavioral sciences; and beyond who advise care–associated infections, “science of safety” the government and others on medical and training programs, and the production and health policy issues. In 1996, the IOM con- dissemination of patient safety toolkits. Free vened a National Roundtable on Health Care software is provided to any hospital to cal- Quality to review health service delivery and culate these quality measures, which in turn identify QoC-related issues. In a consensus are reported to the Centers for Medicare & report, the Roundtable concluded that seri- ­ Medicaid Services (CMS) according to quality ous and widespread quality problems existed regulations. In addition, the AHRQ publishes throughout the American medical care sys- the annual National Healthcare Q­ uality and tem that harmed many people (Chassin and Disparities Report (AHRQ 2017), which is Galvin 1998). It also determined that the mandated by ­ Congress to measure progress in quality of health care can be precisely defined the national priority areas for improving QoC and measured with a degree of scientific and reducing disparities in the care received accuracy comparable to that of most mea- by different racial and socioeconomic popula- sures used in clinical medicine. tion groups. L e v er 2 : I mpro v ing Q u ality of C are 109 Other government and nongovernmental The NQF is best known for its reporting, organizations that are involved in developing first published in 2002, on serious reportable and disseminating quality measures include events (SREs), or “never” events—defined as the National Quality Forum (NQF), the Joint 28 “preventable, serious, and unambiguously Commission of the National Committee for adverse events that should never occur” in a Quality Assurance, and the American Medi- health care setting.7 The Medicare program cal Association’s Physician Consortium for does not provide payment for the 28 “never” Performance Improvement. The NQF is a pri- events because, literally, they should never vate, nonprofit membership organization (and occur. a federal government contractor) that was Although medical professionals once established to coordinate and endorse quality frowned upon the idea of public reporting— standards and quality measures that are pro- and the evidence is still mixed on how it affects duced by other organizations and agencies and QoC—reporting is now a standard practice in used for public reporting. The NQF has more the United States. The QoC provided by indi- than 400 member organizations, representing vidual physicians, physician groups, health consumers, health plans, medical profession- plans, and hospitals is measured and publi- als, employers, the government, pharmaceuti- cized through various organizations and pro- cal and medical device companies, and other grams.8 The hope is that information about quality improvement organizations. The NQF providers’ quality will help protect consumers’ uses a multistep consensus development pro- rights, help them make better choices among cess to ratify measures before they are applied providers, and hold providers accountable for (as shown, for example, in its Ambulatory the safety and quality of their services. In addi- Care Project, described in box 3.2). Its inde- tion to public reporting, major payers (includ- pendence and its ability to convene working ing the government’s payer agency, CMS) have groups of high-level technical authority have added financial incentives for quality perfor- been instrumental to its success. mance to their payment schemes. BOX 3.2  The Ambulatory Care Project of the National Quality Forum Ambulatory care embraces a wide range of health con- through an open call for measures and were actively ditions, services, and settings and is the primary form of sought by NQF staff through literature reviews, a care that patients receive. The demand for performance search of the Agency for Healthcare Research and measures to evaluate all aspects of ambulatory care, Quality’s National Quality Measures Clearinghouse, including various settings of care, is growing rapidly. NQF member websites, and an environmental scan. Priorities are to ensure the safety, appropriateness, Using standardized evaluation criteria ­ i ncluding and effectiveness of outpatient care, the coordina- scientific acceptability, usability, and feasibility, the tion of care, timely communication, pediatric urgent Ambulatory Care Steering Committee e ­ valuated care, and the quality of clinicians’ performance. 27 measures for appropriateness as voluntary consen- Performance measures address issues including timely sus standards for accountability and public reporting. treatment, antibiotic use, patient admission and dis- The chosen measures focus on the care of the fol- charge, and appropriate documentation by staff. lowing conditions and situations in the ambulatory The National Quality Forum’s (NQF) Ambulatory care setting: heart disease, diabetes, hypertension, Care Project is a multistage endeavor that seeks con- obesity, asthma, prevention, depression, medication sensus on standardized measures of performance in management, patient experience with care, and care outpatient care. ­ Proposals for standards were solicited coordination. Source: NQF 2008. 110 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Private corporations are also exploring United Kingdom: Publicly and Centrally how they can support the achievement of bet- Driven Quality Improvement ter quality and lower cost. For example, in The quality improvement agenda in the 1987 an experiment was launched, known as United Kingdom’s health sector has been the National Demonstration Project on Qual- largely led by the government through the ity Improvement in Health Care (NDP), in Department of Health, which acts as a cen- which 21 health care organizations partici- tralized provider of health sector informa- pated in an eight-month study of the appli- tion, services, and policy initiatives. cability of quality improvement methods Extensive changes in health service deliv- ­ typically used in industrial settings. Twenty- ery and quality regulation have occurred in one companies agreed to support the 21 the past few years. Before 2008, the National health care organizations during their stud- Health Service (NHS) focused on building ies by providing free consulting, materials, system capacity and reducing patients’ wait- access to training courses, and reviews. The ing time to ensure free access to needed health companies participating in this experiment care for all. The NHS report, High Quality included many of America’s leading orga- Care for All: NHS Next Stage Review pro- nizations, such as AT&T, Corning, Ford, posed to move beyond the centrally driven Hewlett-Packard, IBM, and Xerox. performance management regime, with its The results were impressive (Berwick and focus on driving activity and meeting targets, others 1990): Fifteen of the health care orga- to a new quality improvement agenda (DH nizations made significant progress. Other 2008). The report set out recommendations hospitals reduced lengths of stay by nearly to tackle variations in the quality of care, half for some procedures, reduced emergency champion best practice, and enable greater room waiting times by 70 percent, and elimi- competition and choice as a way to incentiv- nated 67 percent of the time it took for physi- ize a culture of quality improvement (Keown cal examinations. One facility reduced infec- and Darzi 2015). It presented a vision of tions after surgery by more than half. a service in which quality improvement is The NDP was extended for three more driven by local clinicians armed with better years and evolved into the Institute for data on the effectiveness of their own work, Healthcare Improvement (IHI), a nonprofit spurred on by financial incentives and by organization that provides leadership in the choices of well-informed patients rather redesigning health care to reduce errors, than by top-down targets. Operationally, the waste, delay, and unsustainable costs. The report endorsed more and better information IHI takes an approach it calls the “science about clinical performance and strengthen- of improvement,” which emphasizes rapid- ing of incentives to improve quality (Maybin cycle testing of innovations in the field, fol- and Thorlby 2008). lowed by rollout to generate learning about Two years later, the Department of Health what changes, in which contexts, produce set out the most significant reorganization what types of improvements.9 plan in NHS history (DH 2010). This removed In sum, the quality improvement efforts in responsibility for citizens’ health from the the United States have been driven by the gov- secretary of state for health, ­ abolished the ernment, nonprofit organizations, and the pri- primary care trusts and strategic health vate sector. Adherence to scientific models for authorities (which had been responsible for measuring performance and improvement has enacting the directives and implementing been crucial for obtaining support from health the fiscal policy dictated by the Department professionals and for replicating success. The of Health at the local and regional levels), government is an important partner by finan- and transferred £60–80 ­ billion of health cially supporting the organizations as well as care funds from the primary care trusts to by backing demonstration projects for quality 211 clinical commissioning groups (CCGs) improvement and knowledge creation. L e v er 2 : I mpro v ing Q u ality of C are 111 that are partly run by general practitioners. audit in the world—measures the effective- The new plan had significant implications ness of diabetes care in primary and second- for the regulation of the quality of care ary facilities against clinical guidelines and because the CCGs are to champion three quality standards in England and Wales, and key principles: putting patients at the center it publishes its report online. The audit col- of the NHS, focusing on health outcomes in lects and analyzes data for use by a range of the measurement and review of QoC, and stakeholders to drive changes and improve- empowering health professionals. ments in the quality of services and health The U.K. government has established the outcomes for people with diabetes. Health and Social Care Information Cen- The well-known National Institute for ter as the national provider of information, Health and Clinical Excellence (NICE) data, and information technology systems is another arm’s-length body of the U.K. for commissioners, analysts, and clinicians Department of Health that develops clinical in health and social care. Its website provides guidelines and care standards based on best a comprehensive guide to the NHS Out- evidence and assesses the efficacy and cost- comes Framework (box 3.3).10 Established effectiveness of health technologies and inter- in 2010, this framework is intended to pro- ventions used in the NHS. NICE provides vide national-level accountability for the out- technical leadership in government decisions comes of NHS-provided health care as well on the allocation of resources for health care, as to drive transparency, quality improve- and it supports evidence-based clinical care. ment, and outcome measurement throughout In the United Kingdom, although efforts the NHS. The NHS Commissioning Board’s to improve the quality of care are largely Outcome Indicator Set uses the same five- ­ government-driven and top-down, the under- part structure as the Outcomes Framework lying goals and strategies are similar to those (NHS Commissioning Board 2012). Patient- used in the United States. Several institutions reported outcome measures have been added and mechanisms in the United Kingdom to the quality review; they focus on measur- have a similar structure to their U.S. coun- ing health gains in patients undergoing hip terparts. They include technical bodies that replacement, knee replacement, and varicose- review evidence and develop clinical guide- vein and groin-hernia surgery. lines, develop comprehensive quality frame- Besides collecting and maintaining data works, hold providers accountable, conduct on health care (such as episode statistics for performance reviews and audits, and align hospital-based and general practitioners’ incentives in support of quality improvement. care), the Health and Social Care Informa- In addition, both countries have supported tion Center is also mandated to conduct pilots in organizational innovation designed ­ clinical audits. For example, the National to facilitate quality improvement at the orga- Diabetes Audit—the largest annual clinical nization level. BOX 3.3  U.K. National Health Service outcomes indicator set, 2013/14 1. Preventing people from dying prematurely 4. Ensuring that people have a positive experience of 2. Enhancing quality of life for people with long- care term conditions 5. Treating and caring for people in a safe environ- 3. Helping people to recover from episodes of ill ment and protecting them from avoidable harm health or following injury Source: NHS Commissioning Board 2012. 112 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A A recent successful example of such a pilot the National Health and Medical Research in the United Kingdom is the centralization Council, and the Clinical Excellence Com- of cardiac and stroke services in London, mission. These institutions develop clinical which has led to reductions in mortality and standards and guidelines, monitor and evalu- length of hospital stay (box 3.4). It empha- ate the performance of health care providers, sizes that service reconfiguration can improve assess the value of health interventions and the quality of care by helping to create a criti- technology, and offer quality-related infor- cal mass of high-quality providers, imposing mation to providers and the public. best-practice standards, and facilitating inte- grated pathways. Australian Commission on Safety and Quality in Health Care The Australian Commission on Safety and Australia: Central and State Government Quality in Health Care was established as Collaboration for Quality Improvement a corporate Commonwealth entity after the Like the United Kingdom, Australia has Parliament of Australia passed the National established a robust institutional infrastruc- Health Reform Act of 2011. The commis- ture to promote, support, and guide quality sion is jointly funded by all state and territo- improvement in health care. This includes rial governments on a cost-sharing basis, and the Commonwealth (federal) Department of its annual program of work is developed in Health (including the Medical Services Advi- consultation with the national, state, and ter- sory Committee), the Australian Commis- ritory health ministers. The commission led sion on Safety and Quality in Health Care, the development of the Australian Charter of BOX 3.4  An innovative U.K. model to improve care for stroke patients Receiving appropriate care is the single most important In the London pilot, change was governed by determinant of outcome for patients who have suffered a top-down approach led by the regional health a stroke. In several countries, acute stroke services authority. The Manchester pilot, by contrast, used are being centralized as a means of improving access a more bottom-up, network-based approach led by to critical acute care, including rapid access to brain local providers and commissioners. In London, 8 of imaging and anticoagulant drugs, and to create fewer the original 32 acute stroke service providers were but higher-volume specialist services. Hospitals of dif- converted to hyperacute stroke units (HASUs), and ferent capabilities work together to create a centralized 24 became local stroke units (SUs) that delivered system of stroke care in which patients are taken to post hyper-acute care; stroke services were with- central specialist units rather than to the nearest hos- drawn from 5 hospitals. In Manchester, 3 HASUs pital. Research in Australia, Canada, Denmark, the and 10 SUs were created, and acute care was not Netherlands, and the United States has shown the cost- entirely withdrawn from any hospitals. Both mod- effectiveness of this approach. els resulted in reduced lengths of hospital stay, while In 2010, stroke and major trauma were chosen as London’s also saw reduced mortality. cases for piloting a new health care delivery model to Moving forward, the establishment of specialist improve patient outcomes, for two reasons: (a) good centers for rare diseases will also be considered to evidence on how to improve the quality of stroke improve the coordination of care for patients. As care, and (b) a clinical community that desired part of the new care model, specialized providers change to improve such care. The new service deliv- will be encouraged to develop networks of services ery model split stroke care into hyperacute, acute, over a wider area, integrating different organiza- transient ischemic attack, and community care. tions and services around patient needs. Sources: Morris and others 2014; Rudd 2011; Turner and others 2016. L e v er 2 : I mpro v ing Q u ality of C are 113 Health Care Rights, the National Safety and immediately using existing information sys- Quality Health Service Standards, and the tems. The Australian Institute of Health and National Safety and Quality Framework. Welfare reports these quality indicators pub- Being a high-level coordinating and facili- licly to (a) provide transparency and to inform tating body, the commission is uniquely placed decision making about overall priorities and to advocate collaboration in patient safety and system-level strategies for safety and qual- health care quality. The commission’s Austra- ity improvement, and (b) shape the quality lian Safety and Quality Framework for Health improvement activities of service providers. Care was endorsed by the Australian health ministers in 2010. The framework provides a Medical Services Advisory Committee basis for preparing strategic and operational The Medical Services Advisory Commit- safety and quality plans, sets out guidance on tee (MSAC) is an independent expert group priority areas, stipulates actions for research under the Commonwealth Department of and clinical improvement in safety and qual- Health and plays a role similar to that of ity, and promotes discussion among stake- the NICE in the United Kingdom. It advises holders about ways to improve partnership the health minister on more appropriate, and collaboration. higher-value health care based on up-to-date Under the framework, safe, high-quality evidence on the comparative safety, clinical care should follow three core principles: effectiveness, and cost-effectiveness of new or existing medical services and technolo- • It should be consumer- or patient-centered gies. It is also responsible for informing the (so that people have timely and easy access health minister on whether medical services to care and providers respond to their covered by Australia’s publicly funded, choices and needs). fee-for-service Medicare Benefits Schedule • It should be driven by information (so that (MBS) are sufficiently safe and cost-effective care decisions are guided by knowledge to warrant public subsidies. In addition, it and evidence). produces technical guidelines on therapeutic • It should be organized for safety (making and diagnostic services and provider-patient safety and quality central to how health interaction. facilities are run). In an intensive assessment of health tech- nology in Australia, the MSAC recommended The framework sets out 21 areas for that the government develop a postmarket action that stakeholders can take to improve surveillance system for the MBS (Australian the safety and quality of care provided in all G overnment 2009). This system involves ­ health care settings. identifying priority technologies for review The commission has also funded the regarding appropriateness, efficacy, and National Indicators Project, which developed cost-­effectiveness. Outcomes could include a set of 55 national indicators of safety and delisting from the MBS, reducing payment ­ quality: 13 indicators apply to primary and fees, limiting the frequency or interval of ser- community health services, 25 to hospitals, vices, or amending a service description or 6 to specialized health services, 5 to residen- technology specifications to better capture the tial care for the elderly, and 11 to all types patient groups most likely to benefit from it. of health services.11 The indicators measure The MSAC also leads other reviews on pri- safety, appropriateness, effectiveness, continu- mary health care and Medicare compliance ity, and responsiveness. Most of them focus rules. on the appropriateness of care, responding to the growing emphasis on evidence-based National Health and Medical Research health care and best-practice guidelines, while Council 25 relate to safety (AIHW 2009). Most of Australia’s National Health and Medi- the indicators (40 out of 55) can be reported cal Research Council helps clinicians, 114 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A researchers, policy makers, and consumers to similarities in the strategies applied: health access clinical guidelines via its online portal technology assessment, dissemination of up- launched in March 2010. Within the coun- to-date clinical guidelines, clinical practice cil, the National Institute of Clinical Studies audit and review, public reporting and dis- seeks to help close gaps between knowledge closure of information on the performance and clinical practice in health care. of specific facilities, financial incentives The institute’s publications include sta- directed at providers, and health informa- tistical analysis of clinical practices as they tion and education. These three countries are currently performed and how they could have created institutional arrangements be improved. Based on these reviews, the within the government or in partnership institute works in partnership with clini- with private organizations and academia to cal groups and health care organizations to fulfill these functions. help improve the application of evidence to practice. It provides access to resources and evidence for health professionals, managers, Recommendations to China for researchers, and policy makers. Improving the Quality of Care The challenges China faces with respect to Clinical Excellence Commission QoC can be successfully addressed by creat- Because the Australian states and territo- ing unified leadership, suitable institutional ries have independent decision powers for architecture, stakeholder participation, health policies, several have developed their and implementation tools to foster continu- own institutions to monitor and improve the ous quality improvement at all levels of the safety and quality of care. An example is service delivery system. China may like to the Clinical Excellence Commission (CEC), consider building a comprehensive strategic which was established in 2004 to lead safety framework consisting of three core action and quality improvement in the New South areas: Wales health system. The CEC’s main activities include coor- 1. Strengthening institutional leadership and dination of systemwide analyses of issues system support through audits and reviews, working col- 2. Establishing quality measurement and laboratively with health sector stakeholders, feedback mechanisms and implementing programs, projects, and 3. Transforming organizational management initiatives to address identified issues. The to cultivate continuous quality improvement CEC’s Clinical Practice Improvement Series provides training for clinicians to improve Two additional core action areas—­ the quality of care delivered and to improve provider skills and patient engagement—are patient outcomes, using an approach simi- addressed in chapters 4 and 6. Table 3.3 dis- lar to the U.S. Institute for Health Care plays the core action areas listed above and Improvement’s “science of improvement” corresponding implementation strategies. model. Core Action Area 1: Organizational Conclusion Structure to Create Information Base and Develop Strategies for Quality As shown in these international exam- Improvement ples, several OECD countries are trying to address the gaps in the quality of their care G overnment leadership and steward- to deliver safer, more effective, and higher- ship are vital for building capacity to value care that meets the public’s changing improve the quality of health care. Inter- demands. The Australian, United Kingdom, national experience points to three cat- and United States cases also show striking a ctivity that the government egories of ­ L e v er 2 : I mpro v ing Q u ality of C are 115 TABLE 3.3  Three core QoC action areas and implementation strategies Core action areas Implementation strategies 1: Organizational structure Explore options to cultivate a national coordination architecture to oversee •  to lead creation of an systematic improvements to health sector quality information base and Conduct an in-depth national study of the quality of care and quality improvement •  development of strategies for initiatives at all levels of the system quality improvement Develop a national strategy for quality improvement •  2: Systematic QoC Shift the measurement of quality from structure to process and outcomes •  measurement and Create and maintain an atlas of variation in process quality and outcomes •  continuous use of resulting Use measures of quality to improve performance •  data to support quality Establish an engagement model to support peer learning and energize collective •  improvements quality improvement 3: Transformation of Promote evidence-based standardized care •  management practice to Embed the “quality culture” in the management philosophy of health care •  improve QoC in health organizations and promote modern managerial techniques facilities Use e-health innovations to support quality improvements •  can consider: (a) expanding the mandate • Define treatments and interventions that of current bodies or setting up additional are reimbursable under social health insur- ones to lead, oversee, and implement qual- ance, based on cost-effectiveness analysis ity improvement initiatives; (b) conducting and ethical considerations national reviews; and (c) developing nation- • Assess and promote clinical guidelines wide approaches for quality enhancement. • Conduct research and build the capacity needed to advance the continual improve- Strategy 1: Cultivate a national ment of quality care. coordination architecture to oversee systematic improvements Stakeholder organizations, including the This architecture would be publicly respon- NHFPC, the Ministry of Finance, the Min- sible for coordinating all efforts aimed istry of Human Resources and Social Secu- at quality assurance and improvement in rity, key professional and scientific bodies, health care and would actively engage all private providers, and the public could be stakeholders to facilitate the implementation represented in this coordination architec- of quality assurance and improvement strat- ture.12 The entity could also serve as the plat- egies for this purpose. It would have nine form for tapping international expertise and key functions: sharing knowledge about care improvement. In the long run, it would serve as an impor- • Ensure that national aims for quality are tant source of scientific information on all set quality-related topics for both clinicians and • Establish quality standards and develop the public. It would become the institutional quality measures leader in promoting QoC and ensuring that • Continuously measure and report on prog- evidence-based care is consistently delivered ress toward those standards to the highest standard. • Develop a standardized national medical As discussed earlier, several OECD coun- curriculum, incorporating the best avail- tries have established such institutions over able scientific knowledge the past 15 years: • Ensure that the medical professions are certified to deliver care in accordance with • United Kingdom: The NICE is responsi- these standards ble for developing evidence-based clinical • Oversee efforts to accredit and certify guidelines and pathways and evaluating both public and private providers clinical interventions.13 116 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A • France: The national authority for health, on quality and performance. These reports HAS, is responsible for the assessment of contribute to the collection of reliable infor- drugs, medical devices, and procedures; mation on performance and the analysis of the publication of guidelines; the accredita- problematic areas. They also help focus the tion of health care organizations; and the attention of leaders and professionals on certification of doctors (Chevreul and avoidable shortcomings in quality and on ­others 2010). opportunities to do better for patients and • The Netherlands: The Quality Institute communities. Such studies can bring quality has crafted a mandatory framework for issues to the forefront of the policy debate. the development of care standards, clini- Moreover, by showing the government’s cal guidelines, and performance measures commitment to addressing real needs, they (VWS 2018). can also help to improve public confidence in • United States: The AHRQ supports the the health care system. development of measures of quality, For example, prompted by mounting evi- national reporting on quality, and research dence of quality failures, public demands, on quality.14 and increasing costs, several countries have • Germany: The Institute for Quality and carried out systematic reviews of national Efficiency in Health Care (IQWiG) is approaches to quality, assessed the sta- responsible for reviewing the evidence on tus quo, and proposed recommendations. diagnosis and therapy for selected condi- As mentioned earlier, two seminal IOM tions, providing evidence-based reports reports—To Err is Human: Building a Safer (for example, on drugs, nondrug interven- Health System and Crossing the Quality tions, and diagnostic and screening tests), Chasm: A New Health System for the 21st and developing recommendations on dis- Century —exposed the breadth and depth of ease management programs.15 health care quality issues in the United States and set out a strategy to address these failures Operationally, one option for China would (IOM 2001; Kohn, Corrigan, and Donald- be to broaden the mandate of the MSMGC son 1999). Another example is the “Quality to incorporate additional government and in Australian Health Care Study” commis- nongovernment actors and enhance its capac- sioned by the Australian Ministry of Health, ity to perform the recommended functions. which used retrospective clinical auditing Although the MSMGC already has some of methods to assess adverse events in hospitals these responsibilities, its limited staff (30), (Wilson and others 1995). And in the United lack of stakeholder representation, and nar- Kingdom, “A First-Class Service: Quality in row focus on public hospitals may be insuf- the New NHS” highlighted key mechanisms ficient to perform the proposed functions. for enhancing accountability, performance Another option would be to establish a coor- measurement, and inspection in health care dination architecture directly under the State (DH 1998). Council to ensure the highest-level authority Such studies are not yet available in to mobilize various public, private, and profes- China. China has piloted the collection of sional stakeholders. Importantly, the chosen performance data and the monitoring of institution will need to apply the same quality quality and patient safety in hospitals but has standards to both public and private facilities. not yet published rigorous analyses of these data (Jiang and others 2015). Led by the pro- Strategy 2: Conduct a national study of posed national authority for health care qual- QoC and quality-improvement initiatives ity, similar research in China could system- at all levels atically document quality problems related to In many countries, efforts to improve health structures, processes, and outcomes. Doing system performance have been catalyzed so would help to galvanize quality improve- by comprehensive, evidence-based reports ments throughout the nation. L e v er 2 : I mpro v ing Q u ality of C are 117 To make this happen, an independent The strategy would build on existing work panel including both Chinese and interna- (the national reviews would provide inputs to tional experts on health care quality, together the strategy) and serve as an evolving guide with Chinese academic research institutions, for the nation. It could be revised annually could be commissioned to conduct the pro- with increasing refinements. posed study. The panel would summarize the findings and issue a comprehensive report on Core Action Area 2: Systematic QoC the QoC in China and recommend goals and Measurement and Use of Data to targets for quality improvement and reforms Support Quality Improvements in policy, training, and practice. A notable feature of quality improvement Strategy 3: Develop a national strategy for efforts in the past decade in OECD countries quality improvement is their widespread use of quantitative data on Drawing on the results of the proposed health care processes and outcomes. Thanks study, a strategy could be developed that to both proliferation of data and advance- would describe an acceptable level of quality, ments in statistical methods, reliable indi- set forth quality goals, clarify the roles and cators of quality are much easier to obtain responsibilities of stakeholders, and mandate today than in the past. These measures give activities at different levels. policy makers a powerful tool to benchmark An example along these lines is the U.S. providers’ quality, identify low and high per- National Strategy for Quality Improvement formers, devise incentives to reward higher in Health Care, an annual report launched in quality, and evaluate progress over time. 2011 after enactment of the Patient Protec- tion and Affordable Care Act (HHS 2011). Strategy 1: Shift quality measurement from The strategy articulated three national structure to process and outcomes aims—better care, healthy people/healthy Structural quality is relatively easy to mea- communities, and affordable care—and six sure. Reliable data on infrastructure, equip- priorities: ment, and human resources are readily available in China. But as discussed ear- • Making care safer by reducing harm lier, although adequate structural quality is caused in the delivery of care necessary, it is not sufficient to ensure bet- ­ • Ensuring that each person and their family ter health care outcomes or experiences. are engaged as partners in their care Thus, measures are also needed to capture • Promoting effective communication and the ­processes of care between patients and coordination of care providers. Development of such measures is • Promoting the most effective prevention more complex and should be conducted on and treatment practices for leading causes the basis of the best scientific and clinical evi- of mortality dence or clinical guidelines. For example, to • Starting with cardiovascular disease, make evidence-based care the norm, doctors’ working with communities to promote clinical actions must be measured against wide use of best practices to enable healthy recommended processes. living Changes in the quality of processes of • Making quality care more affordable for care are in turn reflected in changes in out- individuals, families, employers, and gov- comes. Measures of outcome, which center ernment by developing and spreading new on ­survival rates and the extent of health and health care delivery models. functional restoration as a result of health care, are arguably the measures that matter The aims and priorities of the strategy the most to the beneficiaries of any health sys- would form the basis for designing local tem, and as such they are critical to measur- initiatives and for monitoring progress. ing the performance of any patient-centered 118 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A care model. Although data on patient out- differences in patients’ risk, variations in comes like mortality and medical complica- patient outcomes are significant among tions are collected in China, these data are ­ B eijing’s tertiary public hospitals. broad measures and not useful for comparing China may consider developing a Chinese quality across providers. For example, mor- version of the Dartmouth Atlas of geographic tality analysis in China does not typically variations in health care to inform the public consider the differences in health risks among and professionals about differences in prac- the patients admitted to hospitals, leading tice on important health topics. The Dart- to estimates that are not comparable across mouth Atlas of Health Care is a U.S. map health facilities. of regional variations in health care qual- Many OECD countries are making ity, outcomes, costs, and utilization.16 In the efforts to engage patients in quality assess- United Kingdom, the NHS Atlas series offers ment and developing tools to measure similar insights.17 health outcomes from the patient perspec- Measuring regional variations allows lead- tive. Patient-reported outcome measures ers to identify opportunities to improve care (PROMs) and patient-reported experience through standardization. Regional data can measures (PREMs) constitute feedback on help to uncover best practices that should patients’ physical, mental, and social health be spread more widely and can reveal where and on how well they are managing their inappropriate, excessive, or deficient care is chronic diseases or health conditions. As occurring. Under the supervision of the pro- noted in chapter 2, these measures may be posed authority responsible for quality, a des- incorporated into the quality measurement ignated team could create an atlas of varia- frameworks used for both integrated health tion for China. systems and single health providers. Strategy 3: Use measures of quality to Strategy 2: Create and maintain an “atlas improve performance of variation” in process quality and Three ways in which quality measures can outcomes be applied to improve frontline quality are In most nations, China included, the quality accreditation, public reporting, and pay-for- of health care and outcomes varies from one performance incentives. Together they can place to another and even among clinicians provide a comprehensive system for provid- in the same city. This variation derives from ing performance feedback and incentives for differences in professional opinions, habits, improvement. training, and application of scientific stan- Accreditation. In the United States, report- dards. The use of certain clinical procedures ing of quality-related data and m ­ easures is for specific conditions showing these large mandatory for hospital accreditation, which variations are considered “supply-sensitive” in turn is a prerequisite for ­ hospitals to because they are largely due to provider parti­cipate in the public insurance schemes, choices (whether providers deem it necessary Medicare and Medicaid. The Joint Commis- to admit a patient or perform a surgery), not sion, an independent organization respon- science or patient preferences. sible for accrediting health facilities in the Controlling variation begins with under- United States, requires accredited hospitals standing it. For example, significant varia- to report data for at least six core sets of tions in elective surgeries (such as tonsillec- measures for specific conditions or processes tomy or prostatectomy) and hospitalization (such as acute myocardial infarction, perina- associated with chronic diseases have been tal care, stroke, emergency department visits, documented in the United States and inter- surgical improvement projects, and venous nationally (Wennberg 2010). Xu and oth- thromboembolism), drawing from patients’ ers (2015) found that after adjustment for medical charts or electronic medical records. L e v er 2 : I mpro v ing Q u ality of C are 119 For public and private hospitals seeking initiated two P4Q programs: (a) the Hospi- accreditation, China can consider requiring tal Readmission Reduction program, which the reporting of data on quality. links payments to a hospital’s performance Public reporting. Making quality mea- in reducing readmissions for selected high- sures publicly available is an effective way to cost or high-volume conditions like heart create peer pressure among providers and to attack, heart failure, and pneumonia; and encourage them to pursue quality improve- (b) the Hospital Value-Based Purchasing ment by making them aware that they are Program, in which Medicare adjusts a por- being monitored. Public disclosure of qual- tion of its payment to hospitals based on ity measures can also help patients make how well they perform on quality measures informed choices among providers based on and how much progress they make in qual- their safety and quality performance. In the ity improvement. past decade, this has become the norm in The U.K. government in 2004 introduced OECD countries. a pay-for-performance scheme to recognize For example, in the United States, state- quality in family practice, covering the man- level maps for benchmarking quality can agement of chronic diseases, practice organi- be found on the AHRQ website, and infor- zation, and patients’ experience of care. Pay- mation on quality at the level of individual ments under the scheme make up as much as facilities and health plans can be found on ­ 25 percent of a family practitioner’s income multiple websites, including Hospital Com- (Doran and Roland 2010; Kroneman and pare, managed by the CMS, and the sites of others 2013). Some evidence shows that the the National Committee for Quality Assur- impact on quality improvement is enhanced ance and the Joint Commission.18 In France, when public reporting is coupled with P4Q similarly, information on the quality of pro- incentives (Lindenauer and others 2007; Wer- viders is published online on the Scope Santé ner and others 2009). website,19 and in Canada, it is provided by the The concept of pay-for-performance Canadian Institute of Health Information.20 has gained prominence in China in recent The CMS’s Hospital Compare site allows years. Although a payment system could be users to compare three hospitals at a time on implemented based on practitioners’ work- seven quality dimensions: surveys of patients’ loads, service quality, and patient satisfac- experiences, timely and effective care, com- tion, China’s lack of standardized measures plications, readmissions and deaths, use of and the still-dominant fee-for-service incen- medical imaging, payment, and value of care. tives for revenue generation make this chal- Patients using the site may choose the most lenging. P4Q schemes are ideally designed suitable hospital based on their needs and to avoid unintended cost-shifting. This preferences. occurred, for example, when an experiment Pay-for-performance incentives. Pay-for- in Guizhou removed incentives for overpre- quality (P4Q) schemes provide financial scribing medication. Doctors responded by incentives to improve quality. Although their increasing their nondrug services such as impact has been mixed and depends on the injections and unnecessary referrals to hos- design of the incentives, several countries pital care—which in turn raised total health have adopted such schemes. care costs (Wang and others 2013). But In the United States, the CMS began there are promising examples. For exam- in 2004 to financially penalize hospitals ple, in Ningxia Province, an intervention that did not report to CMS the same per- that combined capitation with P4Q incen- formance data they collected for the Joint tives reduced antibiotic prescriptions and Commission; they also decided they would total outpatient spending without signifi- no longer pay for the 28 “never” events cant adverse effects on other aspects of care (as defined earlier). In addition, the CMS (Yip and others 2014). 21 120 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Strategy 4: Establish an engagement model A preliminary evaluation suggests that to support peer learning and energize implementing the pathways reduced patients’ collective quality improvement average hospital stay and curbed unnecessary Besides benchmarking their own quality to services (Cheng 2013). Patients paid less out that of peer organizations, hospitals should of pocket, and a substantial improvement be encouraged to share valuable lessons and in communication and relations between to support each other in transforming orga- patients and providers raised the satisfaction nizations toward better quality and collec- of both groups. tively achieving clearly defined goals. Exam- However, other studies have noted that ples of such mutual support in the United managers and physicians resisted implement- States are the CMS Partnership for Patients ing the clinical pathways because doing so and the hospital engagement networks. Phy- would cause them to lose income. Managers sicians, nurses, hospitals, employers, patients were driven by revenue generation and did and their advocates, and the federal and not see clinical pathways as a useful manage- state governments have joined to form the rial instrument (He, Yang, and Hurst 2015). Partnership for Patients, adopting common China may consider analyzing lessons from goals to make care safer and improve care these experiences to inform the further devel- transitions. opment and adoption of clinical pathways. T he hospital engagement net works China has no standard evidence-based help to identify successful ways to reduce system for ensuring nationwide standardized hospital-acquired conditions and work to ­ care, nor for continuously aligning Chinese spread these approaches to other hospitals guidelines with appropriate worldwide clini- and health care providers. A form of provider- cal standards adapted to China. It is impor- to-provider peer network to share informa- tant to develop a larger set of standardized tion and learning is proposed in chapter 10. clinical pathways and to mandate their use in all hospitals. Under the guidance of the proposed national authority, and with the Core Action Area 3: Transformation of assistance of prestigious Chinese hospitals, Management to Improve QoC professional associations, and clinical leader- Effective organizational management is indis- ship groups, evidence-based care guidelines pensable for safety and quality assurance. can be created or adopted based largely on Even capable health professionals can make international standards and then modified mistakes in hectic, often overcrowded clini- to suit the characteristics of the Chinese cal environments where they are practicing health system. The standards could focus on increasingly complex medical interventions. evidence-based care protocols, use of appro- Managers can use known and tested tools to priate medication, person-centered care, and support quality improvement. skills and methods for continuous quality improvement. Strategy 1: Promote evidence-based standardized care Strategy 2: Embed “quality culture” in Clinical guidelines and pathways are valuable management philosophy and promote tools to standardize care and reduce varia- modern managerial techniques tions in practice. With technical assistance High-quality health care does not arise from from the United Kingdom’s NICE, China’s inspection alone. To ensure safety and sus- Ministry of Health has developed evidence- tain quality improvement requires a “quality based clinical pathways and applied them in culture” and continuous attention to qual- several pilot reforms in rural public hospitals. ity improvement from managers and staff. The intent is to standardize procedures and Important elements of a quality culture limit providers’ discretionary prescription of are an openness toward errors, a relatively services and drugs. flat management hierarchy, collaborative L e v er 2 : I mpro v ing Q u ality of C are 121 teamwork in a learning environment, and a using specific and measurable aims that focus on continuous system improvement. are tracked over time. In contrast, an accountability mechanism that centers on individuals and punishes These and other management approaches them for errors by “naming and shaming” can be combined and flexibly applied. The discourages providers from reporting errors need is to use them to cultivate a sense of con- and reinforces a deeply embedded belief that tinuous attention to improving the quality high-quality care results simply from being of management practices. Some are already well trained and trying hard. Some e ­ vidence being applied in some large Chinese hos- suggests that “naming and shaming” may pitals. For example, Anzhen Hospital has still be a common management practice. applied the PDSA cycle to hospital strategic A survey of employees of six secondary gen- management (Nie, Wei, and Cui 2014), and eral public hospitals in Shanghai in 2013 Peking University People’s Hospital has used found that although hospital staff are gen- TQM with PDSA to improve the efficiency of erally positive about the safety climate in specialist clinic registration (Chen and oth- their workplace, “fear of blame” and “fear ers 2014). Lessons from these experiences of shame” are two important concerns. In should be examined, and similar initiatives the United States, these are among providers’ expanded, throughout China.22 smallest concerns (Zhou and others 2015). Sound scientific evidence exists for treat- Strategy 3: Use e-health innovations to ing many health conditions—evidence that support quality improvements can drive care improvement and, in some Many nations are investing substantially cases, reduce costs. But much of this evidence in (a) electronic health (e-health) and is not fully applied in daily clinical practice. mobile health (m-health) tools, seeking new Identifying and filling the gap between what ­ efficiencies using electronic health records; is known and what is done requires con- (b) computerized decision-support systems; tinuous quality improvement efforts at any (c) ­picture archiving and communication health facility. Health facilities can improve ­ systems; and (d) remote patient monitoring quality by using some of the modern mana- and other technologies. gerial approaches shown to change health Electronic health records. EHRs form the workers’ behaviors and optimize the clini- bedrock of e-health systems because they cap- cal care system (Deming 2000; Langley and ture fundamental patient data, often includ- others 2009): ing images, patient histories, and relevant nonclinical data. EHR platforms allow data • Continuous quality improvement (CQI) to be captured, manipulated, and shared, and total quality management (TQM) potentially reducing errors and improving approaches emphasize a continuous effort efficiencies within and across health facilities by all members of the organization to and systems. EHRs can also form a central meet the needs and expectations of clients; IT hub through which other e-health mech- managers and clinicians work together to anisms (such as computerized order entry identify undesirable variations in the pro- for physicians) can be run. Platforms such cess of care and try to eliminate them. as electronic workstations accomplish simi- • Six Sigma targets aim to reduce error rates lar goals while increasing clinicians’ mobil- to six standard deviations from the pro- ity. In some cases, in addition to improving cess mean, to ensure standardized service efficiency, EHRs help to improve clinicians’ where appropriate. practice and ultimately yield gains in both • The Plan-Do-Study-Act (PDSA) cycle is a quality and safety. mechanism in which clinical teams learn In the United States, for example, an how to apply key ideas for change to their EHR system was shown to have posi- organizations in a series of testing cycles, tive effects on the quality of pediatric care 122 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A (Adams, Mann, and Bauchner 2003). At the platform to secure expert knowledge the Boston Medical Center’s Pediatric Pri- from higher-level health facilities, whose mary Care ­ Center—a clinic with more than staff can view the data uploaded by the GPs 28,000 annual patient visits—­ c linicians and provide feedback. The workstations used the Automated Record for Child have also greatly improved the performance Health (ARCH), a point-and-click inter- of providers at the village level, who have face reminiscent of the paper records with benefited both from the connection to pro- which they were already familiar. Placed viders in THCs and from content sources in each examination room, ARCH allowed such as the standardized formulary and clinicians to record routine health main- care standards that are embedded in the tenance; maintain lists of problems and system. Users have responded positively to medications; monitor patients’ growth; the workstations, noting improved patient record obstetrical history, medical history, satisfaction due to the trusted results of the and family history; view limited laboratory technology. data; link to internet-based resources; and Computerized decision-support systems. print reports. ARCH prompted clinicians CDSSs are electronic platforms that inte- to ask about and record certain risk fac- grate clinical and demographic data to sup- tors that, when tracked over time, provided port decision making by clinicians. They can a longitudinal view of health not provided serve a number of purposes, such as improv- by the paper-based method. The research- ing e-prescribing—sometimes considered its ers found that the clinicians using ARCH own form of CDSS—or providing treatment were significantly more likely than clinicians recommendations. CDSSs generally include using paper-based records to address rou- active or passive prompts to guide clinicians’ tine health maintenance ­ topics such as diet, decisions with the goal of standardizing care sleep, psychosocial issues, smoking in the and reducing errors to improve quality and home, exposure to violence, and behavioral safety (Black and others 2011). Factors that or social developmental milestones. Clini- can influence the usefulness of CDSSs include cians who used ARCH reported that the use clinician training and the quality of custom- of the system had improved overall QoC and ization based on the intended goals of the the guidance they gave families. Although system. In the examples below, CDSSs have some users noted that the use of the system been successfully used to improve quality and reduced eye-to-eye contact with patients, reduce errors. they all recommended its continued use. In Feixi in 2014, the county Bureau of In Huangzhong in 2013, the Qinghai Pro- Health introduced a standardized formu- vincial Health Department equipped each of lary policy to regulate providers’ prescrib- 30 village clinics with a general practitioner ing behaviors and to ensure that medications (GP) workstation: an electronic system that are used safely in THCs and village clin- allows physicians to conduct medical exam- ics. A recommended list of medications for inations, perform tests, download data from 50 common outpatient conditions THC and their current location, and upload data to village clinic providers might encounter, such the regional health system (Meng and oth- as influenza and bronchitis, was integrated ers 2015). The workstations are portable, so into the district’s computerized information practitioners can travel with them and bring system. The system also contains predefined high-quality, reliable health care to patients’ prescription packages. When a village pro- homes. THC personnel perform routine vider inputs a particular diagnosis and related checks on the system every three months symptoms, the computer system proposes and respond to issues reported by village- medications to recommend. Providers must level providers as needed. The workstations choose from among the options presented by have improved information sharing between the system. Feixi’s CDSS has standardized the village and the county level. GPs can use the prescription behaviors of primary health L e v er 2 : I mpro v ing Q u ality of C are 123 care providers and enhanced safety and reli- cooperation groups that were organized by ability for patients. the municipal hospitals. Each hospital pro- Similar gains through CDSSs have been vided a site to house one medical imaging documented in the United States. In one e-consultation center and one electrocardio- example, physicians at a university hospital gram (ECG) e-consultation center; it also clinic received a CDSS, accessible on their provided a director, staff, and supporting handheld personal digital assistants, to guide facilities, along with training and technical the prescription of nonsteroidal inflamma- support for the physicians at the CHCs. tory drugs (Berner and others 2006). The Much of this program was supported by CDSS, called MedDecide, contained a suite implementation plans and technical guid- of clinical prediction rules based on evidence- ance from the Hangzhou government and based literature. The study found that the the Hangzhou Bureau of Health, which sets physicians who used the system prescribed requirements for e-consultation services and more safely than those who did not and that regulates program implementation. Cur- they documented more complete assessments rently, the PACS services are provided free of of the risks to patients. Though it is a nar- charge to CHCs and patients and are funded rowly defined example, this trial highlights by the bureau. The bureau also compensates how mobile technology and CDSSs can sup- municipal hospitals retroactively based on the port physicians in making safer treatment number of e-consultations performed during decisions and that m-health programs at the the year at the following rates: RMB 30 for point of care can improve clinicians’ perfor- each case of ordinary imaging e-­consultation, mance in the ambulatory setting. RMB 50 for each case of CT or magnetic Picture archiving and communication sys- resonance diagnosis e-consultation, and tems. PACS are clinical IT systems that allow RMB 10 for each case of ECG e-consultation. facilities to acquire, archive, process, and By the end of 2013 in the Xiacheng district, distribute digital images (such as radiologi- imaging e-consultation services through cal scans) to improve the quality of patient PACS were available to 46 CHCs, and more care (Black and others 2011). They are often than 300 e-consultation cases had been integrated with EHRs and are designed to evaluated. Through 2014 in the Jiang-gan encourage the scaling of expertise within district, more than 20,755 ECG submissions and across facilities. Through PACS, physi- and 16,425 imaging submissions had been cians can share images with clinicians within reviewed through the e-consultation centers. the facility or externally and receive expert Remote patient monitoring. Remote advice and treatment recommendations via patient monitoring is a broad term encom- the platform. Some countries have success- passing technologies that allow clinicians fully implemented PACS to extend the reach to observe patients outside of conventional of their specialists. settings, such as from the patient’s home or Hangzhou provides a strong example of from a care setting where the physician is not such a system in China. The city began in physically present. Remote patient monitor- 2013 to better integrate its municipal tertiary ing has the potential to extend physicians’ hospitals with CHCs (Yan 2015). PACS were reach, increase the time physicians have used to establish regional e-consultation cen- available to treat patients, and more actively ters for medical imaging and electrocardio- engage patients in their care. A number of grams. The reform was designed to increase examples of remote patient monitoring for the access of CHC-based primary care pro- both inpatient and outpatient care show the viders to the expertise in municipal hospi- positive impact it can have on both quality tals, build higher-quality capacity at those and safety for patients. CHCs, and increase patients’ access to expert In an outpatient example, a U.S. study imaging services. Hangzhou structured shows the potential benefits of a mobile- its e-consultation network in four distinct phone-based self-management aid for 124 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A adolescents suffering from asthma (Rhee and recommendations based on experience in others 2014). Recognizing the potential bene- China and OECD countries. We recom- fits of the self-management of asthma, as well mend that the government continue to as the portability and accessibility of mobile engage all stakeholders in the health sec- phones, the investigators evaluated whether tor to publicly affirm its quality improve- mobile-phone technology could facilitate ment goals, strengthen technical leadership symptom monitoring, treatment adher- through M&E of clinical care quality, and ence, and adolescent-parent partnerships foster improvement at the front line by sup- for asthma. The system allowed adolescents porting innovations in health care delivery to communicate via short message service and spreading successful experiences. The (SMS) text in natural English and to initi- reforms in the delivery of health care pro- ate interactions with an automated support posed here are aligned with the “supply-side system. Participants received reminder texts reforms” that were introduced by the Chi- from the system, such as for taking medica- nese government in 2015 and early 2016, tion. They could also interact with the system which will be a key lever for building the through scheduled or unscheduled communi- PCIC model. cations to share their concerns about symp- toms, medications, or other asthma-related Notes activity. The system would automatically generate responses that were evaluated by 1. For resources on national-level quality indica- certified asthma educators who could step tors, see the following: “Quality Measures,” in and respond through the system. Parents Centers for Medicare & Medicaid Services (https://www.cms.gov/Medicare/Quality- of participants received automatically gener- Initiatives-Patient-Assessment-Instruments/ ated emails from the system informing them QualityMeasures/index.html?redirect=/ of their adolescent’s asthma-control levels, QUALITYMEASURES); the World Health levels of activity, frequency of use of rescue Organization’s Performance Assessment Tool medication (such as albuterol), and use of for Quality Improvement in Hospitals (PATH) control medication. project (http://www.­pathqualityproject.eu and The study found that the response rates Veillard and others [2005]); and the OECD’s from adolescents to the system’s text messages Health Care Quality Indicators (HCQI) proj- were 81–97 percent and that adolescents ini- ect (http://www.oecd.org/els/health-systems/ tiated a message to the system an average of health-care-quality-indicators.htm and Arah 19 times over a two-week trial period. Post- and others [2003]). 2. Barefoot doctors are “farmers who received trial focus groups illuminated how beneficial a short medical and paramedical training, to the m-health intervention had been for ado- offer primary medical services in their rural lescents with asthma, indicating that it had villages” (Yang and Wang 2017). raised patients’ awareness of symptoms and 3. See, for example, the “General Hospital Evalu- triggers, improved self-management and med- ation Standards” (revised version) in 2009; ication adherence, and improved their sense of “Tertiary Hospital Accreditation Standards” in control over asthma. Overall, the mobile sup- 2011 (Weiyiguan Fa 2011, No. 33); and “Sec- port platform improved the quality of asthma ondary Hospital Accreditation Standards” in care for these adolescents, and illustrates how, 2012. “Requirements on Medical Errors and for certain patient populations, m-health is a Adverse Events Reporting” were announced in useful lever for promoting QoC. 2011 (Weiyiguan Fa 2011, No. 4). 4. “Guidelines on Antimicrobial Drug Use,” NHFPC (2012, No. 84); NHFPC 2015. Conclusion 5. “Guiding Opinions on the Implementation of Clinical Pathways During the Twelfth Five- This chapter has reviewed the challenges Year Plan Period,” Department of Medical China faces in enhancing the quality and Service Management, Ministry of Health value of care and has proposed a set of (Weiyizhen Fa 2012, No. 65). L e v er 2 : I mpro v ing Q u ality of C are 125   6. “Guidance of the General Office of the State Measurement,” National Committee for ­ Quality Council on Overall Pilot Reform of Urban Assurance, http://www.ncqa.org / HEDIS​ Public Hospitals,” State Council General QualityMeasurement.aspx; and “Pioneers in Office (Guo Ban Fa 2015, No. 38); and Quality,” Joint Commission, http://www.joint- “Opinions of the State Council on Compre- commission​.org/­accreditation/top_performers. hensively Scaling-Up Reform of County-Level aspx. Public Hospitals,” State Council (Guo Ban Fa 19. For France’s Scope Santé website, see http:// 2015, No. 33). www.scopesante.fr/.  7. Regarding SR Es, see “Serious Report- 20. For the Canadian Institute of Health Infor- able Events,” National Quality Forum mation’s “Health System Performance” Topics (accessed June 28, 2018), http:// re sou rce s , se e ht t ps: // w w w.ci h i.c a /en​ www.­qualityforum.org/topics/sres/serious​ /health-system-performance. _reportable_events.aspx. 21. Capitation is a payment arrangement for   8. See, for example, hospital indicators in the health care providers such as physicians or Medicare Hospital Compare datasets (https:// nurse practitioners. It pays a physician or data.medicare.gov/data/hospital-compare), group of physicians a set amount per period which enable people to compare the quality of time for each enrolled person assigned to of care at more than 4,000 Medicare-certified them, per period of time, whether or not that hospitals across the United States. person seeks care.   9. For more information, see “How to Improve,” 22. Chapter 10 presents an approach for scaling IHI website (accessed June 28, 2018), http:// up care improvement that applies the PDSA www.ihi.org/resources/Pages/HowtoImprove. cycle. 10. For t h e N H S O ut c om e s Fr a m e work website, see https://digital.nhs.uk /data​ - a nd - i n for m at ion /publ ic at ion s /c i - hub​ /nhs-outcomes-framework. References 11. The National Indicators Project was under- Abdullah, A. S., F. Qiming, V. Pun, F. A. Still- taken by the Australian Institute of Health man, and J. M. Samet. 2013. “A Review of and Welfare in close consultation with the Tobacco Smoking and Smoking Cessation commission and a wide range of clinical and Practices among Physicians in China: 1987– other stakeholders. 2010.” Tobacco Control 22 (1): 9–14. 12. Chapter 4 includes a discussion on engaging Adams, W. G., A. M. Mann, and H. Bauchner. patients and the public in the development 2003. “Use of an Electronic Medical Record and reporting of quality measures. Improves the Quality of Urban Pediatric Pri- 13. For details, see the NICE website: https:// mary Care.” Pediatrics 111 (3): 626–32. www.nice.org.uk/. AHMAC (Australian Health Ministers’ Advisory 14. For details, see the AHRQ website: https:// Council). 1996. “The Final Report of the Task- www.ahrq.gov. force on Quality in Australian Health Care.” 15. For details, see the IQWiG website: https:// AHMAC report, Council of Australian Gov- www.iqwig.de/en. ernments (COAG) Health Council, Adelaide, 16. For more i n for m at ion on t h e D a r t- South Australia. mouth Atlas, see the website: http://www​ AHRQ (Agency for Healthcare Research and .­dartmouthatlas.org/. Quality). 2017. “2016 Healthcare Quality 17. For details on the NHS Atlas series, see and Disparities Report.” AHRQ Publication ht t ps: // w w w.eng la nd.n hs.u k /r ig htc a re​ No. 17-0001, U.S. Department of Health and /products/atlas/. Human Services, Rockville, MD. 18. For these quality-related data features, see AIHW (Australian Institute of Health and Wel- “National Healthcare Quality and Dispari- fare). 2009. Towards National Indicators of ties Reports,” AHRQ, http://nhqrnet.ahrq. Safety and Quality in Health Care. Canberra: gov/inhqrdr/state/select?utm_source=AHRQ​ AIHW. - E N & u t m _ m e d i u m = a r t i c l e & u t m​ A llegranzi, B enedet ta, S. B . Nejad, _campaign=SS2015; “Hospital Compare,” C. Combescure, W. Graafmans, H. Attar, CMS, https://www.medicare.gov/hospitalc- L. Donaldson, and D. Pittet. 2011. “Burden ompare/search.html; “HEDIS & Performance of Endemic Health-Care-Associated Infection 126 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A in Developing Countries: Systematic Review Safety of Health Care: A Systematic Over- and Meta-Analysis.” The Lancet 377 (9761): view.” PLoS Medicine 8 (1): e1000387. 228–41. Brook , R . H . 1995. “T he R A N D/ UC L A Andel, Charles. 2012. “The Economics of Health Appropriateness Method.” Publication No. ­ Care Quality and Medical Errors.” Journal of RP-395 (reprinted by RAND Corp.), U.S. Health Care Finance 39 (1): 39. Department of Health and Human Services, Arah, O. A., N. Klazinga, N. S. Klazinga, D. M. J. Washington, DC. Delnoij, A. H. A. Ten Asbroek, and T. Custers. Campbell, Dennis. 2014. “NHS Wastes over £2bn 2003. “Conceptual Frameworks for Health a Year on Unnecessary or Expensive Treat- Systems Performance: A Quest for Effective- ments.” The Guardian, November 5. ness, Quality, and Improvement.” Interna- CDC (Centers for Disease Control and Preven- tional Journal for Quality in Health Care 15 tion). 2018. “National Healthcare Safety (5): 377–98. Network (NHSN): Patient Safety Compo- Australian Government. 2009. “Review of nent Manual.” Division of Healthcare Qual- Health Technology Assessment in Austra- ity Promotion, CDC, Atlanta. http://www​ lia: A Discussion Paper.” Department of .cdc.gov​/ nhsn/pdfs/pscmanual/pcsmanual_­ Health and Ageing, Australian Government, current.pdf. Canberra. Chassin, Mark, and Robert Galvin. 1998. “The Baicker, Katherine, and Amitabh Chandra. 2004. Urgent Need to Improve Health Care ­ Quality: “Medicare Spending, the Physician Workforce, Institute of Medicine National Roundtable and Beneficiaries’ Quality of Care.” Health on Health Care Quality.” JAMA 280 (11): Affairs 23 (3): 184–97. 1000–05. Berner, E. S., T. K. Houston, M. N. Ray, J. J. Alli- Chen, Ming, Hongmei Zhao, Saijun Jia, Lihua son, G. R. Heudebert, W. W. Chatham, and Wang, and Yue Zhao. 2014. “Total Qual- R. S. Maisiak. 2006. “Improving Ambulatory ity Management: Improving the Procedure of Prescribing Safety with a Handheld Decision Additional Registration in Specialist Clinic.” Support System: A Randomized Controlled Chinese Hospitals 18 (1). Trial.” Journal of the American Medical Infor- Cheng, Tsung-Mei. 2013. “A Pilot Project Using matics Association 13 (2): 171–79. Evidence-Based Clinical Pathways and Pay- Berwick, Donald, M. Godfrey, A. Blanton, and ment Reform in China’s Rural Hospitals Jane Roessner. 1990. Curing Health Care: Shows Early Success.” Health Affairs 32 (5): New Strategies for Quality Improvement. San 963–73. Francisco: Jossey-Bass. Chevreul, Karine, Isabelle Durand-Zaleski, Berwick, Donald, Thomas W. Nolan, and John Stéphane Bahrami, Cristina Hernández-­ Whittington. 2008. “The Triple Aim: Care, Quevedo, and Philipa Mladovsky. 2010. Health, and Cost.” Health Affairs 27 (3): “France: Health System Review.” Health Sys- 759–69. tems in Transition 12 (6). Bhattacharyya, Onil, Yin Delu, Sabrina T. Dajiang Net . 2012 . “W ho Will B ear the Wong, and Chen Bowen. 2011. “Evolu- R at i n g C on su mpt ion of I nva l id Te r - tion of Primary Care in China 1997–2009.” tiary Hospitals?” [title in English transla- Health Policy 100 (2): 174–80. doi:10.1016/j. tion]. Opinion letter, jxnews.com, August healthpol.2010.11.005. 2 8 . h t t p : / / j x c o m m e n t . j x n e w s . c o m . c n​ Bi, Yufang, Runlin Gao, Steve Su, Wei Gao, Dayi /­system/2012/08/28/012087450.shtml. Hu, Dejia Huang, Lingzhi Kong, and others. Dayal, Prarthna, and Krishna Hort. 2015. Qual- 2009. “Evidence-Based Medication Use among ity of Care: What Are Effective Policy Options Chinese Patients with Acute Coronary Syn- for Governments in Low- and Middle-Income dromes at the Time of Hospital Discharge and Countries to Improve and Regulate the Qual- One Year after Hospitalization: Results from ity of Ambulatory Care? Policy Brief 4 (1). the Clinical Pathways for Acute Coronary Manila: World Health Organization, on behalf Syndromes in China Study.” American Heart of the Asia Pacific Observatory on Health Sys- Journal 157 (3): 509–16. tems and Policies. Black, A. D., J. Car, C. Pagliari, C. Anandan, Deming, William Edwards. 2000. The New Eco- K. Cresswell, T. Bokun, and A. Sheikh. 2011. nomics: For Industry, Government, Educa- “The Impact of e-Health on the Quality and tion. Cambridge, MA: MIT Press. L e v er 2 : I mpro v ing Q u ality of C are 127 de Silva, D., and J. Bamber. 2014. “Improving the 21st Century. Washington, DC: National Quality in General Practice.” Evidence Scan Academies Press for the Committee on Quality No. 23, The Health Foundation, London. of Health Care in America. DH (Department of Health). 1998. “A First-Class Jiang, Lixin, Harlan Krumholz, Xi Li, Jing Li, Service: Quality in the New NHS.” Consul- and Shengshou Hu. 2015. “Achieving Best tation document, Her Majesty’s Stationery Outcomes for Patients with Cardiovascular Office (HMSO), London. Disease in China by Enhancing the Quality ———. 2008. “High Quality Care for All: NHS of Medical Care and Establishing a Learn- Next Stage Review Final Report.” Consul- ing Health-Care System.” The Lancet 386 tation document, Her Majesty’s Stationery (10002): 1493–1505. Office (HMSO), London. Keown, Oliver P., and Ara Darzi. 2015. “The ———. 2010. Equity and Excellence: Liberating Quality Narrative in Health Care.” The Lan- the NHS. London: Her Majesty’s Stationery cet 385 (9976): 1367–68. Office (HMSO). Kohn, Linda T., Janet M. Corrigan, and Molla Dlugacz, Yosef D., Andrea Restifo, and Alice S. Donaldson, eds. 1999. To Err Is Human: Greenwood. 2004. The Quality Handbook Building a Safer Health System. Washington, for Health Care Organizations: A Manager’s DC: National Academies Press. Guide to Tools and Programs. New York: Kroneman, Madelon, Pascal Meeus, Dionne Sofia John Wiley. Kringos, Wim Groot, and Jouke van der Zee. Donabedian, Avedis. 2005. “Evaluating the Qual- 2013. “International Developments in Rev- ity of Medical Care.” Milbank Quarterly 83 enues and Incomes of General Practitioners (4): 691–729. from 2000 to 2010.” BMC Health Services Doran, Tim, and Martin Roland. 2010. “Les- Research 13: 436. sons from Major Initiatives to Improve Pri- La Forgia, Gerard Martin, and Bernard Coutto- mary Care in the United Kingdom.” Health lenc. 2008. Hospital Performance in Brazil: Af fairs 29 (5): 1023 –29. doi:10.137 7/ The Search for Excellence. Washington, DC: hlthaff.2010.0069. World Bank. Fan, Yunzhou, Zhaoxia Wei, Weiwei Wang, Li Langley, Gerald J., R. Moen, K. M. Nolan, T. Tan, Hongbo Jiang, Lihong Tian, Yuguang W. Nolan, and C. L. Norman. 2009. The Cao, and Shaofa Nie. 2014. “The Incidence Improvement Guide: A Practical Approach to and Distribution of Surgical Site Infection in Enhancing Organizational Performance. New Mainland China: A Meta-Analysis of 84 Pro- York: John Wiley & Sons. spective Observational Studies.” Scientific Laxminarayan, Ramanan, and David L. Hey- Reports 4: 6783. mann. 2012. “Challenges of Drug Resistance He, Jingwei Alex, Wei Yang, and Keith Hurst. in the Developing World.” BMJ 344: e1567. 2015. “Clinical Pathways in China: An Evalu- Li, Yongbin, Jing Xu, Fang Wang, Bin Wang, ation.” International Journal of Health Care Liqun Liu, Wanli Hou, and Hong Fan. 2012. Quality Assurance 28 (4). “Overprescribing in China, Driven by Finan- Hesketh, Therese, Dan Wu, Linan Mao, and cial Incentives, Results in Very High Use of Nan Ma. 2012. “Violence against Doctors in Antibiotics, Injections, and Corticosteroids.” China.” BMJ 345: e5730. Health Affairs 31 (5): 1075–82. HHS (U.S. Department of Health and Human Li, Weiping and Jianxiu Wang. 2015. “China’s Services). 2011. “Report to Congress: National Public Hospital Governance Reform: Dong- Strategy for Quality Improvement in Health yang Case Study.” China National Health Care.” Planning report pursuant to the Afford- Development Research Center, Beijing. able Care Act, Agency for Healthcare Research Liao, Xinbo. 2015 “An Ethical Analysis of Over- and Quality (AHRQ), Rockville, MD. Treatment.” Clinical Misdiagnosis and Mis- IMS Institute for Healthcare Informatics. 2013. therapy 28 (1): 1–5. “Avoidable Costs in U.S. Health Care: The $200 Lindenauer, Peter K., Denise Remus, Sheila Billion Opportunity from Using Medicines More Roman, Michael B. Rothberg, Evan M. Ben- Responsibly.” Study report, IMS Institute for jamin, Allen Ma, and Dale W. Bratzler. 2007. Healthcare Informatics, Parsippany, NY. “Public Reporting and Pay for Performance in IOM (Institute of Medicine). 2001. Crossing the Hospital Quality Improvement.” New England Quality Chasm: A New Health System for Journal of Medicine 356 (5): 486–96. 128 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Liu, Dongying, Qingchun Hou, and Haiyan ———. 2012. “In Amenable Mortality—Deaths Zhou. 2013 “Appropriate Technology for Gen- Avoidable through Health Care—Progress in eral Practice in Township Health Centers.” the U.S. Lags That of Three European Coun- Chinese Journal of General Medical Practice tries.” Health Affairs 10: 1377. 16 (12A): 4027–30. NQF (National Quality Forum). 2008. National Maybin, Jo, and Ruth Thorlby. 2008. “High Voluntary Consensus Standards for Ambula- Quality Care for All.” Briefing, July 2008, The tory Care, Part 1: A Consensus Report. Wash- King’s Fund, London. ington, DC: NQF. McGlynn, E. A., S. M. Asch, J. Adams, J. Keesey, Qian, Xu, Helen Smith, Li Zhou, Ji Liang, and J. Hicks, A. DeCristofaro, and E. A. Kerr. Paul Garner. 2001. “Evidence-Based Obstet- 2003. “The Quality of Health Care Delivered rics in Four Hospitals in China: An Obser- to Adults in the United States.” New England vational Study to Explore Clinical Practice, Journal of Medicine 348 (26): 2635–45. Women’s Preferences, and Providers’ Views.” Meng, Qingyue, Zhang Luyu, Zhu Weiming, and BMC Pregnancy and Childbirth 1 (1): 1. Ma Huifen. 2015. “People-Centered Health Rhee, H., J. Allen, J. Mammen, and M. Swift. Care: A Case Study from Huangzhong County, 2014. “Mobile Phone–Based Asthma Self- Qinghai Province.” China Center for Health Management Aid for Adolescents (mASMAA): Development Studies, Peking University. Case A Feasibility Study.” Patient Preference and study commissioned by the World Bank, Adherence 8: 63–72. Washington, DC. Rudd, Tony. 2011. “The Legacy of NHS London Morris, Stephen, Rachael M. Hunter, Angus I. G. Stroke.” Presentation at “Progressing Health Ramsay, Ruth Boaden, Christopher McKev­ Care in London” conference, October 11. itt, Catherine Perry, Nanik Pursani, and https://www.kingsfund.org.uk/audio-video​ others. 2014. “Impact of Centralising Acute /tony-rudd-legacy-nhs-london-%E2%80%93​ Stroke Services in English Metropolitan Areas -stroke-programme. on Mortality and Length of Hospital Stay: Stranges, Elizabeth, and Carol Stocks. 2010. Difference-in-Differences Analysis.” BMJ “Potentially Preventable Hospitalization for 349: g4757. Acute and Chronic Conditions, 2008.” Sta- NCHS (National Center for Health Statistics). tistical Brief No. 99, Healthcare Cost and 2010. “The National Survey Study on Patient- Utilization Project (HCUP), AHRQ, U.S. Doctor Relationship.” National Health and Department of Health and Human Services, Family Planning Commission of the People’s Rockville, MD. Republic of China, Beijing. Turner, Simon, Angus Ramsay, Catherine Perry, NHFPC (National Health and Family Planning Ruth Boaden, Christopher McKevitt, Stephen Commission). 2014. “The National Health Morris, Nanik Pursani, and others. 2016. and Family Planning Commission Publicly “Lessons for Major System Change: Central- Solicited Opinions on the ‘Measures for the ization of Stroke Services in Two Metropolitan Management of Medical Quality (Draft for Areas of England.” Journal of Health Services Soliciting Opinions)’.” Release, August 5. Research & Policy 21 (3): 156–65. http://www.nhfpc.gov.cn/yzygj/s3586/201405 Veillard, J., F. Champagne, N. Klazinga, V. /4b01f1da84f342cb8357b8ecd77f95c9.shtml. Kazandjian, O. A. Arah, and A.-L. Guisset. NHS (National Health Service) Commissioning 2005. “A Performance Assessment Framework Board. 2012. “The CCG Outcomes Indicator for Hospitals: the WHO Regional Office for Set 2013/14.” Infographic, NHS Commission- Europe PATH Project.” International Journal ing Board, NHS England, Leeds, U.K. for Quality in Health Care 17 (6): 487–96. Nie, Xiaomin, Yongxiang Wei, and Xiaoyan Cui. Wang, X., and H. Xue. 2011. “Analysis of the 2014. “Implementation of Hospital Strategic Stock of Equipment and Human Resources at Management under PDCA Cycle.” Chinese Township Health Centers in Guizhou Prov- Hospital Management 34 (3). ince.” Chinese Journal of Public Health Man- Nolte, Ellen, and Martin McKee. 2011. “Varia- agement 27 (2): 117–20. tions in Amenable Mortality—Trends in 16 Wang, Jin-cai, J. Zhen, W. Fan, Q. Zhang, and High-Income Nations.” Health Policy 103 (1): Guo-li Zhang. 2013. “Investigation and Anal- 47–52. ysis of the Effects of National Essential Drug L e v er 2 : I mpro v ing Q u ality of C are 129 System on Medical Expenses in Different Lev- Yang, Le, and Hongman Wang. 2017. “Medical els of Medical Institutions.” China Pharmac y Education: What about the Barefoot Doc- 23: 2982–84. tors?” The Lancet 390 (10104): 1736. Wei, Jade W., Ji-Guang Wang, Yining Huang, Yin, Wenqiang, Zhongming Chen, Guan Hui, Ming Liu, Yangfeng Wu, Lawrence K. S. Wong, Xuedan Cui, Qianqian Yu, Haiping Fan, Xin and Yan Cheng. 2010. “Secondary Prevention Ma, and Yan Wei. 2015. “Using Entropy of Ischemic Stroke in Urban China.” Stroke 41 Weight RSR to Evaluate Village Doctors’ (5): 967–74. Prescription in Shandong Province under the Wennberg, John E. 2010. Tracking Medicine: Essential Medicine System.” Modern Preven- A Researcher’s Quest to Understand Health tive Medicine 42 (3): 465–67. Care. Oxford: Oxford University Press. Yin, Xiaoxv, Fujian Song, Yanhong Gong, Xiao- Werner, Rachel M., R. Tamara Konetzka, Elizabeth chen Tu, Yunxia Wang, Shiyi Cao, Junan Liu, A. Stuart, Edward C. Norton, Daniel Polsky, and Zuxun Lu. 2013. “A Systematic Review and Jeongyoung Park. 2009. “Impact of Pub- of Antibiotic Utilization in China.” Jour- lic Reporting on Quality of Post-Acute Care.” nal of Antimicrobial Chemotherapy 68 (11): Health Services Research 44 (4): 1169–87. 2445–52. WHO (World Health Organization). 2006. Qual- Yip, Winnie, and William C. Hsiao. 2015. “What ity of Care: A Process for Making Strategic Drove the Cycles of Chinese Health System Choices in Health Systems. Geneva: WHO. Reforms?” Health Systems & Reform 1 (1): Wilson, R. M., P. Michel, S. Olsen, R. W. Gib- 52–61. berd, C. Vincent, R. El-Assady, O. Rasslan, Yip, Winnie, T. Powell-Jackson, W. Chen, M. Hu, and others. 2012. “Patient Safety in Develop- E. Fe, M. Hu, W. Jian, M. Lu, W. Han, and ing Countries: Retrospective Estimation of W. C. Hsiao. 2014. “Capitation Combined Scale and Nature of Harm to Patients in Hos- with Pay-For-Performance Improves Antibiotic pital.” BMJ 344: e832. Prescribing Practices in Rural China.” Health Wilson, R. M., W. B. Runciman, R. W. Gibberd, Affairs 33 (3): 502–10. B. T. Harrison, L. Newby, and J. D. Hamilton. Zhang, Lihua, Jing Li, Xi Li, and Lixin Jiang. 1995. “The Quality in Australian Health Care 2015. “National Assessment of Statin Therapy Study.” Medical Journal of Australia 163 (9): for Patients with Acute Myocardial Infarction 458–71. 2001–11: Insight from the China PEACE- Wu, Xianru, Shuxiu Luo, Bin Chen, Yunkang Lu, Retrospective Acute Myocardial Infarction Cuirong Gan, Yong Yang, and Kaling Wang. Study.” The Lancet 386: S42. 2009. “A Survey of Village Doctors’ Knowl- Zhang, Peipei, Lianyi Zhao, Jing Liang, Yan edge and Clinical Skill in Prevention and Qiao, Quanyan He, Liuyi Zhang, Fang Wang, Treatment of Hypertension.” Internal Medi- and Yuan Liang. 2014. “Societal Determina- cine of China 4 (6): 910–12. tion of Usefulness and Utilization Wishes of Xu, J. 2010. “Investigation of CPR Knowledge Community Health Services: A Population- and Skills Training of Village Doctors.” Medi- Based Survey in Wuhan City, China.” Health cal Innovation of China 7 (21): 18–20. Policy and Planning: czu128. Xu, Y., T. Shu, W. Yang, M. Liang, and Y. Liu. ZJOL (Zhejiang Online). 2011. “Zhejiang 2015. “Variations in Quality of Care at Establishes the First National ‘Medical Large Public Hospitals in Beijing, China: Quality Control and Evaluation Office.’” A Condition-Based Outcome Approach.” Zhejiang Online—Health Network News PLOS One 2 (10). Network, March 2. http://health.zjol.com.cn​ Yan, Fei. 2015. “Integrated Health Services /system/2011/03/02/017334382.shtml. Reform between Community Health Service Zhou, Ping, M . K. Bundorf, Jianjun Gu, Centers and Hospitals in Hangzhou, Zheji- Xiaoyan He, and Di Xue. 2015. “Survey on ang Province.” School of Public Health, Fudan Patient Safety Climate in Public Hospitals University. Case study commissioned by the in China.” BMC Health Services Research World Bank, Washington, DC. 15 (1): 1. 4 Lever 3: Engaging Citizens in Support of the PCIC Model Introduction has sometimes led to violence toward provid- ers (Chen 2012; Yuan 2012). Although the The people-centered integrated care (PCIC) response from health authorities focuses on model seeks to organize primary health care adding security staff to hospitals, outlawing around the health needs of citizens and com- hongbao (red packets given as gifts), and munities of China, not simply around the dis- related stopgap measures, the underlying eases from which they suffer. The model causes of distrust remain unaddressed. hinges on patients’ responsibility and engage- This chapter focuses on core action areas ment with their health, on their confidence in that directly seek to strengthen people’s the system, and on their trust that the system engagement in their health, the health sys- will meet their needs in a responsive, appro- tem, and the patient-provider relationship. priate, and timely manner. At the same time, Patient empowerment and engagement is cen- beneficiaries of the health system need to be tral to any health system reform that aims to empowered with knowledge and understand- improve efficiency and make providers ing of individual health–promoting behaviors accountable for the services they deliver. For that will be amplified through interaction optimal use of resources, decisions about with the formal service delivery system. Such investment and disinvestment in services empowerment and engagement of citizens is a must be shaped by patients’ preferences foremost strategic direction advocated in the (Coulter, Roberts, and Dixon 2013; Mulley, World Health Organization’s (WHO) frame- Richards, and Abbasi 2015). Moreover, dif- work on people-centered and integrated ferent outcomes matter to different patients. health services (WHO 2015b). When clinicians overlook or misunderstand In part because of rising incomes, rapid patients’ preferences, the consequences can urbanization, and increased demand for be as harmful as misdiagnosing disease health services, the Chinese population has (Mulley, Trimble, and Elwyn 2012). high expectations that health system reforms Outside the hospital and other acute-care will improve service delivery. These expecta- settings, much of health care, including dis- tions have only partially been met by ease prevention and health promotion, is a increased access to health care and increased knowledge-intensive service industry where reimbursement of health care costs. In fact, value is coproduced from two-way commu- public dissatisfaction with the health system nication between multidisciplinary clinical 131 132 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A teams and the patients they serve (Mulley Significantly, though, the communiqué 2009). This two-way nature of health care also invokes each individual’s responsibility underscores the need for approaches and to strive for better health: the concept processes that support greater health liter- emphasizes that everybody is responsible acy and sharing of knowledge. Without this for the construction of a Healthy China exchange, decisions are made with avoidable and that each person should strengthen ignorance at the front lines of care delivery, “the management of self-health” and avoid services fall short of meeting needs while sickness as much as possible. Thus, the cre- exceeding wants, and efficiency declines ation of a Healthy China involves not just over time. health care but also changes in people’s Strengthening patient engagement is a goal awareness about health and healthy life- relevant for China, as reflected in several styles through education to realize personal state policies1 that call on the health system “hea lt h m a nagement”— for ex a mple , and its stakeholders to t h roug h b ehav iora l cha nge s such as increased physical activity and improved • Strengthen health promotion, education, eating habits. and dissemination of medical and health On the basis of these principles, the knowledge; advocate a healthy and civi- National Health and Family Planning lized lifestyle; promote rational nutrition Commission (NHFPC) initiated the develop- among the public; and enhance the health ment of a “Healthy China Construction Plan awareness and self-care ability of the (2016–20).” These policies in turn reflected people; several initiatives to improve patient engage- • Build sound and harmonious relations ment, including the following: between health care workers and patients; and • Changshu (Jiangsu Province) has applied • Promote the transparency of hospital diabetes prevention and control measures information through regular disclosure of as part of the WHO Alliance for Healthy finances, performance, quality, safety, Cities, and the approach has shown prom- price and inpatient cost, and so on. ise in addressing the spread of diabetes (Szmedra and Zhenzhong 2013). The most recent state directive explicitly • The NHFPC released the National Health mentions use of media “to publicize disease Literacy Promotion Action Plan, 2014–20, prevention and treatment knowledge . . . as to raise health literacy in China by provid- well as reasonable selection of medical insti- ing information on basic health knowl- tutions” and “more publication” to “increase edge, healthy lifestyles, and basic medical people’s understanding” toward diagnosis skills (NHFPC 2014). and treatment.2 • The erstwhile Ministry of Health (now The 2015 communiqué of the 18th Session NHFPC) and the China Journalists’ of the Central Committee of the Fifth Plenary Association in 2005 launched the China Session of the Communist Party of China Health Communication Awards; each emphasized the need to create a “Healthy year this project develops health commu- China.” Seeking to improve outcomes and nication strategies focused on one selected well-being through a multipronged approach, disease, such as hypertension (2005) or the concept of Healthy China involves cancer prevention (2006). ­ deepening the reform of the health system, • In Shanghai, a self-management program services, rationalizing drug prices, integrating ­ for hypertension (centered on a hyperten- improving coverage of basic health ­ services in sion manual and delivered in the setting of both urban and rural areas, modernizing a community antihypertensive club) hospital management, and implementing a showed promising reductions in blood food-safety strategy. pressure (Xue, Yao, and Lewin 2008). L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 133 The Shanghai Chronic Disease Self- to provide useful inputs to discussions with Management Program improved partici- health providers when making decisions pants’ health behavior, self-efficacy, and about care and hence to the patient’s range of health status and reduced the number of influence on the outcome of such decisions. hospitalizations (Fu and others 2003). The final core action area is “public” in • A recent Health News article argued for nature: creating a supportive environment for the need for shared decision making citizen engagement. Strategies in this area between provider and patient in China to aim to improve the macro-environment to manage and prevent illness (Zhong 2015). support interventions for engaging patients • The National Clinical Information System, and to mobilize societal forces to enable peo- established in 2013, is an official website ple to live a healthy life. Of interest are mod- that provides a platform for news on qual- els such as the WHO’s Healthy Cities and ity control. Villages (CSDH 2008) and environmental “nudges” to improve healthy behaviors. This chapter draws on experience with strengthening patient engagement in health sys- Challenges to Engaging Citizens tems in China and around the world and describes a variety of approaches used to engage Although official policy statements indicating patients. The chapter is organized as follows: a desire to move toward a patient-centered health system are a step in the right direction, • “Challenges to Engaging Citizens” pres- a much-needed comprehensive, systemwide ents evidence regarding the patient-­ approach to engage citizens in health—​ provider relationship in China, which with well-defined roles for patients and needs to be improved urgently. ­ providers—is still missing. China’s health sys- • “Two Routes to Engaging People to tem needs to become more patient-­ centered. Improve Health Outcomes and Restore In part because of rising incomes, rapid Trust” describes “individual” and “pub- urbanization, and increased demand for lic” routes to engaging patients and citi- health services, the Chinese population has zens in health and explains why this high expectations that health system reforms chapter focuses on the latter. will improve service delivery performance. • “Recomme n d ation s: Stre ng the ning It is important to meet these expectations: Patient Engagement in the Patient- public dissatisfaction with the health system Provider Relationship” outlines the con- has sometimes led to violence toward provid- cepts and core action areas for the ers (Chen 2012; Yuan 2012). Recent years individual route to engagement: (1) build- have shown an increasing tendency toward ing health literacy; (2) strengthening self-­ medical disputes in China (China Medical management practices; (3) improving Tribune 2012, 2013; CMDA 2013; Hesketh shared decision making; and (4) using and others 2012; Moore 2012); of these, information and communications technol- roughly a third of the medical disputes caused ogy to strengthen patient engagement.3 d i rec t i nju ries to med ical person nel (Guangzhou Daily 2014). The current The core action areas presented in this patient-physician relationship needs to be chapter complement, build on, and ultimately improved, in particular to avoid violence tar- reinforce each other. For example, shared geting doctors. decision making cannot take place without a Outbursts of anger and frustration are basic level of health literacy among patients— thought to stem from poor care, medical which in turn is linked to and cultivates a errors, and exorbitant costs, but relatively few certain confidence in the patient’s ability to studies examine the direct causes that underlie manage his or her own health. This experi- patient dissatisfaction and ensuing conflict. ence is critical to shaping the patient’s ability A 2010 report by the NHFPC Center for 134 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Health Statistics found that only 59 percent of cash, and gifts received as part of hongbao patients were satisfied with outpatient care from patients. and 55.8 percent with inpatient care (Center Further, hospitals were perceived as system- for Health Statistics 2010). The leading rea- atically refusing care to poor patients and finan- sons for dissatisfaction with doctors included cially devastating families whose members are poor attitudes (51 percent), short consultation suffering from prolonged illness, while rudi- times and lack of effort (43 percent), and over- mentary health insurance schemes alongside an prescription of unnecessary medication or underdeveloped legal infrastructure provided exams (23 percent). Dissatisfaction with few options for the sick who could not afford nurses was due to poor attitude (78 percent), health services (Yip and others 2012). poor nursing skills (24.7 percent), and unpro- Patient-physician trust is an implicit, fun- fessional behavior (23 percent). damental building block of medicine and of Echoing these findings, Sylvia and others achieving better health outcomes. A physi- (2015) documented average patient-physi- cian’s trust of his patients and a patient’s cian interaction time of 7.2 minutes in rural trust of his physician are inherently related, clinics, with half the time spent filling pre- and both are crucial for health care partner- scriptions. On average, clinicians spent only ships. Reciprocal trust establishes a moral 1.6 minutes consulting with patients, in dimension to healing that is related to, but spite of long wait times. Liang and Bao also distinct from, the biomedical aspects of (2012) found that, in Shanghai, patients on eradicating disease. average wait for 13 and 16 minutes for out- Because satisfaction is premised on expec- patient and inpatient registration, respec- tations—the greater the discrepancy between tively; after that, they need to wait an addi- the perceived service and prior expectations, tional 30 minutes to be seen by a doctor. the greater the patient dissatisfaction (Linder- Patients at community health centers Pelz 1982)—policy makers should strive to (CHCs) were more likely to feel satisfied understand and guide the financial, cogni- with the convenience, waiting time, and tive, and emotional expectations of both communication with doctors but less likely patients and doctors. In this context, the to feel satisfied with medical charges, drug findings of the Fifth National Health costs, and medical equipment (Tang and Survey—which report that 76.5 ­ percent of others 2013). From a positive perspective, outpatients and 67 percent of inpatients were the Fifth National Health Services Survey satisfied with their care-­ seeking experiences shows that 76.5 p ­ ercent of the outpatient (NHFPC 2015)—are encouraging. and 67 percent of the inpatient patients were satisfied with their services and experiences (NHFPC 2015). Two Routes to Engage People to Tucker and others (2015) studied incidents Improve Health Outcomes and of conflict at seven hospitals in Guangdong province and found that patient perceptions Restore Trust of injustice stemmed from costs of care A vast body of literature points to how (box 4.1), commercialization of medicine, and strengthened patient and citizen engagement conflicts of interest. Physicians’ intent to heal in health can improve the quality of their and cure was perceived to be compromised by interactions and suggests that satisfaction a wide range of nonsalary incentives, includ- and health outcomes can be improved ing indirect favors and all-expenses-paid trips through interventions that delineate and to tourist sites and conferences; incentives enhance the roles of both provider and from hospitals and clinical departments to patient in joint production of good health. generate revenue; direct cash payments from Patients are thus not just consumers of health pharmaceutical companies based on the num- services but are also empowered to act in ber of branded drugs prescribed; and favors, ways that influence their own health. L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 135 BOX 4.1  Understanding citizen mistrust: Perspectives from patients and providers One study implemented at seven hospitals in systematic, while patients’ information is disorga- Guangdong province recorded 25,000 medical dis- nized. This information inequality can cause many putes in 2013 and found that the origins of patient- conflicts.”—Physician physician mistrust were rooted in strong perceptions • “[Health professional education] is just taught of injustice. Patients felt that drug costs and overall according to the book line by line; it’s very rigid medical costs were inflated and that clinical decisions and dogmatic. For example, patient-doctor com- about diagnostic tests and drug prescriptions were munication isn’t sufficient. Actually at the bedside, skewed toward maximizing revenue instead of improv- we learn a lot of these kinds of communication ing outcomes. skills. But the kinds of communication skills we Several of the remarks by patients and medical person- were taught in school are not the kinds of skills we nel are illustrative: can apply.”—Nurse Physicians also perceived injustices within the • “Now everything is guided by economics. For physi- cians, hospital salaries can’t come close to matching medical system, pointing to intense workloads money from kickbacks and commissions. Maybe (for example, seeing 50 outpatients within a because his wallet grows, he is willing to engage in four-hour outpatient clinic shift) and pressures practices that violate his own professional ethics so from within the hospital to generate revenue in that he can increase his own profits.”—Patient the face of low salaries, high patient expecta- • “You will find that when some patients see the tions, and sensationalist reports from mass doctor, they carry with them an extremely distrust- media. Physicians also noted that the training ful, hostile, and negative tone of speaking. These system produced a limited number of subspecial- patients lay it out: newspapers are talking about the ized experts clustered in urban tertiary care set- violence, hospitals are not to be trusted, and doc- tings (transiently evaluating a large volume of tors are the worst. If you see a hundred patients and patients) rather than a larger number of primary only see one distrustful patient, it matters. It slowly influences your perception of patients.”—Physician care doctors (longitudinally caring for a smaller • “The biggest problem is that the information volume of patients). Finally, medical training pri- between patients and physicians is asymmetric. oritized technical biomedical competence over Physicians have too much information and patients caregiving (cognitive, behavioral, emotional, and have too little. And physicians’ ­information is very moral support) and empathy for patients. Source: Tucker and others 2015. This empowerment can help dislodge a to date have focused on adding security staff deeply entrenched culture or mindset of help- to hospitals, outlawing hongbao, and related lessness and perceptions of lack of control stopgap measures. that lead to the range of adverse outcomes These measures seek to address the symp- observed in China, from patient dissatisfac- toms but not the fundamental issues that tion to violent conflict as a last resort in the underlie mistrust of health services. As seen most extreme cases. Box 4.2 defines the key earlier, drivers of cost and commercialization concepts of empowerment, engagement, and of health care, overuse and underuse of coproduction. ­ services, and suboptimal clinical quality all Acknowledging the worsening of patient- contribute to dissatisfaction with the system. physician relations, national Chinese govern- Chapters 3 (on quality of care) and 6 (on pro- ment leaders have identified citizen mistrust vider incentives) propose solutions for these as a major problem and sounded calls to problems. A case can be made that addressing action (Hesketh and others 2012; Lancet perverse incentives to overprovide care may 2012, 2014; Zhong 2015). Policy responses be a first step to improving provider-patient 136 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A BOX 4.2  Defining empowerment, engagement, and coproduction of health The terms “empowerment,” “engagement,” and participate in strategic decision making on how, “coproduction” are often used interchangeably to where, and on what their health resources should describe policies or interventions that seek to achieve be spent. Engagement is also related to the com- such goals, but in reality they represent distinct, if munity’s capacity to self-organize and to generate overlapping, strategies: changes in its environment. • Coproduction is about care that is delivered in • Empowerment is about supporting people and com- an equal and reciprocal relationship between munities to take control of their own health needs. It (a) the clinical and nonclinical professionals and results in, for example, the uptake of healthier behav- (b) the individuals using care services, as well iors, the ability of people to self-manage their ill- as their families, caregivers, and communities. nesses, and changes in people’s living environments. ­ Coproduction therefore goes beyond models of • Engagement is about people and communities engagement, because it implies a long-term rela- being involved in the design, planning, and deliv- tionship between people, providers, and health ery of health services. This enables them to make systems where information, decision making, and choices about care and treatment options or to service delivery become shared. Source: WHO 2015b, 22. engagement. But much more will still need to The WHO strategy on People-Centered and be done to fix patient-provider interaction and Integrated Health Services states that “at the build trust in the long run. most fundamental level, it is people themselves Broadly, interventions to enhance patient who spend the most time living with and engagement can be organized at two levels responding to their own health needs and will (box 4.3): engaging people as (a) individual be the ones making choices regarding health patients—the “individual” route—or (b) as behaviors and their ability to self-care or care members of the public—the “public” route. for their dependents. Since people themselves These two approaches can also be described tend to know better the motivations that drive as “patient involvement” (which refers to these behaviors, people-centered care cannot people “making decisions about their own be provided without engaging them at a per- health”) and “public involvement” (which sonal level” (WHO 2015b, 22). engages “members of the public in strategic Broadly, at the individual level, patient decisions about health services and policy at engagement encompasses two key aspects: a local or national level”) (Florin and Dixon empowerment and activation. Patients need 2004). The remainder of this section outlines to be empowered with knowledge and infor- the differences between the two approaches mation to make sound health care choices, and presents the rationale for the core actions ranging from generating changes in behaviors, recommended later in the chapter. selecting providers to seek services, and weighing the costs and benefits of surgical ver- ­ sus nonsurgical treatment options to accessing The Individual Route: Engaging Patients timely and effective complaint resolution in the Micro Context of the Patient- mechanisms and addressing potential causes Provider Relationship of ill health in their living environments. Once The first route to engaging patients in health equipped with essential information, patients addresses the relationship between the patient can be “activated” to participate in various and the medical provider as individuals. activities for managing their health and health L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 137 BOX 4.3  Individual and public routes to patient engagement Efforts to engage people in health practice and Public routes to engagement policy can involve them as individual patients or Public participation initiatives associated with health as members of the public. Individual and public endeavor to make the field more citizen-centered in engagement can require quite different things from much the same way that health care provision aims participants. Jones and others (2004) suggest that to be patient-centered. Public engagement exercises patient involvement is essentially “private participa- take a variety of forms: for example, they provide tion” in which individuals promote and protect their participants with information, ask for their opinion, own preferences and values. Public involvement, in and incorporate them as active partners within pol- the context of treatment services and public health, icy formation (Arnstein 1969; Charles and DeMaio can request citizens to “put aside their particularis- 1993; Feingold 1977). tic preferences . . . and participate for the common Several countries have taken the ambitious good” (Tenbensel 2010). approach of involving patients and the wider pub- lic in different levels of the decision-making pro- Individual routes to engagement cess, including health services planning and, at the In the clinic, efforts to engage individuals in their national level, health care policies. Examples include own health are evident in initiatives that pro- Germany’s Institute for Quality and Efficiency in mote person- or patient-centered care as a way of Health Care (https://www.iqwig.de/en), the United refocusing of medicine’s regard for the patient’s “­ Kingdom’s National Institute for Health and Care viewpoint” (Laine and Davidoff 1996). Excellence (https://www.nice.org.uk), and the United The Institute of Medicine has described patient- States’ Patient-Centered Outcomes Research Institute centered care as one of six areas that are central to (http://www.pcori.org). health improvement efforts (IOM 2001). It defines In Australia, the Consumer Health Forum such care as that which “is respectful of and respon- (https://www.chf.org.au) acts as a national voice sive to individual patient preferences, needs, and and collaborative for health consumers. Its mission values and [ensures] that patient values guide all includes advocacy, research, issue identification, and clinical decisions” (IOM 2001). consumer representation related to a large array of Others are more specific regarding the ethi- themes including health literacy, consumer-centered cal content of person-centered care, stating that it regulations and policy making, quality and patient “emphasizes patient autonomy, informed consent, safety, access to information, new technologies, and and empowerment” (Edwards and Elwyn 2009). equitable access to care. Source: Williamson 2014. care, addressing risky behaviors, and safe- individual level. These include building health guarding their living environment. literacy, strengthening self-management, and Health providers play a vital role in patient improving shared decision making with engagement by providing information about enhanced use of technology. These three treatment options, explaining the potential approaches represent recommended core risks and benefits of each option, encourag- action areas for strengthening patient engage- ing patients to deliberate on and express their ment and are described in detail in the preferences, and developing plans for long- “Recommendations” section. A substantial term self-management. Patient engagement in body of evidence highlights the effects of health care thus requires change and effort patient engagement approaches, with benefits from both providers and patients. accruing in the form of improved quality of Health systems use a variety of approaches care, appropriate decisions, and good health to empower and activate patients at the outcomes (box 4.4). 138 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A BOX 4.4  Effects of patient engagement strategies A Commonwealth Fund survey of 11 Organisation multiple settings (Coulter and Collins 2011; Sta- for Economic Co-operation and Development cey and others 2011) and may also successfully (OECD) countries found that engaging patients can increase the use of less-invasive treatments that are improve quality and patient experience, reduce med- often also less expensive (Deyo and others 2000; ical errors, encourage compliance, and ultimately Kennedy and others 2002; Morgan and others lead to better health outcomes with lower cost 2000; National Voices 2014; Wennberg 2010). (Osborn and Squires 2012). Useful patient engage- • Electronic health (e-health) and mobile health ment strategies include the following: (m-health) platforms can be powerful tools for supporting strategies for strengthening patient • Self-management interventions improve not only engagement (Bennett and others 2010; Bove patient knowledge, coping skills, and confidence and others 2013; Chen and others 2008; Nolan to manage chronic illnesses, especially among the and others 2011; Rhee and others 2014; Wong elderly, but also intermediate health outcomes; in and others 2013). In particular, they promote some cases, they even reduce hospitalization rates patient engagement by improving patient knowl- (Picker 2010). edge, increasing patients’ willingness to partici- • Shared decision making has the potential to pate in their own care process, enhancing account- improve patient satisfaction and health care in ability, and enabling self-monitoring. The Public Route: The Macro the clinical environment and individual Environment’s Role at the Local and patient-provider relationships to include the National Levels ecology of individuals, families, communi- ties, and organizations: The first strategy Targeting patient-provider relationships at refers to initiatives such as the WHO’s the micro level may not be sufficient by itself Healthy Cities and Villages; the spirit under- to improve patient confidence and trust in lying these initiatives is that all societal the health system. This is because these rela- forces can be mobilized to create conditions tionships rest in a macro environment influ- that enable people to live a healthy life. The enced and shaped by nationwide and soci- second strategy is informed by insights from etywide forces. As well as being seekers and recent research in behavioral economics. It consumers of health services, patients are entails “nudges” and messages embedded in also citizens of nations and states, and as the physical and social environment that are such can be engaged to exert influence used to cue people to adopt healthier behav- on the broader policy environment that iors. The next section describes each action shapes the health and well-being of the area in detail. ­ societies they live in. This is the “public” route to engagement. Instead of focusing on how patients could Recommendations: be engaged as citizens to improve participa- Strengthening Patient tion in civic processes shaping health ser- Engagement in the Patient- vices in China, we focus on how the macro- environment can be shaped by policies and Provider Relationship collective activities to support better indi- Four core action areas and corresponding vidual engagement in health. Two “public implementation strategies can strengthen route” engagement strategies are recom- c itizen engagement in support of PCIC ­ mended that broaden the perspective from (table 4.1). Core action areas 1–4 build patient L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 139 confidence and trust through patient empow- patients as citizens—and to create an envi- erment and engagement in the individual ronment that sustains and nurtures civic par- sphere of the patient-provider relationship. ticipation in Chinese society at large to reduce “participatory deficit” (W HO • Health literacy (core area 1) can empower 2015a)—may well be long-term goals for patients with information needed to China, but these are topics beyond the scope improve health behaviors and change of a health sector report such as this. To treat expectations about what is acceptable them adequately would require delving into treatment, including harm. It also shapes the political economy of Chinese governance patient expectations for interactions with institutions as well as governance-society medical professionals. relations. • Self-management and shared decision making (core areas 2 and 3) cultivate a Core Action Area 1: Health Literacy culture of comanaging health with the provider. Together, these interventions Health literacy is the ability to understand help improve the patient-provider relation- and act upon health information so that peo- ship, build trust, and empower patients to ple have greater motivation and ability to break patterns of tension and prevent con- control their health. The concept entails the flict. E-health and m-health solutions can ability to understand basic health knowledge facilitate implementation of these core and to use this to make health-related action areas. decisions. • A supportive macro environment (core Health literacy is essential to good health action 4) focuses on broader, collective and fundamental to public health. If people activities to foster an enabling environ- cannot obtain, understand, and use health ment of citizen and patient engagement. information, they will not be able to take care of themselves effectively, to navigate the The recommendations focus mainly on the health system without difficulty, or to make individual route to engagement. Certainly, appropriate health choices for their own, the macro environment affects patients’ indi- their family’s, and their community’s health. vidual experiences as coproducers of health. Adults with limited health literacy report And proposing strategies to define and having less knowledge about their medical strengthen the roles and responsibilities of conditions and treatments, worse health TABLE 4.1  Four core citizen engagement action areas and implementation strategies Core action areas Implementation strategies 1: Health literacy ·   Improve citizen understanding of evidence-based care, the importance of health-related behaviors, and preventive practices ·   Launch public media campaigns to encourage health promotion and prevention activities 2: Self-management practices ·   Train health providers to support and facilitate self-management by patients ·   Educate and support patients on how to self-manage 3: Shared decision making ·   Cultivate an expectation of patient involvement in decisions about their health care ·   Develop and promote use of decision-aid tools at health facilities 4:  Supportive macro environment ·   Improve macro environment for health promotion: develop Healthy Cities (and for citizen patient engagement in Healthy Villages)a health promotion and improvement ·  Create environmental “nudges” to improve health choices a. WHO 1998. Also see “Healthy Cities,” Health Promotion Programmes, World Health Organization Western Pacific Region, http://www.wpro​ .who.int/health_promotion/about/healthy_cities/en/. 140 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A status, less understanding and use of preven- positive role of health education and commu- tive services, and a higher rate of hospitaliza- nication in developing competencies for differ- tion and use of emergency rooms (Berkman ent forms of health action that benefit the and others 2011; Kindig, Panzer, and health of individuals and the population. Nielsen-Bohlman 2004). Surprisingly, as Particularly, the WHO Commission on Social many as half of all adults in the United States Determinants of Health proposed that “health have difficulty understanding and acting literacy implies the achievement of a level of upon health information—a problem that knowledge, personal skills, and confidence to leads to ineffective care (Kindig, Panzer, and take action to improve personal and commu- Nielsen-Bohlman 2004). nity health by changing personal lifestyles and Nutbeam (2008a, 2008b) distinguished living conditions” (CSDH 2007). Gaining two perspectives on health literacy: health lit- health literacy as an asset could fundamentally eracy as a risk factor and health literacy as an address some of the social determinants of asset. These two perspectives differ subtly in health outside the narrowly defined health their approach to the same concept. care system. Health literacy as a risk factor. Under­ Clearly the two approaches are distinctive standing health literacy as a risk factor leads in their clinical versus public health perspec- to a focus on how to mitigate the negative tives, but both are valuable and complemen- effects of low health literacy on health-related tary for guiding policies to promote health lit- behaviors and health outcomes. To this end, eracy. They imply different strategies in the Institute of Medicine defines health liter- response to low literacy that may supplement acy as “the degree to which individuals have each other. In addition to improving access to the capacity to obtain, process, and under- effective school education and providing adult stand basic health information and services education to targeted populations with low needed to make appropriate health decisions” basic literacy (CSDH 2007), health systems (Kindig, Panzer, and Nielsen-Bohlman must also enhance the quality of health com- 2004). Research following this theory has munications and education and provide linked health literacy to a range of health greater support and tailored information to behaviors and outcomes, including effective increase functional literacy to understand and management of chronic disease, compliance use health information for managing health with medication and other health advice, and and diseases (Coulter and Ellins 2007). participation in health and screening programs. Strategy 1: Improve citizen understanding Health illiteracy can also be a demand- of evidence-based care, the importance of side barrier: in particular, low health literacy health-related behaviors, and preventive among the poor and among ethnic or racial practices minority groups is associated with poorer While health literacy is the outcome of a health status, more hospital admissions, more complex array of individual, social, and eco- drug and treatment errors, less use of preven- nomic processes, the health system is a criti- tive services, and poorer adherence to treat- cal intervention point. Patients look to health ment recommendations (Berkman and others providers for information and education on 2011; Kindig, Panzer, and Nielsen-Bohlman how to manage illnesses and long-term con- 2004). Lower health literacy among seniors is ditions. Beyond the information acquired associated with higher mortality (Berkman through one-on-one patient-provider interac- and others 2011). Tackling gaps in health lit- tions, many countries have implemented eracy is considered an important element in f ormal educational approaches to target ­ optimizing clinical effectiveness and reducing d isadvantaged population groups. These ­ health inequities. approaches include training courses for small Health literacy as an asset. By contrast, the groups, colleges, and adult education institu- health-literacy-as-asset approach promotes the tions as well as one-on-one counseling. L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 141 The United Arab Emirates and the United by verbal information from clinicians and Kingdom provide successful examples. by web-based interventions as part of an “Skilled for Health,” a national program run educational program. in part by the U.K. Department of Health, • Newspapers, magazines, and broadcast aimed to help people improve their health media are important vehicles in health while boosting their language, literacy, and education campaigns across the world. numeracy skills. The program was intended Media publicity can be a key component, both to provide useful information and skills for example, to discourage smoking; and to improve people’s confidence to look encourage use of folic acid among after their health. The program targeted par- pregnant women (the Netherlands); raise ­ ticipants’ health knowledge in the areas of awareness of excessive and rising hysterec- healthy eating, smoking, exercise, drinking, tomy rates (Switzerland); diminish stigma and looking after their mental health. associated with depression (United Educational sessions on a range of health Kingdom); promote uptake of immuniza- topics such as healthy eating, exercise, and tion and cancer screening; educate about first aid were delivered to people in deprived human immunodeficiency virus (HIV) areas (ContinYou 2010). risk; and publicize appropriate care for In the United Arab Emirates, student suspected myocardial infarction. “ambassadors” at universities were trained in • Social marketing is used by government the basics of genetic screening and then departments and health authorities to encouraged to spread the word to their peers achieve specific behavioral goals for a about the importance of being screened social good (French and Blair-Stevens (Laurance and others 2014). 2007). It typically involves a systematic approach to health promotion using tried Strategy 2: Launch public media campaigns and tested techniques that are informed by to encourage health promotion and commercial insights (for example, on seg- prevention activities mentation and from marketing theory) Other strategies tackle health literacy across and theories of behavioral change. Such whole populations and focus on improving marketing interventions aim to help peo- the provision of high-quality health informa- ple make healthy choices, adopt healthier tion. Some media-based campaigns focus on lifestyles, or make better use of health both providers and individuals: in Canada, services. They have targeted healthy eat- ­ for example, the National Literacy and ing, substance misuse, physical activity, Health Program promotes awareness among workplace health, and well-being. For health professionals and patients of the links example, social marketing played a key between health literacy and health. Many role in a Chinese campaign to prevent and media-based campaigns use printed materi- control hepatitis B (box 4.5). als, videos, websites, and formal and infor- mal courses. Quality health information that In China, messaging should focus not only is timely, relevant, reliable, and easy to under- on changing expectations about medications, stand is an essential component of any strat- intravenous therapy, and other diagnostics egy to support self-care, shared decision and therapeutics, but also on making citizens making, self-management of long-term con- aware of the harm caused by overuse and ditions, and health promotion.4 The follow- misuse of treatments. A series of messages ing are typical in public media campaigns for and public education efforts should be health education: launched to change public perceptions regarding medications, procedures, and clini- • Informational materials , provided at cal services. It would require a continuous, health facilities or electronically, can be multiyear, multichannel communication pro- tailored to the individual and reinforced gram, and ideally would use the energies of 142 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A health care professionals as well as civil soci- citizens could make in their personal lives ety agencies. The goal would be to help peo- that would lead to a healthier future (for ple understand the salient features of good, example, for smokers this would be smoking evidence-based care. cessation; for alcohol drinkers this would be However, it would also be best for cam- reducing their intake; for overweight or dia- paign planners to draw on research on why betic patients this would be to walk at least a and how people understand and use informa- mile a day). Messages advocating these tion in choosing to seek care in China. In health-enhancing behaviors would be assem- particular, this education effort would need bled into national or provincial campaigns to to decrease the nonscientific overdependence get every citizen to engage in one or more of on procedures such as intravenous infusions, them. Given that people do not have equal medications, and hospital visits and admis- access to information, complementary and sions that the current volume-based payment more targeted interventions may be needed system has encouraged. for low-income, elderly, and ethnic popula- Further, a national appeal to the public to tion groups. engage in the collective pursuit of health One example that could serve as a model could be explored. This would start with pro- for China is the Million Hearts Campaign in ducing a technical review of three to five the United States (box 4.6), a national initia- major evidence-based changes that individual tive that set an ambitious goal to prevent BOX 4.5  Social marketing in China: Prevention and control of hepatitis B China’s anti – hepatitis B campaig n has been and presenting awards to outstanding pieces described as an excellent example of social market- (Cheng and Chan 2009). The Chinese government ing whose design and implementation maximized played a major role in this nationwide campaign, effectiveness as a result of the ample attention paid which was cosponsored by the China Foundation to the social, cultural, and regulatory context. for Hepatitis Prevention and Control and the The first public service advertisement (PSA) Information Office of the Ministry of Health, was aired by a Chinese television station in 1986, with donations of expertise from McCann Health and since then the Chinese government and media China and airtime and space from many media have been hosting annual national PSA campaigns outlets. Source: Cheng, Kotler, and Lee 2011. BOX 4.6  The Million Hearts Campaign The Million Hearts Campaign rallies communities, committed to the campaign goal and to specific health care professionals, health systems, nonprofit activities; (b) promoting optimal care with the ABCS organizations, federal agencies, and private sector strategy (aspirin when appropriate, blood pressure organizations around a common goal: preventing control, cholesterol management, and smoking ces- 1 million heart attacks and strokes by 2017.a A small sation) has achieved some early success. set of changes have served as targeted interventions The campaign has also helped to pass laws that to achieve this goal (figure B4.6.1). create a healthier environment (for example, on So far, the campaign’s results include the smoke-free zones, a sodium reduction program in ­ following: (a) More than 100 partners have formally communities, and elimination of trans fats). (Box continued next page) L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 143 BOX 4.6  The Million Hearts Campaign (continued) FIGURE B4.6.1  Million Hearts targets Million Hearts® Targets By 2017 . . . The number of Reduce smoking American smokers has declined from 26% to 24% NA Americans consume less Reduce sodium than 2,900 milligrams of intake sodium each day Eliminate trans TRANS Americans do not fat intake FAT consume any arti cal trans fat Optimizing Care in the Clinical Setting Aspirin use when appropriate . . . Focus on ABCS Of the people who have had a heart attack or stroke, 70% are taking aspirin Blood pressure control Of the people who have Use health tools hypertension, 70% have adequately and technology controlled blood pressure Cholesterol management Of the people who have high levels of bad cholesterol, 70% are Innovate in care managing it e ectively delivery Cholesterol management Of current smokers, 70% get counseling and/or medications to help them quit Million Hearts® promotes clinical and population-wide targets for the ABCS. The 70% values shown here are clinical targets for people engaged in the health care system. For the U.S. population as a whole, the target is 65% for the ABCS. Source: “Infographic: Million Hearts Targets,” About Million Hearts, https://millionhearts.hhs.gov/about-million-hearts/targets.html. ©U.S. Department of Health and Human Services (HHS). Note: “ABCS” refers to a strategy comprising the following: aspirin when appropriate, blood pressure control, cholesterol management, and smoking cessation. a. Million Hearts is an initiative co-led by the Centers for Disease Control and Prevention (CDC) and the Centers for Medicare & Medicaid Services (CMS). For more information, see the Million Hearts website: https://millionhearts.hhs.gov/about-million-hearts/index.html. 144 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A 1 million heart attacks and strokes by 2017 health services; make more informed by improving access to effective care; raising decisions about office visits, medication, and ­ the quality of care through the ABCS strat- procedures; and practice behaviors that help egy (aspirin, blood pressure, cholesterol, control their conditions. smoking cessation); focusing clinical atten- The prevalence of chronic disease and the tion on the prevention of heart attack and scope of its consequences have created a dra- stroke; activating the public to follow a heart- matically new situation in health care. As healthy lifestyle; and improving prescription Holman and Lorig (2004) describe, patients, and adherence to appropriate medications health professionals, and the health service under ABCS. must now play new roles: S c o t l a n d ’s o n g o i n g E a r l y Ye a r s Collaborative, launched in 2012, is another • The patient—who must be responsible for example of this kind of campaign. 5 In this daily management, behavior changes, collaborative, Scotland is asking all parents emotional adjustments, and accurate nationwide to read their children a bedtime reporting of disease trends and tempos— story each night, which has been shown to becomes the principal caregiver. Expressed improve future literacy and educational in economic terms, health is the product of attainment. China could use these models as health care, and the patient, as a principal examples while tailoring the campaign to the caregiver, is a producer of health (Hart specific Chinese context. 1995). As in any production system, a pro- ducer must be knowledgeable about the product and skilled in the production Core Action Area 2: Self-Management process. Practices • The health professionals, in addition to Barring self-care for instances of short-term being professional advisers and partners in minor illness (such as a cold or other com- the design and conduct of medical man- mon viral infection), much self-care across agement, become teachers in developing the world today consists of the day-to-day the patient’s management skills. In the management of chronic illnesses such as present system, physicians, nurses, and asthma, arthritis, and diabetes. Strictly public health workers are not trained for speaking, people suffering from these condi- this role. tions “self-manage” most of the time: they • The health service becomes the organizer manage their daily lives and cope with the and financial supporter of the new roles effects of their conditions as best they can, for the patient and health professionals, mostly without help from their health care focusing on assuring continuity and inte- providers. gration of care. More technically, self-management is defined as “the individual’s ability to manage All approaches to self-management include symptoms, treatment, physical and social careful elicitation of the patient’s view of his consequences, and lifestyle changes inherent or her problems, concerns, values, and prefer- in living with a chronic condition” (Barlow ences; sensitive sharing of relevant evidence- and others 2002, 178). It is also about based information by health professionals; enabling people “to make informed choices, and discussion to find common ground. For to adapt new perspectives and generic skills patients to self-manage their conditions, they that can be applied to new problems as they and their families need to be systematically arise, to practice new health behaviors, and educated about their conditions, how to to maintain or regain emotional stability” monitor them, and how to incorporate (Lorig 1993, 11). By promoting systems for healthy behaviors into their lifestyles. patient self-management, health systems can When people with chronic diseases seek empower individuals to reduce their use of professional advice, they need appropriate L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 145 help and support to enhance their self-man- • Assess knowledge, behaviors, and confi- agement skills. For example, people with dence routinely asthma must know when to use their inhal- • Advise on the basis of scientific evidence ers; people with diabetes must monitor their and current information blood glucose levels; arthritis patients must • Agree on goals and treatment plan for learn to cope with pain and how to amelio- improving self-management rate it, when possible; and, with self-­ • Assist in overcoming barriers management, chronic obstructive pulmonary • Arrange helpful services. disease (COPD) sufferers can maximize their lung function and improve the quality of Training for health professionals typically their lives (box 4.7). seeks to ensure that, at a minimum, they will Strategy 1: Train health providers to • Inform the patient about the disease, treat- support and facilitate self-management ment, or management options; Cultivating appropriate self-management • Educate the patient about effective practices among patients often requires a cul- self-management; ture shift among health care practitioners. • Train the patient on skills such as how to Professionals are urged to stop believing carry out technical tasks such as testing that their goal is to increase the patient’s blood glucose levels for diabetics, monitor- compliance with what they choose to recom- ing peak flow for asthmatics, and so on; mend and to instead increase the patient’s • Advise on behavioral change—how to mod- capacity to make informed decisions. The ify existing behaviors or adopt new ones; Five A’s Paradigm summarizes this approach • Challenge unhelpful beliefs, including (Glasgow, Emont, and Miller 2006): beliefs about the causes of illness; and BOX 4.7  Self-management of COPD in the U.S. veterans population Management of chronic obstructive pulmonary disease pnea and how functional they can be. The self-man- (COPD) is complicated because patients with this dis- agement module is designed around the principle that ease tend to suffer from many other chronic conditions many of the goals of pulmonary rehabilitation can be that affect their quality of life. A major challenge is to accomplished remotely and that self-management is help patients avoid acute exacerbations that often lead part of this and helps sustain desired behaviors and to hospital admissions. In the military veteran popu- the benefits of the program. This, in turn, reduces the lation covered by the Veterans Health Administration likelihood of the need for hospital admissions. (VHA) in the United States, for example, COPD is To design the module, the VHA analyzed stud- the fourth most common reason for hospitalization. ies and collected advice from pulmonologists across Pulmonary rehabilitation, which includes both exercise the United States to create a coherent packet for care and education, helps patients manage their medications teams to facilitate communication of the best and and compensate for their disabilities. Two problems most important strategies. The module focuses on are that some patients must travel long distances for three areas: improving exercise tolerance and patient rehabilitation and that even among patients who find health status, managing symptoms, and manag- it more convenient to participate, behaviors learned in ing or reducing exacerbations. Within these areas, the program are not always sustained. the module addresses strategies such as medication, To address these problems, the VHA developed a exercise, and smoking cessation. Overall, the tool self-management module “to help COPD patients help is intended to allow patients to work more closely themselves” by considering the many elements that with their physicians to maximize lung function and improve how patients with COPD perceive their dys- improve the quality of their lives. Sources: Ali and Li 2015; Basu 2014. 146 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A • Counsel patients on managing emotions— p e er - le d e du c at ion c ou r s e s i n s el f-­ how to cope with the effect of their illness management, in which people with chronic and its effect on their emotions, including conditions learn from other people with the how to deal with anxiety and depression. same chronic conditions (Lorig and others 2001). These courses are often run by volun- Training in communication, teamwork, tary organizations. Educational courses fol- and relationship-building skills should be lowing this format have been used across a embedded in medical school curricula, post- wide variety of settings, including Australia, graduate clinical training, and continuing Barbados, Chile, Denmark, Japan, the medical education. The widely used Calgary- Republic of Korea, Peru, the United Cambridge framework divides a health care Kingdom, and the United States. consultation into five stages: initiating the ses- Participants learn how to set goals and sion, gathering information, physical exami- make action plans; solve problems; develop nation, explanation and planning, and closing their communication skills; manage their the session, with a list of tasks that must be emotions; pace their daily activities; manage accomplished in each (Kurtz and others 2003). relationships with family, friends, and work Providers’ ability to communicate compe- colleagues; communicate with health and tently with patients should become a condition social care professionals; find other health for qualification to practice, and due attention care resources in the community; understand should be paid to lessons from research on the importance of exercise and healthy eat- interpersonal and communication skills. ing; and manage fatigue, sleep, pain, anger, These skills can be learned and improved. For and depression. example, trainees can be taught how to New technologies have been adopted express empathy (Bonvicini and others 2009), to create interactive approaches delivered how to break bad news (Makoul and others electronically. In the United Kingdom, for 2010), and how to practice shared decision example, the Expert Patient Program is an making (Bieber and others 2009). ­ internet-based resource with e-mail remind- Providers can be trained to use decision ers (Lorig and others 2008). Web-based aids6 (further discussed below) and to be ready packages that combine health information to answer questions, especially when commu- with social support, decision support, or nicating with patients about uncertainty, the behavioral change support have been devel- relative risks of different treatment options, oped for people with chronic diseases such and the specific time frames that define risks as asthma, diabetes, eating disorders, and and outcomes. An evidence-based educational urinary incontinence. approach, the Flinders Program, is oriented to In the United States, health coaching by chronic-care management. It seeks to assess telephone (providing people with advice and and improve the relationships between provid- support over the phone as a component of ers and patients that will lead to patients’ disease-management systems) and telecare actively monitoring their conditions while technologies (including devices to transmit ­ promoting healthy lifestyles (Horsburgh and information over phone lines to sophisticated others 2010).7 The Flinders Program contains machines that monitor the patient’s vital a series of training modules to enhance pro- signs) are also used (Audit Commission viders’ knowledge of chronic-care manage- 2004; Rollnick, Miller, and Butler 2002). ment with a focus on communication skills. Giving patients access to their medical records—by either enabling them to read and Strategy 2: Educate and support patients on review these or by encouraging them to hold how to self-manage their own copy—can also increase a patient’s Instituting a culture of self-management confidence to self-manage. among patients requires education. A typical Self-management education works best format is short (usually six weekly sessions), when integrated into the primary and L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 147 secondary health care systems and when the correct diagnoses based on science, but also learning is reinforced by professionals. The that patients receive “the care they need, and most effective self-management programs are no less; the care they want, and no more” those that are longer, more intensive, and well (Coulter and Collins 2011, vii). Shared deci- integrated into the health system and where sion making reflects the extent to which the learning is reinforced by health profession- citizens feel empowered to engage in their ­ als during regular follow-up. Efforts should health care. focus on providing opportunities for patients There are compelling ethical and practical to develop practical skills and the confidence reasons to engage patients in making shared to manage their own health. Hands-on decisions about their health. Patients may have ­ participatory learning styles are better than expectations and preferences about treatments traditional didactic teaching (box 4.8). and health outcomes that differ from those of their health provider. Recognizing those expectations and preferences is vital to ensur- Core Action Area 3: Shared Decision ing responsive and respectful care. Studies Making show that providers consistently overestimate Shared decision making is a collaborative their ability to predict patients’ preferences. process in which providers and patients work In one study, doctors reported believing that together to identify problems, set priorities, 71 percent of patients with breast cancer establish goals, create treatment plans, and would rate keeping their breast tissue as a top solve issues. As such, it is the essential under- priority, whereas in reality only 7 percent of pinning for delivering truly people-centered patients said so (Lee and others 2010). care. Practicing shared decision making is a In another, informing patients about the trade- way to ensure not only that doctors make off associated with surgery for benign prostate BOX 4.8  Encouraging self-management of health: Examples from India and the United Kingdom The Year of Care in Diabetes in the United Kingdom avoid the need for more expensive help later on. The was a pilot program launched to actively involve community enables members to measure their men- diabetes patients in deciding, agreeing, and working tal health through tests and questionnaires, access on how their condition is managed. The core idea help on guided support programs, get individual live was to transform the annual review (which often therapy over a secure Skype-like connection, and just checks that particular tests have been carried track their progress. Although the focus is on self- out) into a genuinely collaborative consultation by management, the intervention incorporates elements encouraging patients to share information with their of health literacy as well. health care team about their concerns, their experi- The Seven-Day Mother and Baby Health Checklist, ence of living with diabetes, and any services or sup- developed by WHO and implemented in India, helps port they might need. Both the patient and the team mothers identify danger signs in the crucial first week then jointly agree on the priorities or goals and the after birth. Upon discharge from the health facility, actions to take in response to these. a health care worker explains the list to the mother. Another self-management resource in the United Texts and audio messages are sent by mobile phone to Kingdom is the Big White Wall, an online mental remind the mother to check the baby and herself for health community where members can find support danger signs. This intervention has elements of health managing their care from clinicians, family mem- literacy (education on what are the danger signs) and bers, and each other. The initiative provides mem- also develops the capacity for self-management (deter- bers with access to immediate support, which may mining when to seek professional help). Source: Laurance and others 2014. 148 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A enlargement led to a 40 percent reduction in consider their preferences and want to have the number of patients opting for surgery a say in treatment decisions (Coulter and (Wagner and others 1995). Surgery can ame- Magee 2003). For example, a U.K. National liorate urinary symptoms associated with the Health Service survey found that nearly half disease, but many informed patients would of hospital patients wanted more involve- rather forgo surgery to avoid postsurgical sex- ment in their treatment decisions. Providers ual dysfunction. should communicate to patients that they A Cochrane review found that, compared are expected to take an active role in their with usual care, decision aids increased health health care, and patients should understand knowledge, particularly when the decision aid that although they may lack technical provided detailed rather than simple informa- knowledge, they bring an equally important tion (Stacey and others 2011). Exposure to a form of expertise to the decision-making decision aid that displayed probabilities meant process. that patients more accurately gauged the risks Internationally, there are several exam- associated with health interventions. Exposure ples of how to improve patient involve- to a decision aid with explicit value clarifica- ment in health care processes at the facil- tion resulted in a higher proportion of patients ity level: choosing options that were congruent with their values. • In the United States, under the Program Decision aids were also found to improve of All-Inclusive Care for the Elderly patient-provider communication and to (PACE) model, patients and health care increase satisfaction with the decision and the teams collectively design and agree on the health care process. They reduced patients’ patients’ health goals (Ali 2015). Efforts decisional conflicts related to feeling unin- to improve patient-centeredness at the formed and unclear about their personal val- Beth Israel Deaconess Medical Center in ues. The Cochrane Review Group on con- the United States provide another exam- su mers and com mu n ication provides ple (box 4.9). continuous updates to effective interventions • In Shanghai , the family doctor system to enhance patient-provider communication encourages patients and families to jointly and patient engagement for achieving better set treatment goals with their providers, health outcomes. 8 The China Cochrane and monthly patient-satisfaction scores Center in West China Hospital, Sichuan track progress (Ma 2015). University, may expand its clinical reviews to • In Ge r m any , a core feature of the cover high-quality provider-patient interac- Gesundes Kinzigtal system is the joint tion using decision aids. setting and attainment of goals. Shared Just as cultivating the practice of self-­ decision-making tools augment this pro- management builds a patient’s sense of cess, while case managers support the empowerment, shared decision making too patients in changing their health condi- leads to a beneficial redistribution of power tions and behaviors (Hildebrandt and between patient and provider (Coulter 2011). others 2015; Nolte and others 2015). It can be achieved by changing the ethical and • In Denmark , the Integrated Effort for legal requirement of informed consent into a People Living with Chronic Diseases more active standard of informed patient (SI K S) projec t prioriti zes patients’ choice (Wennberg 2010). The strategies below involvement in developing their own outline possible steps toward this goal. treatment plans, setting goals through shared care plans, and providing feed- Strategy 1: Cultivate an expectation of back about whether these goals were met patient involvement in health care decisions in partnership with the care team (Nolte Surveys have found that about three-­ and others 2015; Runz-Jørgensen and quarters of all patients expect clinicians to Frølich 2015). L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 149 BOX 4.9  Improving patient involvement at Beth Israel Deaconess Medical Center, United States At the Beth Israel Deaconess Medical Center in e-mail, and request appointments and prescription Boston, efforts to improve patient-centeredness refills. A “trigger response” system encouraged fam- incorporated elements of shared decision making, ily members who have a serious concern about the patient literacy, and self-management. A patient- patient to request a review by the care team. Patient care committee was established, responsible for set- education was conducted about patients’ right to ting up patient and family advisory councils. The see test results, read the medical notes made by mission was to make sure that the patient’s voice their physicians, and communicate with their phy- is heard, to improve communication, and to foster sicians. Strategies employed included dissemination innovations that enhance the patient’s experience of information packages and provision of support to of care. Patient and family advisers participated in ­foreign-language speakers. focus groups and meetings about proposed design Finally, training and education of staff mem- changes. bers about building a patient-centered environment Beth Deaconess also developed a web-based began at recruitment, when they were asked to work portal that allowed patients to see their test results, ­ through patient-oriented scenarios to learn about communicate with their physician or practice by best practices and Beth Deaconess’s standards. Source: Laurance and others 2014. Strategy 2: Develop and promote use of support (ACP 1992; Entwistle, Sowden, and decision-aid tools at health facilities Watt 1998), but in general they are designed Many health treatment and screening deci- to enable people to sions are complicated for a variety of reasons. In some circumstances, there is no single best • U nderstand the probable outcomes of choice because people vary in the values or in options by providing information relevant the importance that they place on the benefits to the decision; versus the harms associated with different • Consider the personal value they place on treatment or screening options. In other cir- benefits versus harms by helping clarify cumstances, there can be uncertainty regard- preferences; ing the scientific evidence about the benefits • Feel supported in decision making; and harms associated with the different • Move through the steps in making a deci- options. Some clinical practice guidelines rec- sion; and ommend that people convey their values for • Participate in decisions about their health outcomes to their practitioners, particularly care. regarding treatment and screening decisions in which the best course of action depends on Decision aids in the health sector are being the importance the patient places on the ben- developed in several parts of the world, pri- efits versus the harms (ACP 1992; Eddy marily in North America and Europe. Their 1992; Sawka and others 1998). development is motivated by factors that Decision-support interventions are being include the rise of consumerism, with an developed as an adjunct to practitioners’ emphasis on informed choice rather than counseling (B ekker and others 1999; informed consent; the evidence-based practice Estabrooks and others 2000; Molenaar and movement’s expansion of its audience beyond others 2000; O’Connor and others 1997, practitioners to consumers; the use of infor- 1999; RTI 1997). Decision aids vary in their mation strategies targeted at consumers to specific aims and in the types of decision they reduce unwarranted geographic variations in 150 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A clinical practices; the use of decision-analysis doctor and decide how to treat or manage techniques to identify treatment decisions that their condition. are highly sensitive to people’s values; the Benefits achieved from the use of patient interest in cost savings by reserving optional decision aids can be enhanced by patient acti- interventions for those people who agree that vation methods such as health coaching and the treatment benefits outweigh the harms; one-on-one interactive interviews with doc- and the expansion of the criteria used to judge tors as well as with nurses, pharmacists, psy- health care quality to include patient satisfac- chologists, health educators, or genetic coun- tion with the counseling they receive about sellors. These coaching or interview sessions options. Overall, decision aids provide reli- provide opportunities for clarification and able, balanced, and evidence-based informa- decision support, but they also encourage tion outlining treatment options, outcomes, patients to be more confident in managing and uncertainties and risks associated with their own health and to make treatment deci- treatment options, with the goal of helping sions. Patients can also benefit from question patients discuss their preferences with prompts, which are checklists to spark ideas providers. about questions to ask during interactions Patient-decision aids can take a variety of with health professionals. forms, ranging from simple one-page sheets Most health coaches are nurses who have outlining treatment options to more detailed received additional training in motivational leaflets, computer programs (box 4.10), appli- interviewing—which embodies a shift from cations, or interactive websites. An important “monologue to dialogue” between patients feature is that the aids are designed not just and providers—or specific decision-support to inform patients but also to help them think techniques. These approaches avoid direc- about what the different options might mean tive styles of teaching and advice-giving, for them and to shape their preferences on which can generate resistance or a sense of the basis of scientific information. They can hopelessness among those on the receiving be prescribed to patients before a consulta- end. Coaching has also been shown to be tion so that patients can better prepare them- highly important in helping patients navi- selves to discuss their preferences with the gate the health care system so that they can BOX 4.10  Decision aid for stable coronary heart disease by the Informed Medical Decisions Foundation The decision aid for stable coronary heart disease is Among other features, the aid also gives patients an interactive computer-based resource with infor- access to videotaped conversations with other mation tailored to patients’ specific clinical circum- patients who have already lived through various stances. The aid uses predictive models that help treatments and outcomes. This option is intended patients to envisage short- and long-term conse- to help patients who are struggling to assess how quences of their choices. Use of the decision aid helps they might feel in the future about health states that patients to understand that, given its potential com- they have not yet experienced. The tool also gener- plications, surgery can both increase long-term sur- ates printouts that aim to facilitate conversations vival rates and reduce short-term survival rates. Based between patients and caregivers—conversations that on such information, a patient whose only remaining make it easier for patients to clearly express their desire in life is to attend his daughter’s wedding six preferences. months later might choose to forgo the surgery. Source: Mulley, Trimble, and Elwyn 2012. Note: For more information, see the Center for Shared Decision Making website: http://med.dartmouth-hitchcock.org/csdm_toolkits.html. L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 151 actively choose providers based on their Healthy Settings approach, which is clearly health needs, preferences, and knowledge of laid out in the 1986 Ottawa Charter for providers (box 4.11). Health Promotion (WHO 1986). Decision aids differ from health education In addition, behavioral economic research materials because of their detailed, specific, (such as Thaler and Sunstein 2008) has and personalized focus on options and out- shown the importance of the immediate envi- comes for the purpose of preparing people for ronment in influencing people’s behavioral decision making. In contrast, health education choices. Designing “nudges” that are embed- materials are broader in perspective, helping ded in the physical and social environment to patients to understand their diagnosis, treat- cue people toward adopting healthier behav- ment, and management in general terms, but iors may be another promising health promo- not necessarily helping them to make a specific tion strategy. These strategies are discussed personal choice among options. below. Strategy 1: Develop healthy cities (and Core Action Area 4: Supportive healthy villages) Environment for Citizen Engagement In the physical and social contexts in which The conditions under which people live have people engage in daily activities, environmen- a vital influence on their behavior and the tal, organizational, and personal factors state of their health. An informed public is an interact to affect health and well-being. These essential prerequisite for health promotion social determinants of health contribute to and improvement, but knowledge cannot be the distribution of health in the population transformed into actions and sustained over and are important targets for health promo- time without a supportive environment. This tion. With China’s rapid urbanization, a supportive environment pertains not only to series of “urban diseases” have emerged, such clinical settings but also to the ecology of as environmental pollution, traffic jams, individuals, families, communities and orga- housing shortages, insufficient public nizations, and society as a whole. ­ services, unsafe drinking water and food, All societal forces can be mobilized to cre- noncommunicable diseases, increased stress, ate conditions that enable people to live a accidents, and injuries. These environmental healthy life. This important aspect of sup- and societal factors can pose severe threats to porting citizen engagement in health promo- people’s health. Similarly, environmental deg- tion and improvement underlies WHO’s radation and lack of social support in rural BOX 4.11  Health coaching to coordinate care in Singapore To improve the quality and efficiency of care, and families understand the individuals’ conditions, Singapore implemented a national transitional care effectively articulate their preferences, and enable program for elderly adults with complex care needs self-management and care planning. These care and limited social support called the Aged Care coordinators are mostly nurses and medical social Transition (ACTION) Program. It was designed to workers who are hired by the Agency for Integrated improve coordination and continuity of care and Care. The program targeted complex cases: patients reduce rehospitalization and visits to the emergency older than 65; patients with multiple diagnoses department. and comorbidities who take more than five differ- The program trained and deployed dedicated care ent types of medication; or patients with impaired coordinators to provide coaching to help individuals mobility, significant functional decline, or both. Source: Wee and others 2014. 152 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A China are prominent concerns for health. International Healthy Cities Mayors’ Forums Other countries face similar complex in 2008 and 2010 helped to exchange lessons challenges. and experiences. The Chinese government WHO promotes the global Healthy Cities has responded positively to the movement. As movement as a comprehensive strategy to cre- early as 1994, several cities were selected to ate the supportive environment essential for participate in the Healthy Cities collaboration improving health and addressing social deter- project with WHO; more recently, Hangzhou, minants of health problems. The Healthy Dalian, Suzhou, and 10 more ­ c ities have Cities movement envisages cities with a health- joined the Healthy Cities pilot (box 4.12). promoting environment that enables people to A policy being drafted to scale up the Healthy support each other in performing all the func- Cities movement in China will put health at tions of life and developing to their maximum the heart of the local development agenda potential (Hancock and Duhl 1986). and will potentially link local government The key factors affecting health in cities officials’ performance reviews to its progress. can be considered within three broad themes: The University College London (UCL)– the physical environment, the social environ- Lancet Commission on Healthy Cities arrived ment, and access to health and social services at five key recommendations for achieving (Galea and Vlahov 2005). Municipal govern- Healthy Cities (Rydin and others 2012): ments will plan, construct, and manage the city in a way that continuously improves the • City governments should work with a physical and social environment and broad- wide range of stakeholders to build a polit- ens access to public services that promote ical alliance for urban health. In particu- health—for example, by modifying the physi- lar, urban planners and those responsible cal environment (increasing urban green for public health should be in communica- spaces or designing wider bicycle lanes) or tion with each other. regulating public health (for example, ban- • Attention to health inequalities within ning smoking in public areas or requiring urban areas should be a key focus when safety belts for drivers). planning the urban environment, necessi- Building a healthy city is by nature an tating community representation in the intersectoral endeavor. For example, local arenas of policy making and planning. government policies on housing, the housing • Action needs to be taken at the urban scale market, citizen action on housing conditions, to create and maintain the urban advan- and local lead-poisoning control programs tage in health outcomes through changes may all interact to influence the rates of to the urban environment, providing a lead poisoning in a particular city (Galea, new focus for urban planning policies. Freudenberg, and Vlahov 2005); hence, to • Policy makers at the national and urban reduce these rates will involve political com- scales would benefit from undertaking a mitments by local government along with complexity analysis to understand the institutional changes, capacity building, many overlapping relations affecting innovations, and partnership. The Healthy urban health outcomes. Policy makers Cities movement includes a strong focus on should be alert to the unintended conse- citizen empowerment and participation. The quences of their policies. approach promotes participatory governance • Progress toward effective action on urban by empowering individuals and valuing com- health will be best achieved through local munity knowledge in decision making and experimentation in a range of projects, action on health (CSDH 2008, 18). supported by assessment of their practices Globally, the Healthy Cities movement has and of practitioners’ decision-making pro- attracted many cities to participate (De cesses. Such efforts should include practi- Leeuw 2009; De Leeuw and others 2008, tioners and communities in active dialogue 2015; Green, Jackisch, and Zamaro 2015). and mutual learning. L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 153 BOX 4.12  Changshu: One example of a Chinese Healthy City Changshu is in the southeastern part of eastern to increase awareness of chronic disease prevention, ­ China’s Jiangsu province, about 100 kilometers north- management, and control are being applied in clini- west of Shanghai. Situated in the developed Yangtze cal settings. The following directives are included in River Delta, the city has a population approaching every health care institution’s educational mandate: 2 million. To implement the Healthy City concept, the Changshu Bureau of Health started with the intro- • Focus on the development of a healthy city duction of lecturer groups made up of medical profes- through the modernization of rural health and sionals to provide information and education to both peasant health care services. rural and urban residents. Teams made up of medi- • Print and distribute circulars that address health cal doctors and nurses as well as clinical researchers issues, especially those associated with chronic provided health education seminars and screenings at lifestyle illnesses including diabetes, cardiovascu- regularly held community events and health fairs. lar disease, and hypertension. These events demonstrated to citizens that health • Provide the “500 Questions” pamphlet to govern- is something that must be actively pursued and that ment organizations, enterprises, institutions, and the individual is key to controlling his or her future schools as the principal content reference for life- health trajectory. Demonstrating that good health style and chronic diseases. and the avoidance of lifestyle illnesses are the respon- • Conduct special training in disease management sibility of each individual through conscious actions for patients with chronic disease. rather than simply chance, the Healthy Cities pro- • Organize activities promoting chronic-disease gram promoted individual empowerment as the first awareness in government organizations, enter- step toward internalizing health responsibility. This prises, institutions, and schools. is the Healthy Cities message: that all requirements • Use various methods of publicity to exploit differ- for economic growth can be in place, but the lack of ent media types. a healthy population cripples growth in its infancy. A • Include the basic learning aspects of the “500 healthy population is the essential foundation upon Questions” pamphlet in publicity circulars. which to build sustainable economic growth. For its public health cadre, the Changshu As an example of exploiting various media to Ministry of Health developed standard guidelines promote Healthy Cities through improved levels of entitled “Basic Knowledge and Skill of People’s public health, the city of Changshu held “a ceremony Health Literacy” that contained “The 66 Principles for promoting public health lifestyles” in December of Health,” which provided the content of public 2011. Through broadcast media, the vice mayor of health forums. In addition, brochures describing dis- Changshu promoted public awareness of healthy ease management and care, as well as how to avoid ­ lifestyles by encouraging the adoption of a rational disease through lifestyle changes, were distributed diet, engaging in moderate exercise, and quitting during neighborhood visits. To better inform public smoking. He also encouraged rational limits to alcohol health policy, the Ministry of Health posed questions consumption and striking a “psychological balance” that measured the public health knowledge of citi- in life and living. The Ministry of Health awarded zens in the Changshu region. These surveys revealed prizes to “healthy families”: those that demonstrated low levels of information about the 10 most impor- their commitment to practicing healthy ways of living. tant chronic illnesses facing people in the region. Attendees also received small gifts including health- After establishing baseline levels of chronic disease related books as well as cruets and measuring spoons in the population, the Ministry issued a 2009 white to accurately measure food portions. Theatrical paper, “The Changshu Testing Program for Public skits promoting healthy lifestyles were performed. Knowledge of Chronic Noncommunicable Diseases Changshu talk-radio programs invited experts from Prevention and Control.” This document requires municipal hospitals to speak about chronic diseases all health care institutions in the Changshu region such as hypertension, diabetes, and cancer and to to apply the material contained in a pamphlet (titled interact with callers. Further, the Ministry organized “500 Questions to Determine Public Knowledge of alternative methods of promoting awareness such as Chronic Noncommunicable Disease Prevention and free screening and counseling for diabetes to coincide Control”) to ensure that the methods that are thought with World Diabetes Day each November 14. Source: Case study commissioned by the World Bank. 154 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Strategy 2: Create environmental “nudges” smoking to bank the money they would have to improve health choices spent on their habit but only allowing them Most people value their health yet persist in to withdraw it when they test as nicotine- behaving in ways that undermine it. Many free). Table 4.2 gives examples of nudging psychological reasons underlie this gap strategies as opposed to regulatory strategies between value or cognition and behavior, one (Marteau and others 2011). being that people’s behavior can be subcon- Some of these strategies have proven sciously triggered by environmental or emo- highly effective, as follows: tional cues that are driven by default, habits, or perception of social norms (Thaler and • Australia, France, Poland, and Portugal Sunstein 2008). These inherent human biases have adopted “opt-in” as the default for offer an opportunity for noncoercive policy indicating willingness for organ donation interventions to change behavior toward and, as a result, 90–100 percent of their healthier choices. By changing the seemingly citizens are registered donors, compared subtle cues in the physical, social, and policy with only 5–30 percent in countries that environment, so-called “nudging” interven- do not use the donor default strategy tions can signal to people to make better health (Johnson and Goldstein 2003). choices without coercion or material incentive. • In some states in the United States, the Nudges might involve subconscious cues default is that pharmacists can fill written (such as painting targets in urinals to improve drug prescriptions with generic drugs accuracy) or correcting misapprehensions unless the physician opts out by placing about social norms (like telling people that “dispense as written” on the prescription most people do not drink excessively). They (Blumenthal-Barby and Burroughs 2012). can alter the profile of different choices (such • An example of making health messages as the prominence of healthy food in can- more salient to act on is the requirement teens) or change which options are the default for restaurants to put caloric amounts on (such as having to opt out of rather than opt menus. In New York, this requirement into organ-donor schemes). Nudges can also caused people to order meals containing create incentives for certain choices or impose fewer calories and caused restaurants to minor economic or cognitive costs on other lower the calorie content of meals (Rabin options (such as enabling people who quit 2008). TABLE 4.2  Examples of nudges and regulation to change target behaviors Behavior Nudges Regulations Smoking ·   Make nonsmoking more visible through mass media campaigns ·   Ban smoking in public places communicating that most people do not smoke and most ·   Increase price of cigarettes smokers want to stop ·   Reduce cues for smoking by keeping cigarettes, lighters, and ashtrays out of sight Alcohol ·   Serve drinks in smaller glasses ·   Regulate pricing through taxes or minimum ·   Make lower alcohol consumption more visible by highlighting in pricing per unit mass media campaigns that most people do not drink to excess ·   Raise the minimum age for purchase of alcohol Diet ·   Designate sections of supermarket trolleys for fruit and ·   Restrict food advertising in media directed at vegetables children ·   Make salad rather than chips the default side order ·   Ban industrially produced trans-fatty acids Physical activity ·   Make stairs, not elevators, more prominent and attractive in ·   Increase tax on gasoline year-on-year (fuel price public buildings escalator) ·   Make cycling more visible as a means of transport, for example ·   Enforce car drop-off exclusion zones around through city bicycle-hire schemes schools Source: Marteau and others 2011. L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 155 • People also respond to a change in percep- Raising Attainment for All program to form tion of a social norm. The State of the Children and Young People Improvement Montana ran an intensive “Most of Us Collaborative; for more ­ i nformation, see Wear Seatbelts” media campaign from https:// beta.gov.scot /policies/improving​ -­public-services/children-and​-young-people​ 2000 to 2003 in which the Department of -improvement-collaborative/. Transportation let people know that most 6. Decision aids are interventions designed to people (85 percent) wear seatbelts. This help people make specific and deliberative resulted in a significant increase in the choices among options by providing informa- reported use of seatbelts (Linkenbach and tion about the options and outcomes that are Perkins 2003). relevant to a person’s health status. • Finally, a successful technique to increase 7. For more about the Flinders Program, see fruit consumption among school students https://www.flindersprogram.com.au/. is to place fruits and vegetables in promi- 8. For updates on Cochrane Reviews, see http:// nent places in the cafeteria and display cccrg.cochrane.org/our-reviews. them attractively. This demonstrates the behavior-shaping effect of priming cues. References ACP (American College of Physicians). 1992. Notes “Guidelines for Counselling Postmenopausal 1. “Opinion on Deepening China’s Health Care Women about Preventive Hormone Therapy.” System Reform,” State Council (Zhong Fa Ann als of Intern al Medicine 117 (12): 2009, No. 6); “The Twelfth-Five Year Plan of 1038–41. Health Care Development,” State Council Ali, Rabia. 2015. “Programs of All-Inclusive Care General Office (Guo Fa 2012, No. 57); for the Elderly (PACE) in the U.S.: Background “Guidance of the General Office of the State and Rationale.” Case study commissioned by Council on Overall Pilot Reform of Urban the World Bank, Washington, DC. Public Hospitals,” National Health and Ali, Rabia, and Rong Li. 2015. “Patient-Centered Family Planning Commission (Guo Ban Fa Medical Home (PCM H) Model of U. S. 2015, No. 38); “Opinions of the State Council Veterans Health Administration.” Case study on Comprehensively Scaling-Up Reform of c o m m i s s i o n e d b y t h e Wo r l d B a n k , County-Level Public Hospitals,” State Council Washington, DC. (Guo Ban Fa 2015, No. 33); and “Planning of A rnstein, S . 1969. “A L adder of Citizen the National Health Service System (2015– Participation.” Journal of American Institute 2020).” State Council General Office (Guo of Planners 35 (4): 216–24. Ban Fa 2015, No. 14). Audit Commission. 20 04. “I mplementing 2. “Guidance of the General Office of the State Telecare: Strategic Analysis and Guidelines for Council on Promoting Multi-Level Diagnosis Policy Makers, Commissioners and Providers.” and Treatment System,” State Council (Guo Public Sector National Report, London. Ban Fa 2015, No. 70). Barlow, J., C. Wright, J. Sheasby, A. Turner, and 3. Related e-health themes are discussed in J. Hainsworth. 2002. “Self-Management ­chapter 3. Approache s for Pe ople w it h C h ron ic 4. Public education campaigns should be part of Conditions: A Review.” Patient Education and a comprehensive strategy for health promo- Counseling 48 (2): 177–87. tion that includes legislation and regulation— Basu, Sandra. 2014. “VHA Emphasizes Self- for example, legislative action against Management in New COPD Care Tool.” U.S. smoking through banning cigarette advertise- Medicine, May 21. ments, banning smoking in public places, Bekker, H., J. G. Thornton, C. M. Airey, J. B. taxation of cigarette sales, and so forth. A full Connelly, J. Hewison, and M. B. Robinson. treatment of possible options goes beyond the 1999. “Informed Decision Making: A n scope of this report. Annotated Bibliography and Systematic 5. In 2016, the Scottish Government com- Review.” Health Technology Assessment 3 (1): bined the Early Years Collaborative with the 1–156. 156 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Bennett, G. G., S. J. Herring, E. Puleo, E. K. Cheng, H., and K. Chan. 2009. “Public Service Stein, K. M. Emmons, and M. W. Gillman. Advertising in China: A Semiotic Analysis.” 2010. “Web-Based Weight Loss in Primary In  Advertising and Chinese Society: Impacts Care: A Randomized Controlled Trial.” and Issues, edited by H. Cheng and K. Chan, Obesity 18 (2): 308–13. 203–21. Copenhagen: Copenhagen Business Berkman, Nancy D., Stacey L. Sheridan, Katrina School Press. E. Donahue, David J. Halpern, and Karen Cheng, H., P. Kotler, and N. Lee. 2011. Social Crotty. 2011. “Low Health Literacy and Marketing for Public Health: Global Trends Health Outcomes: An Updated Systematic and Success Stories. Boston: Jones & Bartlett Review.” Annals of Internal Medicine 155 (2): Learning. 97–107. China Medical Tribune. 2012. “Medical War in Bieber, Christiane, J. Nicolai, M. Hartmann, K. 2012” [title translated from Chinese]. Blumenstiel, N. Ringel, A. Schneider, M. D e cemb er 16. ht t p: // w w w.cmt.com.c n​ Härter, W. Eich, and A. Loh. 2009. “Training /detail/111139.html. Physicians in Shared Decision-Making: Who ———. 2013. “Violent Events against Doctors in Can Be Reached and What Is Achieved?” the First Half of 2013.” (Accessed August 16, Patient Education and Counseling 77 (1): 2013.) http://www.cmt.com.cn/detail/270013​ 48–54. .html. Blumenthal-Barby, Jennifer S., and Hadley CMDA (Chinese Medical Doctor Association). Burroughs. 2012. “Seeking Better Health Care 2013. “Four Doctors Dead within a Week: Who Outcomes: The Ethics of Using the ‘Nudge.’” Will Guard the Physicians’ Health?” July 14. American Journal of Bioethics 12 (2): 1–10. ContinYou. 2010. “Skilled for Health Has the Bonvicini, Kathleen A., Michael J. Perlin, Carma ‘Heineken’ Effect.” ContinYou, Coventry, U.K. L. Bylund, Gregory Carroll, Ruby A. Rouse, Coulter, Angela. 2011. Engaging Patients in and Michael G. Goldstein. 2009. “Impact of H e a l t h C a re . L o n d o n : M c G r aw- H i l l Com mu nication Training on Physician Education. E xpression of E mpathy in Patient Coulter, Angela, and Alf Collins. 2011. Making E n c ou nt e r s .” Pa t i e n t E d u c a t i o n a n d Shared Decision-M aking a Realit y: No Counseling 75 (1): 3–10. Decision about Me, Without Me. London: The Bove, A. A., C. J. Homko, W. P. Santamore, M. King’s Fund. Kashem, M. Kerper, and D. J. Elliott. 2013. C o u lt e r, A n g e l a , a n d J o E l l i n s . 2 0 0 7. “M a nag i ng Hy per tension i n Urba n “Effectiveness of Strategies for Informing, Underserved Subjects Using Telemedicine: A Educating, and Involving Patients.” British Clinical Trial.” American Heart Journal 165 Medical Journal 335 (7609): 24. (4): 615–21. Coulter, Angela, and H. Magee, eds. 2003. The Center for Health Statistics. 2010. “The National European Patient of the Future. Maidenhead, Survey Study on Patient-Doctor Relationship.” U.K.: Open University Press. National Healt h and Fa m ily Plan n ing Coulter, A., S. Roberts, and A. Dixon. 2013. Commission, Beijing. “Delivering Better Services for People with Charles, C ., and S. DeMaio. 1993. “Lay Long Term Conditions: Building the House of Participation in Health Care Decision Making: Care.” Report, The King’s Fund, London. A Conceptual Framework.” Journal of Health CSDH (Commission on Social Determinants Politics, Policy and Law 18 (4): 881–904. of Health). 2007. “Achieving Health Equity: Chen, F. 2012. “At Shenzhen Pencheng Hospital a From Root Causes to Fai r Outcomes. Nurse Was Attacked by Knife” [title translated Commission on Social Determinants of from Chinese]. Shenzhen News, September 4. Health, Interim Statement.” CDSH, World h t t p : / / h e a l t h . s o h u . c o m / 2 0 1 2 0 9 0 4​ Health Organization, Geneva. /n352262739.shtml. ———. 2008. Closing the Gap in a Generation: Chen, Z.-W., L.-Z. Fang, L.-Y. Chen, and H.-L. Health Equity through Action on the Social Dai. 2008. “Comparison of an SMS Text Determinants of Health. Geneva: WHO. Messaging and Phone Reminder to Improve De Leeuw, Evelyne. 2009. “Evidence for Healthy Attendance at a Health Promotion Center: A Cities: Reflections on Practice, Method and Randomized Controlled Trial.” Journal of Theory.” Health Promotion International 24: Zhejiang University Science B 9 (1): 34–38. i19–i36. doi:10.1093/heapro/dap052. L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 157 De Leeuw, Evelyne, Agis D. Tsouros, Mariana Zhu Shitai, Ding Yongming, and Wei Zhihua. Dyakova, and Geoff Green, eds. 2008. Healthy 2003. “Implementation and Quantitative Cities: Promoting Health and Equity — Eva lu at io n of C h ro n i c D i s e a s e S e l f- Evidence for Local Policy and Practice. Management Programme in Shanghai, China: Summary Evaluation of Phase V of the WHO Randomized Controlled Trial.” Bulletin of the E uropean Healthy C ities Network. World Health Organization 81 (3): 174–82. Copenhagen: W HO Regional Office for Galea, Sandro, and David Vlahov. 2005. “Urban Europe. Health: Evidence, Challenges, and Directions.” De Leeuw, Evelyne, Geoff Green, Mariana Annual Review of Public Health 26: 341–65. Dyakova, Lucy Spanswick, and Nicola Palmer. Galea, Sandro, Nicholas Freudenberg, and David 2015. “European Healthy Cities Evaluation: Vlahov. 2005. “Cities and Population Health.” Conceptual Framework and Methodology.” Social Science & Medicine 60 (5): 1017–33. Health Promotion International 30 (Suppl. 1): Glasgow, R. E., S. Emont, and D. C. Miller. 2006. i8–i17. “Assessing Delivery of the Five ‘A’s’ for Patient- Deyo, R. A., D. C. Cherkin, J. Weinstein, J. Centered Counselling.” Health Promotion Howe, M. Ciol, and A. G. Mulley Jr. 2000. International 21 (3): 245–55. “Involving Patients in Clinical Decisions: Green, G., J. Jackisch, and G. Zamaro. 2015. Impact of an Interactive Video Program on Use “Healthy Cities as Catalysts for Caring and of Back Surgery.” Medical Care 38: 959–69. Supportive Environments.” Health Promotion Eddy, D. M. 1992. A Manual for Assessing International 30 (Suppl. 1): i99–i107. Health Practices and Designing Practice Gu angzhou Daily . 2014. “Doctor-Patient Policies: The Explicit Approach. Philadelphia: Disputes Increase 10 Times in 10 Years: American College of Physicians. Doctors Average 2.4 Minutes to See 1 Patient” Edwards A., and G. Elwyn. 2009. Shared [title translated from Chinese]. June 30. http:// Decision-Making in Health Care: Achieving gd.sina.com.cn​ / news/m/2014-06-​ 3 0/0726​ Evidence-Based Patient Choice. 2nd ed. 109553.html. Oxford: Oxford University Press. Hancock, T., and L. Duhl. 1986. “The Healthy Entwistle, V. A., A. Sowden, and I. Watt. 1998. City: Its Function and its Future.” Health “Evaluating Interventions to Promote Patient Promotion 1 (1). Involvement in Decision Making: By What Hart, J. T. 1995. “Clinical and Economic Criteria Should Effectiveness Be Judged?” Consequences of Patients as Producers.” Journal Journal of Health Services Research and of Public Health Medicine 17: 383–86. Policy 3: 100–7. Hesketh, T., D. Wu, L. Mao, and N. Ma. 2012. Estabrooks, C. A., V. Goel, E. Thiel, S. P. Pinfold, “Violence against Doctors in China.” BMJ C . S awk a , a nd J. I . W i l l ia m s. 20 0 0. 345: e5730. “Consumer Decision Aids: Where Do We Hildebrandt, Helmut, Alexander Pimperl, Oliver Stand? A Systematic Review of Structured Gröne, Monika Roth, Christian Melle, Timo Consumer Decision Aids.” Technical report, Schulte, Martin Wetzel, and Alf Trojan. 2015. Institute for Clinical Evaluative Sciences, “Gesundes Kinzigtal: A Case Study on People Toronto. Centered/Integrated Health Care in Germany.” Feingold, E. 1977. “Citizen Participation: A Case study commissioned by the World Bank, Review of the Issues.” In The Consumer and Washington, DC. the Health Care System: Social and Managerial Holman, Halsted, and Kate Lorig. 2004. “Patient Perspectives, edited by H. M. Rosen, J. M. Self-Management: A Key to Effectiveness and Metsch, and S. Levey, 153–60. New York: Efficiency in Care of Chronic Disease.” Public Spectrum. Health Reports 119 (3): 239. F lor i n D., a nd J. Di xon. 20 0 4. “P ubl ic Horsburgh, M., J. Bycroft, F. Mahony, D. Roy, D. Involvement in Health Ethics.” BMJ 328: Miller, F. Goodyear-Smith, and E. Donnell. 159–61. 2010. “The Feasibility of Assessing the Flinders French, J., and C. Blair-Stevens. 2007. Big Pocket Program of Patient Self-Management in New Book: Social Marketing. London: National Zealand Primary Care Settings.” Journal of Social Marketing Center. Primary Health Care 2 (4): 294–302. Fu, Dongbo, Fu Hua, Patrick McGowan, Shen IOM (Institute of Medicine). 2001. Crossing the Yi-e, Zhu Lizhen, Yang Huiqin, Mao Jianguo, Quality Chasm: A New Health System for the 158 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A 21st Century. Washington, DC: National Linder-Pelz, Susie. 1982. “Toward a Theory of Academies Press for the Committee on Quality Patient Satisfaction.” Social Science & of Health Care in America. Medicine 16 (5): 577–82. Johnson, E. J., and D. Goldstein. 2003. “Do Linkenbach, J., and H. W. Perkins. 2003. “Most Defaults Save Lives?” Science 302 (5649): of Us Wear Seatbelts: The Process and 1338–39. Outcomes of a 3-Year Statewide Adult Seatbelt Jones, I. L., M. Berney, M. Kelly, L. Doyal, C. Campaig n in Montana.” Presentation, Griffiths, G. Feder, S. Hillier, G. Rowlands, National Conference on the Social Norms and S. Curtis. 2004. “Is Patient Involvement Model, Boston, July 17. Possible When Decisions Involve Scarce Lorig, K. 1993. “Self-Management of Chronic Resources? A Qualitative Study of Decision Illness: A Model for the Future.” Generation Making in Primary Care.” Social Science and 17: 1–14. Medicine 59: 93–102. Lorig, Kate R., Philip L. Ritter, Ayesha Dost, Kennedy, A. D., M. J. Sculpher, A. Coulter, N. Kathryn Plant, Diana D. Laurent, and Ian Dwyer, M. Rees, and K. R. Abrams. 2002. M c N ei l . 2 0 0 8 . “T he E x p er t Pat ient s “Effects of Decision Aids for Menorrhagia on Programme Online, a 1-Year Study of an Treatment Choices, Health Outcomes, and Internet-Based Self-Management Programme Costs: A Randomized Controlled Trial.” for People with Long-Term Conditions.” JAMA 288: 2701–08. Chronic Illness 4 (4): 247–56. Kindig, David A., Allison M. Panzer, and Lynn Lorig, Kate R., Philip Ritter, Anita L. Stewart, Nielsen-Bohlman, eds. 2004. Health Literacy: David S. Sobel, Byron William Brown Jr., A Prescription to End Confusion. Washington, Albert Bandura, Virginia M. Gonzalez, Diana DC: National Academies Press. D. Laurent, and Halsted R. Holman. 2001. Kurtz, Suzanne, Jonathan Silverman, John “Chronic Disease Self-Management Program: Benson, and Juliet Draper. 2003. “Marrying 2-Year Health Status and Health Care Content and Process in Clinical Method Utilization Outcomes.” Medical Care 39 (11): Teaching: Enhancing the Calgary–Cambridge 1217–23. Guides.” Academic Medicine 78 (8): 802–09. Ma, Jin. 2015. “Primary Care–Centered Care in Laine, C., and F. Davidoff. 1996. “Patient- Two Districts in Shanghai.” Shanghai Jiao Centered Medicine: A Professional Evolution.” Tong University. Case study commissioned by JAMA 275 (2): 152–56. the World Bank, Washington, DC. Lancet. 2012. “Ending Violence against Doctors Makoul, Gregory, Amanda B. Zick, Mark in China.” The Lancet 379 (9828): 1764. Aakhus, Kathy J. Neely, and Phillip E. Roemer. ———. 2014. “Violence against Doctors: Why 2010. “Using an Online Forum to Encourage China? Why Now? What Next?” The Lancet Reflection about Difficult Conversations in 383 (9922): 1013. Medicine.” Patient Education and Counseling Laurance, Jeremy, Sarah Henderson, Peter J. 79 (1): 83–86. Howitt, Mariam Matar, Hanan Al Kuwari, Marteau, Theresa M., D. Ogilvie, M. Roland, Susan Edgman-Levitan, and Ara Darzi. 2014. M. Suhrcke, and M. P. Kelly. 2011. “Judging “Patient Engagement: Four Case Studies that Nudging: Can Nudging Improve Population Highlight the Potential for Improved Health Health?” BMJ 342: 263–65. Outcomes and Reduced Costs.” Health Affairs Molenaar, S., M. A. G. Sprangers, F. C. E. 33(9): 1627–34. Postma-Schuit, E. J. Rutgers, J. Noorlander, Lee, Clara N., Rosalie Dominik, Carrie A. Levin, and J. Hendriks. 2000. “Feasibility and Effects Michael J. Barry, Carol Cosenza, Annette M. of Decision Aids.” Medical Decision Making O’Connor, Albert G. Mulley Jr., and Karen R. 20 (1): 112–27. Sepucha. 2010. “Development of Instruments Moore, M. 2012. “Female Doctor Axed to Death to Measure the Quality of Breast Cancer i n Ch i nese Hospit a l.” T he Teleg raph , Treatment Decisions.” Health Expectations 13 November 29. (3): 258–72. Morgan, M. W., R. B. Deber, H. A. Llewellyn- Liang, Ying, and Yong Bao. 2012. “Analysis of Thomas, P. Gladstone, R. J. Cusimano, and K. waiting time in medical visits of Shanghai resi- O’Rourke. 2000. “Randomized, Controlled dents.” Journal of Shanghai Jiaotong University Trial of an Interactive Videodisc Decision Aid (Medical Science) 32 (10): 1368–72. for Patients with Ischemic Heart Disease.” L e v er 3 : E ngaging C iti z en s in S u pport of the P C I C M odel 159 Journal of General Internal Medicine 15 (10): of E f f ic ac y a nd Pol ic y I mpl ic at ions.” 685–93. Monograph of the National Cancer Institute Mulley, A. 2009. “Inconvenient Truths about 25: 67–80. Supplier Induced Demand and Unwarranted Osborn, Robin, and David Squires. 2012. Variation in Medical Practice.” BMJ 339: “I nternational Perspec tives on Patient b4073. E n g a g e m e n t R e s u l t s f r o m t h e 2 011 Mulley, Albert, Tessa Richards, and Kamran Commonwealth Fund Survey.” Journal of Abbasi. 2015. “Delivering Health with Ambulator y C are M an agement 35 (2): Integrity of Purpose.” BMJ 351: h4448. doi: 118–28. 10.1136/bmj.h4448. Picker (Picker Institute Europe). 2010. “Invest in Mulley, A., Chris Trimble, and Glyn Elwyn. 2012. Engagement: Self-Management.” Patients’ Preferences Matter: Stop the Silent Rabin, R. C. 2008. “New Yorkers Try to Swallow Misdiagnosis. London: The King’s Fund. Calorie Sticker Shock.” MSNBC, July 16. National Voices. 2014. “Prioritising Person- http://www.nbcnews.com/id/25464987/ns​ Centered Care—the Evidence.” National /health-diet_and_nutrition /t/new-yorkers​ Voices, London. - t r y - s w a l l o w - c a l o r i e - s t i c k e r - s h o c k / #​ NHFPC (National Health and Family Planning .WzawCdJKiM8. Commission). 2014. “National Health Literacy Rhee, H., J. Allen, J. Mammen, and M. Swift. Promotion Ac tion Plan (2014 –2020).” 2014. “Mobile Phone-Based Asthma Self- Planning document, NHFPC, Beijing. Management Aid for Adolescents (mASMAA): ———. 2015. “Report on Development of A Feasibility Study.” Patient Preference and Deepening Health Reform to the 18th meeting Adherence 8: 63–72. of the Standing Committee of the 12th Rollnick, S., W. R. Miller, and C. C. Butler. 2002. National People’s Congress.” NHFPC report Motivational Interviewing: Preparing People presented December 22, Beijing. for Change. New York: Guilford Press. Nolan, R. P., R. E. Upshur, H. Lynn, T. Crichton, RTI (Research Triangle Institute). 1997. “The E. Rukholm, D. E. Stewart, and M. H. Chen. Effects of Informatics Tools and Decision Aids 2011. “Therapeutic Benefit of Preventive to Support Shared Patient Decision Making Telehealth Counseling in the Community about Medical Screening and Treatment.” Outreach Heart Health and Risk Reduction Report to the Agency for Health Care Policy Trial.” American Journal of Cardiology 107 Research, RTI (formerly Research Triangle (5): 690–96. Institute) International, Durham, NC. Nolte, Ellen, Anne Frølich, Helmut Hildebrandt, Runz-Jørgensen, Sidsel Marie, and Anne Frølich. Alexander Pimperl, and Hubertus J. Vrijhoef. 2015. “SIKS—The Integrated Effort for People 2015. “Integrating Care: A Synthesis of Living with Chronic Diseases: A Case Study on Experiences in Three European Countries.” People Centered/Integrated Health Care in European Observatory. Case studies commis- Denmark.” The Research Unit for Chronic sioned by the World Bank, Washington, DC. Conditions, Bispebjerg Hospital, Denmark. Nutbeam, D. 2008a. “The Evolving Concept of Case study commissioned by the World Bank, Health Literacy.” Social Science & Medicine Washington, DC. 67 (12): 2072–78. Rydin, Yvonne, Ana Bleahu, Michael Davies, Julio ———. 2008b. “Health Literacy: Perspectives D. Dávila, Sharon Friel, Giovanni De Grandis, from Australia.” Society for Academic Primary Nora Groce, and others. 2012. “Shaping Cities Care—Health Literacy Group, London. for Health: Complexity and the Planning of O’Connor, A. M., E. R. Drake, V. J. Fiset, J. Page, Urban Environments in the 21st Century.” The D. Curtin, and H. A. Llewellyn-Thomas. 1997. Lancet 379 (9831): 2079–2108. “Annotated Bibliography of Studies Evaluating Sawka, C. A., V. Goel, C. A. Mahut, G. A. Taylor, Decision Support Interventions for Patients.” E. C. Thiel, and A. M. O’Connor. 1998. Canadian Journal of Nursing Research 29 (3): “Development of a Patient Decision Aid for 113–20. Choice of Surgical Treatment for Breast O’Connor, A. M., V. Fiset, C. De Grasse, I. D. Cancer.” Health Expectations 1 (1): 23–36. Graham, W. Evans, and D. Stacey. 1999. Stacey, Dawn, C. L. Bennett, M. J. Barry, N. F. “Decision Aids for Patients Considering Col, K. B. Eden, M. Holmes-Rovner, H. Options Affecting Cancer Outcomes: Evidence Llewellyn-Thomas, A. Lyddiatt, F. Légaré, 160 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A and R. Thomson. 2011. “Decision Aids for Wennberg, John E. 2010. Tracking Medicine: A People Facing Health Treatment or Screening Researcher’s Quest to Understand Health Decisions.” Cochrane Database Systematic Care. Oxford: Oxford University Press. Reviews 4: CD001431. doi:10.1002/14651858. WHO (World Health Organization). 1986. “The Sylvia, Sean, Yaojiang Shi, Hao Xue, Xin Tian, Ottawa Charter For Health Promotion.” Huan Wang, Qingmei Liu, Alexis Medina, and Charter document presented at the First Scott Rozelle. 2015. “Survey Using Incognito I n t e r n at i o n a l C o n f e r e n c e o n H e a lt h Standardized Patients Shows Poor Quality Promotion, Ottawa, November 21. Care in China’s Rural Clinics.” Health Policy ———. 1998. “Health Promotion Glossary.” and Planning 30 (3): 322–33. doi:10.1093/ WHO/HPR/98.1, WHO, Geneva. heapol/czu014. ———. 2015a. “People-Centered and Integrated Szmedra, Philip, and Li Zhenzhong. 2013. Health Services: An Overview of the Evidence.” “Implementing a Healthy Cities Plan in China: WHO/HIS/SDS/2015.7, WHO, Geneva. T he I mpac t of Focused P ublic Health ———. 2015b. “WHO Global Strategy on People- Education on a Diabetes Epidemic.” Southeast Centered and Integrated Health Services.” Re vie w of A si an Studies (SE R A S) 35: WHO/HIS/SDS/2015.6, WHO, Geneva. 230–40. Williamson, L . 2014. “Patient and Citizen Tang, C., Z. Luo, P. Fang, and F. Zhang. 2013. Par ticipation in Health: T he Need for “Do Patients Choose Community Health Improved Ethical Support.” American Journal Services (CHS) for First Treatment in China? of Bioethics 14 (6): 4–16. doi:10.1080/1526516 Results from a Community Health Survey in 1.2014.900139. Urban Areas.” Journal of Community Health Wong, C. K., C. S. Fung, S. S. Siu, Y. Y. Lo, K. W. 38 (5): 864–72. Wong, D. Y. Fong, and C. L. Lam. 2013. Tenbensel, T. 2010. “Public Participation in “A Short Message Service (SMS) Intervention Health Policy in High-Income Countries— to Prevent Diabetes in Chinese Professional Why, Who, What, Which, and Where?” Social Drivers with Pre-Diabetes: A Pilot Single- Science and Medicine 71 (9): 1537–40. Blinded Randomized Controlled Trial.” Thaler, R., and C. Sunstein. 2008. Nudge: Diabetes Research and Clinical Practice 102 Improving Decisions about Health, Wealth, (3): 158–66. and Happiness. New Haven, C T: Yale Xue, F., W. Yao, and R. J. Lewin. 2008. “A University Press; London: Penguin Books. Randomised Trial of a 5 Week, Manual Based, Tucker, J. D., Yu Cheng, Bonnie Wong, Ni Gong, Self-Management Programme for J i n g - B ao N i e , We i Z hu , M e g a n M . Hypertension Delivered in a Cardiac Patient McLaughlin, and others. 2015. “Patient– Club in Shanghai.” BMC Cardiovascular Physician Mistrust and Violence against Disorders 8: 10. doi:10.1186/1471-2261-8-10. Physicians in Guangdong Province, China: Yip, Winnie, Chi-Man Yip, William C. Hsiao, A Qualitative Study.” BMJ Open 5 (10): Wen Chen, Shanlian Hu, Jin Ma, and Alan e008221. doi:10.1136/bmjopen-2015-00822. Maynard. 2012. “Early Appraisal of China’s Wagner, Edward H., Paul Barrett, Michael J. Huge and Complex Health-Care Reforms.” Barry, William Barlow, and Floyd J. Fowler. The Lancet 379 (9818): 833–42. doi:10.1016/ 1995. “The Effect of a Shared Decision- S0140-6736(11)61880-1. Making Program on Rates of Surgery for Yuan, T. 2012. “Man with a Knife Stabbed Four Benign Prostatic Hyperplasia: Pilot Results.” Doctors in China” [title translated from Medical Care 33 (8): 765–70. Chinese]. ChinaNews.com, March 24. http:// Wee, Shiou-Liang, Chok-Koong Loke, Chun www.chinanews.com/fz/2012/03-24/3769439​ Liang, Ganga Ganesau, Loong-Mun Wong, .shtml. and Jadon Cheah. 2014. “Effectiveness of a Zhong, Nanshan. 2015. “Shared Decision Making National Transitional Care Program in Is the Core of Humanistic Spirit.” Health News, Reducing Acute Care Use.” Journal of the June 19. http://www.jkb.com.cn /medical​ American Geriatrics Society 62 (4): 747–53. Humanities/2015/0619/372485.html. 5 Lever 4: Reforming Public Hospital Governance and Management Introduction One of the core objectives for reforming public hospitals is the separation of govern- Reforming hospitals is an integral part of m e nt ad m i n i s t r at io n f ro m ho s pit a l reforming service delivery in favor of people- ­ management. The central government envis- centered primary ­ care. Hospitals will con- ages public hospitals as independent entities tinue to play an important role, but one that with legal ­ personality. Hence, policy direc- will become less financially dominant and tives aim to grant hospitals greater manage- focus more on providing only the specialized rial autonomy from direct hierarchical con- offer. services that hospitals alone can ­ trol by the government administrative As primary care is strengthened and the ­ apparatus. However, hospitals are to retain patient-centered integrated care (PCIC) their public institutional identity and main- model is put in place, a wide range of care processes will need to be shifted from hospi- tain their accountability to government pri- tals to ambulatory settings (for example, cer- orities, particularly in terms of acting in the tain surgeries and diagnostics, chemotherapy) public ­interest. and primary care ­ facilities. Hospitals will While relinquishing direct control over hos- become centers of excellence, but with ade- pitals, government agencies would play vital quate volume to deliver high quality ­ c are. roles in regulating, sector planning, standard They can perform important training and setting, and monitoring and evaluation workforce development ­ functions. They can (M&E) of hospital ­ performance. At the same also focus more on biomedical research and time, policy directives aim to improve mana- on providing clinical support to lower-level gerial practices in ­ hospitals. They promote ­ providers. As described in chapter 2, many of professionalized management and endorse these changes are under way in ­ China. State strengthening of managerial functions such as Council directives since 2009 have empha- cost accounting, clinical ­ management, logis- sized the importance of governance and man- tics and materials management, patient flows, agement reforms as part of a comprehensive and nursing ­ management. strategy that includes reforms in financing, Strong hospital governance and manage- provider payment methods, pricing, and care ment will be needed to alter the roles and ­integration.1 responsibilities of hospitals and strengthen 161 162 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A their links to the broader health care delivery • “Hospital Managerial Practices” exam- system. Governance and management are ­ ines the challenges and lessons learned the least understood drivers of hospital from managerial practices in China and p erformance and the hardest to isolate ­ ­internationally. analytically. Emerging evidence from China ­ • “Recommendations for Moving Forward and elsewhere suggests that elements of good with Public Hospital Reform” specifies governance include (a) clearly specified and the core action areas for reform and the enforced accountabilities to payers and gov- corresponding strategies for improving ernment, (b) a degree of autonomous decision governance and managerial ­ practices. making, (c) aligned incentives, and (d) effec- tive organizational forms that interface with government (as owners) and hospital Overview and Trends management. Management also matters: just ­ Hospitals account for a high proportion of as better-managed firms tend to display China’s total health ­ spending. According to better performance, the same is true for government statistics, hospitals absorbed hospitals. Managers put in place the organi- ­ 70 percent of total health spending in zational behaviors and practices that are ­2013. 3 Using a different methodology, the aligned with the broader governance and Organisation for Economic Co-operation incentive ­environment. and Development (OECD) estimated hospital This chapter focuses on two major compo- expenditures to account for 54 percent of nents of the public hospital reform agenda— total health ­ spending.4 Both figures exceed governance and managerial practices—and is the average ratio for OECD countries of organized as follows:2 38 ­percent. Of more concern is the high growth rate of • “Overview and Trends” presents a brief China’s hospital spending, which reflects sector. overview of China’s hospital ­ expanding supply and ­ utilization. While total • “Public Hospital Governance” examines health spending grew 2 ­ .7-fold between 2005 hospital governance challenges and lessons and 2013, hospital spending surged threefold learned from reform in China and interna- (figure 5.1). tionally, drawing on available literature as Over that same period, the number of hos- well as cases and surveys commissioned pitals increased by nearly one-third (from report. for this ­ 18,703 to 24,709); the number of beds increased by 83 percent (from ­ 3 .37 to 6.18 million); and admissions grew ­ ­ 1.7-fold FIGURE 5.1  Growth in total health and hospital spending in China, (from ­ ­ 40.1 ­ 51.1 to 1 million). China currently 2005–13 has more beds per 1,000 population than do 3.5 several OECD countries, including Canada, the United Kingdom, and the United States 3.0 (­figure 5.2). Although beds-per-population Expenditure, RMB, trillions 2.5 ratios are, on average, decreasing throughout 2.0 the OECD countries, they continue to rise in 1.5 ­China. With their dominance in the health care 1.0 landscape, hospitals are the point of entry 0.5 into the health system for most Chinese seek- 0 ing ­care. More than half of patients’ first con- 2005 2006 2007 2008 2009 2010 2011 2012 2013 tacts with the delivery system for an illness Total hospital expenditure Total health expenditures episode, and 40 percent of all outpatient vis- its, occur in ­hospitals. In urban areas, more Source: China Health Statistical Yearbook, National Health and Family Planning C ­ ommission. than half of outpatient visits occur in L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 163 hospitals, compared with less than 20 percent FIGURE 5.2  Trends in hospital beds per 1,000 population in China in rural ­ areas. and selected OECD countries, 2000–13 Within the hospital subsector and broader delivery system in China, public hospitals are 4.5 the major providers of health ­ care. However, 4.0 official statistics show that in recent years United States United Kingdom Beds per 1,000 population nearly all hospital growth has occurred in the 3.5 private sector, while the number of public hospitals has declined ­ (figure 5.3). Even so, 3.0 Canada most of the gains in bed numbers have 2.5 occurred in public facilities (figure 5.4). Although public hospitals make up just 2.0 China under half of all hospitals ( ­47.4 percent), they account for the vast majority of beds 1.5 (80.6 percent, compared with ­ ­ 19.4 percent 1.0 in private f­ acilities). Similarly, most inpa- 06 07 08 09 10 11 12 13 00 01 02 03 04 05 tient admissions (88 percent) occur in 20 20 20 20 20 20 20 20 20 20 20 20 20 20 public hospitals, which account for more than 85 percent of health ­ p rofessionals. Source: “Statistical Bulletin of Health and Family Planning Development in 2014,” National Health and Family Planning C ­ ommission . Thus, most public hospitals are relatively Note: OECD = Organisation for Economic Co-operation and ­ Development. large facilities, averaging 310 beds in 2014, while private facilities are much smaller, FIGURE 5.3  Growth in number of hospitals in China, by ownership averaging 67 ­ beds. type, 2005–13 Hospital dimensions are trending upward, especially in tertiary hospitals, which are 18 15,483 mainly public and located in large urban 16 14,309 13,850 13,396 areas (figure 5.5). These trends are mirrored 14 Hospitals, thousands 11,313 by statistics on utilization: admissions are 12 growing faster at tertiary hospitals than at 10 8 7,068 secondary hospitals (­ figure 5.6). 5,403 6 The trend of ever-larger public tertiary 4 hospitals is driven by a number of factors 2 330 (some of which this report addresses else- 0 where), including the following: 2005 2008 2010 2013 Public Private • Social insurance systems that favor inpa- tient over outpatient care ­ ommission. Source: China Health Statistical Yearbook, National Health and Family Planning C • Demand by a rising middle class for spe- cialty and high-tech care (KPMG 2010), As reviewed in chapter 3, information on which is heavily concentrated in tertiary the quality of hospital care is s ­ parse. Even hospitals less is known about hospital efficiency and • Need for improvement in primary care productivity. Although the average lengths of ­ networks that are staffed by less-well- stay are declining, they remain longer in qualified medical professionals China than in most OECD countries • Need to strengthen functional links (10.4 days and ­ ­ 8.9 days in China’s secondary (including referral systems) between hospi- and tertiary facilities, respectively, compared tals and primary care providers with ­8.1 in OECD ­ countries).5 Moreover, of • Investment planning practices that favor the 34 countries tracked by the OECD, only hospital construction and expansion 7 registered longer lengths of stay than (Huang ­2009) C hina. Occupancy rates are generally ­ 164 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 5.4  Growth in number of hospital beds in China, by Weifang prefecture (Audibert and others ownership type, 2005–13 2013), Shenzhen city (Yang and Zeng 2014), and a panel of regional hospitals across 30 4.5 provinces (Hu, Qi, and Yang ­ 2012). In addi- 3.87 4.0 tion, Chu, Zhang, and Chen (2015) applied a Hospital beds, millions 3.5 3.01 directional-distance-function approach to 3.0 2.61 2.5 2.30 measure technical ­ efficiency. 2.0 In general, these studies found huge vari- 1.5 ations in technical efficiency,7 with higher- 1.0 0.71 level public hospitals, hospitals in large cit- 0.5 0.14 0.27 0.37 ies (Beijing, Guangdong, and Shanghai), 0 and private hospitals demonstrating higher 2005 2008 2010 2013 efficiency ­s cores. Technical efficiency in Public Private public hospitals is yet to be improved, while scale was the source of low efficiency for Source: China Health Statistical Yearbook, National Health and Family Planning C ­ ommission. private hospitals (because of their small ­s ize). 8 However, Yang and Zeng (2014) FIGURE 5.5  Trends in average number of beds in Chinese reported that scale inefficiency was also secondary and tertiary hospitals, 2008–14 the main source of inefficiency for large public hospitals, suggesting that the afore- 1,200 mentioned large expansion of hospital size may reduce ­ productivity. (That is, there is a 1,000 961 U-shaped cost curve indicating scale ineffi- 905 935 ciencies at both ends of the size ­range.) Avg. number of beds per hospital 875 829 800 719 767 The efficiency scores in Chinese hospitals are much lower than those of hospitals in OECD countries measured using similar 600 ­ m ethods. China’s introduction of health insurance is seen to have contributed to effi- 400 278 291 300 ciency gains, while the use of budgetary 247 264 210 231 subsidies has been associated with lower 200 ­ efficiency. Improvements in efficiency were also found to be related to technological 0 change. More research with larger data sets ­ 2008 2009 2010 2011 2012 2013 2014 is needed before conclusive findings can be Tertiary hospitals Secondary hospitals ­advanced. Source: “Statistical Bulletin of Health and Family Planning Development in 2014,” National Health and Family Planning ­ Commission . Public Hospital Governance: Challenges and Lessons high—at over 100 percent in tertiary facili- from Reform in China and ties, suggesting overcrowded ­ c onditions. Secondary and primary hospitals register Internationally lower occupancy rates of 88 percent and Governance has many meanings, and the 60 percent, ­respectively. term is used differently across contexts and Micro studies applying robust methods to organizational ­ s ettings. Hospital gover- inconclusive. measure hospital efficiency are ­ nance has been defined as “a set of pro- For example, using variants of data envelope cesses and tools related to decision making analysis,6 researchers examined hospital effi- in steering the totality of its institutional ciency in Guangdong province (Ng 2011), activity, influencing most major aspects of L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 165 organizational behavior and recognizing FIGURE 5.6  Growth in admissions to Chinese tertiary and the complex relationships between multiple secondary hospitals, 2006–14 st a keholders” (Du rá n , S a lt ma n , a nd Dubois 2011, ­ 38). 8,000 7,000 Organizationally, governance consists of 6,000 Admissions an organization’s structures and functions 5,000 4,000 that set and enforce policies and exercise 3,000 the ultimate authority for decisions made 2,000 on behalf of the organization and its 1,000 0 owners. Important functions of hospital ­ 2006 2008 2010 2012 2014 governance include defining and reviewing Secondary Tertiary its mission, role, and goals; providing financial stewardship; formulating future ­ ommission. Source: China Health Statistical Yearbook, National Health and Family Planning C strategy; appointing and evaluating the chief executive officer (CEO); ensuring clin- ical efficiency and quality; and representing hospitals,9 this section examines four major the hospital’s stakeholder groups (Coile elements of public hospital reform: ­1994). Reform of public hospital governance • Accountability mechanisms to ensure that aims to better align the policies and perfor- hospitals perform well and to align their mance objectives of the government (as performance with public objectives hospital owner) with the behaviors of hospi- ­ • Incentives facing the organization to sup- tal managers by providing incentives port accountability and accountabilities, which are usually exe- • Degree of autonomy or decision-making cuted through organizational forms such as authority granted to managers governing boards and ­ c ouncils. Although • Organizational makeup and legal status of public hospital reform is highly complex and governance ­models context-specific, globally the trend has been ­ to move away f rom cent rali zed , A fifth element—the quality of managerial command-and-control, direct administra- ­ practices to implement decisions and respond tion of hospitals by government ministries to accountabilities and incentives—is taken and toward more “arm’s-length,” indirect up in the concluding “Recommendations” oversight, which allows more independent ­section. decision making by the hospitals themselves (Huntington and Hort 2015; La Forgia and Hospital Governance Challenges Couttolenc 2008; Preker and Harding 2003; in China Saltman, Durán, and Dubois ­ 2011). Granting greater autonomy to public hos- Public hospital reforms in Shanghai, pitals requires altering how the government Zhejiang, and Sanming aim to affect hospital engages with ­ t hem. It involves putting in behaviors by linking hospital director income place (and enforcing) a new set of account- performance. However, insufficient infor- to ­ ability mechanisms and crafting incentives to mation is available to judge the impact of this support these a ­ ccountabilities. Taken performance assessment system and how it together, these accountabilities and incentives differs from routine systems to evaluate man- foster the alignment of hospital behaviors agers’ ­performance. with government objectives while respecting A ne c dot a l e v id e nc e su g ge s t s t h at the increased decision-making autonomy of Sanming is better at hospital reform imple- ­hospitals. mentation because its directors’ positions Following a framework developed to ana- are at risk based on performance assess- lyze practices in the governance of public ments. Dongyang’s board has established 166 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A a comprehensive hospital-based perfor- matters, yet responsibility for setting prices mance assessment system that embraces and payment mechanisms, allocating human the financial, efficiency, quality, patient resources and capital investments, and pur- satisfaction, and safety ­ d omains. Unlike chasing services is divided among other Dongyang, Shanghai and Zhenjiang do not ministries. independently assess hospital performance T he main form of oversight of the or compliance with rules and standards, bureaus is hierarchical; it is usually applied and they appear to piggyback on supervi- through directives known as “red letters” sory practices performed by government that instruct hospitals to implement public ­agencies. policies and to follow relevant public Sanming’s leadership group carefully administration rules such as for human supervises the implementation of human resource management, use of funds, use of resource, compensation, and pricing ­ reforms. public assets, or product ­ procurement. In Nevertheless, some observers suggest that practice, these directives often provide influencing managers’ behaviors may be dif- ambiguous and sometimes conflicting guid- ficult because managers appear more ance because (a) the functions, responsibili- accountable to the higher-level leaders who ties, and accountabilities of public hospitals appointed them than to the government agen- are not clearly defined and (b) the agencies cies responsible for reform implementation or themselves have unaligned policies and on-the-ground ­performance. diverse interests (Yip and others ­2012). Nor Public hospitals in China require stronger are the directives rigorously enforced— governance arrangements if they are to drive partly because supervision itself is divided improvements in quality and efficiency, pro- across different a­ gencies. mote service integration, and act in the public Although financial reporting is strong, interest rather than respond to vested inter- public hospitals face weak requirements from ests related to revenue generation (Allen, Cao, government and social insurers to improve and Wang 2013; He 2011; Tam 2008; WHO their safety processes, quality, patient satis- and World Bank ­ 2015).10 This subsection faction, or ­efficiency. Improvements along summarizes the main governance challenges these lines are generally not a priority (Tam observed in China’s public ­hospitals. 2008). Hospital directors are rarely moni- ­ tored or sanctioned for noncompliance with Accountability government directives or for failure to meet Mechanisms to hold hospital managers agreed-upon targets (He and Qian ­ 2013). accountable for efficient and high-quality ­ services or for fulfilling social functions need Incentives to be ­developed. Given many Chinese hospi- Hospitals earn a large share of their revenues tals’ underlying incentives to enhance reve- by selling services to social insurers and self- nues via treatment choices and drug provi- paying individuals, usually through fee-for- sion, their managers are oriented toward service payment ­ systems. Surpluses are dis- augmenting service volume and expanding tributed to staff through nontransparent infrastructure, including acquiring high-tech bonus schemes that are based on service pro- ­equipment. duction and revenues, usually at the depart- Diffuse lines of accountability make it dif- ment ­level. ficult to counterbalance these ­ i ncentives. Under these conditions, hospitals and their Hospital directors are in principle account- clinicians have strong incentives to maximize able to multiple government agencies at local revenues by raising service volumes, ordering government ­ levels. For example, the National expensive procedures, selling pharmaceuti- Health and Family Planning Commission cals, providing unnecessary care, generating (NHFPC) is responsible for health-related admissions, and extending patients’ hospital L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 167 ­ tays.11 Given the incentives to capture more s boards or councils have been created or patients, hospitals have little interest to inte- given responsibility to oversee and monitor grate with, or shift care to, lower levels of the hospital activities and performance related ­ unctions. health care system or to fulfill social f to quality, efficiency, or fulfillment of social Meanwhile, hospitals’ revenue-seeking ­functions. behaviors have led to considerable citizen The central government envisages public ­discontent. hospitals as having full decision-making authority over management and operations Autonomy so that hospital operation and management Public hospital autonomy in China has few are separated from the governments’ owner- parallels ­internationally. Most of the hospi- ship rights (State Council 2015a, ­ 2 015b). tals enjoy considerable autonomy in financial Importantly, the central government also and asset management, retaining their finan- envisages putting in place an accountability cial surpluses, opening and closing services, framework by requiring M&E of several expanding or contracting their physical domains of hospital performance (for exam- plants and equipment, and entering into and ple, functions, quality, spending, patient sat- servicing d­ ebts. However, the legacy of “com- isfaction, access, and e ­ fficiency). However, mand-and-control” remains, both with the as in the case of previous reform directives, appointment of senior managers and in the it is not clear how such a framework will be management (conducted directly by local designed or applied, nor who will have government leaders or agencies) of “quota” the authority to assess (and enforce) perfor- personnel. This legacy means that hospital ­ mance and ­ compliance. International experi- managers lack full decision-making authority ence may provide some guidance on reform- to hire, dismiss, and set compensation for all ing governance arrangements for public staff. It may also limit the quality of manage- ­ ­hospitals. ment ­practices. Nevertheless, in light of the weak account- Hospital Governance Challenges and abilities and distorted incentives, some local Lessons from International Experience government officials consider, probably cor- rectly, that merely granting public hospitals In high- and middle-income countries, a more autonomy—or similarly, freeing them range of modalities for hospital governance from the vestiges of hierarchical government have emerged that differ in their legal provi- control—will result in chaos (WHO and sions, financing schemes, accountability World Bank ­ 2015). arrangements, and decision-making ­ rights. All of them operate within and respond to Organizational arrangements specific incentive environments, which vary Most public hospitals in China are governed considerably across countries (Nolte and directly by government b ­ ureaus. As men- Pitchforth 2­ 014). The foundation trusts of tioned earlier, oversight is exercised through the National Health Service (NHS) in the directives issued by different ­ bureaus. It is United Kingdom, nonprofit private founda- each bureau’s responsibility to assess and tions in the Netherlands, regional health enforce implementation and to engage enterprises in Norway, public health care directly with hospital ­ m anagers. Some companies and foundations in Spain, and bureaus may be reluctant to sanction public social health organizations in Brazil, to hospitals for lack of compliance because name a few, reflect different aspects and con- they see the hospitals as extensions of the figurations that public hospitals have government administrative ­ a pparatus. adopted (Durán and Saltman 2015; La Except in a limited number of pilots, no Forgia and Couttolenc 2008; Mossialos and independent supervisory structures such as others 2015; Preker and Harding ­ 2003). 168 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A The impetus for developing these mod- arrangements (such as boards); establishing a els arose from concerns over avoidable more arm’s-length relationship between gov- distortions that resulted from command­ ­ -and- ernment and hospital; making use of indirect control political authority, including the tools of accountability (such as performance following: reviews, compliance monitoring, external audits, use of contracts, and contract man- • Overlap or conflicts of interest between agement) and provisions for their enforce- ownership, regulation, and management ment; and aligning incentives with public (functions that were often consolidated in objectives (Deber, Topp, and Zakus 2004; La a single government agency) Forgia and Couttolenc 2008; Preker and • Political interference in hospital opera- Harding 2003; Saltman, Durán, and Dubois t ion s , e sp e c i a l ly hu m a n re s ou rc e ­2011). management China’s experience along these lines is • Restrictive or inflexible administrative nascent at best (Allen, Cao, and Wang 2013; rules applied to all inputs World Bank ­ 2010). The following discussion • Undersupply of some services and over- summarizes the international experience supply of others with organizational arrangements, auton- • Inefficiencies and budgetary overruns ­ncentives. It focuses omy, accountability, and i • Low quality on three countries: Brazil, the United • Patient dissatisfaction Kingdom, and ­ Spain.12 • Inability to respond to technological change or more generally to change the Organizational arrangements model of care Nearly all reforms are based on or at least • Weak links to the broader delivery ­ system initiated through legislation or ­ regulation. The breadth and depth of legislation or regu- Another driver of greater hospital inde- latory change varies across countries and pendence was a reform movement known as over t ­ ime. In some countries, such as Brazil “new public management” (Greenwood, and the United Kingdom, framework laws Pyper, and Wilson ­ 2002). Reformers sought were issued that supported a single gover- to modernize hospitals’ organizational cul- nance modality that applied to all hospitals ture by introducing a market or quasi-market participating in the r ­ eform. In São Paulo, environment that would enhance competition Brazil, public hospitals with reformed gover- and motivate entrepreneurship and thereby nance were incorporated under civil law as improve efficiency and responsiveness to nonprofit social organizations of “public patients. In some countries, such as Spain ­ interest,” known as social health organiza- and the United Kingdom, reforms that tions (OSSs), following statutes that were set granted greater autonomy to public hospitals out in a 1998 state ­ l aw.13 The United were in part a political compromise to avoid Kingdom used a two-step process for legally privatization. In several countries, the public ­ granting autonomy to public hospitals: self- hospital reforms were part of broader public governing trust status in the mid-1990s administration reforms (as in Brazil) or were superseded by the establishment of founda- a component of health system reforms such 2000s. tion trusts in the early ­ as the separation of purchasing from provi- In contrast, Spain enacted different laws sion and the introduction of alternative pro- for different governance ­ modalities. India vider payment ­ mechanisms. and Panama limited their reforms to a hand- Reforms that shift decision rights to public ful of hospitals that benefited from specific ­ omplex. International experi- hospitals are c laws enacted to make them independent pub- ence suggests that to succeed, they require lic ­entities. putting in place a legal or regulatory frame- Hospital governance models take a wide work; set ting up new organizational range of legal and organizational forms and L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 169 corresponding nomenclatures (table 5.1) and TABLE 5.1  Hospital governance models in are specified in regulatory ­frameworks. The selected countries models emerged in the 1990s and early 2000s Country Hospital governance model and vary considerably in terms of the organi- Brazil • Social health organizations (OSSs) zational structures (and the degree of inde- Czech Republic • Limited liability companies pendence granted to hospitals) that they • Joint-stock companies established to replace hierarchical govern- Estonia • Joint-stock companies ment ­administration. • Foundations Most countries legislated some form of Norway • State enterprises board or council that serves as the unit of responsibility between hospital management Portugal • Public enterprise entity hospitals and government ­ owners. In general, boards Spain • Public health care companies are expected to set overall policies and strate- • Foundations • Consortiums gies, approve and oversee business plans and • Administrative concessions (to a financial matters, monitor performance private firm) against objectives or targets, appoint manag- Sweden • Public-stock corporations ers, and ensure that the hospital acts in the United Kingdom • Self-governing trusts public ­interest. • Foundation trusts Boards can take many roles, forms, and Sources: La Forgia and Couttolenc 2008; Saltman, Durán, and compositions, and they can be responsible for Dubois ­2011. a single hospital, a group of hospitals, and even regional networks of facilities, as several examples show: was formed for a regional network con- sisting of a hospital and ambulatory care • In the United Kingdom, the board of gov- facilities run by a private firm under ernors (BoG) for the NHS foundation ­concession. trusts (F Ts) consists of elected and • In Norway, hospitals are organized as appointed ­ m embers. The BoG in turn independent trusts with boards consisting appoints the hospital governance board of appointed elected politicians and consisting of hospital executives and non- regional health ­ authorities. executives representing various profes- sional ­interests. Autonomy • In Brazil, the Secretariat of Health of the Although in many countries the new gover- State Government of São Paulo (SES) con- nance modalities for public hospitals have tracts with nonprofit organizations granted their managers considerable decision (NPOs) to manage public ­ hospitals. Each rights, few public hospitals can be considered NPO is required to have a board that is fully autonomous and comparable to inde- legally accountable to the ­ government. pendent private ­ entities. Across the world, Board members can be public officials, even substantially autonomous public hospi- representatives of private entities, or pri- tals lack the decision-making flexibility of vate citizens selected by the N­ PO. Boards private ­hospitals. Experience has shown that have not been formed in specific h ­ ospitals. decision-making boundaries are a moving • In Spain, governance boards of various target and depend on shifting political and types have been established in hospitals, financial c­ onditions. Some well-intended and members are appointed by the reforms were not fully implemented, the gov- government. Depending on the modality, ­ ernments proving unwilling or unable to board members are high-ranking public relinquish bureaucratic and political control officials from a specific region, or they can (Huntington and Hort 2015; Preker and be a mix of public officials and private Harding 2003; Saltman, Durán, and Dubois members. In one model in Spain, a board ­ ­2011). 170 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A The FTs in the United Kingdom and the structures that are used to reimburse hos- OSSs in São Paulo, Brazil, enjoy considerable pitals for care or ancillary services such as autonomy in hiring, firing, and compensating medical ­ education. OSSs do not set prices staff as well as in managing inputs, procure- and are not permitted to charge fees to ment, and finances as long as their finances “public” ­ patients. More recently, however, are under control and regulators find their OSSs have negotiated fees with private health care quality and performance to be insurance plans to provide services to acceptable (for the FTs) and compliant with insurance plan ­ members. the SES contractual terms (for the ­ O SSs). They can also retain and invest surpluses and In Spain, an administrative concession to a borrow ­commercially. private joint-venture company constitutes a In both countries, these hospitals are free model that probably gives public hospitals to hire their CEOs and do so based on merit more autonomy than in any other model in ­criteria.14 CEOs are not appointed by the Europe. The hospital has the right to decide ­ ­ government. As public facilities, however, the on all its inputs (including capital invest- hospitals cannot choose their patient mix, ments) and on expanding its services, though and those in both countries must seek gov- its profit margins are capped at 7 ­ ercent. ­ .5 p ernment permission to make major shifts in At the time of this reform, staff of the for- provide. Similarly, plans for the services they ­ merly government-run hospital received the infrastructure expansion and purchases of right to remain civil servants or to become expensive equipment require government nonstatutory ­ staff. All new staff are nonstat- ­approval. utory, with compensation and benefits set by Some differences between the FT and OSS the private company awarded the ­ concession. models are worth noting: In the other governance modalities in Spain, hospitals enjoy less ­ autonomy. For • Infrastructure ­planning. For FTs but not example, foundations and consortiums have the OSSs, infrastructure expansion usu- decision rights over inputs and investments ally requires a multiyear planning exercise but follow public procurement ­ r ules. They that is set by the economic regulator and have freedom to hire and fire staff but have ultimately the national Department of only limited authority to determine workers’ ­Health. income levels other than bonuses at the • Business ventures and a ­ ssets. FTs can set ­margin.15 Spain’s public health care compa- up joint ventures and subsidiary busi- nies have limited ­ autonomy. They employ nesses but cannot sell core assets such as statutory workers who are subject to civil ser- land and buildings, because these assets vice ­rules. Because their boards mainly com- are locked to prevent privatization and prise regional health officials, regional health cannot be used to guarantee debt or be departments still appear to exercise consider- sold to pay c ­ reditors. OSSs are not permit- able ­control. ted to sell shares or seek ­ investors. • Labor ­c ontracts. FTs have the right to Accountability depart from nationally determined labor Putting in place sound accountabilities is contracts and pay scales for their medical arguably the essential tenet of governance professionals and unionized staff, though reforms and is an important driver of ­ results. none have done ­ so. OSSs are free to set the To be sure, the granting of greater autonomy labor contracts and compensation for all needs to be accompanied by the implementa- their ­staff. tion of strong and enforceable accountability • Price ­setting. For service price setting, FTs mechanisms to orient hospital behaviors supposedly have more freedom, but in toward improved performance, compliance practice they are price takers of the cen- with social functions, and alignment with trally determined tariffs and other price gover n ment ­ p r ior it ie s. I nter nat iona l L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 171 experience suggests that the success—judged Commission, is responsible for ensuring com- by the achievement of public policy goals—of pliance with regulatory standards for quality any public hospital reform involving auton- and patients’ safety in all public and private omy depends on the effectiveness of account- health care ­facilities.16 FTs are also answer- ability ­mechanisms. able to other regulatory bodies in specific Typically, hospitals are held accountable areas such as financial management, medical by a number of agencies (such as government education, and fertility ­treatment. owners, payers, and regulators) in several Annual ­reports. Finally, FTs are required areas (such as use of resources, performance, to produce publicly available annual reports compliance with rules, procedures, and stan- on financial status, patient engagement activ- dards) through specific checks and balances ities, and a range of quality measures cover- (such as reporting requirements, inspections, ing adverse events, infection rates, mortality involvement). It audits, contracts, and citizen ­ rates, patient feedback, staff views, and per- is worth repeating that the key to checks and formance against ­ t argets. With few excep- balances is their rigorous enforcement, tions, public hospitals in England have a which, as explained below, requires a robust strong focus on public service and on fulfill- information s ­ ystem. Boards are the usual ing social functions, though this focus is interface between oversight agencies and the more implicit and embedded in social hospitals and are ultimately held accountable arrangements than ­ codified. for hospital behaviors and ­performance. Spain United Kingdom In Spain, boards and managers are held In the United Kingdom’s FTs, oversight accountable by regional health authorities focuses on board performance and account- through audits, inspections, and reporting ability and is conveyed through three mecha- requirements on financial status, quality, and nisms: the BoG, Monitor, and annual ­ reports. other i­nformation. Compulsory minimum Board of g ­ overnors. The BoG for FTs data sets established by legislation for all holds the nonexecutive directors (NEDs) on hospitals (autonomous and nonautonomous) the hospital governance board (including the are a main feature of information tracking chairman) accountable—both individually and reporting in Spain that incorporates and collectively—for performance, financial indicators of activities, accessibility, and reporting, quality, and other ­aspects. In turn, p erformance. The detailed topics covered ­ the NEDs hold the hospital executive, includ- and the periodicity and comprehensiveness ing the CEO, to a ­ ccount. The roles of board of reporting vary across regions according to chairman and CEO are separate, which is an rules and procedures set by regional public essential feature of British governance in authorities. Most autonomous models apply ­ both the public and private ­ sectors. ­ a program contract arrangement (contrato Monitor. The government created two programa) between the regional health ser- agencies. One is an FT-specific eco- oversight ­ vice executive and facilities as the preferred nomic regulator, known as Monitor, that is mechanism for funding and ­ accountability. responsible for licensing FTs, monitoring their Contractual terms specify reporting require- financial performance, assessing their achieve- ments and other documentation that hospi- ment of nationally set targets (such as for tals must send to the regional health service waiting times) and compliance with FT laws, ­executive. and gauging their quality of g ­ overnance. Accountability arrangements also vary by Monitor does not oversee (or prescribe) model; more explicit accountability measures how the targets are m ­ et. Monitor also has have been put in place for those hospitals responsibilities to foster competition, set sys- granted greater a­ utonomy. For example, in temwide prices, and ensure continuity of c ­ are. the concession model the contract between Another oversight agency, the Care Quality the private operator and the government 172 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A requests specific information from the opera- standardized information system that tor including well-structured reports on mon- links all OSSs with the SES management itoring indicators, detailed business and unit to facilitate performance monitoring; financial records, and clinical reports based much information is placed in the public on a large set of i ­ndicators. Reporting ­domain. requirements in less autonomous models such • Independent review c ­ ommission. The SES as public health care companies and founda- established an independent review com- tions are related to performance assessments mission consisting mainly of civil society for payment p ­ urposes. Public health care representatives to conduct “social audits” companies, foundations, and consortiums of OSS operations to ensure that hospitals use some intermediate economic indicators are fulfilling their social obligations (such based on statements of income and expendi- as not charging patients, not denying care, tures, but they give a strong simultaneous and not referring patients for unnecessary role to budget ­monitoring. Consortiums pro- services) and that their behaviors are vide monthly reports on waiting lists and aligned with the mission of a public quarterly reports on financial ­ status. ­hospital. Brazil Incentives Accountability mechanisms for Sáo Paulo’s The behavior of any hospital is driven by OS S s consist of several i nterlock i ng incentives, whether monetary or nonmone- components: tary. Some incentives are embedded in the way hospitals and staff are paid, but others • Performance ­m easures. Production tar- are ingrained in the system c­ ulture. Examples gets, quality benchmarks, and rules are include the centrality of free care and dedica- specified in a legally binding “manage- tion to public service observed in the United ment contract” between the SES and each Kingdom’s ­N HS. OSS; OSS boards are legally accountable and supervise hospitals’ compliance with United Kingdom contractual ­terms. The United Kingdom attaches considerable • Purchasing and contract ­ m anagement. weight to the use of “codes of ­ b ehavior.” The SES created a dedicated service pur- Though these codes appear voluntary and chasing and contract management unit are self-policed, they are laid on top of that reviews and analyzes OSS perfor- many rules—for example, on care ­ quality— mance and rule compliance and negotiates that have the force of l ­aw. Clinical regula- ­budgets. tions are strong and e ­ nforced. FTs are at • Contract and performance ­ e nforcement. financial risk for budgetary ­ overruns. If an The contract management unit enforces FT is perceived by Monitor, the economic compliance with the management contract regulator, to have become financially unvi- and performance measures through finan- able, it is taken under central ­ c ontrol. sanctions. Poor performance can lead cial ­ Importantly, some of the most prestigious to withholding of funds and cancellation hospitals in the United Kingdom have not of the government contract (and retender- become FTs, in part because of a poor track ing to another ­N PO). record of staying within their financial • Audits. The SES conducts internal audits, ­envelopes. and the state’s comptroller general con- All FTs are paid using fixed prices under ducts external audits to verify the perfor- an activity tariff for standard procedures, mance data and financial statements together with a series of other payment mech- submitted by the ­OSSs. anisms for specialized activity, emergency • Standardized information ­ s ystem. The care, medical education, and earnings from SE S developed and installed a research. The standard tariff—calculated ­ L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 173 from national average retrospective cost the vast majority of primary care referrals are information for diagnosis-related groups either to the local facility or, for patients with (DRGs)—is the dominant portion of the rev- more serious conditions, to the nearest spe- enues of district general h ­ ospitals. More- cialty unit, so it is unlikely that patient choice complex tertiary units earn much more of is a big driver in ­performance. their income from specialized tariffs and from ­teaching. Such hospitals also inevitably Brazil attract the bulk of research funding, whether São Paulo’s OSSs face powerful incentives to privately sponsored or public, and benefit meet performance (and productivity) targets, from other allowances for such factors as improve quality, and align their behavior location in major city ­ centers. with public ­priorities. Importantly, OSSs are To the extent that standardized national at financial risk for budgetary overruns and DRG-based prices underlie the tariff, there is poor ­performance. no incentive—indeed the opposite—to favor The global budget is performance-driven expensive treatments, though it remains true in two ­respects. First, 90 percent of it is paid that the greater the activity, the greater the monthly and is linked to meeting production revenue. That said, both self-governing trust ­ or volume targets for specific services (such and FT hospitals have framework contracts as inpatient, outpatient, diagnostic, and sur- with their local purchasing agencies that cap gical procedures) as specified in the manage- their total revenues (hence the fact that so ment ­ contract. If hospitals skimp on produc- many ­ U.K. hospitals are now running at a tion, they are financially ­ s anctioned. For financial ­loss). example, if a hospital only meets 70 percent In FTs, hospital professional staff, particu- of a production target for, say, surgical larly senior ones, have reasonably high sala- procedures, its financing is reduced by ries that are little affected by their ­ activity. 30 ­percent. In contrast, hospitals have few In general, the FTs (or more broadly, the incentives to oversupply services because they NHS) have few incentives to order extra do not receive a financial benefit if they diagnostic tests, buy more drugs, or carry out exceed production targets (except under interventions judged unnecessary on clinical extenuating circumstances such as an g rounds. Staff are paid for extra sessions ­ epidemic). Second, 10 percent of the budget is ­ beyond the standard in their contract, and placed in a retention fund and paid in quar- modest bonuses may be paid, but generally terly allotments against achievement of clinicians face a broadly flat income through- benchmarks for efficiency and quality (such N HS. out the ­ as for infection control, mortality rates, Hospital staff also have options to treat length of stay, and ­readmissions). private patients within and outside the Importantly, both sets of benchmarks hospital. A study found that, on average, ­ are enforced, and monthly and quarterly income from private patients provides budgetary allocations are reduced for an about 2 percent of an FT’s total income— OSS hospital that fails to meet ­ t hem. The suggesting that profit-making activities are state government pioneered the use of stan- minimal at ­ b est.17 Equally, staff do not dardized cost accounting systems that were share in profits or surpluses created at the installed in each OSS hospital with a vir- level of the h ­ ospital. When a change in tual link to the SES’s contract management internal management or a gain in efficiency ­ u nit. The cost data allow hospital manag- reduces costs for a department, most of the ers to monitor the costs of all inputs in each resulting surplus will stay on the hospital’s department and allow the contract man- central ­books. agement unit to compare costs across all Nor do patient choice and competition facilities and services, analyze efficiency appear to be strong incentives for individual and productivity, and negotiate global clinicians or for o ­ rganizations. For example, ­budgets. 174 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A The availability of cost data has shifted according to established fees (based on the nature (and transparency) of annual bud- ­DRGs). get formulation away from the more or less The capitation payment provides a strong arbitrary setting of ceilings to a calculus incentive to strengthen primary care, avoid based on volume and c ­ osts. Having informa- unnecessary hospital admissions and ser- tion on volume and costs also allows the SES vices, and integrate hospitals with primary to monitor potentially opportunistic behav- providers to provide cost-effective care at the iors such as reducing high-cost ­ services. appropriate ­location. Given that the conces- OSSs cannot provide bonuses to their sionaire is at financial risk for covering managers, who receive fixed salaries; they the cost of patients seeking care outside can provide bonuses to staff, but none have the region, it faces a strong incentive to done s­ o. There is no evidence that OSSs pay strengthen quality and safeguard patients’ higher salaries than traditional public ­experiences. hospitals. OSS managers prefer to recruit ­ high-quality staff who “fit” well with the International experience: Ingredients for organizational culture, and they quickly dis- successful public hospital reforms miss nonperformers (World Bank ­ 2006). International experience suggests that reforms of public hospital governance appear Spain to work best under the following conditions: In Spain, the hospitals managed under the concession model are the only ones that face • A social consensus on the role of public significant financial r ­ isk. The Alzira region in health care Valencia provides a noteworthy e ­ xample. • A strong legal framework that defines the This initiative involves a public-private part- organizational forms of governance, includ- nership in which the concessionaire, a private ing roles, functions, and accountabilities company, operates a 300-bed public hospital • Full decision rights for hospital managers along with 40 public primary health care to hire, fire, promote, and shape workers’ centers in a region, serving approximately incentives (through remuneration, work 250,000 people (NHS European Office hours, rewards, and sanctions) and to ­2011). manage all other inputs The concessionaire must offer universal • S ou nd ac cou nt abi l it y me cha n ism s access and is responsible for nearly all care together with effective enforcement mea- provided to the population; it receives a fixed sures to ensure performance and compli- per capita payment from the regional ance with public objectives government. Service obligations and other ­ • The right incentives to align hospital and responsibilities are specified in a ­ contract. staff behaviors with desired performance The government has the right to audit and and outcomes and, most critically, to avoid inspect facilities to ensure compliance with adverse actions such as private rent-­ regulation and contractual terms as well as to seeking via increasing clinical activity, impose ­ sanctions. The concessionaire must sales diagnostic tests, or drug ­ cover all expenses from the capitation pay- ment, including amortizations, investments, The trend in the United Kingdom has payroll, and other operating expenses for the been to promote greater independence of network of f ­acilities. Medical staff receive public hospitals but with strengthened salaries that contain a fixed (80 percent) and accountabilities. The sustainability of Brazil’s ­ variable (20 percent) ­ component. The latter is OSSs and of the two Spanish models (conces- paid depending on achievement of targets for sions and consortiums) may be related to the access and ­ quality. If patients seek care out- strong contractual relationships with private side the region, the concessionaire is respon- partners, which involve transfers of financial sible for paying the full cost of that care risks. In contrast, public health companies ­ L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 175 and foundations have been closely linked to efficiency, in part because they granted hos- politicians and public administrators through pitals only partial decision-making author- board membership; over time, political inter- ity; in these cases, political meddling contin- ference in the management of these hospitals ued and accountability was lax (Saltman, increased, legislation was amended, and Durán, and Dubois ­ 2011). But despite their managerial autonomy ­ decreased. limited effect, the European reforms were In the United Kingdom, where many hos- deemed “reasonably successful” (Saltman, pitals are in financial difficulty, the eco- Durán, and Dubois 2011, 71), and they have nomic regulator, Monitor, is placing less been more or less embraced by governments, emphasis on converting self-governing trusts managers, and ­ citizens. To be sure, in many to FT ­status and more emphasis on support- countries, the movement toward greater hos- ing ­hospitals in financial ­d istress. Monitor pital independence will be difficult to and other government agencies have relied ­reverse. on   “ N H S prov id e r s”— m e mb e r s h ip -­ representative organizations of public and Emerging Models of Hospital private ­hospitals—that actively support hos- Governance Reform in China pitals in the early stages of their transition to FTs to help with the challenges of organiza- Typical of broader Chinese reform measures, tional redesign and to provide training to local experiments in public hospital reform board ­members. were to serve as the basis for formulating pol- In general, though the evidence is not icies and the subsequent rollout of successful definitive, public hospital reforms that models to address the aforementioned embraced the above components have been challenges. Seventeen cities were identified to ­ shown to increase efficiency, quality, and launch public reform pilots in ­ 2010. patient satisfaction (La Forgia and Couttolenc This section examines the opportunities 2008; McKee and Healey 2002; McPake and and constraints in hospital governance in others 2003; Preker and Harding 2 ­ 003). China, drawing on selected cases commis- A recent study examining hospitals in sioned for this ­ report. Following the analyti- England and the United States found a strong cal framework presented earlier, table 5.2 relationship between the governance prac- shows the major components and character- tices of hospital boards and quality processes istics of governance models being piloted in and ratings (Tsai and others ­ 2015). Recent four mostly urban areas: Shanghai-Shenkang, analyses also show that the Alzira conces- Z h e nji a n g - K a n g f u , D o n g ya n g , a n d sionaire model in Spain provides more ­S anming.18 The findings are presented by efficient and better-quality care than com- ­ ­component.19 parator hospitals, as measured by lower read- mission rates, shorter waiting times, higher Organizational arrangements productivity, and higher patient satisfaction The Shanghai and Zhenjiang models are typ- (NHS European Office ­ 2011). ical of governance reforms observed in the Other studies found that governance hospitals that are piloting reform in China, reforms involving autonomy had little effect building upon initiatives that were launched (Allen and others 2012; Govindaraj and in Wuxi, Wifang, and other cities in the early Chawla 1996), although it is possible that and mid-2000s (World Bank ­ 2010). These the reforms featured in these studies involved agencies— cities legislated the creation of new ­ only limited autonomy and little strengthen- usually referred to as hospital management ing of accountability mechanisms to improve centers or councils (HMCs)—that are led by performance (Castaño, Bitrán, and Giedion high-level municipal officials and consist of 2004; Preker and Harding ­ 2003). Indeed, representatives of public agencies that are several European public hospital reforms involved in health sector operations or in have been found to have little effect on ­oversight. 176 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 5.2  Characteristics of selected reform models for Chinese public hospital governance, 2015 Component Shanghai (Shenkang) Zhenjiang (Kangfu) Dongyang Sanming Organizational arrangement Organizational unit or Multihospital Multifacility Hospital board Prefecture health reform name management center “network” council leadership group Jurisdiction Municipal Municipal Municipal Prefecture Number of hospitals 24 5a 1 22 Autonomy Hiring and firing hospital Partialb No Yes Yes director Hiring and firing quota staff No No No No Flexibility in setting No No Yes Yes remuneration for quota staff Pricing No No No Yes Residual claimant Noc Noc Yes Partial Asset management Partiald Partiald Yes Yes Accountability mechanisms Performance assessment of Yes No Yes Yes hospital directorse Performance assessment of No No Yes No hospital(s) Review and enforcement of No No Yes No safety and quality standards Compliance supervision Similar to other Similar to other Yes Partial public facilities public facilities Incentives Realigning staff incentives Partial No Yes Yes Governance unit: members’ No No No No position at risk Hospital directors’ position No No Yes Yes at risk Sanctions for Similar to other Similar to other Yes Yes noncompliance with rules or public facilities public facilities for low performance 2014. ­ Sources: Li and Jiang 2015; Li and Wang 2015; Ma 2015; Ying ­ a. Network includes nine ambulatory ­centers. b. Can recommend to g ­ overnment. ­ level. ­ c. Retained at hospital ­ d. In consultation with government ­agencies. director. e. Usually involves signing a “responsibility agreement” between governance unit and hospital ­ Staffed by civil servants, the HMCs were and performance of the participating granted legal personality, but their member hospitals. The one in Zhenjiang shares this ­ hospitals also maintained their original objective but also aims to promote greater personalities. The goals of the HMCs legal ­ vertical integration among a mix of facili- in these t wo cities differ. The one in ties at the tertiary, secondary, and primary Shanghai aims to improve the operations ­levels. L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 177 The Dongyang pilot involves a single policies for hospital directors and all medical h ­ ospital that formed an independent board ­ professionals. For example, the government with government and nongovernment directly pays the salaries of hospital direc- ­ participation. The board consists of represen- tors, and physicians’ income is no longer tatives of government agencies, private cor- linked to revenue-based ­ bonuses. The LG porations, and local and foreign medical also makes all decisions regarding major schools. The hospital has special legal status, ­ asset investments, prices, and reimbursement and its statutes are similar to those in the cor- rates for social insurance (Ying ­2014). porate governance models observed in private ­hospitals.20 The pilot aims to create a corpo- Accountability rate governance arrangement that will Three of the governance models aim to influ- improve the efficiency, capacity, and quality ence hospital behaviors by linking hospital of services while maintaining the hospital’s directors’ income to ­ performance. However, public nature through “social responsibility” it is uncertain how the performance assess- (Li and Wang ­ 2015). ment system in force differs from routine sys- Sanming has not created a new agency but tems and whether it is sensitive and specific has decreed a fully empowered leadership enough to capture improvements in an objec- group (LG) to enact health system reforms tive and reliable ­ w ay. Another question with an initial focus on the prefecture’s 22 relates to the observation that hospital direc- tertiary and secondary hospitals (Ying tors appear more accountable to the higher- ­2014).21 level leaders who appointed them than to the government agencies that are responsible for Autonomy reform implementation or on-the-ground In Shanghai and Zhenjiang, key decisions on performance (Qian ­ 2015). human resource management, staff compen- Among the four cases studied here, only sation, and service pricing remain with gov- Dongyang’s board has established a compre- ernment ­ a gencies. Hospitals have largely hensive hospit a l-based per for ma nce -­ retained their residual claimant status and assessment system that spans the financial, their ability to manage their own ­ assets. In efficiency, quality, patient satisfaction, and Zhenjiang, entities were unwilling to relin- safety domains (Li and Wang 2 ­ 015). In fact, quish control to the HMC, which spans inde- Dongyang’s is the only governance arrange- pendent municipal and district public admin- ment within the group to mandate a continu- istrative units (Li and Jiang ­ 2 015). As a ous quality improvement program—which result, the district administration has retained has entailed establishing a quality manage- control of asset management in primary care ment department, a committee to control units and some secondary hospitals, while medical records, and a clinical assessment the municipal administration, together with system for physicians and ­ nurses. hospital management, controls asset manage- The Shanghai and Zhenjiang HMCs do ment in larger ­ hospitals. not independently assess hospital perfor- In contrast, Dongyang’s independent hos- mance or compliance with rules and stan- pital board and Sanming’s LG exhibit consid- dards, and they appear to piggyback on the erably more decision ­ rights. Dongyang still supervisory practices performed by govern- abides by government pricing policies and ment agencies (Li and Jiang 2015; Ma rules that govern the hiring and firing of 2 015). Dongyang’s board and Sanming’s ­ quota staff, but the board determines the LG are fully empowered to apply sanctions full compensation package for all staff ­ themselves. In practice, however, Sanming’s (Li and Wang 2 ­ 015). Sanming’s LG has sanctions have centered on noncompliance assumed full decision-making authority with LG-initiated r ­ eforms. The LG care- except over the recruitment and dismissal of fully supervises the implementation of the quota s ­ taff. It has altered compensation hu m a n re sou rc e , comp ensat ion , a nd 178 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A pricing reforms it has crafted, but it relies organizational arrangements has emerged, on government agencies to supervise other but most are pilot i­ nitiatives. Each has aimed domains (Ying ­2014). to consolidate decision making within a sin- gle entity by coordinating the actions and Incentives policies of diverse government departments Dongyang and Sanming have placed physi- responsible for the health ­ sector. The failure cians on salaries and delinked their bonus to rigorously evaluate these models, and income from the revenues the hospitals derive public hospital reform interventions in gen- from sales of drugs, medical supplies, and eral, limits the ability to draw lessons from diagnostic t­ests. The salaries contain fixed the pilot experiences (box 5.1). and variable components, with the latter Variants of the H MC model imple- being unrelated to revenues but linked to a mented in Shanghai and Zhenjiang are combination of indicators of productivity, common in other pilots in China (World cost control, quality, and patient satisfaction Bank ­ 2 010). In practice, the HMC mem- (Sanming Prefecture Government ­ 2015). As bers are all current or former government suggested above, altering physician compen- officials and appear to behave more as sation was a major accomplishment and has extensions of government than as indepen- addressed a cost-escalating distortion widely dent ­agents. Additional decision rights have observed in public ­ hospitals. not been transferred to the HMCs from Shanghai has placed a hard budget con- government ­ a dministration. Meanwhile, straint on total personnel spending, but this hospitals under HMC governance maintain measure has not clearly affected the bonus their financial autonomy and do not appear system. Surplus revenues are still distributed ­ to be answerable to the HMCs on financial to physicians at the discretion of hospital and m atters. Arm’s-length tools and mecha- ­ departmental directors (Ma ­ 2015). nisms to foster accountability have not In Shanghai and Zhenjiang, partly because been developed, and the HMCs mostly rely most of the governance units consist of gov- on the direct supervisory and oversight ernment officials, member facilities’ positions mechanisms traditionally operated by the are not at risk if they perform poorly or do relevant government a ­ gencies. Nor do the not comply with government policies (Li and H MCs have a robust track record in Jiang 2015; Ma ­ 2 015). In Dongyang and realigning ­i ncentives. Sanming, by contrast, hospital directors can In contrast, Dongyang Hospital has many be dismissed (by the board or the leadership of the features of corporate governance group, respectively) for poor performance (Li observed ­ internationally. It is operated by an and Wang 2015; Ying ­ 2014). independent board with members from both In terms of residual claimant status, the government and nongovernment bodies and Sanming LG has set caps on reimbursements enjoys considerably more decision-making for inpatient stays and outpatient visits; if authority than the ­HMCs. The board has cre- hospitals spend below the caps, the savings ated strong accountability mechanisms and are shared with the social insurance schemes incentives to control costs and improve quality (Ying 2014) and patient ­ experience. However, the origins of the Dongyang governance model, which involved external partners that provided The Limits of Hospital Governance financing and technical support, may be diffi- Models cult to reproduce elsewhere in ­ China. After These four cases suggest that public hospital 20 years of existence, the model has yet to be governance reforms in China have made ­ replicated. As in the OSS model in Brazil and important advances but have not conclu- the concessions and consortiums models in sively separated hospital management from Spain, having an external partner may help to government ­ administration. An array of defend (and advance) the reform ­ model. L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 179 BOX 5.1  Impacts of public hospital reform in China Given that none of the following cases has been rig- ator hospitals both before and after the intro- orously evaluated, it is difficult to distinguish the duction of the zero-markup policy in ­ 2012. Its results of public hospital reforms from the results of average outpatient and inpatient costs were sig- other health reforms—such as the policy specifying nificantly lower than those in a sample of com- zero markup on drug ­ sales. parator hospitals over the same p ­ eriod. Restric- Based on available but limited administrative tive use of antibiotics has exceeded national data, salient results of each case are as follows: ­ s tandards. The hospital reports a surgical-site • Shanghai: Compared with similar hospitals in infection rate of 1 percent and an overall noso- Shanghai, the HMC’s member facilities showed comial infection rate of 7 ­ p ercent. These rates no difference in trends in service volume (inpatient monitored. are closely ­ discharges), lengths of stay, or asset expansion • Sanming: For outpatient visits and inpatient from 2003 to ­ 2013. admissions, government reports show signifi- • Zhenjiang: Compared with provincial and national cantly lower growth in Sanming’s hospital spend- averages between 2009 and 2013, ­ Zhenjiang was ing between 2009 and 2013 than the average for able to contain charges and spending, outpatient vis- Fujian province and the national ­average. The cost admissions. The reduced growth in its, and inpatient ­ of an average inpatient stay fell by ­ percent 3.9 ­ spending may relate to the consolidation of steriliza- in Sanming, while it rose by ­ 11.6 in Fujian and tion, pathology, and logistics services for all network by ­14.7 percent ­nationally. Drug revenues as a members by the hospital management ­ council. proportion of Sanming’s total hospital revenues • Dongyang: Dongyang’s income from drug sales decreased from 47 percent to 28 percent between was 10–15 percent lower than that of compar- 2011 and ­ 2013. 2014. Sources: Li and Jiang 2015; Li and Wang 2015; Ma 2015; Ying ­ Facing a financial crisis, high-level leaders ­ ustainability. Despite intense promotion by s in Sanming created the LG with a broad the central government, the Sanming model reform mandate, though with an initial focus China. has yet to be replicated in ­ on addressing cost escalation in the prefec- hospitals. The LG can best be described ture’s ­ as an arrangement for health system gover- Hospital Managerial Practices: nance rather than as a model of hospital Challenges and Lessons from governance. The LG was granted full auton- ­ omy and authority to alter accountabilities China and Internationally and incentives to foster more efficient use of How hospital managers respond to the resources. To its credit, the LG took hospital ­ accountabilities and incentives embedded in on the complex and deep-rooted issues that their governance and organizational environ- distorted the incentives facing h­ ospitals. The ments is a key determinant of hospital hospitals themselves were not granted greater ­ performance. Managerial practices can be decision-making authority, nor was a plat- defined as “the set of formal and informal for m of ac cou nt abi l it y me ch a n i sm s rules and procedures for selecting, deploying, formalized. The LG did consolidate decision ­ and supervising resources in the most effi- making across multiple institutional a ­ ctors. cient way possible to achieve institutional The group lacks institutionalization, and in objectives” (Over and Watanabe 2003, the future, administrative rotation of key 1 22–23). Hospital management entails a ­ officials may erode its effectiveness and wide range of clinical and nonclinical 180 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A functions related to selecting, using, and public health care companies in Spain and supervising ­resources. may have led Spanish regional governments The effectiveness of managerial practices to backtrack on governance reforms (Durán ­ erformance. is a key determinant of hospital p ­2015). This section first briefly reviews the interna- In sum, although the evidence is incon- tional literature on the relationship between clusive, there is a broadly held belief that hospital management and performance and management matters for improving hospi- then examines what is known about manage- tal performance and that developing mana- rial practices in Chinese public ­ hospitals. It gerial capacities is an important goal for concludes with a brief review of innovative health systems and ­ h ospitals. How to managerial practices in small subset of develop such capacities and create the private hospitals in ­ ­ China. enabling organizational environment for their effective application is an important emerging concern of health systems in Management Matters: International high-income countries (Lega, Prenestini, Evidence and Experience and Spurgeon ­ 2013). Studies of hospital management in several In additional to traditional managerial countries have shown that better manage- training, managerial measurement and orga- ment practices are associated with better nizational tools borrowed from other indus- outcomes, quality of care, and financial per- tries are increasingly used in hospital settings formance (Bloom and Van Reenen 2010; internationally to improve ­ p erformance. Lega, Prenestini, and Spurgeon 2013; These tools include total quality management McConnell and others 2013; Tsai and oth- (TQM); the plan-do-check-act (PDCA) cycle; 2015). In a recent systematic review of ers ­ “lean management”; the balanced scorecard the literature, Parand and others (2014) (BSC); and the 5S cycle (discard, arrange, found a positive relationship between man- clean, standardize, and ­discipline). While the agement and promoting improvement in evidence of their effects is mixed, hospital quality and patient safety, but more research managers consider these tools effective for is needed to better understand how this fostering organizational and behavioral occurs in ­practice. For example, the impact change (La Forgia and Couttolenc 2008; of managerial practices on outcomes may be Mazzocato and others 2012; Naranjo-Gil mediated by other factors such as the degree ­2009). of physician engagement with management, leadership styles, commitment, and organi- Managerial Practices in Public Hospitals zational culture (Curry and others 2011; in China Kirkpatrick and others 2009; Mannion, Davies, and Marshall 2005; Parand and Information on managerial practices in others ­2014). Chinese hospitals is ­scarce. Available studies A recent empirical study of hospitals in are generally qualitative, small in scale (usu- England and the United States has shown ally based on a single hospital), or focused on that good governance is associated with a single managerial function such as staff better managerial practices (Tsai and others performance, patient-flow management, or 2015). Professional management was intro- ­ application of managerial tools such as bal- duced into English hospitals in the ­ 1980s. anced ­ scorecards. This section reviews the High managerial competence contributed to literature on hospital management practices, the success of England’s FTs and Brazil’s drawing on surveys and case ­ studies. 22 We OSS-operated hospitals (Edwards 2011; La first examine what is known about manage- Forgia and Couttolenc ­ 2008). In contrast, ment in public hospitals and then turn to low managerial capacity may be a contrib- management practices in successful private uting factor in the poor performance of ­hospitals. L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 181 A recent pilot study aimed to systemati- • Management practices appear reactive in cally measure management practices in a the sense that hospitals lack systems to small sample of secondary (35) and tertiary find and prevent potential problems or to (75) public hospitals across 27 provinces continuously improve their processes and (Liu ­2015). 23 Researchers applied a meth- ­services. odology known as the World Management • Because of a lack of autonomy in staffing, Survey (WMS), which was originally devel- managers have little authority to dismiss oped to measure managerial and organiza- low performers; talent management is not tional practices in manufacturing but has a high priority, and there are few conse- subsequently been applied to and validated quences for poor ­ performance. in hospitals in several countries (Bloom and • Hospitals do not systematically analyze Van Reenen 2007, 2010; Bloom and others performance data or use data to provide 2010; McConnell and others 2 ­ 013). The feedback for ­ improvement. survey consists of questions on 20 manage- • Lack of standardization of care may indi- ment practices across four major manage- cate deficient clinical management, which ment domains: standardizing care and can negatively affect quality and o­ utcomes. operations, target setting, performance • Performance management is mainly used monitoring, and talent ­ m anagement. The to allocate staff bonuses, not to improve research team interviewed 291 department individual or hospital ­ p erformance. directors and head n ­ urses. Following the Interestingly, autonomy (defined as author- WMS methodology, practices were scored ity to make decisions on human resource, on a scale of 1 to 5 for each of the asset, and financial management) was asso- 20 ­practices. A higher score indicates better ciated with higher management ­ scores. ­performance. The weighted average management score The research also sought to analyze the in China was found to be 2 ­ .68, with a highly determinants of score variation, accounting dispersed distribution ranging from ­ 1.85 to for the following factors: competition (num- 3.35. Compared with OECD countries where ­ ber of hospitals within a 30-kilometer the WMS has been applied, China is an aver- radius); hospital characteristics (age of facil- age performer, scoring lower than the United ity and average number of beds); location (by Kingdom ­ (2.86) and the United States ­ (3.0) region); economics (per capita gross domes- but higher than France ­ ( 2.4) and Italy tic product [GDP]); and autonomy (decision- ­(2.48).24 making authority on human resource, asset, Figure 5.7 displays the average scores for and financial ­ m anagement). The results each management practice across the four showed that bed count, competition, and domains. Not surprisingly, secondary hospi- ­ autonomy were associated with higher man- tals scored significantly lower ( ­2.66) than agement s ­ cores. How competition may tertiary facilities ( ­ ­2.90), and considerable improve managerial practices was not variation in scores was observed across directly ­examined. ­ provinces. Hospitals scored the highest in use Research on specific management prac- of human resources, promotion of high tices supports a subset of the findings from ­ performers, performance review, and attract- the WMS-based ­ study. For example, surveys ing talented staff, but they scored the lowest show that BSCs are widely applied in China, in standardization and protocols, continuous usually by tertiary public hospitals affiliated improvement, consequence management, with medical schools (Lin, Yu, and Zhang rewarding high performers, and removing ­ 2014). Typical of BSC application interna- poor ­performers. tionally, Chinese hospitals reported that they The scores, combined with findings from used BSCs to develop performance measures interviews, highlighted several managerial along four dimensions: financial, patient sat- shortcomings: isfaction, service quality, and research and 182 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 5.7  Average managerial practice scores of Chinese hospitals, 2015 a. Standardizing care and operations b. Performance monitoring 5.0 5.0 4.5 4.5 4.0 4.0 3.5 3.5 3.0 3.0 Score Score 2.5 2.5 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0 0 ck nce vi e g ce en e e s t of di ci f oc n so m f em ou re anc em nc ar du e o re hu e o ot io alo n ien ut t ow za ng Go ls ur an Pe nt ov u di rma ag ue tra ma Pe ng Pe ew Co ue pr at nd tro nal us o rm pr tin at ayo d diz an q St tion s i r od ce sta in tio m nse rfo rfo rfo im Con an dar L Ra an tp ou c. Target management d. Talent management 5.0 5.0 4.5 4.5 4.0 4.0 3.5 3.5 3.0 3.0 Score Score 2.5 2.5 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0 0 ce Tim tion t et t Re rm igh nt nt nt rg f of rab and om m r M orm igh ec ge str arge ta n o Pr for poo ale ale ale lan ch m ers o ers ag ers et rfo h nn Tar rg y rf h izo ta ilit pa ty pe ding gt gt gt T ba pe ting pe ing s m ari or et in in in et eh ov co Cl in ct r rg wa r tra ta rco an Ta Re Re At te in Source: Liu ­2015. Note: Scores obtained using the World Management Survey methodology, which scored practices of 110 hospitals on a scale of 1 to 5 for each practices. A higher score indicates better ­ of the 20 ­ performance. ­ raining. BSCs were found to contribute to t management tools to improve their opera- improved organizational performance and tions and ­quality. For example, the provin- worker s ­ atisfaction. However, another sur- cial hospital in Guangdong has used “lean vey, albeit small in scale, found that the use management” techniques to improve the of BSCs in Chinese hospitals focuses on efficiency and throughput of operating financial performance rather than on quality rooms (Guo, M a, a nd Zha ng ­ 2 014). and patient satisfaction (Gao and Gurd Dongyang Hospital has applied manage- ­ 2014). The tool has mainly been used for ment tools such as TQM, quality-control assessing physicians’ performance to deter- circles, and the PDCA cycle to improve clin- mine their ­ bonuses. ical and nonclinical processes through- Evidence, mainly from single-site case out the hospital (Li and Wang 2 ­ 015). studies, shows that hospitals in China Although TQM and PDCA have been a re usi ng i nter nat iona l ly recog n i zed applied in ­ s pecific Ch inese facilities L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 183 (Chen and others 2014), the extent to which Though not discussed in the summaries, unknown. they are used elsewhere is ­ each case presents evidence that management As mentioned earlier, public hospital practices affect quality, patient satisfaction, executives are appointed by higher-level and ­finances. However, the evidence has not authorities of the party and government, been independently ­ verified. Similar manage- and the appointments are not merit b ­ ased. rial practices may also be evident in public Even within hospitals, promotions are usu- hospitals, but research is needed to document ally based on years of tenure and are not management ­ i nnovations. For example, as determined ­ c ompetitively. Most hospital previously mentioned, the public hospital in managers have received lit tle formal Dongyang has pioneered a continuous qual- t raining. For example, a 2004 study of ­ ity improvement program, and a number of managers in 96 hospitals across 21 prov- public hospitals use BSCs to monitor inces found that less than one-third had ­performance. received short-term professional training, while more than half had learned manage- Aier Eye Hospital Group ment through their work experience (World The Aier Eye Hospital Group (AEHG) is a Bank ­ 2010). Presidents of public hospitals medical care company operating 70 ophthal- are typically responsible for all managerial, mological hospitals across China in 2014 and clinical, and academic activities and tend to employing more than 5,0 0 0 medical manage during their “spare time” or to del- p rofessionals (physicians, nurses, and ­ egate managerial functions to junior s ­ taff. t echnicians). The A EHG has adopted ­ There are no standards or qualification c orporate-governance arrangements and ­ s ystems for hospital managers, and most ­ management structures typical of private see managerial know-how as something hospital systems in OECD countries, in that requires investment by government which major decisions and several key func- authorities rather than by the hospitals tions (such as procurement, human resource ­themselves. management, and asset management) are centralized at the headquarters level (Chen, Gao, and Wang ­ 2015).25 Managerial Practices in Private A board of directors that oversees all Hospitals operations of the conglomerate has four ­ Several elite private hospitals have adopted specialized committees: strategic, audit, com- ­ management practices and service models pensation, and managerial s ­upervision. that set them apart from other private (and Management is structured in seven main public) hospitals in China and make them domains: strategic investments, medical more similar to hospitals in OECD ­ countries. devices, marketing, medical management, Three examples are reviewed here: the Aier operations, human resources, and f ­inance. Eye Hospital Group, Foshan Chancheng Individual hospital management teams are Hospital, and Wuhan Asia Heart ­ Hospital. appointed by the board and are responsible To maintain and expand their market posi- for developing and implementing business tion, these facilities have created enabling plans that are aligned with corporate objec- organizational and managerial environments tives and ­ policies. The AEHG also applies a of continuous improvement to provide high- “unified management framework” in which quality care ­ efficiently. Though they are not individual facilities submit standardized representative of private hospitals in China, reports, including financial statements, cus- these facilities can serve as learning platforms tomer analysis, and production and quality, for improving clinical and nonclinical man- ­ ntervals. Monthly audits are conducted at set i agement in public and other private ­ hospitals. on five randomly selected ­ hospitals. Noteworthy management practices are high- AEHG management emphasizes high- lighted ­below. qualit y care and patient ­ s atisfaction. 184 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A For example, each of the group’s hospitals is established a “head nurse home” program in required to operate six quality-related com- which head nurses engage with patients and mittees, covering medical quality control, their families upon admission, advising about quality of nursing management, ethics, phar- their conditions, treatments, and care pro- maceutical management, hospital-infection cesses and responding to any ­ concerns. Head control, and medical-records ­ management. nurses also make arrangements for postdis- These committees are responsible for moni- charge care in outpatient ­ settings. toring and assessing relevant practices, and The hospital also introduced a “one-card they conduct inspections twice a year as well solution”: a smart card linked to a mobile as monthly ­ reviews. Every month each hospi- application (WeChat) that facilitates a num- tal submits a report on medical quality to the ber of processes including one-stop payment, leading quality group at ­ headquarters. Each appointment making and reminders, access department operates a quality-control team to diagnostic examination results, use of composed of a director, deputy director, and mobile drug-dispensing machines, and inqui- head nurse; the team’s responsibilities include ries or questions to medical p ­ rofessionals. reviewing and reporting adverse e ­ vents. The According to FCH managers, the one-card AEHG also has crafted standard clinical solution has significantly reduced waiting pathways and operating procedures, includ- time for all outpatient services, including ing checklists for diagnostic services, surgical d iagnostics. A free shuttle bus transports ­ procedures including for presurgical and fragile elderly patients to and from their postsurgical care, infection control, prescrip- appointments and therapy ­ sessions. tions, and medical ­records. The FCH uses a BSC system to assess staff Several measures are in place to enhance and departmental ­ performance. This system patient satisfaction and communication, incorporates a wide range of indicators including internet-based appointment sched- related to economic efficiency, patient trust, uling; flexible, walk-in appointments for the quality of care, patient safety (reduction of frail and elderly; WeChat registration; a limit medical errors), and staff professional of 40 outpatients per doctor per day; a limit growth. Physicians’ and nurses’ bonus com- ­ of 20 minutes on patients’ waiting time; pensation is tied to their BSC ­ scores. In addi- access to a 24-hour hotline; free shuttle ser- tion, FCH takes medical complaints from vice for people with disabilities; and a patients and families s ­ eriously. It has insti- patient-service team to inform patients about tuted a root-cause analysis program that risks and involve them in their treatment and investigates complaints, errors, accidents, and ­recovery. factors that contributed to the event and rec- ommends and enforces corrective ­ measures. Foshan Chancheng Hospital In 2004, Foshan Chancheng Hospital (FCH) Wuhan Asia Heart Hospital was converted from a public hospital to a Wuhan Asia Heart Hospital (WAHH) oper- nonprofit ­facility. Shares are divided among ates a 759-bed facility specializing in the employees and private ­ i nvestors. Although treatment of cardiovascular ­ d iseases. In FCH is a general hospital, it is known for 2014, it had a staff of 2,000, including 450 specialty care in maternity, pediatrics, ortho- physicians and 850 n ­ urses. Partly to keep pedic surgery, and ­ urology. In 2013, the FCH within the reimbursement rates of social had 700 beds and 1,300 ­ staff. insurance schemes, WAHH has established The FCH has a number of initiatives ori- cost-control measures (Chen and Gao 2015): ented toward making it more “people-­ oriented and patient-centered,” by developing • Uniform bar coding and tracking of all a hospitalwide culture of fostering positive drugs and consumables, records of which experiences for ­ patients. For example, it has are affixed to a patient’s medical record L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 185 • Standardized cost accounting, which Recommendations for Moving determines costs per procedure Forward with Public Hospital • A tracking system that identifies devia- tions of 20 percent or more from stan- Reform dard costs and triggers an audit in such There is broad agreement in China that cases (management aims to keep such deeper reforms are needed to improve hospi- deviations to fewer than 5 percent of tal performance in cost control, quality of patients) care, and patient ­ s atisfaction. Alternative • Package pricing for ambulatory and short- models of governance and improvement in stay surgical procedures managerial practices are only two pieces of a • Application of clinical pathways complex hospital reform puzzle that involves • Strict pharmaceutical procurement reforms in financial arrangements, human and management, which keeps total resources, planning, and service ­ integration. drug spending to less than 14 percent These latter themes are taken up in other and 25 percent of treatment costs for volume. chapters of this ­ inpatients and outpatients, ­ ­ respectively. International and Chinese experience sug- gests that there is no single path to public Both WAHH and the AEHG use a “dif- hospital reform, but emerging models have ferentiated management model” in which common elements: clinical and academic responsibilities rest with the facility president and nonclinical • Establishing (and enforcing) accountabil- matters are handled by a general ­ m anager. ity mechanisms In public hospitals, the hospital president • Crafting strong incentives to align behav- has both clinical and nonclinical iors with performance objectives and pub- responsibilities. This pattern is replicated ­ lic priorities at the departmental level: each departmen- • Developing sound organizational arrange- tal clinical director has an administrative ments for governance manager to support nonclinical activities • Increasing the decision rights of hospital and facilitate coordination with other managers d epartments. Each department also has a ­ • Strengthening managerial ­capacities medical assistant—usually a nurse—who is responsible for communication with This section recommends specific imple- patients, post-discharge follow-up, and mentation strategies in each of these core promotion of patients’ ­ s elf-management. areas, drawing on the Chinese (Shanghai- For example, WAHH has established a Shenkang, Zhenjiang-Kangfu, Dongyang, patient service center that makes appoint- and Sanming) and international (Brazil, ments, provides medical advice, under- Spain, and United Kingdom) case studies as takes triage, and assists patients with navi- literature. The core actions well as the general ­ g at ion a mong t he ho spit a l’s s er v ic e and corresponding implementation strategies ­departments. are summarized in table 5.3. For long-ter m postd ischa rge c a re , Like the other types of reforms recom- WAHH operates a “life link” follow-up mended in this report, public hospital reform program in which about 60 percent of dis- requires a unitary vision of comprehensive charged patients ­ p articipate. It has also ­ reforms. The international models described developed standard operating procedures here evolved over t ­ime. In the United to facilitate patient flows and the division Kingdom, hospital governance reform bene- of staff responsibilities in outpatient clin- fited from previous health system reforms, ics, diagnostics, wards, and operating including for the professionalization of man- ­rooms. agement, the separation of purchaser from 186 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 5.3  Five core action areas and implementation strategies for improving public hospital governance and management Core action areas Implementation strategies 1: Strong accountability • Specify the rules, reporting requirements, and other mechanisms to foster strong mechanisms for autonomous hospital accountability to government, including contracts, financial management, public hospitals to strengthen audits, patient-safety processes, and performance requirements performance • Set up institutional arrangement to support monitoring and oversight • Determine the information to be publicly disclosed • Establish effective enforcement mechanisms 2: Incentives aligned with public • Gradually place hospitals at financial risk for budgetary overruns and low objectives and accountabilities performance (for example, quality, efficiency, and patient satisfaction) • Develop payment systems that promote cost consciousness • Install standardized cost-accounting systems in hospitals and use the results in budget setting 3: Sound organizational • Develop organization form(s) that support the proposed governance model(s) arrangements for public hospital • Codify the model within a legal framework governance • Set the functions, roles, and composition of governance entities such as boards and councils 4: Gradual delegation of decision • Establish a formal application and approval process for hospitals to achieve (more) rights to hospitals autonomous status under the to-be-determined governance model • Set a phased timetable for transferring functions that are currently managed by government bureaus but are to be shifted to the (approved) autonomous hospitals • Create formal mechanisms to assist hospitals in developing governance arrangements and subsequently to assess effects and provide feedback for adjustments 5: Managerial capacity building • Assess the skills of hospital managers, the quality of managerial practices, and their effects on the quality and efficiency of hospital operations and services • Study and adapt managerial practices implemented in leading public and private facilities • Establish an executive management program for upgrading skills along several dimensions • Support demonstration projects that address specific managerial challenges • Develop a career path for professional hospital managers and integrate managerial and leadership competencies into recruitment and promotion practices • Work with academic institutions to strengthen and expand degree programs in hospital management and ultimately to establish centers of excellence in management and leadership development • Create a benchmarking system for hospital management that periodically tracks indicators of various dimensions of management and links them to important indicators of hospital performance—not to evaluate management but to proactively find problems and improve practices as a means to improve hospital performance p ­ rovider, and the introduction of managed learned from failures in earlier hospital gov- pricing. It also benefited from high clinician ­ ernance reforms and involved strengthening remuneration in hospitals without either fee- the state’s mixed delivery system, consisting for-service or profit making and from strong of public purchasing of private ­ provision. clinical regulation and ­ enforcement. In São Even with these advantages, São Paulo took Paulo, Brazil, the OSS model was introduced more than five years to introduce the account- as part of broader public administrative ability arrangements and incentives that sup- reforms. Its design took account of lessons ­ port the governance ­model.26 L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 187 Though it is difficult to disentangle any experience both in China and internation- one of the five core action areas from the oth- ally suggests that this would not be the case ers, accountabilities and incentives are clearly if sound indirect mechanisms for account- crucial. Establishing robust accountabilities ­ abilit y a re established and sk illf u lly and powerful incentives to strengthen perfor- ­ deployed. It is important that regulations mance and align hospital behaviors with pub- and rules specifying accountability mecha- lic objectives—Core Action Areas 1 and 2, nisms are strong and enforceable, and that respectively—underpin the remaining core measures are put in place to support their action areas, which relate to putting in place ­enforcement. effective organizational governance models Recommended strategies to strengthen (Core Action Area 3), strengthening auton- accountability include the establishment of omy (Core Action Area 4), and improving rules and other mechanisms to foster managerial practices (Core Action Area ­ 5). accountability; the implementation of institu- Without strong (and enforceable) account- tional arrangements to support monitoring abilities and appropriate incentives, it is and oversight; the determination of publicly unlikely that emerging organizational disclosed information; and the establishment arrangements will represent the interests of of enforcement ­mechanisms. government or ­ p atients. Instead, greater autonomy may stimulate deviant behaviors Strategy 1: Specify rules, reporting and greater distancing from public priorities, requirements, and other mechanisms to foster and there will be little demand for better strong hospital accountability to government managerial ­ practices. Finding a workable “Arm’s-length” accountability mechanisms balance between decision-making autonomy applied to autonomous public hospitals usu- and accountability is no easy ­task. Indeed, no ally entail rules and compliance ­ monitoring. hospital, whether public or private, can act They include internal and external audits of outside the interests of its ­owners. Planners board appointments and operation, account- must find a pragmatic formula for combining ing and financial reporting, quality and these elements while accounting for local safety standards, patient outcomes, and ful- context and ­ capacities. Implementers must fillment of social ­ functions. also display a willingness to make the neces- Reporting requirements, performance tar- sary ­adjustments. gets, and other checks and balances are increasingly embodied in legally enforced con- tracts between the hospital board (as owner) Core Action Area 1: Strong and government (as the service ­ purchaser). Accountability Mechanisms For example, as evident in the British, A fundamental component of hospital reform Brazilian, and Spanish cases examined earlier is to put in place sound accountability mech- in this chapter, contracts often are the instru- anisms to orient hospital behaviors toward ment used to allocate resources, set perfor- improved performance, compliance with mance requirements, assess compliance with social functions, and alignment with govern- government regulations, and mandate the ment ­ priorities. International experience sug- integration of care with lower-level ­providers. gests, for example, that the success of any The overall content, terms, and management public hospital reform involving greater of contracting mechanisms can be specified in autonomy depends on the effectiveness of regulations. However, as discussed below, ­ accountability ­mechanisms. good contracts require strong contract man- In China, many observers consider that agement, monitoring, and e ­ nforcement. granting public hospitals more autonomy, or, similarly, freeing them from direct Strategy 2: Set up institutional administrative control, would result in even arrangements to support monitoring and mor e u n c o n s t r a i n e d ­ b e h av ior s . B ut oversight 188 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Oversight agencies or similar units within of monitoring indicators as well as financial current agencies play key roles in fostering ­statements. accountability and enforcing ­ c ompliance. Some countries have found it necessary to Strategy 4: Establish effective enforcement establish new public units to deploy newly mechanisms established tools for indirect accountabil- The effectiveness of accountability mecha- ity; some have built contract management nisms depends on their effective ­enforcement. units to negotiate, monitor, and enforce In São Paulo, every year, the SES refuses pay- accou ntabilit y requ i rements derivi ng ment of all or part of the retention funds from the receipt of public funding; and (10 percent of a hospital’s agreed financing some have established units to support the envelope) for OSSs that have not achieved establishment and operation of hospital agreed-upon performance ­ b enchmarks. ­boards. Continued poor performance spanning two In the United Kingdom, as noted earlier, years results in cancellation of the SES con- Monitor oversees financial performance, tract with the NPO that operates the OSS legal and regulatory compliance, and the hospital and the selection of another NPO achievement of national targets, while the through an open tendering ­ process. Care Quality Commission reviews compli- In China, many of these tools for indirect ance with standards and policies for quality accountability are in use, though apparently and patient ­safety. In São Paulo, the SES’s in only a handful of ­hospitals. For example, purchasing unit manages and monitors com- the director of Dongyang Hospital signs pliance with contracts, while an indepen- a performance agreement with the board, dent review commission comprising citizens, wh ich l i n ks t he d i rec tor’s sala r y to academics, and government representatives performance. Continued underperformance ­ conducts social audits of OSS operations would put the director’s position at ­ r isk. that focus on compliance with social Financial accounts are audited internally by ­functions. the hospital board and externally by the Dongyang Audit ­ Bureau. The board assesses Strategy 3: Determine the information to be the hospital’s performance on a series of indi- publicly disclosed cators reflecting cost containment, quality, In the United Kingdom, the FTs are required and ­efficiency. to produce publicly available annual reports on financial status; patient-engagement activ- Core Action Area 2: Incentives ities; and a range of quality, patient safety, Aligned with Public Objectives and and performance measures including adverse Accountabilities events, infection rates, mortality rates, patient feedback, staff views, and perfor- Hospital behaviors are shaped by incentives, mance against ­ targets. which are usually embedded in how hospitals In São Paulo, the OSSs report on unit and staff are ­ paid. As discussed in chapter 6, costs and service production by department, all payment systems come with a set of under- and on metrics related to quality, patient lying ­incentives. However, hospitals may also ­ fficiency. The state government safety, and e respond to nonfinancial incentives that are has installed information systems in autono- embedded in the culture and behaviors of mous public hospitals to enable validated medical care organizations and the broader reporting of performance and costs, and it delivery ­ s ystem. For example, the United makes all data publicly available on the SES Kingdom’s NHS contains embedded incen- ­website. tives, apparently shared by all or most of its In Spain, hospitals governed under the personnel, that support a culture of public concession model must provide the govern- service, free care, and to a certain extent, eth- ment with detailed and updated dashboards practices. Such a culture may not be so ical ­ L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 189 evident in other s ­ ystems. Professional staff that is performance-driven and sets targets also respond to incentives related to growth for volume, quality, and e ­ fficiency. Hospitals and career ­paths. have no incentive to either oversupply or The following strategies are recommended undersupply ­ services. For example, if hospi- to support the alignment of incentives with tals skimp on production, they are financially public objectives: (a) gradually placing hospi- sanctioned; if they exceed production targets, tals at financial risk for budgetary overruns they are not financially compensated except and low performance in quality, efficiency, under extenuating circumstances such as an and patient satisfaction; (b) developing ­ epidemic. As mentioned earlier, 10 percent of p ayment systems that promote cost con- ­ a hospital’s budget is placed in a retention sciousness; and (c) creating the institutional fund and paid in quarterly allotments against capacity in government to monitor perfor- meeting efficiency and quality benchmarks mance and enforce ­ sanctions. (such as for infection control, mortality rates, length of stay, and readmissions). These Strategy 1: Place hospitals at financial risk for benchmarks are strictly ­ enforced. budgetary overruns and low performance To facilitate oversight, each OSS has a The United Kingdom’s FTs and São Paulo’s standardized cost-accounting system with a OSSs are at financial risk for budgetary over- virtual link to the SES’s purchasing and con- runs and for meeting performance ­ goals. tract management ­ unit. This unit helps hos- However, each country takes a different pital managers to monitor the costs of all ­approach. inputs in each department and uses the data Like many other OECD countries, the to compare costs across all facilities and ser- United Kingdom pays its hospitals, including vices, analyze efficiency and productivity, FTs, using a DRG-based payment system and negotiate global ­ budgets. The availabil- spending. DRGs with a cap or ceiling on total ­ ity of cost data has shifted the nature (and bundle discrete services provided to a patient transparency) of annual budget formulation with a specific diagnosis into a single pay- away from more or less arbitrary setting of ment and thereby promote the efficient use of ceilings to a calculus based on volume and services. Use of this system gives little incen- ­ costs. Countries using DRG-based systems ­ tive to offer more intensive treatment, extend have also strengthened and standardized a patient’s length of stay, or favor expensive their cost-accounting systems, and France t herapies. A ceiling on total spending also ­ and Germany provide financial incentives to provides a disincentive for a hospital to hospitals to comply with cost-accounting increase the number of patients (and thus the standards (Busse and Quentin ­ 2011). volume of DRG ­ payments).27 Several countries have also integrated cri- Core Action Area 3: Sound teria on quality and outcomes into DRG sys- Organizational Arrangements for tems to allow payments to vary according to Public Hospital Governance p erformance. In the United Kingdom, for ­ example, a percentage of a hospital’s annual Organizational arrangements or forms such income (from DRG payments) is adjusted as boards or councils serve as the organiza- according to the hospital’s performance in tional interface between government (as own- meeting the quality goals specified in its con- ers) and hospital management and staff (as tract with regional authorities (Mason, Ward, service ­providers). Though they may vary and Street ­ 2011). As reviewed in chapter 6, considerably in their composition, size, roles, DRG-like initiatives are already under way in functions, and degree of independence from ­China. government, they constitute the key organi- São Paulo’s OSSs have taken a different zational element of the broader accountabil- approach to aligning financial ­incentives. The ­ overnance. ity framework for public hospital g state government establishes a global budget The nature of any organizational form is 190 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A woven into the fabric of the governance legislated the OSS as a new governance model itself and codified in a legal or regula- model, constituting it as a nonprofit organi- tory ­framework. Boards or councils legally zation of “social interest” (utilidad publica)— exercise the ultimate authority for decisions a form of public benefit ­ organization. In made by or on behalf of the o ­ wners. In Spain, whose government is heavily decen- China, developing such arrangements is nec- tralized, multiple models for the governance essary to achieve reform objectives related to of public hospitals have emerged, supported the “separation of government administra- by regional legal ­ frameworks. In many coun- tion from hospital management” (Burns and tries, individual public hospitals have been Liu ­2017). granted independence through special and Recommended strategies in this core facility-specific ­legislation. following: action area are the ­ Internationally, governance reform mod- els specify an organizational form that is 1. Develop organizational form(s) that sup- formally and legally responsible for hospi- port the proposed governance ­model(s) tal behaviors and ­ p erformance. Nearly 2. Codify the model within a legal f ­ ramework all countries have legislated a hospital-level 3. Set the functions, roles, and composition body—and, in some cases, a multihospital of governance entities such as boards and body—to replace the direct government ­councils. ad m i n i s t r at io n of publ i c ­ h o spit a l s . Composition, tenure, and selection criteria (and process) are also specified in law or in Strategy 1: Develop organization forms that corresponding ­regulations. support the proposed governance models The degree of separation from the govern- Legislation usually specifies the major ment administrative apparatus varies, as evi- elements of a new public hospital model ­ denced by board ­ membership. In Spain, gov- including the components outlined in this er n ment of ficials dom i nate boa rd chapter: organizational forms, accountabili- membership in the public health companies ties, incentives, and ­autonomy. It character- and foundations, suggesting strong govern- izes the legal structure of the model as a legal ment participation in decision m ­ aking. The entity separate from ­government. governance bodies of the OSSs in Brazil and Depending on national law, different legal the concession models in Spain mainly draw forms have emerged in different countries, their members from private companies and including nonprofit corporations, public ben- nongovernmental o ­ rganizations. Other coun- efit corporations, nongovernmental organiza- tries have a mix of public, private, and citizen tions, joint ventures, public foundations, and representation, as in the FTs in the United t rusts. In nearly all cases, hospitals under ­ Kingdom and Dongyang Hospital in ­ China. these legal models are considered government The FTs represent a special case in which entities but are governed separately and inde- local people, staff, and patients can become pendently from the ­ government. The degree FT members and elect board members, while of gove r n m e nt p a r t i c ip at ion va r i e s the other board members are appointed by ­significantly. ­government.28 In China, most hospital boards established Strategy 2: Codify the governance model thus far (such as those for hospital manage- within a legal framework ment councils) consist exclusively of public Legislation in the United Kingdom created a ­ officials. However, a recent State Council new single national model of hospital gover- policy directive requires that governance nance, the FT, which is established as a pub- boards or councils should consist of a broader corporation. In Brazil, in the con- lic benefit ­ range of participants, including representa- text of a national legal framework for public tives of government agencies, delegates of the administrative reform, the State of São Paulo Pe ople’s C ong re s s , m e mb er s of t he L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 191 Communist Party of China, and representa- implementation is highly c ­ omplex. Decision- stakeholders.29 tives of relevant ­ making boundaries vary considerably, can shift over time, and depend on political and Strategy 3: Set the functions, roles, and conditions. Relative to autonomous financial ­ composition of governance entities public hospitals elsewhere, public hospitals in Finally, the functions and responsibilities of China already enjoy a high degree of finan- hospital boards or councils must be clearly cial autonomy, but they have less freedom of ­ delineated. The vast literature in hospital action in human resource ­ management. trade and practitioner journals includes man- Some observers consider autonomy to be uals and codes that provide guidance on the “acid test” of public hospital governance board functions, responsibilities, and opera- because it involves drawing a line between, tional procedures (Bjork 2006; Bley and on the one hand, higher-level policy decisions Shimko 1987; Cuervo-Cazurra and Aguilera and, on the other hand, frontline operational 2004; Gage, Camper, and Falk 2006; Gill, decisions involving quality, efficiency, and Flynn, and Reissing 2005; Holland, Ritvo, responsiveness that should be exempt from and Kovner 1997; Rice ­ 2003). higher-level scrutiny, oversight, and political The major board functions include (a) pro- interference (Saltman, Durán, and Dubois viding financial stewardship, including ­ 2011). As seen in the Brazilian and British approval of business plans, budgets, and cap- cases, implementation of autonomy requires ital spending; (b) formulating strategies; (c) years of negotiation and adaptation while evaluating management; (d) monitoring hos- putting the accountability and incentive pital performance (for example, to ensure framework in ­ place. efficient service production and clinical qual- Recommended strategies to implement ity and compliance with social functions); greater decision-making autonomy include and (e) responding to stakeholders such as the following: ­ government, payers, and patients (Coile 1994; Edwards ­ 2011). 1. Establish a formal application and It is important that the regulatory frame- approval process for hospitals to achieve work as well as hospital governance statutes (more) autonomous ­ status. specify responsibilities such as determining 2. Set a phased timetable for transferring the organization’s mission and objective; functions currently managed by govern- selecting the CEO and assessing the CEO’s ment bureaus to the (approved) autono- performance; ensuring effective planning and mous ­hospitals. adequate resources; monitoring programs 3. Create formal mechanisms to assist hospi- and services; ensuring ethical integrity; main- tals in developing governance arrange- taining accountability to government and ments and subsequently to assess effects other stakeholders; and recruiting, training, and provide feedback to shape the inevi- and assessing new board members (Ingram table ­adjustments. 1996). It is recommended that governance ­ bodies (boards or councils) recruit and select Strategy 1: Establish an application and hospital managers based on merit criteria and approval process for (more) autonomous according to processes that are specified in status regulations and hospital ­statutes. Once the governance model and correspond- ing legal form(s) are in place (see Core Action Area 3), China may consider developing a Core Action Area 4: Gradual Delegation formalized process for hospitals to apply for of Decision Rights to Hospitals more autonomous status according to the International experience suggests that grant- governance model ­ adopted. This process will ing public hospitals greater autonomy is not specify which functions are to be transferred an all- or-nothing endeavor and that from the public administrative apparatus to 192 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A the (approved) hospitals; the timetable for accountability mechanisms than they are their transfer; and approval procedures by accustomed to—particularly at the early govern ment agencies and ­ r egulators. ­ stages. Support will be needed to establish Functions can include hiring, firing, and the design and composition of the governance compensating staff; managing inputs; open- organization (board or council), member ing and closing services; procurement and selection and tenure and preparation of inter- financial management; investments; borrow- nal statutes (aligned with the legal and regu- forth. ing; and so ­ latory ­framework). The government may also want to con- Strategy 2: Set a timetable for transferring sider developing programs for induction of, functions from the government to and continuing support for, board ­ members. autonomous hospitals Such programs would focus on major Of equal importance, it is the government’s governance themes including strategic and ­ role to set the prerequisites that a hospital business p ­ ­ lanning, performance monitoring, must satisfy to gain fully autonomous s ­ tatus. financial management, conducting legal and China may want to examine processes used regulatory responsibilities, and interacting for this purpose in other ­ countries. For exam- with ­managers. ple, the United Kingdom has established an China could consider examples of how application and approval process for hospi- OECD countries provide technical support tals to achieve FT status, which qualifies and training for governance arrangements them for greater decision-making ­ autonomy. in public and nonprofit ­ o rganizations. The process specifies a three-phased approach In the United Kingdom, NHS Providers involving approval procedures and responsi- (a membership organization and trade asso- bilities of government agencies and regulators ciation of NHS health care organizations) ­(Monitor). provides technical and training support for The process is based on the FT model cod- the FTs, including training programs for law. Applicants are appraised against ified in ­ NHS FT governors, preparation programs criteria in domains that include the quality of for hospitals seeking FT status, and activi- care, financial viability, business strategy, ties and courses to help FT boards fulfill governance structure, service performance, their functions and ­ responsibilities. 30 and stakeholder ­ relations. Once a hospital In the United States, a company called the has met the criteria, a formal agreement is Governance Institute provides training, signed setting out the accountabilities, roles, workshops, and “how-to” materials to help and responsibilities of the FT, government, board members successfully perform their and ­regulators. roles at ­U.S. nonprofit ­ hospitals. 31 And in China will need to set a selection and New Zealand, a specialized group within the approval process and to define stages or Treasury Department supports the effective- degrees of autonomy, with corresponding cri- ness of newly appointed board ­ members. It teria for ­each. Transparency in selection and advises and manages the selection and induc- approval is ­critical. tion process for board members of state-run enterprises through its “Appointment, Strategy 3: Create mechanisms to help Induction, and Professional Development hospitals develop governance arrangements, Program” (COMU ­ 2011). assess effects, and provide feedback Finally, the government may consider provid- Core Action Area 5: Managerial ing active technical support to hospitals to Capacity Building assist with their introduction of new gover- nance arrangements, helping them to prepare Hospitals in China face challenges to improve the way for more autonomous decision mak- quality. China is moving their efficiency and ­ ing and to comply with more rigorous forward with reforming hospital governance L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 193 and separating hospital operations from the Table 5.4 provides examples of best mana- government’s administrative ­ a pparatus. hospitals. gerial practices for ­ Reforms that seek gains in efficiency and quality are unlikely to succeed without Strategy 3 (short-term): Establish an ­ h igh-quality hospital management (and executive management program for ­leadership). upgrading skills Hospital managers require strong skills Skills should be upgraded along several in planning; setting organizational goals dimensions: (a) standardizing care (for and annual and multiyear plans; allocating example by using checklists, handoff resources efficiently; monitoring perfor- p rotocols, and discharge protocols); (b) ­ mance; setting a functional command chain refining target setting (for example, scope with corresponding accountabilities; and of targets, links among targets, and diffi- ensuring effective systems for managerial culty of achievement); (c) measuring perfor- functions related to financing, human mance (for example, by monitoring errors resources, information and data flows, and adverse events and using continuous logistics and material management, and performance-improvement processes); and quality ­ assurance. Management can be pro- (d) improving talent management (by fessionalized through a variety of short- and assessing senior managers and policies for long-term measures, many of which can be internal recruitment, retention, dismissal, implemented in parallel ­fashion. 32 The fol- and p ­ romotion). The development of capa- lowing are specific actions to professional- bilities applies to both clinical and nonclin- ize management and improve managerial ical executive managers, both of whom ­practices. need first-rate managerial and leadership ­skills. Strategy 1 (short-term): Assess the skills of hospital managers, the quality of Strategy 4 (short-term): Support managerial practices, and their effects on demonstration projects that address specific the quality and efficiency of hospital managerial challenges operations and services Managerial challenges to be addressed The aforementioned World Management include care standardization, infection con- Survey, as well as other available instru- trol, and materials ­ management. Pilot proj- ments, can be applied for this ­ p urpose. ects can apply tools such as the PDSA cycle, These surveys would provide valuable TQM, and “lean management” to improve information to shape government commit- efficiency, raise quality, and improve patients’ ment to managerial improvement and set ­satisfaction. the stage for corresponding strategies and ­actions. Strategy 5 (long-term): Develop a career path for professional hospital managers Strategy 2 (short-term): Study and adapt Recruitment and promotion practices should managerial practices implemented in also integrate managerial and leadership leading public and private facilities ­competencies. For example, case work commissioned for this study examined managerial practices Strategy 6 (long-term): Create a benchmarking in high-end private hospitals that intro- system for hospital management duced a variety of such practices to deliver A benchmarking system periodically tracks high-­ c are. Much can quality and efficient ­ indicators of various dimensions of manage- be learned from the innovations in these ment and links them to important indicators h ospitals. Many of the same skills and ­ of hospital ­performance. It should be used practices used in private hospitals are not to evaluate management but to proac- appropriate for their public c ­ ounterparts. t ively f i nd problems a nd i mprove 194 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 5.4  Best practices in hospital management Management area Sample of specific themes For more information Training and • Accounting for staff demands for training • Becker’s Hospital Review (2012) development • Clinical training for different learning styles programs • Computer-based modules • Interactive programs • Developing physician leaders Building leadership • Identifying capability gaps • HRET (2014) • Expanding the management team • Experimenting with different organizational approaches • Promoting clinician engagement in quality and efficient improvement Talent management • Link between talent management, candidate • NCHL (2010) and succession assessment, and succession planning planning • Use of leadership competency model • “Stretch” job assignments • Professional development plans • Rigorous and repeated assessment of leadership against key metrics Health care • Reducing waiting time • Brandenburg and others (2015) scheduling, • Designing patient flow pathways • Willyard (2006) appointments, and • Maximizing appointment utilization • Becker’s Hospital Review (2010) patient flows • Promoting effective dialogue with patients • Drazen and Rhoads (2011) Clinical management • Integrated approach to improving clinical quality, • Potash (2011) efficiency, and service excellence • AHRQ (2017) • Establishing goals and setting priorities (n.d.) • Jefferson University Hospitals ­ • Using data and strategies • Improving patient safety Patient-centered • Improving responsiveness to patients’ needs • Aboumatar and others (2015) hospital care • Improving discharge experience • Improving pain management Medical teams in • Providing evidence-based feedback of team performance • Salas and others (2008) hospitals • Creating a supportive learning environment • Cocchi (2012) • Enhancing physician-nurse teamwork and relationships • Rice (2014) management practices as a means to improve Notes hospital ­performance. 1. T he d i re c t ive s i nclude S t at e C ou nc i l (2012); “Guidance on Comprehensive Pilot Strategy 7 (long-term): Work with academic R e for m of U rb a n P ubl i c H o spit a l s ,” institutions to strengthen and expand National Health and Fa m ily Plan n ing degree programs in hospital management Commission (Guo ban fa 2015, N ­ o. 38); Ultimately, the goal is to establish centers of a n d “G u i d a n c e o n C o mp r e h e n s ive l y excellence in management and leadership Scaling-Up Reform of County-Level Public ­ development. This may entail revising and Hospitals,” State Council ( Guo ban fa updating curricula, introducing internships 2015, ­N o. ­33). and in-service training for recent graduates, 2. Other components of the public hospital a nd developi ng comp e tencie s across reform agenda, related to payment methods, ­ r ecognized management and leadership financing, and human resources, are the themes of chapters 6 and ­ 7. ­domains. L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 195 3. Except where noted, the sources of the data in 11. The incentive structure facing providers is the this section are China Health Statistical subject of chapter ­ 6. Yearbook (National Health and Family 12. The discussion draws on cases prepared by Planning Commission, various years) and the Durán (2015); La Forgia and Couttolenc “Statistical Bulletin of Health and Family (2008); Saltman, Durán, and Dubois (2011); Planning Development in 2014” (NHFPC and Wright ­ (2015). ­2015). 13. The state law was based on a national frame- 4. Estimate of health expenditures as a share work law that aimed to reform the Brazilian of total health spending is from OECD administrative ­ apparatus. The goal was to Health Statistics (database), OECD, Paris, transfer to private managerial functions for ­h ttp://w w w.oecd.org /els/ health-systems​ social activities that had been wholly and par- /­health-data.htm. ­ tially funded by the state government, though 5. Data on average lengths of hospital stays managed directly by the state ­ administration. by cou ntr y are from OEC D Health 14. English hospitals have hired (and fired) CEOs Statistics (database), OECD, Paris, ­ h ttp:// since the 1980s, and FTs continue this ­ practice. www.oecd​. org /els ​/ ­h ealth-systems/health​ 15. Management teams were allowed to make -data.htm. decisions regarding performance-related 6. Data envelope analysis is a commonly used incentives amounting to a substantial part of method for estimating efficiency that involves the salary (for clinical staff, amounting on the use of linear programming to rank firms average to 8 percent in consortiums and according to a score of relative e ­ fficiency. It is 15 percent in public health care companies, based on the idea that production units—or and, for administrative staff, up to 40 percent decision-making units—seek to maximize salary). of their ­ their output for a given quantity of inputs or, 16. The Department of Health (and its secretary alternatively, to minimize the quantity of of state) has overall and political responsibil- inputs for a given o ­ utput. direction. ity for strategic ­ 7. Technical efficiency is associated with internal 17. The Department of Health’s own numbers factors such as management and control of the suggest that, on average, hospitals derive ­ 0.9 production ­process . It is generated with a percent of their income from private patients, given quantity and combination of inputs— though some derive much more (four units in or, the fewer inputs that are used to turn out a cancer care, eye care, and children’s care given quantity of product, the more efficient is derive more than 10 ­ p ercent). ( Source: the process in the technical ­ sense. “Private Services in Foundation Trusts,”   8. Scale efficiency is determined by operational NHS Privatisation (blog) [accessed April 7, size or ­ scale. Small hospitals are usually inef- 2015], ­https://nhsprivate.wordpress.com​ ficient because small scale results in higher /­executive-summary/). ­ unit ­costs. 18. As already suggested, a number of local gov-   9. The framework to analyze governance prac- ernments have launched pilots involving alter- tices is adapted from La Forgia and others native governance arrangements for public ­(2013). ­ hospitals. Some are official pilots sanctioned ­ 10. In addition, see “Guidance of the General by the central government; others are ­ not. Office of the State Council on Promoting Shanghai and Zhenjiang are official ­ pilots. Multi-level Diagnosis and Treatment System,” Dongyang and Sanming are local ­ initiatives. State Council (Guoban fa, ­ No. ­ 70). The issue 19. Unfortunately, verifiable data on the effects of of misaligned incentives is discussed in chap- the pilots are ­unavailable. None of these cases ter ­ 6. Although public hospitals are owned (or any of the official pilots) has been rigor- and operated by the government and financed ously or independently ­ evaluated. through a combination of direct subsidies, 20. This arrangement was part of a special agree- social insurance payments, and user fees, ment between city leaders and a Taiwanese there has been a wholesale transfer of decision businessman who made a substantial dona- making on income generation to hospital tion to rebuild the hospital in ­1993. management and clinicians, who tend to act 21. A prefecture is an administrative unit com- as private agents to maximize their own mon to all China’s provinces and usually con- ­income. sisting of both urban and rural ­ areas. Located 196 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A in Fujian Province, the Sanming Prefecture 29. “Guidance on Comprehensive Pilot Reform of has a population of about ­ 2.7 million and con- Urban Public Hospitals,” National Health sists of 2 districts and 10 ­ counties. The prefec- and Family Planning Commission (Guo ban ture had 166 health care facilities in 2013, of No. ­ fa 2015, ­ 38). which 22 were secondary and tertiary hospi- 30. For more information, see the NHS Providers tals, with a total of about 8,000 ­ beds. website (accessed January 30, 2016), ­ https:// 22. Single-site studies do not provide sufficient www.nhsproviders.org. information or analysis of the overall quality 31. For more information, see the Governance of management practices, and therefore cau- Institute website (accessed January 30, 2016), tion is advised in making inferences about the ­http://www.governanceinstitute.com. hospital ­system. ­ 32. The measures to strengthen and professional- 23. The survey was commissioned by the World ize management and improve managerial Bank, and the preliminary findings are reported practices are based on Frenk and others ­here. (2010); Lega, Prenestini, and Spurgeon (2013); 24. Country comparisons should be interpreted MSH (2005); Shaw (1998); and case studies with caution; 79 percent of the hospitals that commissioned for this ­report. the researchers originally contacted in China refused to p ­ articipate. This may have contrib- uted to a sampling bias in which the surveyed References hospitals were those with the best manage- Aboumatar, Hanan J ­ . Chang, Jad Al ­ ., Bickey H ment p ­ ractices. The researchers did not exam- Danaf, Mohammad Shaear, Ruth Namuyinga, ine the association between management Sathyanarayanan Elumalai, Jill ­ A. Marsteller, scores and hospital performance indicators and Peter ­ P ronovost. ­ J. ­ 2 015. “Promising because the latter were impossible to ­ validate. Practices for Achieving Patient-Centered 25. Eighty percent of the procurement of goods Hospital Care: A National Study of High- and services is managed centrally and the Performing ­U.S. ­Hospitals.” Medical Care 53 remainder by the ­ hospitals. (9): ­758–67. 26. For example, the establishment of a purchas- AHRQ (Agency for Healthcare Research and ing and contract management unit within the ­ Quality). ­ 2017. “Improving Patient Safety in SES that effectively separated purchasing from Hospitals: A Resource List for Users of the provision was a major facilitating step for the AHRQ Hospital Survey on Patient Safety OSS ­ reforms. The hospital “administrative” Culture.” Reference document, AHRQ, U ­ ­ .S. unit that directly managed non-OSS hospitals Department of Health and Human Services, was incapable of making the transition to Washington, ­DC. arm’s-length oversight and ­ monitoring. ­ Allen, Pauline, Qi Cao, and Hufeng ­ 2013. Wang. ­ 27. Countries using DRGs have also introduced “Public Hospital Autonomy in China in an measures to counteract two perverse incen- International ­Context.” International Journal tives that are embedded in D ­ RGs. They are of Health Planning and Management 29 (2): strengthening auditing systems to detect ­141–59. ­doi:10.1002/hpm.2200. hospitals’ gaming or upcoding to increase Allen, ­ J . Keen, ­ P., ­ J . Wright, P­ . Dempster, r evenues. A nd to deter hospitals from ­ J. Townsend, A ­ ­ . Hutchings, and R Verzulli. ­ .­ skimping on care to reduce costs or refusing 2 012. “Investigating the Governance of ­ care to high-cost patients, most countries, Autonomous Public Hospitals in England: i n c lu d i n g t h e S p a i n a n d t h e U n it e d Multi-Site Case Study of NHS Foundation Kingdom, provide additional payments to ­Trusts.” Journal of Health Services Research cover high-cost patients and other services & Policy 17 (2): ­ 94–100. and activities such as intensive care, special Audibert, Martine, Jacky Mathonnat, Aurore therapies, and medical education (Cots and Pelissier, Xiao Xian Huang, and Anning ­ Ma. others 2011). ­ 2 013. “Hea lt h I nsu ra nce Refor m a nd 28. For example, FT board members consist of a Efficiency of Township Hospitals in Rural chairman, hospital CEO, deputy CEO, chief China: An Analysis from Survey ­ Data.” China nurse, medical director, finance director, Economic Review 27: ­ 326–38. workforce directors, and five nonexecutive Becker’s Hospital ­ Review. ­ 2010. “Quint Studer: elected ­members. Four Best Practices for Improving Emergency L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 197 Department ­R esults.” January ­2 5. ­ https:// and European Observatory on Health Systems w w w.b e cker sho spit a l re v ie w. c om /ne w s​ and ­Policies. -­a nalysis/quint-studer-four-best-practices-for​ Castaño, Ramón, Ricardo Bitrán, and Ursula -improving-emergency-department-results. ­ Giedion. ­ 2004. “Monitoring and Evaluating html. ­ Hospital Autonomization and its Effects on ———. ­ 2012. “7 Best Practices for Hospitals’ Priority Health S ­ ervices.” Paper for the Training and Development P ­ rograms.” December Partners for Health Reformplus project of the ­12. ­https://www.beckershospitalreview​.com​ ­ U.S. Agency for International Development, /­hospital-management-administration/7-best-­prac prepared by Abt Associates, Bethesda, ­ M D. tices-for-hospitals-training-and-development-pro​ Chen, Ming, Hongmei Zhao, Saijun Jia, Lihua grams.html. Wang, and Yue ­ Z hao. ­ 2014. “Total Quality Bjork, David A ­ .­ 2006. “Collaborative Leadership: Management: Improving the Procedure of A New Model for Developing Truly Effective Additional Registration in Specialist ­ Clinic.” Relationships between CEOs and ­ Trustees.” Chinese Hospitals 18 ­ (1). Collaborative Leadership Tools for CEOs Chen, Vivian, and Yue Xia ­ Gao. ­ 2015. “Wuhan brief, Center for Healthcare Governance, Asia Heart Hospital Case ­ Study.” Unpublished ­Chicago. report commissioned by the International Bley, Charles ­ M., and Cynthia ­ T. ­ Shimko. ­ 1987. Finance Corporation, World Bank, B ­ eijing. “A Guide to the Board’s Role in Hospital Chen, Vivan, Yue Xia Gao, and Yuan ­ Wang. F inance.” Chicago: A merican Hospital ­ ­ 2 015. “A ier Eye Hospital Case S ­ tudy.” ­Publishing. Unpubl i she d rep or t com m i s sione d by Bloom, Nicholas, and John Van ­ Reenen. ­ 2007. the International Finance Corporation, “Measuring and Explaining Management World Bank, ­ Beijing. Practices across Firms and C ­ ountries.” Chu, Kejia, Ning Zhang, and Zhonfei ­ C hen. Quarterly Journal of ­ E conomics 122 (4): 2015. “The Efficiency and its Determinants for ­ 1351–1408. China’s Medical Care System: Some Policy ———. ­ 2010. “Why Do Management Practices Implications for Northeast A ­ sia.” Sustainability Differ across Firms and Countries?” Journal of 7: ­14092–14111. Economic Perspectives 24 (1): ­ 203–24. Cocchi, ­ R enee. ­ 2 012 . “B est Practices for Bloom, Nicholas, Rebecca Homkes, Raffaella Enhancing Doctor/ Nurse R ­ elationships.” Sadun, John Van Reenen, Stephen Dorgan, H e a l t h c a r e B u s i n e s s & Te c h n o l o g y and Dennis ­ L ayton. ­ 2010. “Management in (e -­n e w s l e t t e r) , M a y ­7. ­h t t p : / / w w w​ Healthcare: W hy Good Practice Really .­h ealthcarebusinesstech.com/best-practices​ ­M at ters.” Resea rch repor t , McK i nsey -for- ­enhancing​-doctornurse​-relationships/. & Company and the Centre for Economic Coile, ­R. ­1994. The New Governance: Strategies Performance, London School of Economics for an Era of Health ­ Reform. Ann Arbor, MI: and Political ­ Science. Health Administration ­ Press. Brandenburg, Lisa, Patricia Gabow, Glenn Steele, COMU (Crown Ownership Monitoring ­ Unit). John Toussaint, and Bernard ­ Tyson. ­ J. ­ 2015. 2011. “Board Appointments, Induction, and ­ “Innovation and Best Practices in Health Care Professional ­ D e v e l o p m e n t .” C O M U , S cheduling.” Discussion paper, Institute of ­ New Zealand Treasury, ­ Wellington. Medicine, National Academy of Sciences, Cots, Francis, Pietro Chiarello, Xavier Salvador, Washington, ­DC. Xavier Castrells, and Wilm ­ Quentin. ­ 2011. “DRG- Burns, Lawton Robert, and Gordon ­ G. Liu, ­eds. Based Hospital Payment: Intended and Unintended ­2017. China’s Healthcare System and ­ Reform. ­Consequences.” In Diagnostic-Related Groups in Cambridge: Cambridge University ­ Press. Europe, edited by Reinhard Busse, Alexander Busse, Reinhard, and Wilm ­ Q uentin. ­ 2 011. Geissler, Wilm Quentin, and Miriam Wiley, “Moving towards Transparency, Efficiency, 75–92. Berkshire, U ­ ­ .K.: Open University Press, and Quality in Hospitals: Conclusions and World Health Organization, and European ­R ecommendations.” In Diagnostic-Related Observatory on Health Systems and ­ Policies. Groups in Europe, edited by Reinhard Busse, Cuervo-Cazurra, Alvaro, and Ruth ­ V. ­Aguilera. Alexander Geissler, Wilm Quentin, and 2004. “Codes of Good Governance Worldwide: ­ Miriam Wiley, ­ 149–74. Berkshire, ­U.K.: Open What Is the Trigger?” Organization Studies University Press, World Health Organization, 25 ­(3). 198 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Curry, Leslie ­ A ., Erica Spatz, Emily Cherlin, Gage, Larry ­ S ., Anne ­ B . Camper, and Robert Jennifer ­ W. Thompson, David Berg, Henry ­ H. Falk. ­ ­ 2006. “Legal Structure and Governance Ting, Carole Decker, Harlan ­ M . Krumholz, of Public Hospitals and Health S ­ ystems .” and Elizabeth H ­ . ­ B radley. ­ 2 011. “What National Association of Public Hospitals and Distinguishes Top-Performing Hospitals in Health Systems, Washington, ­ DC. AMI Mortality Rates? A Qualitative ­ Study.” Gao, Tian, and Bruce G ­ urd. 2­ 014. “Meeting the Annals of Internal Medicine 154: ­ 384–90. Challenge in Performance Management: The Deber, ­ R., ­A. Topp, and D ­ .­Zakus. ­ 2004. “Private Diffusion and Implementation of the Balanced Delivery and Public Goals: Mechanisms for Scorecard in Chinese ­ H ospitals.” Health E n su r i ng t h at Ho spit a l s Me e t P ubl ic Polic y a n d Pl a n n i ng 30 (2): ­ 2 3 4 – 41. ­ Objectives.” Background paper commissioned ­doi:10.1093/heapol/czu008. by the World Bank, Washington, ­ DC. Gill, Mel, Robert ­ J . Flynn, and Elke ­ Reissing. Drazen, Erica, and Jared ­ R hoads. ­ 2011. “Using 2 005. “The Governance Self-Assessment ­ Tracking Tools to Improve Patient Flow in Checklist: An Instrument for Assessing Board Hospitals.” Issue brief, California Health Care ­ ­E ffectiveness.” Nonprofit Management and Fou nd at ion (C HC F), Oa k la nd a nd Leadership 15 ­ (32). Sacramento, ­CA. Govindaraj, Ramesh, and Mukesh ­ Chawla. ­ 1996. Durán, ­ A ntonio. ­ 2015. “New Roles for Hospitals “Recent Experiences with Hospital Autonomy in China; Lessons from Public Hospital in Developing Countries: What Can We Governance Reforms in OECD: ­ Spain.” Case Learn?” Comparative case study report, Data study commissioned by the World Bank, for D e cision M a k i ng (DDM) Proje c t , Washington, ­DC. Department of Population and International Durán, ­ A., and ­ R. ­B. ­Saltman. ­ 2015. “Governing Health, Harvard School of Public Health, Public ­ H o s p i t a l s .” I n T h e Pa l g ra v e ­Boston. International Handbook of Healthcare Policy Greenwood, ­ J., ­R . Pyper, and ­ Wilson. ­ D. ­ 2002. and Governance , edited by ­ E . Kuhlmann, New Public Administration in B ­ ritain. R . Blank, ­ ­ I. ­L . Bourgeault, and C ­ . Wendt, London: ­Routledge. 4 43– 61. London and New York: Palgrave ­ Guo, Jinshuai, Shijun Ma, and Xun ­ Zhang. ­ 2014. ­McMillan. “Lean Management to Transform a Chinese Durán, Antonio, Richard ­ B. Saltman, and Hans ­ F. ­H ospital.” Planet Lean: The Lean Global ­ W. ­Dubois. 2 ­ 011. “A Framework for Assessing Network Journal, September ­ 17. Hospital ­ G overnance.” In Governing Public He, Jingwei ­ A lex. ­ 2011. “Combating Healthcare Hospitals: Reform Strategies and the Movement Cost Inflation with Concerted Administrative towards Institutional Autonomy, edited by Actions in a Chinese ­ P rovince.” Public Richard ­ B . Saltman, Antonio Durán, and Administration and Development 31: ­214–28. ­ H ans. ­ F. ­ W. Dubois, ­ 35–53. Copenhagen: ­doi:10.1002/pad.602. World Health Organization on behalf of the He, A lex Jing wei and Jiwei Q ­ ian. ­ 2 013. European Observatory on Health Systems and “Hospitals’ Responses to Administrative Cost- ­Policies. Containment Policy in Urban China: The Case Edwards, ­ Nigel. 2 ­ 011. ­ “England.” In Governing of Fujian ­ P rovince.” China Quarterly 216: Public Hospitals: Reform Strategies and the ­946–69. ­doi:10.1017/S0305741013001112. Movement towards Institutional Autonomy, Holland, Thomas ­ P., Robert ­ A . Ritvo, and edited by Richard ­ B. Saltman, Antonio Durán, Anthony ­R . ­Kovner. ­1997. Improving Board and Hans ­ W. Dubois, ­ F. ­ 113–40. Copenhagen: Effectiveness: Practical Lessons for Nonprofit World Health Organization on behalf of the He alth C are ­ O rganiz ation s . C h ic ago: European Observatory on Health Systems and American Hospital P ­ ublishing. ­Policies. HRET (Health Research & Educational ­ Trust). Frenk, Julio, Lincoln Chen, Zulfiqar ­ A . Bhutta, ­ 2 014. “Building a Leadership Team for the Jordan Cohen, Nigel Crisp, Timothy Evans, Health Care Organization of the ­ F uture.” a nd H a r vey F ­ i neb er g. ­ 2 010. “H e a lt h Signature Leadership Series report, HRET, Professionals for a New Century: Transforming ­Chicago. Education to Strengthen Health Systems in an Hu, Hsin-Hui, Quinghui Qi, and Chih-Hai ­ Yang. Interdependent ­World.” The Lancet 376 2 012 . “A nalysis of Hospital Technical ­ (9756): ­1923–58. Efficiency in China: Effect of Health Insurance L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 199 ­R eform.” C hin a Economic Revie w 23: c o m m i s s i o n e d b y t h e Wo r l d B a n k , ­865–77. Washington, ­DC. Huang, ­ Yanzhong. 2009 “An Institutional Li, Weiping, and Jianxiu ­ Wang. ­ 2015. “China’s A nalysis of China’s Failed Healthcare P ubl ic H o spit a l G over n a nc e R efor m : ­Reform.” In Socialist China, Capitalist China: Dongyang Case S ­ tudy.” China National Social Tension and Political Adaptation under Health Development Research Center, ­ Beijing. Economic Globalization, edited by Guoguang Lin, Zhijun, Zengbiao Yu, and Liqun Z ­ hang. Wu and Helen Landsowne, ­ 75–86. New York: 2014. “Performance Outcomes of Balanced ­ ­Routledge. S corecard Application in Hospital Huntington, Dale, and Krishna Hort, ­ 2015. eds. ­ Administration in ­ China.” China Economic Public Hospital Governance in Asia and the Review 30 (C): ­ 1–15. ­Pacific. Comparative Country Studies ­ Vol. 1, Liu, G ­ ordon. 2 ­ 015. “Quality of Hospital ­ No. ­ 1. Geneva: World Health Organization, Management Practices in China— China Asia Pacific Observatory on Health Systems Hospital Management Survey ( ­CHMS).” and ­Policies. Unpublished manuscript, World Bank, Ingram, ­T. ­T. ­1996. Ten Basic Responsibilities of Washington, ­DC. Nonprofit ­Boards. Washington, DC: National Ma, ­ J in. 2­ 015. “China’s Public Hospital Center for Nonprofit ­ Boards. Governance Reform: Case Study of ­ Shanghai.” Jefferson University ­ ­ .d. “Keeping Our Hospitals. n Case study commissioned by the World Bank, Patients S ­ afe.” Thomas Jefferson University Washington, ­DC. Hospitals, ­P hiladelphia. ­h ttps://hospitals​ Mannion, ­ R ., ­H. ­ T. O ­ . Davies, and M ­ . N ­ . .­j efferson.edu /quality-and-safety/keeping​ M arshall. ­ ­ 2005. “Cultural Characteristics of -our-patients-safe/. ‘High’ and ‘Low’ Performing ­ H ospitals.” Kirkpatrick, ­ P. ­ I ., ­ K. Jespersen, ­ I. M . Dent, and ­ Jo u r n a l of He a lt h O rga n i z a t i o n a n d Neogy. ­ ­ 2009. “Medicine and Management in a Management 19 (6): ­ 431–39. Comparative Perspective: T he Case of Mason, Anne, Padraic Ward, and Andres ­ Street. Denmark and ­ England.” Sociology of Health 2011. “England: The Health care Resource ­ and Illness 31 (5): ­ 642–58. ­G roup.” In Diagnostic-Related Groups in KPMG. ­ 2 010. “The Changing Face of Health Europe, edited by Reinhard Busse, Alexander Care in China: Changing Public Policies and Geissler, Wilm Quentin, and Miriam Wiley, Resulting O ­ pportunities.” Infrastructure, 197–200. Berkshire, ­ ­ U.K.: Open University G overn ment & Healthcare repor t, Press, World Health Organization, and Publication ­ N o. HK-Pl10 - 0005, K M PG, European Observatory on Health Systems and Hong ­Kong. ­Policies. L a Forg ia, G era rd M a r ti n , a nd B er na rd Mazzocato, ­ P., ­R. ­ J. Holden, ­ M . Brommels, H ­ . ­C outtolenc. ­2008. Hospital Performance in Aronsson, ­ U. Bäckman, ­ M. Elg, and ­ T hor. J. ­ Brazil: The Search for E ­ xcellence. Washington, ­ 2012. “How Does Lean Work in Emergency DC: World ­ Bank. Care? A Case Study of a Lean-Inspired La Forgia, Gerard, April Harding, Loraine Intervention at the Astrid Lindgren Children’s Hawkins, and Eric ­ Roodenbeke. ­ 2 013. “A Hospital, Stockholm, ­ Sweden.” BMC Health Framework for Developing and Analyzing Se r vices Research 12: ­ 2 8. ­d oi:10.1186​ Public Hospital Reforms in Developing /1472-6963-12-28. Countries that Involve ­ Autonomy.” Working McConnell, ­ K. ­J., ­ R. ­ C . Lindrooth, ­ R. Wholey, D. ­ draft, World Bank, Washington, DC T. ­ ­ M . M addox , and ­ N. ­ B loom. ­ 2 013. Lega, Federico, Anna Prenestini, and Peter “Management Practices and the Quality of ­ Spurgeon. ­ 2013. “Is Management Essential to Care in Cardiac ­ U nits.” JA M A Internal Improving the Performance and Sustainability Medicine 173 (8): ­ 684–92. of Health Care Systems and Organizations? McKee, Martin, and Judith Healy, ­ 2 002. e ds. ­ A Systematic Review and a Roadmap for Hospitals in a Changing E ­ urope. European Future ­Studies.” Value in Health 16 (1 Suppl): Observatory on Health Care Systems ­ S eries. ­S46–S51. Buckingham, ­ U . K . : Wo r l d H e a l t h Li, Weiping, and Mengxi ­ Jiang. ­ 2015. “China’s Organization and Open University ­ Press. Public Hospital Governance Reform: Case McPake, Barbara, Francisco Jose Yepes, Sally Study of ­ Zhenjiang.” Unpublished case study Lake, and Luz Helena S ­ anchez. ­2003. “Is the 200 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Colombian Health System Reform Improving Healthcare Financial Management 65 (10): the Performance of Public Hospitals in 94–98, 100, ­ 102. Bogota?” Health Policy and Planning 18 (2): Preker, Alex, and ­ A . Harding, e 2 003. ­ ds. ­ ­182–94. Innovations in Health Service Delivery: The Mossialos, Elias, Martin Wenzl, Robin Osborn, C o r po ra t i z a t i o n of Pu bli c ­ H os pi t a l s . and Chloe Anderson, ­ eds. ­ 2015. “International Washington, DC: World ­ Bank. Profiles of Health Care Systems, 2014: Australia, Qian, ­ Jiwei. ­2015. “Reallocating Authority in the Canada, Denmark, England, France, Germany, Chinese Health System: An Institutional Italy, Japan, The Netherlands, New Zealand, ­Perspective.” Journal of Asian Public Policy Norway, Singapore, Sweden, Switzerland, and 8 (1): ­ 1 9–35. ­ d oi:10.1080/17516234.2014​ the United S ­ tates.” Annual publication, The .1003454. Commonwealth Fund, New ­ York. R ice, ­ e d. ­ J ., ­ 2 003. “Hospital Boards: A n MSH (Management Sciences for H ­ ealth). ­2005. I nternational Jou rney for World- Class M anagers Who Lead: A Handbook for Governance Effectiveness: Global Trends Improving Health ­ Services. Cambridge, MA: Survey ­2003.” International Health Summit, Management Sciences for H ­ ealth. ­Minneapolis. Naranjo-Gil, ­ D. ­ 2009. “Strategic Performance in Rice, ­ Sabriya. ­ 2014. “Top-Performing Hospitals Hospitals: The Use of the Balanced Scorecard Know How to Use ­ Teamwork.” Modern by Nurse M ­ a n a g e r s .” H e a l t h C a r e Healthcare, November ­ http://www.modern​ 8. ­ Management Review 34 (2): ­161–70. healthcare.com /article/20141108/M AGA​ NC H L ( Nat iona l C enter for Hea lt hc a re ZINE/311089980. ­ L eadership). ­ 2010. “Best Practices in Health Salas, Eduardo, Cameron Klein, Heidi King, Leadership Talent Management and Succession Mary Salisbury, Jeffrey S ­ . Augenstein, David Planning: Case S ­ tudies.” Case study report, ­ J . Birnbach, Donald ­ W. Robinson, and NCHL, ­Chicago. Christin ­ Upshaw. ­ 2008. “Debriefing Medical NHS (National Health Service) European O ­ ffice. Teams: 12 Evidence-Based Best Practices and 2011. “The Search for Low-Cost Integrated ­ ­Tips.” Joint Commission Journal on Quality Healthcare: The Alzira Model—from the and Patient Safety 34 (9): ­ 518–27. Region of ­ Valencia.” Research paper, NHS Saltman, Richard ­ B ., Antonio Durán, and Hans Confederation, ­Brussels. ­F. ­W. Dubois, ­e ds. ­2011. Governing Public Ng, ­ ­ .­ Y. C 2011. “The Productive Efficiency of Hos pit al s: R efo r m S t rat eg i e s an d the Chinese ­Hospitals.” China Economic Review Movement towards Institutional ­ A utonomy. 22: ­428–39. Copenhagen: World Health Organization on Nolte, Ellen, and Emma P ­ itchforth. ­2014. “What behalf of the European Observatory on Health Is the Evidence on the Economic Impacts of Systems and ­ Policies. Integrated Care?” Policy Summary 11, World Sanming Prefecture ­ Government. ­ 2015. “Summary Health Organization Regional Office for of Public Hospital Comprehensive Reform in Europe, Copenhagen (acting as the host orga- Sanming M ­ unicipality.” Case study commis- nization for, and secretariat of, the European sioned by the World Bank, Washington, ­ DC. Observatory on Health Systems and ­ Policies). Shaw, ­J ane. ­1 998. Hospital M anagement: Over, ­ M., and N ­ .­ Watanabe. ­ 2003. “Evaluating the Training and Professional D ­ evelopment . Impact of Organizational Reforms in H ­ ospitals.” Geneva: World Health ­ Organization. In Innovations in Health Service Delivery: The State ­ Council. ­ 2012. “The Twelfth Five-Year Plan Corporatization of Public Hospitals, edited for Health Care ­ D evelopment.” Guo ban fa by Alex Preker and ­ A . Harding, ­ 1 05–51. 2012 , ­N o. ­5 7. ­h ttp://www.wpro.who​.int​ Washington, DC: World ­ Bank. /­health_services/china_nationalhealth​plan.pdf. Parand, Anam, Sue Dopson, Anna Renz and Tam, ­ Wai-keung. ­ 2008. “Failing to Treat: Why Charles ­ Vincent. ­ 2014. “The Role of Hospital Public Hospitals in China Do Not ­ Work.” Managers in Quality and Patient Safety: China Review 8 (2): ­ 103–30. A Systematic ­ R eview.” BMJ Open 4 (9): Tsai, ­ T. ­C ., ­ A. ­K . Jha, ­ A. ­A . Gawande, ­ R. ­ S. ­e005055. ­doi:10.1136/bmjopen-2014-005055. Huckman, ­ N . Bloom, and ­ ­ adun. ­ R. S 2 015. Potash, David ­ L. ­2 011. “Accountable Clinical “Hospital Board and Management Practices Management: An Integrated A ­ pproach.” are Strongly Related to Hospital Performance L e v er 4 : R eforming P u b lic H o s pital G o v ernance and M anagement 201 on Clinical Quality ­ Metrics.” Health Affairs Foundation Trust M­ odel.” Case study commis- 34 ­(8). sioned by the World Bank, Washington, ­DC. WHO (World Health Organization) and World Yang, Jinqiu, and Wu ­ 2014. “The Trade- Z eng. ­ Bank. ­ ­ 2015. Technical Roundtable ­ Discussions. Offs between Efficiency and Quality in the Joint Health ­Study. Hospital Production: Some Evidence from Willyard, ­David. ­2006. “Hospital Revenue Cycle Shenzhen, ­China.” China Economic Review Best ­Practices.” The Health Care Biller ­(Oct. 31: ­166–84. 2006): ­1+. Ying, ­ Yachen. ­2 014. “Achieve Win-Win-Win World ­ Bank. ­ 2006. ­“ B ra z i l. E n h a nc i ng Situation by Jointly Reforming the Drug Performance in Brazil’s Health Sector: Lessons Sector, Health Service Sector, and Health from Innovations in the State of São Paulo and Insurance Sector: Field Study Report on Public the City of ­Curitiba.” Brief, Report ­ No. 35691​ Hospital Reform in ­ S anming.” Case study, -BR, World Bank, Washington, ­ DC. He a lt h D evelopment Re se a rch C enter ———. ­ 2010. “Fixing the Public Hospital System (NHFOC), ­Beijing. China.” China Health Policy Notes Series, in ­ Yip, Winnie Chi-Man, William ­ C . Hsiao, Wen ­ No. 2, World Bank, Washington, ­ DC. Chen, Shanlian Hu, Jin Ma, and A lan Wright, ­ 2015. “New Roles for Hospitals in Stephen. ­ M aynard. ­ ­ 2012. “Early Appraisal of China’s China: Lessons from Public Hospital Governance Huge and Complex Health-Care R ­ eforms.” Reforms in OECD, MICs, and China: Analysis The Lancet 379 (9818): ­ 833–42. ­doi:10.1016​ of Governance Arrangements in England: the /­S0140-6736(11)61880-1. 6 Lever 5: Realigning Incentives in Purchasing and Provider Payment Introduction Incentives in the health sector refer to “all the rewards and punishments that providers People centered integrated care (PCIC), face as a consequence of the organizations in which lies at the core of China’s efforts to which they work, the institutions under build a high-quality and value-based health which they operate and the specific interven- system, has three key elements (as discussed tions they provide” (W HO 2000, 61). in chapter 2): Defined thus, incentives are conditions within the work environment of health professionals • It is organized around the population’s that enable and motivate them to stay in their needs. jobs, come to work every day, and devote the • It promotes primary care as patients’ first necessary time and effort as dictated by the point of contact for most of their health demands of their profession and the needs of care needs and is thus based on a strong their clients. primary health care foundation. Provider payment methods create powerful • It prioritizes integration and coordination incentives affecting provider behavior and the of services across all levels of care, from efficiency, equity, and quality outcomes of promotion and prevention to curative and health finance reforms. The amount and type palliative needs, to reduce fragmentation of payment to health care practitioners and and wasteful use of resources across a organizations affect the amount and type of health system. health goods and services received by consum- ers and, ultimately, the aggregate costs to all Realigning the purchasing and provider payers, including the government, insurers, pay ment i ncentives to rei n force a nd employers, and individuals. The amount and strengthen PCIC development and adoption type of payment also have an impact on the therefore entails modifying how providers behavior of health care organizations and are paid in ways that correspond with, con- individuals with respect to the quality of care form to, and result in these three fundamen- they deliver to consumers. At the same time, tal PCIC outcomes. This is the focus of this the amount and type of payments made by chapter. consumers for health care affects their health 203 204 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A and preventive-care-­ seeking behavior, choice payers (related to determination of benefits of health professional, and quantity of care and resource management) and purchasers they seek. (related to financial management, benefit The most critical determinant of incen- design, contracting, and monitoring)—affect tives is whether the provider payment method provider incentives and behavior. is retrospective or prospective, because it The rest of this chapter is organized as affects who bears the risk of the cost of pro- follows: vision, which in turn has implications for treatment choices. Other determinants • “Evolution of Health System Incentives” include (a) the breadth of provider payments briefly discusses the evolution of health (that is, whether the providers are paid nar- system incentives over the years. rowly for their own services or broadly for • “Challenges and Lessons Learned from bundles of related services such as laboratory Select Studies” presents lessons learned tests and other provider services) and (b) the from select case studies in China that were generosity of the payments (that is, whether carried out in order to assess opportunities the payments are low or high). and challenges in purchasing practices and Collectively and individually, these dimen- identify roadmaps for strengthening capaci- sions affect providers’ incentives and thus ties of social insurance agencies. their behavior regarding a host of variables, • “Provider Payment Reforms in China” such as volume and intensity of care rendered discusses challenges and lessons from (including preventive care), efforts to keep recent provider payment reform pilots. costs down, treatment of low- and high-risk • “Core Action Areas and Implementation patients, number of referrals, provision of Strategies to Realign Incentives” con- quality care, and so on. cludes with a discussion of four core action Incentives are also affected by the deci- areas and associated strategies to deepen sions of payers—defined as entities that pay reform through realignment of incentives. not only premiums (such as individuals, busi- nesses, and the government) but also those who control the premiums before they are Evolution of Health System paid to the providers (that is, insurance com- Incentives panies and governments). Although patients From Socialism to a Market Economy: and businesses function as ultimate purchas- 1949–2009 ers, insurance companies and the government serve a processing or payer function. In the first decades after the People’s Republic Purchasers also worry about financial of China was established in 1949, the health management, including withholding a certain care system was built within the socialist percentage of premiums to provide a fund for planned economy and characterized by pub- committed but undelivered health care, set- lic production, with public financing in urban ting aside resources for uncertainties such as areas and community financing in rural longer hospital utilization levels than areas. The rural collectives and urban work expected, overutilization of referrals, acci- units (danwei) purchased and provided health dental catastrophes, and other financial lia- care for their members. bilities. Likewise, payers are concerned with In an environment marked by a shortage benefits (that is, specific areas of insurance of doctors and medicines, the health care sys- coverage such as outpatient visits, hospital- tem made innovative use of part-time doctors ization, and so forth) that make up the range to provide extensive preventive and primary of medical services that a payer markets to its health care (Manuel 2010). Fees in public subscribers. hospitals were set at low levels, and because All these functions—carried out in the the hospitals were protected against deficits execution of normal business functioning of through a flexible government budget system, L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 205 they had no incentives to provide unneces- health system, contributing to escalating sary care that would increase the economic costs, medical impoverishment, and large- burden on patients (Jing 2004). This period scale public discontent (Blumenthal and saw extraordinary improvements in popula- Hsiao 2005; Liu and Mills 1999; Yip and tion health achieved at very low levels of Hsiao 2008). spending (Liu 2004). These policies created a bias in China’s Two other policies, undoubtedly intro- financing and service delivery system toward duced with good intentions, met with unfor- more costly services delivered in hospitals. tunate consequences. First, to motivate high The pricing systems of provincial pricing performance, bonus schemes were introduced bureaus favored hospitals and reinforced the that linked physician incomes with generated overuse of more expensive care settings and revenues. Second, to improve access, basic higher-priced technology. They also exacer- medical services and pharmaceuticals were bated the general trend among providers to priced artificially low while expensive proce- invest, upgrade facilities, and provide higher dures and drugs were marked up for high levels of care in ways that were more respon- profit margins. sive to potential profits than to health care The consequences of these inappropriate needs and potential improvements in health. incentives are by now well-known: to maxi- The fee-for-service provider payment mecha- mize incomes, physicians resorted to demand nisms stimulated service delivery and fueled inducement to generate higher revenue; and rapid growth of health spending. Massive to maximize profits, hospitals began encour- public investments enabled national insur- aging overprescription of drugs and expen- ance schemes to rapidly enroll people (Meng sive diagnostic tests. These reinforcing and others 2012), and at the same time, reim- actions led to massive inefficiencies and fur- bursement of individual services by fees ther increased the financial burden on encouraged health institutions to provide patients (Chen 2009; Feng, Lou, and Yu patients with more health care than required. 2015; Ma, Lu, and Quan 2008). These and Further, reimbursement rates were increased related misaligned incentives became embed- to lower out-of-pocket payments (copays), ded in the health system, contributing to which led to greater consumer demand escalating costs, medical impoverishment, (Babiarz and others 2012; Lei and Lin 2009; and large-scale public discontent. Liu and Zhao 2014; Meng and others At the same time, wanting health facilities 2012; Yip and others 2012; Yu and others to survive financially, the government priced 2010; Zhang, Yi, and Rozelle 2010). The some new, high-tech procedures above cost result was double-digit spending growth. and allowed hospitals a 15 percent markup on drugs. The consequences were unsurpris- Health Care Reform: 2009–Present ing: to generate higher revenues, hospitals began encouraging overprescription of drugs Responding to the rapidly rising costs and and expensive diagnostic tests, and physi- demand for high-quality, affordable health cians in turn prescribed more drugs and pro- care, the government of China launched one vided more high-profit services. As public of the biggest health policy interventions facilities turned into profit-seeking entities, a in recent times in 2009. Targeted to reach perverse dynamic ensued in which a “pro- 1.3 billion people, the reform invested more vider [had] to dispense seven dollars’ worth than RMB 3 trillion into the health system of drugs to earn one dollar of profit” (Yip between 2009 and 2014 to support five pil- and Hsiao 2008, 462). These reinforcing lars of change: expanding coverage of social incentives and actions led to massive ineffi- insurance schemes, establishing a national ciencies and further increased the financial essential medicines system, advancing public burden on patients. These and related mis- hospital reforms, improving the primary care aligned incentives became embedded in the system, and increasing the equality and 206 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A availability of public health services (Chen nationwide. Moreover, a mixed payment 2009). This investment doubled government system has been established to introduce spending on health and signaled the govern- case-based capitation and per diem payment ment’s increased role in financing health care. methods. The payment system incentivizes As noted, the 2009 health care reform the health providers to control costs and pro- made public hospital reforms one of five pil- vide quality health care services to allocate lars of change. Specific areas targeted for resources efficiently. reform included hospital governance, hospi- tal financing, provider payment methods, Persistent Weaknesses in Incentives and, more recently, encouraging the growth of the private hospital sector. Subsequently, Despite impressive gains in achieving near- the Twelfth Five-Year Plan (TFY) (covering universal insurance coverage in a short time, 2011–15) stated clearly that the objective of a China still needs to further correct the under- public hospital was to pursue the public inter- lying incentive system that governs provider est and introduced a package of policies behavior and influences the nature and scope aimed at realigning incentives in public hos- of purchased health goods and services. The pitals with serving the public interest and large-scale reforms initiated in 2009 have not increasing efficiency (State Council 2011). aggressively attempted to correct the mis- These policies included delinking hospital aligned supply-side incentives that have car- income and staff remuneration from drug rev- ried over from the past three decades. enues, changing provider payment methods, Therefore, hospitals seek help from physicians promoting rational drug use, testing alterna- by offering them bonus schemes linking their tive governance structures, improving human performance with hospital revenues. The net resource management, and adopting more result is a situation almost diametrically oppo- efficient internal management. A zero-markup site to PCIC, in that health sector resources drug policy was introduced, which delinked get reallocated to profit centers for hospitals income of public hospitals from drug sales. and away from patient-centered provision and However, the policy only applied to drugs on physicians get drawn into revenue generation, the Essential Drug List (EDL), and hospitals which influences their treatment choices. continued to charge a markup on drugs not on The fee-pricing schedule widely used by the EDL. To compensate hospitals for the loss purchasing agencies prices some services in drug revenue, local governments were below cost (such as health promotion, pre- encouraged to increase their direct subsidies to vention, and consultations) and other services public hospitals. In addition, the schedule of above cost (such as expensive diagnostics). hospital prices was revised, and prices for This motivates oversupply of services with high-tech diagnostic tests were lowered while higher price margins and steers public pro- the prices of more labor-intensive services, viders away from prioritizing the public which were previously underpriced, were interest. Because this fee schedule yields the raised. lowest profit for providers of health promo- The Ministry of Human Resources and tion and prevention services, such services get Social Security (MoH RSS) issued the neglected and physicians favor overprescrip- “Opinions on Further Improving the Reform tion of antibiotics and intravenous injections, of Health Insurance Payments” and the even for simple health problems, to generate “Opinion on Implementing the Control of revenues. Unsurprisingly, 75 percent of Total Medical Insurance Payment” in 2011 patients suffering from a common cold and and 2012, respectively, which state the objec- 79 percent of all hospital patients in China tives and implementation pathways for the are prescribed antibiotics—numbers that are global budget payment method.1 In line with more than twice the international average the MoHRSS guidelines, a global budget and that have contributed to growth in health payment method has been implemented spending (Zhou n.d.). L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 207 At the same time, the fee schedule incen- the share of township hospitals fell from tivizes even public hospitals to profit from 11 percent to 6 percent, and the share of expensive procedures, and thus hospitals ambulator y health facilities fell from invest heavily in new technologies and medi- 21 percent to 9 percent during this period cal devices and get a high initial stock of (figure 6.1). patients to defray the fixed costs. Reinforced In the absence of a strong primary care by the profit motivation, higher-level hospi- system and an effective referral system, tals, which are at an advantage for capital patients could themselves choose at which investment, keep expanding and drawing in level and where in the hospital chain they more and more physical, financial, and could seek care. Copayments are differenti- human resources at the expense of lower- ated across levels, with higher reimbursement level hospitals, which cannot compete at the at lower levels (that is, reimbursement is same level of technology. higher at secondary facilities than at tertiary The net result is a resource-rich tertiary facilities for the same procedure), but the dif- hospital base that stands together with ference is not sufficient to deter patients from poorly resourced lower-level facilities, a bypassing the secondary facilities for tertiary situation that adversely affects the ability levels because of the perceived quality of lower-level hospitals to provide quality difference. medical services and motivates doctors to And finally, because higher-level facilities seek employment in tertiary facilities where typically attract more specialists and are bet- their income prospects are brighter. Patients ter equipped with high-technology devices, get directed to higher-level facilities, result- patients show a strong preference for s ­ eeking ing in an inefficient situation in which con- even basic care at these high-level facilities. gested higher-level facilities coexist with The net result of this choice process is con- idle resources in lower-level hospitals. gestion, long waiting times, higher marginal Unsurprisingly, while hospitals’ share of cost of production, shorter physician time, total health spending in China rose from more high-tech diagnostics, and related 56 percent in 1990 to 63 percent in 2012, inefficiency- and cost-enhancing outcomes. FIGURE 6.1  Composition of total health expenditure in China, by facility or provider type, 1990–2012 100 Share of total health expenditure, percent 90 80 70 60 50 40 30 20 10 0 1990 1995 2000 2005 2010 2012 City hospitals County hospitals Township health centers Community and other hospitals Ambulatory facilities Pharmacies Public health, admin, and others Source: China Health Statistical Yearbook 2013, Ministry of Health. 208 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Challenges and Lessons Learned URBMI is designed to cover children, stu- From Selected Case Studies dents, the elderly, people with disabilities, and other nonworking urban residents. The Overview of Health Insurance and scheme is financed mainly by the insured Assistance Schemes along with a government subsidy, which As described in chapter 1, the medical secu- increased from RMB 40 in 2007 to RMB rity system in China consists of three basic 420 in 2016. Local governments have the health insurance schemes: Urban Employee authority to decide premium and benefit lev- Basic Medical Insurance (UEBMI), Urban els for different populations according to Resident Basic Medical Insurance (URBMI), affordability and public financing.5 Initially, and the New Cooperative Medical Scheme URBMI covered inpatient care only, but local (NCMS). A medical financial assistance governments have now established pooled (MFA) program for the poor includes urban funds for outpatient services. On average, and rural MFA schemes. By the end of 2015, patients can get 45–50 percent of total inpa- UEBMI, URBMI, and NCMS covered 289 tient expenses reimbursed. million, 377 million, and 670 million resi- The NCMS provides insurance coverage dents, respectively.2 to rural residents. Premiums are heavily sub- In addition, the Insurance Program for sidized (RMB 420 in 2016), while individual Catastrophic Diseases (IPCD) provides finan- contributions are much lower (RMB 120). cial protection for the high medical expenses Most NCMS models initially covered only of patients with catastrophic diseases, cover- inpatient services, although some covered ing 50 percent or more of eligible expenses inpatient plus several selected outpatient ser- over and above the ceiling of basic insur- vices for major acute illnesses, and a small ance. 3 Finally, the Emergency Rescue and proportion covered both inpatient and outpa- Financial Assistance System for Disease, tient services. In the western and central established in 2013, provides emergency regions, the most commonly found model health expenses for patients who lack identity combines a medical savings account (MSA) documents as well as for patients who have for outpatient services and high-deductible identity documents but who cannot pay.4 catastrophic insurance for inpatient services. The three health insurance schemes differ Since 2010, a growing number of regions in how they are financed and operate. have ended MSAs and begun to reimburse UEBMI is a compulsory scheme to which both outpatient and inpatient services from employers and employees contribute 6 per- the pooled fund, based on predefined drug cent and 2 percent of the employees’ wages, and service formulas. On average, the NCMS respectively, to enroll. Contributions are reimburses 55 percent of inpatient costs and divided into two parts: two-thirds of the 50 percent of outpatient costs. employer’s contribution is allocated to a risk- All three schemes have their own formulas pooled fund (the “social pooled fund”), while for reimbursing drugs and services. UEBMI the remaining one-third is combined with the and URBMI are the most generous, covering 2 percent from employees and deposited in an more than 2,000 drugs. The NCMS is much individual savings account owned by the less generous than the other two. The reim- employee. In some cases, the savings account bursed lists are similar for UEBMI and is used to pay for outpatient services while URBMI, though UEBMI reimburses larger the risk-pooled fund is used to pay for inpa- amounts. The EDL (520 drugs in 2012) is tient care. In other cases, funds in the savings included in the drug list of the three insurance account can pay both inpatient and outpa- schemes and has a higher reimbursement rate tient deductibles before using risk-pooled than other drugs. Primary health care institu- funds. On average, patients can get 65–70 tions can only provide and sell essential drugs, p erc ent of tot a l i npat ient ex p en s e s which limits the actual benefit packages at reimbursed. these institutions. L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 209 The rest of this section presents results of a leverage strategic health purchasing and specially commissioned study conducted in stimulate changes in provider behavior, the selected provinces and counties, representing performance of the purchasing agencies con- each insurance scheme and each geographic tinues to be assessed mainly on balancing region of the country, to assess opportunities revenues with claims and on ensuring fund and challenges in purchasing practices and safety while increasing reimbursement rates. identify road maps for strengthening the The obligations and functions of an capacities of social insurance agencies. NCMS management office are to develop Strategic purchasing functions include strate- local NCMS policies; organize and mobilize gic planning and policy development, financial implementation under the framework of pro- management, benefits design, contracting with vincial NCMS policies, project revenues, and providers, and performance monitoring. expenditures; enroll the target population; The sample included the following cities or and train subordinate entities in the towns to counties, by region: collect revenues, examine provider qualifica- tions, monitor health services, audit expenses, • East: Beijing, Hangzhou in Zhejiang prov- and publish financial reports. NCMS offices ince, Changshu and Jiangyin counties in in most counties are thinly staffed and lack Jiangsu province, Shanghai, Tianjin, and the capacity to perform all of these functions. Zhuhai in Guangdong province As a result, about one-fourth of all NCMS • West: Jiulongpo county in Chongqing, management offices in the country have hired Puding county in Guizhou province, commercial insurance companies to do some Yanchi county in Ningxia province, and routine tasks. Local NCMS offices have more Xining county in Qinghai province authority for strategic planning and policy • Central: Xi county in Henan province. making under the national and provincial requirements released by the National Health Most reforms have been accompanied by and Family Planning Commission (NHFPC). investments in the capacity of purchasing The study of selected provinces shows that in agencies, especially for the main functions of Puding, Xi, Jiangyin, and Changshu coun- maintaining fund balance and increasing cov- ties, for instance, local NCMS offices make erage and reimbursement rates. However, pur- decisions on issues related to the benefits chasing agencies still have limited capacity to package, provider payment, and monitoring leverage contracting, develop and manage and contracting rules. effective provider payment systems, and moni- tor provider performance. Many purchasing Revenues Not Always Easy to Forecast and provider payment innovations build on historical claims and rely on the technical con- In general, all three schemes have similar tribution of external experts rather than on financial management processes: projecting routine in-house capacity, and purchasing the budget, collecting revenue, allocating and agencies continue to have limited ability to using funds, dealing with surplus and debts, monitor the case mix and quality of services. auditing, and monitoring. At the end of a fis- cal year, each purchasing agency is required to project revenues and expenses for the next Weak Purchasing Capacity in Basic year, and the projections need to be approved Medical Insurance Agencies by the Financial Department. The mandate of the purchasing agencies of NCMS and URBMI revenues are difficult China’s three insurance schemes is to balance to predict because enrollment is voluntary revenue with expenditure while maintaining and government subsidies are not announced and expanding financial protection for the until the beginning of the new fiscal year. insured. Although the 2009 and 2012 reform Revenues are not all received until the middle guidelines acknowledged the need to better of the year because contributions come from 210 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A different sources and the appropriation of offices to cover the 22 catastrophic diseases government subsidies takes time. (including leukemia, multidrug-resistant In contrast, because UEBMI premiums are tuberculosis [MDR-TB], and breast cancer) mandatory and automatically collected from and has set the minimum reimbursement rate employers’ income and employees’ wages, at 70 percent. The central government also they are usually received in a stable, predict- sets the number of reimbursable drugs (500– able, and timely way. Revenue forecasts for 800 for township health centers, 800–1,200 UEBMI are based on predicted economic for county hospitals) and has set reimburse- growth, insurance coverage, employees’ aver- ment rates 5–10 percent higher for essential age wages, and contribution rates. Estimated drugs. payouts are based on the age structure of the MoHRSS determines the national basic insured, the spectrum of disease, inflation, insurance drug list. Provincial and local benefit packages, and fund balances. MoHRSS bureaus define the formularies for All agencies analyze expenditures during health services and medical technology and the year—usually monthly, but sometimes in equipment. The provincial bureaus can also real time in places with modern information adjust the national drug list (only for systems, such as Shanghai, Tianjin, and Category B drugs, though) according to the Changshu county. Surpluses and debts are local context. Both the national and provin- managed according to guidelines issued by the cial basic insurance drug lists are reviewed central government, which suggest a balance by an expert panel, members of which have of at least 15 percent of annual requirements a pharmaceutical, clinical, pharmacoeco- in NCMS and of six to nine months of pay- nomic, or health care management back- ments in UEBMI and URBMI. Payments to ground. The drugs are listed by their generic providers are usually regular but may be sus- names, and province-based competitive bid- pended if reserves are inadequate (as occurred ding and bulk purchasing systems run by the in Puding county and Beijing in 2013) or if the provincial NHFPC are used to determine the payer needs more time for performance assess- actual drug types, volumes, and providers. ment (as was the case in Yanchi county). The provincial and the lower-level basic med- ical insurance schemes (BMIs) have the right to decide on disease scopes and cost-sharing Benefit Packages Fragmented Across arrangements, which usually exempt the Insurance Schemes elderly and patients with chronic or cata- Each NCMS office can decide on benefits, strophic diseases with higher reimbursement cost sharing, and other arrangements, and so rates and higher ceilings. many different models exist. The provincial Since 2012, the central government has health bureau generally takes responsibility mobilized each province to initiate a new for designing the disease scope, health care insurance scheme in addition to URBMI. package, drug list, and reimbursement rates Operated by commercial insurance institu- for each level of providers. The local county tions, this scheme targets the urban unem- office can adjust the packages and set the ployed who have high out-of-pocket expendi- cost-sharing arrangements under the leader- tures and offers them an extra minimum 50 ship of the county government and the county percent reimbursement. This scheme is also health bureau to reflect the population’s suggested to be adopted in addition to the needs and preferences. When deciding basic NCMS. coverage of services, NCMS should take into account the burden of disease, cost- Selective Contracting Not Used effectiveness, and provider capacities. Effectively The services and medicines in the benefit packages reflect national policy priorities and In theory, both the NCMS and the urban resource constraints. In recent years, the BMIs can selectively contract with public and NHFPC has called upon pooled NCMS private providers annually to deliver the L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 211 defined list of services and drugs to the insured auditing of claims and paying of bills; they do population. In practice, however, all public not routinely collect data on performance. providers in the city are contracted, and their There are exceptions, however. Ningxia contracts are renewed almost every year, rais- province’s Yanchi and Haiyuan counties rou- ing the question of whether URBMI uses tinely measure outcome and system indica- selective contracting effectively as an incentive tors, such as the antibiotics usage rate, and mechanism. withhold 40 percent of the capitation fee Besides, the contents of agreements are usu- ­ contingent on suitable performance. ally the same for each provider and are in the Most BMIs have the authority to set stan- form of broad principles, which are more like dards, verify that standards are followed, and general policies without specific terms and determine the consequences for poor quality. payment figures. In this way, providers are Jiangyin and Changshu counties not only “forced” to undertake the agreement rather monitor providers’ performance but also help than accept it after mutual negotiation. them to improve. The performance assess- Finally, the legal positions of the purchaser ment system in Yanchi county focuses on and the providers and the regulatory founda- quality, clinical outcomes, and a competitive tions are not clear, compromising enforcement pay-for-performance scheme that is profes- of the contracts. sionally managed by an external expert team. Until a few years ago, provider payment Most BMIs strictly monitor the indicators under UEBMI and URBMI was largely fee- that readily flow from the information sys- for-service. In 2012, MoHRSS launched tem, such as budget execution, actual patient provider payment reforms—shifting post- reimbursement, expenditure per visit or payment to prepayment and introducing admission, length of stay, total volume, ratio blended ­ payment methods—to control pro- of total hospital visits to total patients, read- gram expenditure, improve efficiency, and mission rate, and so on. They also monitor enhance quality. Payment rates were set on expenditures on drugs and consumables the basis of growth-adjusted historical prices because providers can make high profits on and negotiations with providers. Providers these items. In addition, they monitor certain were allowed to keep residuals, while pur- efficiency indicators—for example, the extent chasers and providers shared reasonable lev- of “upcoding” (assigning an inaccurate bill- els of overspending. ing code to a medical procedure or treatment Within two years, more than 80 percent of to increase reimbursement) in Xi County and the regions had initiated pilots and adopted the expenditure consumption index (expendi- capitation or global budgets for outpatient ser- ture for treating the same disease category) vices and case-based or per diem payment for and case-mix index in Beijing—to ensure selected inpatient care. However, without that the diagnosis-related groups (DRGs) analysis of volume and quality, it is difficult to provide appropriate payment (that is, there is determine whether these reforms have yielded some, but not excessive, cost variation within the desired benefits, either in controlling groups). expenditures or in enhancing quality of care. Insurance agencies also pay attention to prescribing practices—an important cost driver. The most common indicator they Provider Performance Inconsistently track is drug expenditure as a proportion of Monitored total expenditure. Some counties track other Most NCMS purchasers play a passive third- indicators as well, such as the standard pre- party payer role, and although there have scription rate (Xi county); the antibiotic use been some efforts to introduce performance rate and rate of “split prescriptions” to track evaluation, the indicators are largely tied to single prescriptions that are intentionally split the number of visits and volume of services into two or more separately billed orders provided and less to the quality of these ser- (Yanchi and Changshu); prescription doses, to vices. In general, NCMS offices emphasize track illegal acquisition of prescription drugs 212 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A for personal use or profit (Beijing and models that incentivize providers (hospitals, Jiulongpo); the intensity of antibiotic use, the physicians, and other professional health care hormone use rate to determine whether the providers) on the basis of quality, outcomes, lowest necessary dose of hormones is being and cost containment. In promoting patient used for the shortest needed time, and the fluid value and efficiency, these payment methods infusion rate (Changshu); and whether com- shift some risk to the provider. (Box 6.1 munity health centers use only NCMS-listed ­ contains a typology of payment methods.) medicines (Jiangyin). Several local experiments are also going Several counties and provinces monitor on in China that use payment incentives in quality and clinical outcome indicators, such support of some aspect of PCIC. Some of as coherence of clinical practice with guide- these experiments are aimed at integrating lines (Xi); positive rates of medical tests care between town and village providers, (Xi, Hangzhou, Jiangyin, Changshu); diag- while others focus on creating incentives for nostic accuracy and prescription accuracy county hospitals to become effective gate- (Changshu); treatment according to indica- keepers. Lessons from these pilots are rele- tion (Hangzhou); hospitalization in accor- vant to the discussion of PCIC. dance with indication (Jiangyin); cesarean Provider payment reforms in China started section rate (45 percent or less in tertiary hos- over 15 years ago when, in a policy issued in pitals and 20 percent or less in community 1999, the former Ministry of Labor and centers [Jiangyin]); mortality (Yanchi); read- Social Security (MoLSS) promoted global mission rate (Hangzhou, Xining, Shanghai, budgets, fee-for-service, and per diem pay- Changshu); and referral rate (Xining and ment methods for UEBMI.6 Five years later, Changshu). in 2004, the Ministry of Health (predecessor Different performance indicators are used to the NHFPC) introduced case-based pay- in different ways. When the proportion of sur- ment in seven pilot provinces.7 plus or overspending is set as a performance Innovations in payment mechanisms indicator in a prospective payment system (as started after 2009, when the State Council in Xi, Hangzhou, Xining, and Zhuhai), it is issued an opinion on deepening health usually used to decide how much the providers reforms and encouraged pilots on case-based can be paid when there is a surplus or deficit. payment, capitation, and global budgets.8 Some agencies (Jiulongpo, Beijing, Tianjin, In 2011, MoHRSS issued specific policy and Zhuhai) use indicators of efficiency guidelines on provider payment reform, offer- and cost containment to adjust the budget at ing a road map for achieving a series of the beginning of the next year. Others (Xi, national requirements:9 Yanchi, Jiangyin, and Changshu) set aside a fixed proportion of the final budget to link • Expenditure control, based on insurance with performance—usually quality, efficiency, fund revenue and expenditure projections clinical outcomes, or patient satisfaction indi- • Global budget prepayment for specific cators. In Beijing and Changshu, the copay providers, considering institutional char- rate over the budget cap also depends on the acteristics and service volumes assessment results, and the agency only shares • Capitation for outpatient services the surplus with providers who perform well • Case-based payment for inpatient and cat- in the performance assessment. astrophic outpatient services, or per diem payments for inpatient bed-days in areas Provider Payment Reforms in where case-based payment or capitation for out pat ient c a re cou ld not b e China: Lessons From Recent Pilots implemented Globally, payers are moving away from fee- • A negotiation mechanism between insur- for-service, volume-driven health care ser- ance funds and providers to decide pay- vices and toward value-based payment ment rates. L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 213 BOX 6.1  A typology of health care provider payment methods Payment methods can be summarized as follows: involves managing the rates of utilization of medical services by a defined population. Providers typically • Per time period, that is, a fixed payment per year, have considerable control over utilization of services, such as in a salary particularly unnecessary services, as well as over the • Per beneficiary, or capitation, in which the pro- quality and services they provide. vider receives a fixed amount per beneficiary Under risk-based contracting, the performance • Per recipient , which is like capitation except risk is shifted wholly or partially to the provider. that it covers specific services a patient might Risk-based models include “upside risk” if the pro- receive vider only shares in the savings and not the risk of • Per episode, that is, bundled payments or prospec- loss. A “downside risk” arrangement is one in which tive, diagnosis-related group–based payments providers share in the savings and are responsible for • Per day , in which a provider receives a fixed a portion of any losses. amount per day irrespective of services provided • Per service, that is, payment according to number Bundled payments of services • Per cost, or cost-based reimbursement Bundled payments are a type of prospective payment. • Per charge, that is, cost plus a set profit markup Under bundled payments, providers share one pay- ment for a specified range of services as opposed to Risk-based payment models each provider being paid individually. The intent of Risk-based arrangements are based on estimates bundled payment is to foster collaboration among of expected costs to treat a particular condition or multiple providers to coordinate services and control patient population; they include capitation, bundled costs, reducing unnecessary utilization and improving payments, and shared savings arrangements. These patient care. models put the onus on providers to manage utiliza- Bundled payments share risk between the payer tion and treatment expenses. and providers and are the middle ground between Two types of risks are inherent in a risk-based fee-for-service (in which the payer assumes the risk) contract: insurance risk and performance and utili- and capitation (in which the provider assumes the zation risk. Insurance risk entails the financial costs risk). Providers tend to be a little apprehensive about of diseases, accidents, or injuries spread out over bundled payments because heterogeneity might not a covered population. Carriers of insurance risk be fully reflected in reimbursements and because the typically hold sufficient financial reserves to cover lack of accurate cost data at the condition level could the insurance risk. Performance or utilization risk create financial exposure. Sources: Frakt 2016 and “Alternative Payment Models: Frequently Asked Questions,” American Academy of Pediatrics (accessed January 19, 2016), https://www.aap.org/en-us/professional-resources/practice-transformation/getting-paid/Pages/Payment-Models.aspx. The regulations encouraged establishment global cap for payment to providers by the of new payment mechanisms with reference BMIs, determined on the basis of a number payment rates based on historical fees, fund of factors, including premium collection, affordability, and current payment policies. fund risk considerations, price level, and his- They also suggested rate adjustments based torical utilization of health care.10 In the on social economic development, provider same year, State Council policy directives service capacity, suitable technology applica- mandated that facilities implement payment tion, the consumer price index, and medical reforms involving global budgets, case-based input price changes. Further, the regulations payments, or per diem payments.11 suggested a global cap on all payment Issued over the years, these directives have arrangements for different providers. In spawned local experiments involving a switch 2012, MoHRSS, the Ministry of Health, and from fee-for-service payment to global bud- the Ministry of Finance issued a policy on a geting, capitation, case-based payment, per 214 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A diem payment, or pay for performance, as a patient satisfaction differed little between the result of which financial risks have begun to fee-for-service and capitation models. shift to providers. Case-Based Payments Global Budgets Despite a decade of experimentation, the use Global budgets that support integration of of case-based payments (that is, negotiated care and cap expenditures are being imple- rates per treated case) remains limited. mented in a few provinces and counties, Further, rate setting commonly reflects past and they have been effective in slowing trends in caseloads, service volumes, and the growth of insurance expenditures. In prices under the fee-for-service system, Shanghai, for instance, UEBMI switched which also remains the basis for setting from fee-for-service to global caps in 2003 patient copays. and introduced mixed methods, including There are notable exceptions, however. In fee-for-service, per diem payment for mental Beijing, UEBMI pioneered the first DRG sys- diseases and case-based payment for diseases tem in China in six hospitals in 2011, cover- or treatment procedures, to make settlements. ing 108 groups. An evaluation using hospital Global budget prepayment was adopted for discharge data from the six pilot hospitals all providers in Shanghai in 2009. and eight other hospitals that continued to Likewise, Hangzhou determines the global use fee-for-service (and served as controls) budget of each hospital based on its historical found that DRG payment led to reductions of fee claim data, institutional level, and service 6.2 percent in health expenditures and 10.5 characteristics, with adjustments for inflation percent in out-of-pocket payments per hospi- and policy changes. The profits and losses of tal admission. However, hospitals continue to the prepaid budget are shared between use fee-for-service payments for older patients UEBMI and providers. and patients with more complications. In Shanghai, the insurance agency pays the provider a fixed case rate regardless of Capitation actual expenses. An evaluation of the There have been several promising pilots Shanghai experiment showed that to safe- with capitation as well. In Zhenjiang guard profits, hospitals engaged in several (Jiangsu province), capitation is set for pri- opportunistic behaviors, including reducing mary care providers under the budget cap, patients’ length of stay. Hospitals also used based on yearly treatment costs, including cost-shifting tactics, resulting in uninsured medicines and tests. An incentive rule is set patients incurring higher expenditures to up for primary care providers, and full pay- compensate for reduced revenues from ment is made only when the fee for chronic insured patients. treatment reaches 70 percent of the chronic In some instances, global budgets included capitation. a pay-for-performance (or activities) element, Likewise, Changde in Hunan province uses as in Changshu and Hangzhou and in the capitation for inpatient services even in ter- Jiulongpo district of Chongqing, but without tiary hospitals. URBMI allocates 87 percent addressing the possible negative impact on of the fund for capitation payments to provid- quality of care. ers, and the balance is kept as a reserve and risk adjustment fund. A 2008–10 evaluation Per Diem Payments found that this reduced inpatient out-of- pocket costs by 19.7 percent, the out-of-pocket Some counties are experimenting with per share by 9.5 percent, and the length of stay by diem payments. One example is Shenzhen in 17.7 percent. However, the total inpatient cost, Guangdong province, which pays per diems the drug cost ratio, treatment outcomes, and for inpatient services. The total payment is L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 215 determined as a rate per inpatient day, The main indicators used to assess perfor- adjusted for inpatient volume calculated as mance in Xi include antibiotic and steroid use real inpatient volume multiplied by the compliance with regulations (more than inpatient-outpatient ratio. The gap between 95 percent), positive rate of tests done on the payment rate and the fee calculated using large equipment (more than 70 percent), the fee schedule is shared. ­ compliance with hospitalization diagnostic Another example is Changshu in Jiangsu criteria (more than 90 percent), and patients’ province, where URBMI has set up specific satisfaction level (more than 95 percent). per diem rates based on disease severity, Experts are invited to assess provider perfor- treatment period, and institutional level. The mance, and institutions with assessment rates for surgeries vary with presurgical scores of 85 or higher are paid 100 percent of hospitalization, surgical procedure, and post- ­ their claimed payments. Institutions scoring surgical care, and the rates decrease when 80–85 points receive 95 percent payment, inpatient days increase. scores bet ween 75 – 80 points receive 90 ­percent payment, and so on. Institutions scoring 60–65 points receive only 75 percent Mixed Methods of their claims and a warning. Those that Some counties are experimenting with mixed score below 60 three years in a row are not provider payment methods. Guizhou prov- contracted by the Xi NCMS again. ince introduced a salary-plus-bonus payment for village doctors instead of fee-for-service Performance Indicators and and removed the incentives for overprescrib- Monitoring Systems ing medications. An evaluation showed that both outpatient costs and spending on drugs In Xining county (Qinghai province), fell, but doctors increased nondrug services URBMI collects information on efficiency such as injections and had more incentive to indicators (such as rehospitalization rates, refer patients to hospital care, which in turn referral and transfer rates, average inpatient increased total health care costs. bed days, and large equipment cost ratios); In Ningxia province, an intervention tar- patient economic burden indicators (such as geted at primary care providers combined the share of medicines in total expenditures capitation with pay-for-performance incen- and share of out-of-pocket expenditures); tives. An evaluation showed that both anti- and patient satisfaction rates. If hospitaliza- biotic prescriptions and total outpatient tion rates or referral and transfer rates exceed spending declined without major adverse ­ 1 percent of the standard, annual payment is effects on other aspects of care. reduced by 0.2 percent; if the medicine cost In Xi county (Henan province), the NCMS share and out-of-pocket proportion exceed categorizes all diseases treated by public 1 percent of the standard, the fund payment county hospitals into three groups (A, B, C) is reduced by 1 percent. There are similar according to the clinical characteristic of the penalties for other indicators. diagnosis and case mix. For frequent and less- Jiangyin county and Changshu (Jiangsu severe cases (groups A and B), case-based pay- province) have also defined performance prescribed ment rates are set to include services ­ indicators for outpatient and inpatient in clinical guidelines. The case payment rate is ­ services. Besides indicators of diagnostic and based on actual average fee-­ for-service rates examination accuracy, as in Jiangyin, the over the past three years and is negotiated Changshu NCMS uses outside treatment with providers. More complicated cases referral rates as an indicator of provider ser- (group C) remain under the fee-for-service vice capacity and quality. It also focuses on payment. The NCMS prepays 40 percent of prescription behavior, looking at antibiotics total payments to providers, with the remain- and combined use, hormone use, intravenous ing 60 percent based on performance. fluid use, and divided prescriptions. 216 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Use of technology and information man- which separated hospital revenue from agement in performance monitoring has prescriptions improved in Jiulongpo, Zhuhai, Xi, and • Increased hospital fees , as a result of Xining counties and in the Beijing, Hangzhou, which the 22 public hospitals recouped and Shanghai municipalities. In particular, 87 percent of their losses due to the zero Xi county has developed a well-functioning markup on prescription drugs provider performance monitoring system that • Raised the incomes of hospital direc- combines routine performance indicators tors and health professionals by increas- generated by the information technology (IT) ­ ing salaries as well as performance system with random clinical audits. Progress bonuses in most agencies has been constrained, how- ever, by fragmented IT systems, limited staff As a result of these interventions, health capacity, and poor connectivity. Some of the expenditures fell across the board, UEBMI more innovative purchasing agencies reported made a surplus (after years of deficits), and that they have built capacity through a incomes of health professionals increased. combination of well-trained staff and more ­ Some notable results are as follows: demanding recruitment requirements, strate- gic investment in IT and data analytics capac- • Expenditures on prescription drugs fell ity, and exposure to Chinese and international 18 percent, from RMB 680 in 2011 to learning and exchange programs. RMB 560 in 2013, in the 22 public hospi- tals. The average cost of drugs per outpa- tient visit fell below provincial and Incentive Reforms in Sanming: national averages. A Case Study • Expenditures on inpatient services fell by Sanming, a city in northwest Fujian province, 6.6 percent in 2013 compared with 2011, has a population of 2.74 million. It is an old after having grown at an average annual industrial municipality and has a high per- rate of 6.2 percent in the preceding five centage of retirees in its population. The years. health expenditures of Sanming’s 22 public • Expenditures on outpatient services saw hospitals increased steadily from RMB 856 no significant change between 2011 and million in 2008 to RMB 1.69 billion, as a 2013, but their rate of growth was the result of which UEBMI in Sanming has been slowest in the whole province. running a deficit since 2009 (which had accu- • UEBMI gained surpluses (after several mulated to RMB 17.5 million by 2011). years of deficits) of RMB 26.3 million in Against this backdrop, Sanming launched 2012 and RMB 116.8 million in 2013. widespread reforms in February 2012, nota- • Revenue from drugs and consumables bly the following (Ying 2014): dropped from 60.1 percent of all revenues in 2011 to 38.3 percent in 2013 across the • Established the Sanming Public Health 22 public hospitals. Insurance Fund Management Centre • Revenue from health care services rose (PHIFMC) to manage important decisions from 39.9 percent of all revenues in 2011 including drug procurement, utilization to 61.7 percent across the 22 public review, and cost monitoring hospitals. • Centralized provincial drug procurement • Salaries of health care professionals under the PHIFMC, with drug suppliers increased significantly. The total salary chosen through an online bidding process bill of the hospitals increased by 42 per- managed by the PH I FMC , not the cent, from RMB 498.4 million in 2012 to hospital RMB 709.2 million in 2013. Physicians’ • Implemented a zero-markup policy for average annual income increased by prescription drugs and consumables, 48 percent, to RMB 99,800. L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 217 Core Action Areas and generates huge transaction cost problems for Implementation Strategies to a market system and challenging incentive specification problems for a centrally planned Realign Incentives system (Hodgson 2008). Implementing health financing reforms can In China, it is widely accepted that effec- be challenging, even in the most advanced tively leveraging the power of strategic pur- and organized of economies. One common chasing, contracting, and paying providers feature of most health systems is that of flow could improve the value of the government’s of finances from the population and the gov- large investment in the health sector. ernment, via a variety of agencies, to a large, Significant steps have been taken in recent disparate group of providers of health goods years to build the role of health purchasing and services. Each transfer involves a trade of agencies, develop their institutional capaci- sorts, an exchange of trust, in which the ties, and test innovative contracting and pro- payer entrusts money to an agent in exchange vider payment approaches. Now deeper for some desired aspect of health care deliv- reforms are required to overcome the legacies ery. Sometimes the trade is instantaneous, as of earlier policies, support the PCIC when an individual goes to the pharmacy and approach, and yield better value for money. buys an over-the-counter drug. Many times, China can build on the experiences of the the exchange takes the form of a promise, as many successful pilots and experiments both when households entrust their money to an within and outside the country to leverage agency that they trust will buy and deliver the power of strategic purchasing and put in the best health care for them whenever in the place a set of incentives that motivates pro- future they may need it. viders at all levels to provide the best-quality Each transfer in the health system gives health services at lowest cost for all citizens rise to a principal-agent situation in which of the country. one party is being paid by another to do The proposed health financing reforms fall something, a situation that can rapidly turn into four broad thematic categories, each into a problem if the agent is motivated to act with its own distinct objective and associated in its own best interest instead of that of the action points (table 6.1): principal. This is relevant to several key health financing elements: raising finance, 1. Implement provider payment reforms in transferring funds to providers at different support of PCIC. levels, and purchasing of health services 2. Bring about coherence and consistency in by different agents, to name just a few. incentive mechanisms and strengthen Each poses a real danger that the interests of integration of care. the actors in the health system—especially 3. Rationalize distribution of services by ­consumers—are compromised. level of facility. The abnormal economic features of the 4. Strengthen capacit y of purchasing health sector further complicate the situation agencies. (Hsiao 1995). The problems related to uncer- tainty, externality, and information asymme- Translating these strategic intents into try are well known. Indeed, the peculiarities individual operating unit plans requires iden- of health care provision are so unusual that tification of critical drivers and decomposi- the usual tenets of neoclassical economics— tion of the complex health financing system assumptions of rational, utility maximizing into discrete, manageable elements across behavior and stable preference functions— functions and across time. The resulting road are not very useful. And finally, the high map, presenting the interaction of multiple degree of variety and idiosyncrasy in the forces over time, will help bring together the acts of seeking health care and of producing many diverse issues that need to be managed and delivering health goods and services through various processes. 218 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 6.1  Four core action areas and implementation strategies to realign health system incentives Core action areas Implementation strategies 1.  Provider payment reforms in support of PCIC ·   Evaluate ongoing reform experiments with prospective payments, and systematically replicate successful efforts in all provinces and cities ·   Switch from fee-for-service to prospective payments for the portion of expenditure that is borne directly by patients ·   Put in place mechanisms for concurrent evaluation of ongoing and new provider payment reforms 2.   Coherent, consistent incentives and stronger ·   Analyze incentive mechanisms across different insurance schemes within each integration of care province to understand areas of consonance and dissonance ·   Develop a strategy for vertical and horizontal consolidation as necessary ·   Establish a designated unit at central and provincial levels to oversee implementation and concurrent evaluation 3.   Rational distribution of services by facility level ·   Determine, standardize, and list procedures at their commensurate level of care (community, township, county, and level 2 and 3 city hospitals) ·   Reassess copayments across different levels and set significantly higher deductibles and out-of-pocket payments for basic procedures that are being demanded at the tertiary level 4.   Capacity building of insurance agencies ·   Develop staff expertise and data analytics capability ·   Empower purchasing agencies and hold them accountable for results ·   Strengthen cost accounting systems Note: DRG = diagnosis-related group; PCIC = patient-centered integrated care. This process is laborious but useful and generate incentives that increase or lower necessary for implementers and decision health spending, improve or compromise effi- makers in the health system and the govern- ciency, enhance or worsen equity, and posi- ment in China to determine which aspects of tively or adversely affect consumer satisfac- the internal and external environments need tion. Which approach to use in which context to be brought together to initiate coherence in requires a deep understanding of the market ideas and actions. Health financing reforms in which providers work, sound judgment, are complex by nature, and organizing and strong administrative and supervisory frameworks help to align activities and steer capacity of regulators. actions necessary for achieving the desired objectives. Strategy 1: Evaluate reform experiments with prospective payments and systematically replicate successful efforts Core Action Area 1: Provider Payment in all provinces and cities Reforms in Support of PCIC China’s contracting and provider payment Provider payment reforms are a central part innovations provide an experience base that of health financing reforms, principally needs to be harmonized and deepened to cre- because payment methods create powerful ate an effective incentive environment across incentives that influence provider behavior levels of care (box 6.2). and have a direct bearing on the efficiency, The experience of ongoing pilots suggests equity, and quality outcomes of health that a combination of case-mix-adjusted, finance reforms. volume-­ based global budgets for inpatient There is no one optimal method of paying admissions; capitation payment for primary providers: depending on the market and insti- care with performance incentives; and tutional context, provider payment methods capped, episode-based payment for outpatient L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 219 BOX 6.2  Examples of provider payment reforms in China Global caps outlays on uninsured patients to compensate for reduced revenues from insured patients. • In Shanghai, Urban Employee Basic Medical • In Beijing, UEBMI pioneered China’s first diagno- Insurance (UEBMI) switched from fee-for-­ service sis-related group (DRG) system in China in six hos- to global caps in 2003 and introduced mixed pitals in 2011, covering 108 groups. An evaluation methods, including fee-for-service, per diem pay- using hospital discharge data from the six pilot hos- ment for mental diseases, and case-based payment pitals and eight other hospitals, which continued for diseases or treatment procedures, to make set- to use fee-for-service and served as controls, found tlements. Global budget prepayment was adopted that DRG payment led to reductions of 6.2 percent for all providers in 2009. and 10.5 percent, respectively, in health expendi- • Hangzhou determines the global budget of a single tures and out-of-pocket payments by patients per hospital based on historical fee claim data, institu- hospital admission. However, hospitals continued tional level, and service characteristics, adjusted to use fee-for-service payments for patients who for inflation and policy considerations. The profit were older and had more complications. and loss of the prepaid budget are shared between UEBMI and providers. Per diem payment Capitation • Shenzhen (Guangdong province) pays for inpa- • In Zhenjiang (Jiangsu province), capitation is set tient services by per diem payment. The total pay- under the budget cap and is based on yearly treat- ment is determined by a rate per inpatient day and ment costs, including medicines and tests. An incen- adjusted inpatient volume (calculated as real inpa- tive rule is set up for primary care providers, and full tient volume multiplied by inpatient-outpatient payment is made only when the fee for chronic treat- ratio). The gap between the payment rate and the ment reaches 70 percent of the chronic capitation. real fee (based on fee schedule) is shared. • Changde (in Hunan province) uses capitation for • In Changshu (Jiangsu province), URBMI has set inpatient services even in tertiary hospitals. Urban up a specific per diem rate based on disease sever- Resident Basic Medical Insurance (URBMI) uses ity, treatment period, and institutional level. In the 87 percent of the fund as the capitation to provid- case of surgeries, the rate varies among presurgical ers, and the balance is kept as a reserve and risk hospitalization, surgical procedure, and postsurgi- adjustment fund. A 2008–10 evaluation found cal care, and decreases as inpatient days increase. that this payment reform reduced inpatient out- Pay for performance of-pocket cost by 19.7 percent, the out-of-pocket ratio by 9.5 percent, and the length of stay by • Guizhou province introduced a salary-plus-bonus 17.7 percent. However, the total inpatient cost, payment method for village doctors in lieu of fee-for- drug cost ratio, treatment effect, and patient service payment and removed the incentives for over- satisfaction showed little difference between fee- prescribing medications. An evaluation showed that for-service and capitation models. both outpatient costs and drug spending fell, but doc- Case-based payment tors increased nondrug services such as injections and gained more incentives to refer patients to hospital • In Shanghai, the insurance agency pays the care, which in turn increased total health care costs. provider a fixed case rate regardless of actual • In Ningxia province, an intervention targeted at expenses. An evaluation of the Shanghai experi- primary care providers combined capitation with ment shows that to safeguard profits, h ­ ospitals pay-for-performance incentives. An evaluation engaged in several opportunistic behaviors, showed that both antibiotic prescriptions and including reducing patients’ length of stay. Hospi- total outpatient spending declined without major tals also engaged in cost-shifting tactics by raising adverse effects on other aspects of care. Sources: Feng and Hairong 2014; Gao, Xu, and Liu 2014; Hong 2011; Hu 2013; Jian and others 2015; Jiang and others 2011; Liang, Wang, and Jing 2013; Liu and others 2012; Wang 2011; Wang and others 2013; Yip and others 2014; Zhang 2010; Zhang and Wu 2013; Zhang and Xu 2014; Zhen 2009. 220 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A specialty services may be the most appropri- Strategy 2: Switch from fee-for-service to ate approach to manage cost escalation, limit prospective payments for the portion of unnecessary referrals, and shift basic services expenditure borne directly by patients delivery to primary care in the short and In implementing global budgets, one pressing medium term. priority is the development of the case-mix- Replicating the successful efforts sys- adjusted, volume-controlled approach to tematically across China’s provinces and effectively control the growth of hospital cities would not only incentivize improved expenditures. This will enable insurance physician-patient contact at the primary agencies to shift savings into primary care. care level but would also strengthen pri- One possibility is to develop a simple set of mary care overall. DRGs for case-mix adjustments to minimize Provider payment systems are undergoing provider cherry picking, complemented by a paradigm shift globally as well, and health rigorous monitoring and pay-for-­ performance care payers are moving away from passively elements that ensure access and quality of reimbursing providers to pursuing a variety care. The approach could build on the experi- of policies to improve the quality and effi- ence of Beijing, where the DRG system is ciency of care. In Organisation for Economic used to adjust the global budgets of all Co-operation and Development (OECD) 263 health care institutions for their case countries, for example, payments for primary mix. In addition, local pilots across the coun- care have become more blended, with coun- try offer a wide range of experiences and tries using capitation and budgets and adding lessons in monitoring and creating incentives ­ elements to drive qualit y or increase for appropriate provider behavior. productivity. It is important that contracting and pay- OECD countries use a variety of ways of ment methods share financial risks between setting fees for primary care providers insurers and providers while improving the (table 6.2). In countries like Chile, Ireland, quality of care and safeguarding the financial and the Netherlands, fees are set unilaterally protection of patients. As a first step, rate set- by the central government. In Greece, ting for provider payments needs to shift Sweden, and Switzerland, fees are set by key from historical claims to cost information. purchasers. In other countries, such as the The way that provider payment rates are set Czech Republic, Germany, the Republic of influences how services are produced, so Korea, and the United Kingdom, fees are linking payment rates to costs can drive more negotiated at the central level between the efficient cost structures. Setting payment purchaser and provider groups. Chile uses a rates above costs for high-priority services third-party negotiator, while Australia, and below costs for low-priority services, for Austria, and Canada negotiate at the example, can improve the efficiency of the regional level. service mix. Inpatient care in many OECD countries is There has been increasing experimenta- paid on the basis of DRG payments alongside tion with new ways of paying providers, espe- other policies to constrain overall budgets cially payment systems that span across levels and stimulate competition, with a focus on of care. One option that China may like to productive efficiency. In many countries, pay- consider is that of bundled payments ment systems have evolved beyond fee-for- (box 6.3), especially since no single method service and budgets and shifted to finance directly rewards improving the value of care. services on the basis of activity using DRGs. Capitation, especially global ­capitation—that Furthermore, payment methods vary by hos- is, a single payment to cover all of a patient’s pital type (public or private) and whether the needs—decouples payment from what pro- payment is being made by social health insur- viders can directly control and rewards ance (table 6.3) or tax-based systems providers for spending less but not specifi- (table 6.4). cally for improving outcomes or value. L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 221 TABLE 6.2  Primary care remuneration systems in OECD countries, 2014 Remuneration of provider setting Remuneration of physicians Predominant primary care Countries Cap FFS P4P GB Other Salary FFS Cap Other Private group Australia ○ ● ● ○ ○ ○ ● ○ ● staffed by ● ● ○ ○ ○ ○ ● ● ○ Denmark physicians and other health Ireland ● ● ○ ○ ○ ○ ● ● ○ professionals Japan ○ ● ○ ○ ○ ○ ○ ○ ○ Netherlands ● ● ● ● ○ ○ ● ● ● New Zealand ● ● ● ○ ○ ● ● ○ ● Norway ● ● ○ ○ ○ ○ ● ● ○ Poland ● ○ ○ ○ ○ ○ ○ ● ○ United ● ● ● ○ ● ● ● ● ● Kingdom United States ● ● ● ○ ● ● ● ● ● Private group staffed Canada ● ● ○ ○ ○ ● ● ● ● by physicians ● ○ ○ ○ ○ ○ ○ ● ○ Italy Private solo practice Austria ○ ● ○ ○ ○ ○ ● ○ ○ Belgium ● ● ○ ○ ○ ○ ● ● ○ Czech ● ● ● ○ ○ ○ ● ● ○ Republic Estonia ● ● ● ● ○ ○ ● ● ● France ○ ● ● ○ ● ○ ● ○ ● Germany ○ ● ○ ○ ○ ○ ● ○ ○ Greece ○ ● ○ ○ ○ ○ ● ○ ○ Korea, Rep. ○ ● ● ○ ○ ○ ● ○ ○ Luxembourg ○ ● ○ ○ ○ ○ ● ○ ○ Slovak ● ● ○ ○ ○ ○ ● ● ○ Republic Switzerland ● ● ○ ○ ○ ○ ● ● ○ Public primary Chile ● ● ○ ○ ○ ● ○ ○ ○ group staffed by ○ ○ ○ ● ○ ● ● ○ ○ Finland physicians and others Hungary ● ○ ● ● ○ ○ ○ ● ○ Iceland ○ ○ ○ ● ○ ● ○ ○ ○ Israel ● ○ ○ ● ○ ● ○ ○ ○ Mexico ● ○ ○ ● ○ ● ○ ○ ○ Portugal ● ○ ● ● ○ ● ○ ○ ○ Slovenia ● ● ○ ○ ○ ● ○ ○ ○ Spain ● ○ ● ○ ○ ● ○ ● ○ Sweden ● ● ● ○ ○ ● ○ ○ ○ Turkey ● ○ ● ○ ○ ● ○ ○ ○ Source: Questions 27 and 33, OECD Health System Characteristics Survey 2012 and Secretariat’s estimates. (For database and questionnaire, see http://www.oecd.org​ /els/health-systems/characteristics-2012-results.htm.) Note: Cap = capitation; FFS = fee-for-service; GB = global budget; OECD = Organisation for Economic Co-operation and Development; P4P = pay for performance. 222 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 6.3  Hospital remuneration systems in OECD countries with social health insurance Country Public hospitals Private not-for-profit hospitals Private for-profit hospitals Austria DRG DRG DRG Belgium Prospective global budget Prospective global budget n.a. Chile DRG DRG Line-item remuneration Czech Republic DRG DRG DRG Estonia DRG n.a. n.a. France DRG DRG DRG Germany DRG DRG DRG Greece DRG DRG Procedure service payment Hungary DRG DRG Procedure service payment Israel Procedure service payment Procedure service payment Procedure service payment Japan DRG DRG n.a. Korea, Rep. Procedure service payment Procedure service payment n.a. Luxembourg Prospective global budget Prospective global budget n.a. Mexico Prospective global budget Procedure service payment Procedure service payment Netherlands DRG DRG n.a. Poland DRG DRG DRG Slovak Republic Procedure service payment Procedure service payment Procedure service payment Slovenia DRG DRG DRG Switzerland DRG DRG DRG Turkey Prospective global budget Prospective global budget Prospective global budget United States (Medicare) DRG DRG Procedure service payment Source: Question 31, OECD Health System Characteristics Survey 2012 and Secretariat’s estimates. (For database and questionnaire, see http:// www.oecd.org/els/health-systems/characteristics-2012-results.htm.) Note: DRG = diagnosis-related group; OECD = Organisation for Economic Co-operation and Development; n.a. = not applicable. TABLE 6.4  Hospital remuneration systems in OECD countries using a tax-based system Country Public hospitals Private not-for-profit hospitals Private for-profit hospitals Australia DRG Procedure service payment Procedure service payment Canada Prospective global budget Prospective global budget Prospective global budget Denmark Prospective global budget n.a. DRG Finland DRG n.a. n.a. Iceland Prospective global budget n.a. n.a. Ireland Prospective global budget Prospective global budget n.a. Italy Prospective global budget DRG DRG New Zealand Prospective global budget n.a. n.a. Norway Prospective global budget Prospective global budget DRG Portugal Prospective global budget Procedure service payment Procedure service payment Spain Line-item remuneration Prospective global budget n.a. Sweden Prospective global budget Prospective global budget Prospective global budget United Kingdom DRG Procedure service payment Procedure service payment Source: Question 31, OECD Health System Characteristics Survey 2012 and Secretariat’s estimates. (For database and questionnaire, see http:// www.oecd.org/els/health-systems/characteristics-2012-results.htm.) Note: DRG = diagnosis-related group; OECD = Organisation for Economic Co-operation and Development; n.a. = not applicable. L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 223 BOX 6.3  Bundled payment models The proliferation of bundled payment models is • In the United States, bundled payments are used transforming the way in which care is delivered. extensively for organ transplant care, as in the Governments, insurers, and health systems in many University of California, Los Angeles kidney countries are trying bundled payment approaches, as transplant program. in these examples: Some U.S. employers have also embraced bun- • In Sweden, the Stockholm County Council dled payments, including Walmart, which intro- adopted bundled payments in 2009 for all total duced a program that encourages employees who hip and knee replacements. The result was lower need cardiac, spine, and selected other surgery to costs, higher patient satisfaction, and improve- obtain care at one of just six providers nationally, ment in some outcomes. all of which have high volumes and track records • Germany uses bundled payments for hospital of excellent outcomes. The hospitals are reim- inpatient care, which has helped control the rise bursed in a single bundled payment that includes in spending on inpatient care. (No additional pay- all physician and hospital costs associated with all ment is made for rehospitalization related to the inpatient and outpatient preoperative and postop- original care.) erative care. Source: Porter and Lee 2013. Fee-for-service rewards providers for increas- levels, conduct periodic review at the provin- ing volume, but that does not necessarily cial and national levels to monitor progress, increase value. Bundled payments that cover and establish credible external review mecha- the full care cycle for acute medical condi- nisms. This review and evaluation will tions, overall care for chronic conditions for a require a complex set of coordinated actions defined period, or primary and preventive at multiple state levels within the provincial care for a defined patient population, are per- and central governments. This process is nec- haps best aligned with value. essary to arrive at an impartial and scientific assessment in support of the scaling-up of a Strategy 3: Put in place mechanisms for reform. concurrent evaluation of ongoing and new Insurance agencies also need to integrate provider payment reforms learning from successful provider payment Rigorous scientific assessment of provider experiments as they are scaled up. For payment reform pilots is essential to judging example, global budgets have been intro- the effectiveness and replicability of the duced to control the growth of expenditures reform in other parts of the country. China in hospitals but have not shifted significant may like to commission a systematic evalua- resources into primary care. Capitation pay- tion of the various reform initiatives in differ- ments have been introduced in many ent parts of the country. instances for outpatient care but not always Such an evaluation will require an inde- with pay-for-performance components to pendent, dedicated, and fully funded autono- mitigate adverse impacts on service volumes mous mechanism established at an arm’s and quality. Further, rate setting in new length from the government to finalize the mechanisms has commonly reflected past list of indicators, gather the relevant data, trends in caseloads and service volumes. evolve sustainable systems for tracking and This historical approach is one reason why data collection, maintain baseline assessment pilots of global budgets have had limited of the indicators at the provincial and central effect on cost containment. 224 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A At the same time, flaws persist that dis- methods much more powerful. If one payer tort payment methods, old and new. Most changes the way it pays providers, the pres- importantly, in the absence of proper cost ence of multiple payers allows providers to information, rates continue to reflect pricing transfer costs to other payers not included in systems that reinforce the overuse of more the payment reform (Yip and Hsaio 2008). expensive care settings and higher-priced For many health conditions, treatment technology and exacerbate the general trend involves providers at different levels at differ- among providers to invest in and upgrade ent stages of the treatment cycle. If the sys- facilities. For example, global budgets tem’s incentive structure motivates each level are typically calculated based on historical of provider to hold on to patients rather than rate systems that pay higher rates for the refer them to the appropriate level of care same service to more specialized service pro- based on clinical need, patients could poten- viders (such as secondary and tertiary care tially receive suboptimal care. A situation of hospitals). Together with reforming finan- internally inconsistent incentives could also cial incentives for providers, efforts need to emerge if different methods being used to pay focus on changing the financial incentives providers at different levels of care result in facing patients so as to steer them away from conflicting or mutually reinforcing but per- higher levels of care. verse incentives. For example, primary care providers who are paid by capitation may be motivated to refer high-cost patients to spe- Core Action Area 2: Coherent, cialists who, if paid under a fee-for-service Consistent Incentives and Stronger system, may readily welcome them. Integration of Care It is important, therefore, to ensure coher- How health financing is arranged across dif- ence and internal consistency in the way that ferent levels of the system may also be a the incentive structures are set up across source of inefficiency, especially if it is the health system. As Shortell (2013) notes, marked by fragmentation, contradictory “the largest limiting factor is not lack of incentives at different points of interaction money or technology or information or peo- between the providers and patients, and high ple, but rather the lack of an organizing prin- administrative costs. Establishing a coherent ciple that can link money, people, technology incentive environment that is internally con- and ideas into a system that delivers more sistent is a necessary step for steering payers cost-effective care (i.e., more value) than cur- and providers of health care toward greater rent arrangements.” efficiency and quality. Strategy 2: Develop a strategy for vertical Strategy 1: Analyze incentive mechanisms and horizontal consolidation across insurance schemes to understand Horizontal and vertical consistency and areas of consonance and dissonance coherence—within and across a facility alli- Health providers in China receive payments ance or network—increase the likelihood from multiple sources: out-of-pocket pay- that payment mechanisms will achieve the ments by patients who pay on a fee-for-­ desired changes in provider behavior. service basis; health insurance payments, also Provider payment mechanisms work best largely on a fee-for-service basis, although when they are defined and applied consis- payment mechanisms are changing rapidly; tently across the full continuum of health and direct government funding linked to care production and delivery, from primary public health goods and input-based subsi- care to tertiary interventions, and are com- dies. If a common provider payment mecha- patible in the sense that all providers— nism were used for the first two of these rev- including hospitals; physicians; and town, enue streams, it would make the positive community, and village health centers—face incentivizing effect of prospective payment similar types of incentives. L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 225 Within the proposed organized networks hospitals and primary care providers to work or alliances for PCIC implementation at the together to implement PCIC. county and district levels, for example, net- Another option could be to consider incen- works could receive a prospective global bud- tive payments outside of the global budget get based on capitation plus other revenues. (such as additional funding) that would need The global budget will necessarily entail a to be earned (box 6.4). Hospital performance hard budget constraint along with measures indicators could focus on patient safety, qual- to avoid cost shifting by providers to patients. ity, and efficiency improvements. Measures The global budget may be based on current of this sort would promote the integration of spending levels initially but have a focus on services across the health system and would controlling future spending growth across also incentivize the network to direct the the entire network. The global budget could flow of patients to the appropriate levels of include a “withhold” of a predefined percent- care. Any savings generated by the network age of funding, which could be paid upon could be shared by hospitals and primary compliance with PCIC-related indicators care providers within that network. such as quality improvement, integrated care, The insurance agencies may also benefit reducing unnecessary care, and shifting inap- from taking a closer look at their budgets, propriate care out of the hospital. benefits, and fees to better balance hospital- Such consolidation would require that the based care for catastrophic diseases with pri- network redefine hospital and primary pro- mary care for common diseases. Specific vider roles and establish formal links. measures are needed to limit the overuse of Network management would need to channel hospitals and to shift resources into primary incentives to hospitals and primary care pro- care as its delivery capacity is developed. viders through, for example, risk-adjusted, A combination of regulatory measures and facility-specific global budgets. This would be payment incentives is needed to shift service especially important to align incentives of delivery to the appropriate level of care. BOX 6.4  Quality-compatible modified global payment systems in the United States In the United States, two health care insurance in the alternative quality contract earned significant providers have successfully implemented payment quality bonuses in the first year. schemes among networks of providers to improve quality and reduce waste and unnecessary utiliza- Patient-centered medical home tion. Both programs offer useful lessons for China. CareFirst’s patient-centered medical home (PCMH) Alternative quality contract program began in 2011, and within three years, over 80 percent of all primary care providers in In January 2009, Blue Cross Blue Shield of Mas- the CareFirst service area—including parts of sachusetts launched a new payment arrangement Northern Virginia, the District of Columbia, called the alternative quality contract. The contract and ­ M aryland—had begun to participate in the stipulates a modified global payment arrangement: program. Since the program began, CareFirst’s fixed payments for the care of a patient during a overall rate of increase in medical care spending specified time period. The model differs from past for its members has slowed from an average of models of fixed payments or capitation because it percent per year in the five years before the pro- 7.5 ­ explicitly connects payments to achieving quality gram’s launch to 3.5 percent in 2013. In addition, goals and defines the rate of increase for each con- CareFirst members under the care of participating tract group’s budget over a five-year period, unlike PCMH physicians fare well when measured on key typical annual contracts. All groups participating quality indicators. Sources: Chernew and others 2011; Murray 2015. 226 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A International experience indicates that as providers and creating incentives to long as hospitals are permitted to deliver enhance provider performance. Moreover, basic outpatient services—and to financially as local government institutions, insur- benefit from such service delivery—shifting ance agencies and public health care the delivery system to a PCIC delivery model providers are part of the same legal entity, ­ is nearly impossible. Options include a mix of which compromises the legal enforceabil- regulatory, administrative, and contracting ity of contractual relationships. Under instruments. Regulations could specify which these arrangements, local insurance agen- services can (and cannot) be delivered at each cies tend to concern themselves with level of the delivery system and could install balancing their revenue with competing ­ primary care providers as gatekeepers. demands from public providers for Budgetary allocations could be earmarked increases in funding. for primary, secondary, and tertiary care to • Efforts need to be made to streamline the direct volume and resources toward lower incentives from different flows of public levels of care. And contracting arrangements funds so as to shape provider behavior. could specify which services will be reim- Local governments, for instance, provide bursed at each provider level. direct subsidies to public providers, often in the form of block grants paying for the Strategy 3: Establish a designated unit salaries of regular employees, among other at central and provincial levels to items. oversee implementation and concurrent • Insurance agencies need urgently and evaluation aggressively to build capacities to monitor Fuenzalida and others (2010) note that effec- service access, quality, and provider tive coordination across various subfunctions responsiveness. Currently, insurance agen- of health financing is necessary for successful cies are primarily invested in building reform design and implementation. The three capacity to process claims and detect subfunctions that are key for successful fraud. implementation of financing reforms in • It is important that all insurance agencies China’s health sector are alignment between build strong monitoring capacity and revenue collection and pooling, alignment establish a core set of indicators across key between revenue collection and purchasing, performance dimensions. and alignment between pooling and purchas- ing. Several important observations made Integrating, or at least harmonizing, the earlier regarding these important subfunc- varied benefits, scheme management, and tions and internal alignment are summarized purchasing practices—as well as channeling here for ready reference: supply-side subsidies through the insurance system—is essential to improving pooling of • Empowering insurance agencies to act as funds, enhancing equity, and creating stream- strategic purchasers of health services can lined incentives for providers. The integration significantly improve the returns on the of the NCMS and URBMI schemes has government’s large investment in the started with some pilots, and Hangzhou’s health sector. China has already taken HRSS, like several others, is managing all many important steps to strengthen the three schemes and harmonizing purchasing institutional capacities of health insurance practices. Such experiences could be studied agencies, but deeper reforms are necessary further, and the emerging trends in integrat- to yield better value for money. ing and harmonizing health insurance • The legal mandate and legislative frame- schemes and reducing supply-side subsidies work governing the insurance agencies could be accelerated. must explicitly state that the agencies are In China, the fragmentation of insurance responsible for selectively contracting agencies across and within schemes limits L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 227 their leverage over the behaviors of provider In January 2016, the State Council released organizations. Within each scheme, the pur- the “Opinions” on the merger of URBMI and chasing function is typically decentralized to NCMS to ensure equal access to medical ser- the municipal or county level, as a result of vices and reimbursement among urban and which literally thousands of insurance agen- rural residents as well as to improve overall cies contract and pay health care providers service efficiency in the insurance system. across the country. For example, in Changshu Governments at the provincial and municipal (Jiangsu province), URBMI pays providers a level were required to make specific plans for capitation fee for general outpatient care, the integration before the end of June 2016. with 5 percent of the payment subject to China may also like to look at some other meeting quality of care standards, while experiences with these kinds of reforms. One UEBMI beneficiaries pay providers on a fee- example is Korea, which merged all statutory for-service basis out of personal MSAs. health insurance funds into a single insurer in Greater coordination across the three insur- July 2000 (Kwon 2015). Before this, there ance agencies would help remove many of was no competition among the insurance these kinds of incongruities in purchasing. funds to enroll the insured, and each insur- Merging the three insurance schemes is not ance fund covered a well-defined population necessarily a solution because it would create group. Except for reviewing and assessing a large single unit, which would be difficult claims submitted by providers, health insur- to manage and administer. In fact, the bene- ance funds did not actively exercise their pur- fits of merging could equally be obtained chasing power and did no selective contract- from enhancing coordination of purchasing ing with providers. The 2000 merger not across the three agencies­—without creating a only introduced a new single insurer large insurance monopoly. agency—the National Health Insurance The government is committed to integrat- Company, which later changed its name to ing the three health insurance schemes, the National Health Insurance Service although the huge variations in financing, (NHIS)—but also created a new insurance benefit packages, reimbursement rates, and review agency, the Health Insurance Review management structures among different and Assessment Service (HIRA). schemes in different areas make this difficult. The NHIS handles premium collection, URBMI and the NCMS have similar financ- fund pooling, and reimbursement to provid- ing sources (insured and public finance), ers. HIRA makes decisions related to pur- benefit packages (inpatient and catastrophic ­ chasing (such as claim reviews) and to the outpatient services), and reimbursement design of the benefits package and provider rates, but the differences from UEBMI make payment system. After the merger, the NHIS integration with the latter problematic. expanded health insurance benefits and cov- Nevertheless, integration of URBMI and the erage, started covering cancer screening, NCMS into UEBMI has begun, and pilots reduced coinsurance rates for catastrophic are under way in Chongqing, Guangdong conditions, introduced ceilings for cumula- province, Ningxia province, and Tianjin. tive out-of-pocket payments for covered ser- URBMI has two forms of integration: (a) vices for every six months, and so on. The overall integration of premium collection, savings in administrative expenses due to benefit packages, reimbursement rates, fund economies of scale in management after the pooling, payment methods, and management merger helped enable the benefits expansion authority and (b) partial integration , in (although it is difficult to confirm the causal which the two schemes are managed by the relation). same authority but with different fund pool- Taiwan, China, also consolidated its frag- ing and benefit policies. In some URBMI mented labor insurance market into a single- schemes, different premium levels have been payer system and set up the National Health set up for the insured to choose among.12 Insurance Administration (NHIA) covering 228 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A health care for all (Leung 2015). With the the last few decades of the 20th century, formation of the NHIA, significant invest- incentivized organizations to offer a full line ments were made to develop the resources of services in their communities. In the and technology to enable efficient adminis- absence of rationalization of the overall net- tration and effective policy making that work, this system duplicated expensive ser- ­ propelled progress in building purchasing vices across its operations, overinvested in functions and capacities. The reform bene- hospital capacity, and underinvested in deliv- fited from strong commitment and advocacy ery of primary care. In switching to a value- on the part of top political leaders, a growing based system, it will be important for China economy, and a team of knowledgeable and to reduce inappropriate utilization and pro- informed health policy advisers. The strong vide patients with the right care at the right organization and infrastructure become the place and at the right time to meet quality, bedrock for building the NHIA’s strategic cost, and access targets. capacities for influencing the use and delivery of health services in Taiwan, China, which Strategy 1: Determine, standardize, and list include the ability to design the benefit pack- procedures at their commensurate level of age, develop provider payment systems, and care by facility level monitor provider performance. The starting point of this rationalization is to The monopsony power to set fees, draw define the scope of services that could be pro- up and enforce contracts, and monitor pro- vided at different levels of care across prov- viders, as granted by law, has enabled the inces, counties, and cities. For lower levels of NHIA to manage health care resource alloca- care, this would imply exiting from complex tion while controlling costs and improving service lines where available expertise and the quality of health care. The single-payer infrastructure may compromise ability to approach not only has the administrative deliver high-quality care. For tertiary levels advantage of uniform systems and proce- of care, this would imply minimizing routine dures but also offers equitable access to services, delivered at much higher cost than quality health care for patients across all ­ community providers. This would result in demographics and geographies. A large providers at all levels concentrating on deliv- amount of resources was devoted to develop- ering care at the best combination of cost and ing the NHIA’s claims system. Once set up, quality. the system enhanced the NHIA’s ability to monitor health care utilization and spending, Strategy 2: Reassess copayments and set reduced resource requirements for claims significantly higher deductibles and out-of- processing, and paved the way for other pro- pocket payments for basic procedures being gram and technology developments. Up-front demanded at the tertiary level investments in technology systems and infra- For services covered by the BMI system, China structure increased the efficiency and effec- may like to consider setting up reimbursement tiveness of the administrative processes and rates for specific services according to the cost enhanced the purchaser’s capacities for man- of producing and delivering those services at aging the insurance program. the agreed-upon and designated level of care. In other words, if a certain service is deemed best delivered at the district hospital level, and Core Action Area 3: Rational distribution the district hospitals have the capacity to of Services by Facility Level deliver, the case-mix-adjusted per case rates The production and delivery of health care in estimated for that level could be applied uni- China is characterized by an “outsized” ser- versally across the hospital system. However, vice distribution system, which creates high if only the highest tertiary-level hospital has costs and compromises its value base. The the capacity to deliver that service, a prospec- fee-for-service payment system, dominant in tively determined case-mix-adjusted rate is set L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 229 and paid to that hospital but under an agreed- moved them all to the University College upon ceiling determined by the global budget. facility. Later, they consolidated and moved For services not covered by health insur- all emergency vascular surgery and complex ance, payment methods need to be revised to aortic surgery to the Royal Free hospital, a have a much closer relationship with costs. different facility. The results were immediate: This is consistent with the May 2015 govern- The number of stroke cases treated at ment policy directive that requires health University College climbed from about 200 insurance to cover most medical expendi- in 2008 to more than 1,400 in 2011. tures and sets the target for out-of-pocket Mor talit y associated with strokes at payments paid by each patient at below University College fell by about 25 percent, 30 percent by 2017. and costs per patient dropped by 6 percent A big advantage of this arrangement is (Porter and Lee 2013). that health services would be delivered at the Capital investments also need to be better locations where the value is highest. The aligned with incentives to deliver fewer high- lower-cost facilities at the primary and sec- frequency services at tertiary levels and to ondary levels would take care of relatively push a larger share of services to lower levels. less complex conditions and routine services, International experience indicates that the with charges set accordingly. High-intensity shift of service delivery to primary care is services, such as cardiac and oncological often accompanied by large investments in care, would be delivered at well-resourced, infrastructure and human resources at the well-staffed facilities by subject-matter primary care level to create a “push-pull” ex p er t s , a nd ch a r ge s wou ld b e s e t effect, so that as incentives push services to accordingly. the primary care level, better capacity pulls In several examples globally, matching patient demand at the same time. Developing complexity and needed skills with institu- and strengthening the quality of human tions’ resource intensity has resulted in huge resources is key: China has already invested a value improvements. In the United States, for lot in infrastructure in recent years, and add- instance, Children’s Hospital of Philadelphia ing more infrastructure without concomitant shifted routine tympanostomies (insertion of increases in the quality of human resources tubes into children’s eardrums to reduce fluid will not be very productive. collection and risk of infection) and simple hypospadias repairs (a urological procedure) Core Action Area 4: Capacity Building from its main facility to suburban ambula- of Insurance Agencies tory surgery facilities. This move cut costs and freed up operating rooms and staff at the Strengthening the institutional and human teaching hospital for more complex proce- resources capacity of purchasing agencies is dures, resulting in estimated savings of critical to transform them from passive pay- 30–40 percent (Porter and Lee 2013). ers of health care to active and strategic pur- There are several examples as well of con- chasers of health goods and services on solidation of services in fewer locations. In behalf of the covered population. 2009, 32 London hospitals had hyperacute stroke units staffed by dedicated state-of-the- Strategy 1: Develop staff expertise and data art teams that include neurologists to take analytics capability care of stroke patients. Having so many facil- The experience of purchasing agencies in ities offering the same service did not allow China and globally demonstrates four of the any one or two to amass a high volume and most effective steps to build capacity: enjoy scale benefits. UCL Partners, a delivery system comprising six well-known teaching • Develop staff expertise. Ensuring that the hospitals that serve North Central London, staff recruited to the purchasing agencies decided to consolidate the stroke units and are well versed and trained in the specific 230 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A duties of their positions and also in the automated claims system could be theoretical and practical underpinnings of exploited better to develop comprehensive health financing and purchasing policy is automated provider monitoring systems key. A standardized set of training mod- that generate routine reports on volume, ules could be developed and adapted for service intensity, referrals, adherence to ongoing training at the local level. clinical guidelines, and financial protec- • Establish data analytics capability. tion, and the information could be used to Strategic investments in IT and data ana- inform adjustments in purchasing and pro- lytics capacity to manage information and vider payment strategies. data in-house would help inform decision making. Indeed, investment in IT infra- Strategy 2: Empower purchasing agencies structure, software development, and pro- and hold them accountable for results gramming, as well as staff training in data To improve the value of the Chinese govern- analytics and research, is one of the most ment’s enormous investments in the health critical aspects of purchasing capacity sector, the power of health purchasing needs development. to be more effectively leveraged. The pur- • Learn from domestic and international chasing agencies need to have to have the experiences. Some of the most effective mandate, capacity, and market power to purchasing agencies globally, such as those exert real influence over provider behavior in Korea and Thailand, have developed or and the volume and quality of services deliv- are closely affiliated with highly sophisti- ered. The overarching mandate and account- cated research institutes dedicated to ability of China’s BMI schemes need to shift studying the operation, impact, and toward strategic purchasing and getting value improvement of their health insurance sys- for money. Although previous reform guide- tems. Exposure to Chinese and interna- lines have acknowledged the need to better tional learning and exchange programs is leverage strategic health purchasing and stim- necessary to stay abreast of best practices. ulate changes in provider behavior, purchas- Many of China’s own ad hoc learning and ing agencies should be held accountable for exchange programs could be built upon results that go beyond their fiduciary respon- and made available to a wider range of sibilities to ensure the safety of insurance purchasing agencies more systematically. funds and maintain or increase enrollment In addition, partnerships with local and reimbursement rates, to improving the research institutions and think tanks are quality of care and patient health outcomes. a common source of capacity building In this context it is important that the and should continue to be promoted and purchasers of health services differentiate ­ possibly funded through grants or other ­ between performance and quality. There are financial resources. many examples of pay-for-performance in • Develop capability to assess the quality of China, but because in most cases “perfor- care. Finally, the ability and expertise to mance” still equates to “activities” and is not assess quality is fundamental to the design related to enhancing quality of care or to of any performance-oriented payment improving patient health outcomes, the cur- mechanism. Providers need to be held rent pay-for-performance mechanisms have accountable for delivering high-quality put in place strong incentives to increase services efficiently and without shifting volume. costs to patients. Monitoring efforts by purchasing agencies have focused on frag- Strategy 3: Strengthen cost mented aspects of provider performance, accounting systems including claims expenditures against Significant long-term returns can be expected caps, drug expenditures, and some limited from investing in cost accounting systems. clinical process indicators. The highly Accurate cost information is vital to set L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 231 realistic payment rates, which is critical to Department of China Planning Committee minimize provider cherry picking. Indeed, it (2001, No. 1560). would allow the purchasers of health services  6. “Opinion on Enforcement of Payment to set prices below or above average costs, Management for UEBMI,” MoLSS (1999, No. 23). thus creating incentives to improve the effi-   7. “Notice on Pilots of Cased-Based Service ciency of the service mix and encourage Pricing Management,” Ministry of Health delivery of services in the most appropriate (Peking Union Medical College University care setting. Cost information also helps pro- Press, 2004). viders plan budgets, benchmark within and   8. “Opinion on Deepening China’s Health Care across health care institutions, and monitor System Reform,” State Council (Zhong fa the delivery of services. 2009, No. 6). Almost all OECD countries have cost   9. “Opinion on Promoting Payment Reform of accounting systems that generate cost infor- Medical Insurances, MoHRSS (2011, No. 63); mation used for setting payment rates. “Opinion on Setting Up Pooling Fund for Typically, cost data are collected and pooled Outpatient Services for URBMI,” MoHRSS (2011, No. 59). from a selected number of providers that 10. “Opinion on Carrying Out Global Control on use comparable cost accounting systems Basic Medical Insurance Payment,” MoHRSS that meet predefined quality standards. (2012, No. 70). In addition, these countries make cost 11. “Guidance on Promoting Reform of the accounting mandatory or use cost account- NCMS Payment System,” State Council (Wei ing guidelines to encourage providers to nong wei fa 2012, No. 28). account for their costs and use them for 12. “Opinion on Promoting Payment Reform management. of Medical Insurances,” MoHRSS (2011, No. 63). Notes References 1. “Opinions on Further Improving the Reform of Health Insurance Payments” (Renshenbu fa Babiarz, Kimberly S., Grant Miller, Hongmei Yi, 2011, No. 63) implements payment reforms Linxiu Zhang, and Scott Rozelle. 2012. i n t he U E BM I s cheme. “O pi n ion on “China’s New Cooperative Medical Scheme Implementing the Control of Total Medical Improved Finances of Township Health Insurance Payment” (Renshebu fa 2012, No. Centers but Not the Number of Patients 70) suggests implementation of payment Served.” Health Affairs 31 (5): 1065–74. reforms across all insurance schemes to con- doi:10.1377/hlthaff.2010.1311. trol the increase of medical expenditures. Blumenthal, David, and William Hsiao. 2005. 2. Coverage statistics from China Health and “Privatization and Its Discontents—The Family Planning Statistical Yearbook 2016 Evolving Chinese Health Care System.” New (Beijing: Peking Union Medical College England Journal of Medicine 353 (11): Press). 1165–70. 3. “Notification of the Opinion to Comprehen- Chen, Zhu. 2009. “Launch of the Health-Care sively Scale Up Urban and Rural Residents’ Reform Plan in China.” The Lancet 373 Catastrophic Medical Insurance,” General (9672): 1322–24. Office of the State Council (Guo ban fa 2015, Chernew, Michael E., Robert E. Mechanic, Bruce No. 57). E. Landon, and Dana Gelb Safran. 2011. 4. “Opinion on Establishing an Emergency “Private-Payer Innovation in Massachusetts: Medical Assistance System,” General Office of The ‘Alternative Quality Contract.’” Health the State Council (Guo ban fa 2013, No. 15). Affairs 30 (1): 51–61. 5. “Opinion on Inclusion of College Students in Feng, Jin, Pingyi Lou, and Yangyang Yu. 2015. U R BM I ,” State Council Office (20 08, “Health Care Expenditure over Life Cycle in No. 119); “Guidelines on Cost Allocation the People’s Republic of China.” Asian Methods in Health Services Pricing,” Pricing Development Review 32 (1): 167–95. 232 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Feng, Lin, and Hairong Wang. 2014. “Exploration China: Does More Coverage Mean More of Chronic Disease Management and Payment Service and Better Health?” Health Economics Reform of Medical Insurances in Zhenjiang.” 18 (S2): S25–S46. doi:10.1002/hec.1501. Chinese Health Resources 3: 211–12. Leung, Alexander T. 2015. “Health Services Frakt, Austin B. 2016. “A Typology of Payment Purchasing: A Review of Taiwan’s Development Methods, Part 1.” AcademyHealth (blog), and Experience.” Case study commissioned by Ja nu a r y 19. ht t p: //ac ademyhe a lt h .or g​ the World Bank, Washington, DC. /node/1836. Liang, Hong, Luan Wang, and Li-mei Jing. 2013. Fuenzalida, H., S. O’Dougherty, T. Evetovits, C. “T he Process and Enlightenment from Cashin, G. Kacevicius, and M. McEuen. 2010. Payment Reform of the Basic Medical “Purchasing of Health Care Services.” In Insurance for Urban Workers in Shanghai Implementing Health Financing Reform: City.” Chinese Health Resources 4: 265–67. Lessons from Countries in Transition, edited Liu, Hong, and Zhong Zhao. 2014. “Does Health by J. Kutzin, C. Cashin, and M. Jakab, 155–86. Insurance Matter? Evidence from China’s Copenhagen: World Health Organization. Urban Resident Basic Medical Insurance.” Gao, Chen, Fei Xu, and Gordon G. Liu. 2014. Journal of Comparative Economic s 42 (4): “Payment Reform and Changes in Health Care 1007–20. doi:10.1016/j.jce.2014.02.003. in China.” Social Science & Medicine 111: Liu, X., and Anne Mills. 1999. “Evaluating 10–16. Payment Mechanisms: How Can We Measure Hodgson, Geoffrey M. 2008. “An Institutional Unnecessary Care?” Health Policy and and Evolutionary Perspective on Health Planning 14 (4): 409–13. Economics.” Cambridge Journal of Economics Liu, X., Zhan Chang-chun, Zhou Lv-ling, 32 (2): 235–56. doi:10.1093/cje/bem033. and others. 2012. “Empirical Research on Hong, Yao. 2011. “Discussion on Case-Based the Effect of Payment System on Medical Payment Practice in Shanghai.” Health Insurance Expense Control Sampled with Economic Research 4: 30–31. Zhenjiang City.” Chinese Health Service Hsaio, W. C. 1995. “Abnormal Economics in the Management (12): 909–12. Health Sector.” Health Policy 32 (1–3): 125–39. Liu, Y. 2004. “Development of the Rural Health Hu, Shanlian. 2013. “Piloting the Global Insurance System in China.” Health Policy and Control.” China Social Security 2: 73. Planning 19 (3): 159–65. Jian, Weiyan, Ming Lu, Kit Yee Chan, Adrienne Ma, Jin, Mingshan Lu, and Hude Quan. 2008. N. Poon, Wei Han, Mu Hu, and Winnie Yip. “From a National, Centrally Planned Health 2015. “Payment Reform Pilot in Beijing System to a System Based on the Market: Hospitals Reduced Expenditures and Out-Of- Lessons from China.” Health Affairs 27 (4): Pocket Payments per Admission.” Health 937–48. Affairs 34 (10): 1745–52. Manuel, Ryan. 2010. “China’s Health System and Jing, F. 2004. “Health Sector Reform and the Next 20 Years of Reform.” In China: The Reproductive Health Services in Poor Rural N e x t Tw e n t y Ye a r s o f R e f o r m a n d China.” Health Policy and Planning 19 Development, edited by Ross Garnaut, Jane (Suppl 1): i40–i49. Golley, and Ligang Song, 363–91. Canberra: Ji-yuan, Jiang, Sun De-yao, Qi Qi, and Sun Yi. ANU E Press. 2011. “Research on Payment System Based on Meng, Qun, Ling Xu, Yaoguang Zhang, Juncheng Diagnosis Related Groups in Beijing.” Chinese Qian, Min Cai, Ying Xin, Jun Gao, Ke Xu, J. Journal of Hospital Administration 27 (11): Ties Boerma, and Sarah L. Barber. 2012. 587–89. “Trends in Access to Health Services and Kwon, Soonman. 2015. “Institutional Capabilities Financial Protection in China between 2003 for Health Service Purchasing—Case Study of and 2011: A Cross-Sectional Study.” The Korea’s National Health Insurance Service Lancet 379 (9818): 805–14. doi: http://dx.doi​ (NHIS) and Health Insurance Review and .org/10.1016/S0140-6736(12)60278-5. Assessment Service (HIRA).” Case study com- Murray, Robert. 2015. “The CareFirst Blue Cross– missioned by the World Bank, Washington, DC. Blue Shield Patient-Centered Medical Home Lei, Xiaoyan, and Wanchuan Lin. 2009. “The Model, Maryland, U.S.A.” Case study commis- New Cooperative Medical Scheme in Rural sioned by the World Bank, Washington, DC. L e v er 5 : R ealigning I ncenti v e s in P u rcha s ing and P ro v ider P ayment 233 Porter, Michael, and Thomas Lee. 2013. “The Practices in Rural China.” Health Affairs Strategy that Will Fix Health Care.” Harvard 33 (3): 502–10. doi:10.1377/hlthaff.2013.0704. Business Review, October 13. Yip, Winnie, Chi-Man Yip, William C. Hsiao, Shortell, Sam M. 2013. “Bridging the Divide Wen Chen, Shanlian Hu, Jin Ma, and Alan between Health and Health Care.” JAMA 309 Maynard. 2012. “Early Appraisal of China’s (11): 1121–22. Huge and Complex Health-Care Reforms.” State Council. 2011. “The 12th Five-Year Plan for The Lancet 379: 10. doi:10.1016/S0140​ - 6736​ Economic and Social Development of the (11)61880-1. People’s Republic of China (2011–2015).” Yu, Baorong, Qingyue Meng, Charles Collins, Translated by the Compilation and Translation Rachel Tolhurst, Shenglan Tang, Fei Yan, Bureau, Central Committee of the Communist Lennart Bogg, and Xiaoyun Liu. 2010. “How Party of China. Beijing: Central Compilation Does the New Cooperative Medical Scheme & Translation Press. Influence Health Service Utilization? A Study Wang, Kun, Wu Hua-zhang, Du Yu-kai, and in Two Provinces in Rural China.” BMC o thers. 2013. “Analysis on the Modes of ­ Health Services Research 10: 9. Medical Insurance Payment of Mental Health Zhang, J. 2010. “The Impact of a Diagnosis- Institutions.” Chinese Journal of Social Related Group-Based Prospective Payment Medicine (1): 57–59. Experiment: The Experience of Shanghai.” Wang, Xiang. 2011. “The Strategy Choices of Applied Economics Letters 17 (18): 1797– Payment Reform: Based on 17 Years Reform in 1803. doi:10.1080/13504850903317347. Zhenjiang.” China Health Insurance 7: 35–36. Zhang, Linxiu, Hongmei Yi, and Scott Rozelle. WHO (World Health Organization). 2000. The 2010. “Good and Bad News from China’s World Health Report 2000: Health Systems— New Cooperative Medical Scheme.” IDS Improving Performance. Geneva: WHO. Bulletin 41 (4): 10. Ying, Yachen. 2014. “Achieve Win-Win-Win Z ha ng , Yi ng , a nd Wu Yong-l i ng. 2013. Situation by Jointly Reforming the Drug “Changshu URBMI (NCMS) Payment Reform Sector, Health Service Sector, and Health Practice and Exploration.” Business 22: Insurance Sector: Field Study Report on Public 238–39. Hospital Reform in Sanming.” Case study, Zhang, Zai-sheng, and A i-hao Xu. 2014. He a lt h D evelopment Re sea rch C enter “Evaluation of Capitation Reform of Payment (NHFOC), Beijing. System on Diabetes Mellitus.” Chinese Rural Yip, W., and W. C. Hsiao. 2008. “The Chinese Health Service Administration 9: 1058–60. Health System at a Crossroads.” Health Affairs Zhen, Jie. 2009. “Multiple Payment Practice of 27 (2): 460–68. Medical Insurance in Beijing.” China Health Yip, Winnie, Timothy Powell-Jackson, Wen Chen, Insurance 5: 34–35. Min Hu, Eduardo Fe, Mu Hu, Weiyan Jian, Zhou, R. n.d. “Responsible Utilization of Ming Lu, Wei Han, and William C. Hsiao. Antibiotics Should Be a Societal Responsibility.” 2014. “Capitation Combined with Pay-For- Sanmenxia Central Hospital Paper No. 223 Performance Improves Antibiotic Prescribing (in Chinese), Sanmenxia, China. 7 Lever 6: Strengthening the Health Workforce Introduction Third, the PCIC model emphasizes patient engagement and self-management. In this Transforming China’s health service delivery more collaborative approach, providers and system to provide people-centered integrated patients work together to identify problems, care (PCIC) has several implications for set priorities, establish goals, create treatment China’s health workforce. plans, and solve issues. This implies a new First, in the PCIC model, primary health way for providers and patients to interact care (PHC) plays a central role in providing a that requires clinicians to have better com- continuum of services for patients who have munication and interpersonal skills when noncommunicable diseases. The current PHC dealing with patients. workforce in China would need to be Adopting a PCIC approach will also have strengthened and expanded to enable PHC profound implications for the training, facilities to act as gatekeepers and to coordi- deployment, composition, and manage- nate health care within the entire health ment of the Chinese health workforce. system. International experience suggests that Second, the PCIC model changes the way ­ t ransforming the health workforce to meet health care is organized and managed. It the challenges of PCIC would include rede- relies on general practitioner (GP)-based fining the scope of practice and functions of multidisciplinary teams; strong links to different categories of health workers; com- community-based and social care; effective posing new teams representing a variety of coordination across providers at different lev- specializations; developing new sets of skills els of the health care system; and a strong and competencies; achieving a well-balanced focus on prevention, risk stratification, and distribution of the workforce across different health maintenance. China’s significant levels of care; improving performance man- imbalance in its workforce composition and agement systems; instituting appropriate considerable shortage of these health profes- incentive structures; and transforming pre- sionals pose enormous challenges that will service and in-service training. Applying require medium- and long-term strategies to these core principles to guide health work- overcome. force reform in China implies changing the 235 236 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A institutional, financial, and management sys- China started an ambitious reform of the tems governing the health workforce. world’s largest educational system for health This chapter examines the main chal- care professionals. Universities of western lenges related to the health workforce in medicine that were previously governed by China in the context of PCIC and recom- the Ministry of Health were merged into uni- mends policies to successfully address these versities supervised by the Ministry of challenges. Education. China has also gradually estab- The rest of the chapter is organized as lished a comprehensive medical education follows: system of undergraduate and postgraduate education and continuing professional devel- • “The Chinese Health Labor Market” opment: students enroll directly from high assesses the current health labor market school into clinical medicine courses of either challenges in the context of China’s institu- three years (for a diploma), five or six years tional and policy environment and identi- (for a bachelor’s degree), seven years (for a fies sources and causes of disequilibrium. master’s degree), or eight years (for a medical • “Recommendations for Human Resources doctorate). Reform” presents four core action areas This system produced more than twice as and associated strategies for implementa- many graduates overall in 2013 than in 2003 tion of the recommended reforms. and more than four times as many graduates • “An Implementation Road Map” con- of higher medical education (Liu 2015). cludes by setting forth short- and medium- About half a million nurses graduated in t e r m s t e p s for i mpl e m e nt i n g t h e 2013, the large majority of them with voca- recommended reforms as well as comple- tional degrees (table 7.1). mentary measures that may be needed to The rapid expansion of health professional support these actions and their desired education has raised important concerns. outcomes. These include lack of coordination between medical education and labor market demands; rapid expansion in the numbers of The Chinese Health Labor students without corresponding efforts to Market: Trends and Challenges strengthen the quality of training; and curri- The Health Professional Education cula that focus narrowly on biomedicine, System medical technology, and hospital care with little exposure to community and primary The health labor market in China has experi- care or rural practice. There is a need for bet- enced huge changes in recent years. In 1998, ter cooperation between the Ministry of TABLE 7.1  Enrollment of medical students and medical graduates in China, 2013 Students enrolled Graduates Higher Vocational Higher Vocational Student category education education Total education education Total Clinical medicine (including TCM) 124,279 108,947 233,226 107,904 108,382 216,286 Public health 9,593 — 9,593 6,793 — 6,793 Nurses and midwifery 42,646 521,159 563,805 34,145 534,502 568,647 Pharmacists (including TCM) 38,904 27,216 66,120 31,404 24,249 55,653 Medical technicians 21,167 52,353 73,520 11,091 40,568 51,659 Source: Liu 2015. Note: TCM = traditional Chinese medicine; — = not available. L e v er 6 : S trengthening the   H ealth   W or k force 237 Education, which is now in charge of health quality of clinical care. In December 2013, professional education, and the National the NHFPC and six other ministries issued a Health and Family Planning Commission joint policy directive, “Guidance . . . on the (NHFPC). A study found that medical uni- Establishment of a Standardized Resident versities decide on their enrollee numbers not Training System,” which requires medical on the basis of need or demand for health graduates to complete a three-year standard- professionals but mostly based on the num- ized residency in an accredited institution ber of teachers or even dormitory rooms (Liu (after five years of medical university study).1 2015). Faculty numbers have not kept pace Funding for training institutes and subsidies with the expanding numbers of medical stu- for resident trainees are provided by the cen- dents (figure 7.1) tral and local governments. The massive increase in admissions has considerably increased student-teacher ratios Imbalances in Workforce Composition and shortages of clinical internship positions (Daermmich 2013), which has likely affected The government’s efforts to increase the the quality of medical education. Medical training of health professionals have pro- training focuses on clinical biomedicine and duced a steady expansion of the health hospital practice, with little exposure to com- workforce and significantly changed its munity and primary care or rural practice composition. The compensation structure (Hou and others 2014). has also changed, notably as a result of the Until recently, China did not have a stan- new zero-markup policies on drug prescrip- dardized resident training system at the tion. At the same time, the regulatory frame- national level; medical graduates usually work has been modified to enhance the went directly to work in a hospital or PHC mobility of health workers across regions facility after graduation. Internationally, the and providers. These changes reflect institu- norm is that medical graduates undergo tional transformations in the Chinese labor training in clinical practice (usually for three and health care markets that combine fea- years) as a resident doctor in a hospital before tures of a dynamic and fast-growing market they can practice medicine independently. economy with rigid state control and This is considered important to ensure the intervention. FIGURE 7.1  Numbers of medical students and faculty and faculty-student ratios in China, 1998–2012 2,500 0.14 0.12 2,000 Faculty or students, thousands 0.10 Faculty -student ratio 1,500 0.08 0.06 1,000 0.04 500 0.02 0 0 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 Faculty-student ratio Student Faculty Source: Hou and others 2014, 823. 238 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A However, health workforce challenges (and assistant physicians) and 2 registered continue to pose a major challenge in China’s nurses (NHFPC 2014), which is below the bid to strengthen its public and primary Organisation for Economic Co-operation health care services (Yip and others 2010). and Development (OECD) country average Specialists outnumber GPs, and there are few of 3.2 doctors and 9.6 nurses per 1,000 pop- doctors at the PHC level. Compensation lev- ulation (figure 7.2). els are unattractive, and the underlying incen- Second, among all health workers, there is tives in physician contracts with hospitals are a huge shortage of nursing staff across the perverse. The governance structure of the country. Despite the doubling in total num- health workforce is characterized by the bers and a huge increase in the recruitment of headcount quota system, and physician nursing students, the ratio of nurses to doc- licensing is linked with facilities, introducing tors in China is only 1 to 1, significantly rigidities and limiting mobility. Managerial lower than the average of 2.8 nurses per doc- autonomy in hiring health workers at the tor in OECD countries (Qin, Li, and Hsieh facility level is low, resulting in a mismatch 2013). between staffing needs and available skills. What is encouraging, however, is the This chapter will discuss these issues in increase in recruitment of nursing students: detail. in 2013 alone, nursing schools recruited 0.56 Despite the remarkable increase in the million students, 62 percent of them in three- total supply of health workers, many chal- year vocational nursing schools. As a result, lenges remain in the composition of the the number of nurses per 1,000 population health workforce. increased from 0.98 in 2003 to 2.04 in 2013 First, even though the total number of (figure 7.3). health workers in China increased from 6.2 Third, China has a critical shortage of sev- million in 2011 to 7.21 million in 2013, the eral key specialties: pediatricians, psychia- number of health care professionals per trists, and GPs constitute only 3.9 percent, 1,000 population in China at the end of 2013 0.9 percent, and 5.2 percent, respectively, of was only 5.3, including 2 licensed physicians all licensed physicians. The relatively lower FIGURE 7.2  Ratio of nurses to physicians in OECD countries, 2012 (or nearest year) 5 4.4 4.3 4.3 4.3 4.3 4.0 3.9 4 3.8 3.2 3.2 Nurse-physician ratio 3.0 3.0 3 2.8 2.7 2.7 2.5 2.3 2.2 2.2 2.0 2.0 1.9 1.9 1.9 2 1.7 1.7 1.6 1.6 1.6 1.5 1.5 1.4 1.4 1.1 1.0 1 0 ia a er ) Fin nd Lu Ice nd m nd Ire rg Ne Nor d er y Be nds ite Slo um ng a Ge dom Sw any M Fr en te ce Po ro d ec om 8 ep ia Hu blic Se ry M a Cr a va Est tia ep ia hu ic Cy ia Au us via ia, y Po FYR Sp l Bu ain Tu ria ey ga itz 09 th wa on tal Ki ni i alt Cz R EU2 lan lan Lit ubl h R an rb k R on an a g u on an pr rk ed a oa la la xe la str at d ve rtu ng Sw (20 i ne bo u la rm ed I lg lg L k ar nm ac Slo De Un M Sources: Eurostat, OECD Health Statistics 2014, and WHO European Region “Health for All” databases. Note: OECD = Organisation for Economic Co-operation and Development; EU28 = 28 member states of the European Union; WHO = World Health Organization. a. Austria reports only nurses employed in hospitals. L e v er 6 : S trengthening the   H ealth   W or k force 239 FIGURE 7.3  Number of health professionals per 1,000 population in China, 2003–13 2.2 2.0 1.8 Number per 1,000 population 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Licensed physicians Registered nurses Lab technicians Licensed assistant physicians Pharmacists Village doctors Source: Liu 2015. income and higher occupational risk of these 2.5 in 2013 (figure 7.4; also see annex 7A, posts make them less attractive than other table 7A.2). In 2013, urban areas had 2.3 specialties. The 5.2 percent share of GPs is times as many physicians per 1,000 popula- well below international norms of 30–60 tion than rural areas and 1.8 times as many percent. nurses. Likewise, the shortage of nurses is A 2011 State Council policy proposes to especially severe in rural areas and at the establish a GP-based primary health care PHC level, where the nurse-physician ratios delivery system with two to three GPs per are 0.82 to 1 and 0.55 to 1, respectively. 10,000 population (international practice is The distribution of health workers across one GP per 2,000 population).2 To meet this the public and private sectors has also target, China needs at least 300,000 to changed in recent years (Liu 2015). The share 400,000 GPs, or more than double the cur- of health professionals practicing in the pri- rent GP workforce (172 ,597 in 2014) vate sector increased from 14 percent in 2010 (NHFPC 2015). In China, as in many other to 15.8 percent in 2013 (figure 7.5). The cat- countries, being a GP is less attractive than egory with the largest proportion in the pri- being a specialist. GPs’ lower social status vate sector is assistant licensed physicians and income and limited career development (20.1 percent); technicians have the smallest path affects the recruitment of new medical (12.3 percent). graduates. By 2016, the government had These imbalances in the distribution of the already enrolled 70,000 medical graduates Chinese health workforce result in critical into the standardized three-year residency shortages at the PHC level and in rural areas. training programs, with 8,637 slots for GP The number of health professionals working training. in PHC settings has been increasing, but Fourth, despite an increase in recent years overall the PHC workforce is still limited. In in the number of health workers practicing in 2013, only 30 percent of all health profes- rural areas, health professionals are still sionals and 21 percent of registered nurses heavily located in cities. The urban-rural worked in PHC settings (including township ratio of health workers per 1,000 population health centers [THCs] in rural areas and increased from about 2.0 in 2005 to about community health centers [CHCs] or stations 240 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 7.4  Health professionals per 1,000 population in China, by rural or urban location, 2003–13 10 9 Number per 1,000 population 8 7 6 5 4 3 2 1 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Urban Rural Source: Liu 2015. FIGURE 7.5  Proportion of health professionals in the private sector, 2005–13 25 Share of workers in private sector, percent 20 15 10 5 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 Licensed physicians Registered nurses Technicians Licensed assistant physicians Pharmacists Total health professionals Source: Liu 2015. in urban areas). Although the overall health education than those in higher-level facilities workforce expanded, the share of PHC work- and in towns and cities. In 2012, 20 percent ers declined, from more than 40 percent of the health professionals in urban areas and in 2009 to less than 30 percent in 2013. only 6 percent in rural areas had a bachelor’s A recent survey revealed that nearly 50 per- degree or above. 3 Most PHC workers in cent of the assistant physicians and registered CHCs and THCs have received only post- nurses who left THCs went to work in hospi- high-school training and secondary school tals at the county level or above. training, respectively. PHC facilities and poor In addition, health professionals at the rural areas continue to have difficulty PHC level and in rural areas have less r ecruiting and retaining qualified health ­ L e v er 6 : S trengthening the   H ealth   W or k force 241 professionals. The lack of qualified health significant variation, with PHC workers professionals at the PHC level, especially in earning the least. Average earnings in China’s the rural areas, is the major reason why regulated health sector rank ninth among all patients bypass PHC and seek care directly at sectors, only 13 percent above the economy- hospitals. wide average (figure 7.6), and annual rates of increase have been low relative to those in other sectors. Unattractive Compensation Levels and The compensation structure is also an Perverse Financial Incentives issue: on average in 2012, the basic salary in One possible explanation for the persistent the health sector accounted for 23 percent shortcomings in China’s primary health of total compensation, allowances for workforce is unattractive compensation. 20 ­p ercent, and performance bonuses for Earnings in the health sector—which typi- 57 percent.4 Health care workers—especially cally include a basic salary, a performance doctors—are encouraged to seek additional bonus, and a hardship allowance—show income from bonuses based on the overall FIGURE 7.6  Annual average wages of urban employees in China, by sector, 2005 and 2012 Agriculture, forestry, animal, and shery Hotels and catering services Management of water, environment, and public facilities Service to households, repair, and other services Construction Manufacturing Primary education Public management, social security, and organization Wholesale and retail trades Real estate Average Secondary education Education Health and social service Health 53,067 29,339 Leasing and business services Transport, storage, and post Culture, sports, and entertainment Mining Production and supply of electricity, heat, gas, and water Higher education Scienti c research and technical service Information transmission, software, and IT Financial intermediation 0 15,000 30,000 45,000 60,000 75,000 90,000 RMB 2012 2005 Source: NHDRC 2014a. 242 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A revenue that hospitals receive from the ser- with the government’s reform strategy of vices they provide (such as admissions and strengthening PHC. Low compensation, medical procedures); commissions for pre- combined with other factors, makes it diffi- scribing drugs and ordering tests; informal cult to recruit and retain good, well-qualified payments from patients (hongbao, or “red health workers to work in PHC settings. envelopes”); and private practice (“moon- Additionally, medical schools reportedly lighting”) (Woodhead 2014; Yip and others face difficulties in attracting students who 2010). In responding to these perverse incen- achieve high scores on the national university tives, physicians generate demand for their entrance exam (gao kao). Instead, they often services and overprescribe diagnostic tests attract those whose first career choice was and expensive branded drugs. not medicine (Hou and others 2014; Compensation levels also differ widely Waldmeir 2013). A 2013 survey of students across levels of care. Earnings in public with the highest national exam scores from hospitals—especially urban public hospitals— ­ 1977 through 2012 revealed that less than are higher than in PHC facilities and in rural 2 percent chose to major in health or medi- areas (figure 7.7). The average compensation cine; in contrast, 38 percent chose economics in urban public hospitals is 1.6 times the sec- or business management.5 tor average. Staff working in PHC institutions Moreover, university recruitment of medi- and THCs earn 76 percent and 72 percent, cal students dropped significantly in 2014. respectively, of the health sector average. The For example, Southern Medical University differentials across cadres of health workers planned to enroll 200 students but had only are smaller than what their different education 50 applicants. Guangdong Traditional requirements would suggest. For example, Chinese Medical University planned to enroll doctors earn on average 1.1 to 1.96 times 1,807 but had only 485 applicants. The more than nurses. Chinese Medical Doctor Association’s regu- Large earnings differences between hospi- lar survey of physicians includes the question, tals and PHC settings are incompatible “Would you like your children to go to FIGURE 7.7  Health worker compensation across levels of care, China 2013 120,000 100,000 80,000 RMB 60,000 40,000 20,000 0 All public Urban public County public All PHC Urban CHCs THCs Public health hospitals hospitals hospitals facilities institutions Facility type, level, and location Average compensation for each setting Sector-wide average compensation Source: NHDRC 2014a. Note: CHC = community health center; PHC = primary health care; THC = township health center. “PHC facilities” refers to urban CHCs and rural THCs. “Public health institutions” refers to centers for disease control and hygiene inspection stations. L e v er 6 : S trengthening the   H ealth   W or k force 243 TABLE 7.2  Share of Chinese doctors who hope more than 50 0 beds (see an nex 7A , their children will attend medical school, 2002–11 table 7A.5).8 For THCs, staff quotas are, in Percentage of respondents principle, 1 percent of the total population in Survey year Yes No the THC catchment area; the actual number is calculated considering the ease of transport 2002 10.9 54.0 to the THC and the fiscal capacity of the 2004 10.4 63.0 local government. The current quota in a sur- 2009 11.9 62.5 vey of 12 counties in six provinces covering 2011 6.8 78.0 the east, central, and west regions is 1.2 per Source: National surveys of the Chinese Medical Doctor 1,000 population. The quota standard for Association, http://www.cmdae.org. CHCs is two to three GPs and one public Note: Survey question: “Would you like your children to go to medical school”? health specialist per 10,000 residents, as well as a GP-nurse ratio requirement of 1 to 1. medical school?” The percentage of negative Many issues related to quotas have been responses has been above 50 percent and reported from across the country. The quota increasing (table 7.2). system has not kept pace with the needs to The public perception of medical practice improve health service provision and to con- has been deteriorating, and the recent surge stantly adapt to growing demand and chang- of violence against health professionals (expe- ing technology. For example, headcount stan- rienced by a third of doctors, as further dis- dards for various hospitals, including cussed in chapter 5) is also making medical maternity and child care hospitals, were for- professions progressively less attractive.6 mulated in the 1980s and never updated. Headcount standards no longer meet the changing needs of health facilities. Restrictive Headcount Quota System The result is significant quota shortages in The policy framework for managing the almost all health facilities regardless of level health workforce follows the governance or location. Facilities have to hire additional structure in place for all public service units health workers who are not on quota. (PSUs), which is centered on the headcount A recent survey of health facilities in 10 prov- quota system.7 The headcount quota system inces by the Health Human Resources defines the total number of personnel Development Center (under the Ministry of assigned to a PSU; it is a special human Health) and Shandong University found that resources management arrangement for civil 15 percent of the employees in CHCs, 11 per- servants and public institutions. The quota cent in maternity and child care institutions, framework is formulated by the government’s and 8 percent in THCs are not quota-based. Post Establishment Office (PEO). PSUs must For example, PHC facilities in Yunnan have a quota approved by the PEO, which is province had to hire so many temporary ­ also the basis on which the Bureau of Human nonquota) health workers that, in 2013, (­ Resources and Social Security (and not the 31 percent of all their health workers (13,502 health facility manager) establishes posts. of 43,595) were not on quota. Quota-based employees have permanent jobs Hospitals must pay their salaries and ben- with pensions and other social security bene- efits out of their own revenues. This adds fits, and it is difficult for health facility man- pressure on hospitals to generate extra reve- agers to fire them. nue and creates a “second class” of employees Staffing quotas for public hospitals are with different contracts and compensation based on the number of beds. The national systems but often undertaking the same roles guideline sets the bed-staff quota at 1 to and tasks. It is broadly reported that non- 1.3–1.4 for hospitals with fewer than 300 ­ quota staff are paid less and get fewer bene- beds; 1 to 1.4–1.5 for hospitals with 3 ­ 00–500 fits; for example, they do not receive a beds; and 1 to 1.6–1.7 for hospitals with pension. 244 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Limited Mobility and Market Entry To address this constraint, the government launched the pilot Physician Dual Practice Two processes—qualification certification and Policy in 2009 (further discussed in box 7.1). practice licensing—govern entry into health This new policy allows a physician to register service provision by physicians and by assis- to practice in up to three hospitals or clinics tant physicians, nurses, pharmacists, and rural in the same city if the physician (a) obtains village doctors in China.9 Health professionals the agreement of the first facility; (b) obtains who meet the eligibility requirements can take approval from the local health administra- the annual national qualification examina- tion authority; and (c) signs a legal agreement tion, and those who pass the examination with all health facilities where he or she obtain a qualification certificate as a physi- works regarding malpractice disputes and cian, nurse, or pharmacist. (There is no litigation.10 national exam for village doctors.) Those who Although the government expanded the obtain the certificate can then apply for a pilot in 2011, the policy has not achieved its practice license from the local health author- intended results. By June 2011, only 166 doc- ity, which grants permission to practice. tors had registered for multiple practices. In The recertification process has some limita- Guangdong province (the initial pilot site), tions. Current regulations specify that nurses about 100 physicians had registered. Three need to update their licenses every two years reasons explain the low participation rate: and that licensed doctors need to attend in- service training every two years, though doc- • Public hospitals are overloaded, so their tors face no specific requirement regarding physicians do not have time to work else- recertification. Several quality assurance ini- where as well. tiatives are in place. For example, the Chinese • The policy requires physicians themselves Medical Doctor Association has created a pro- to bear the risk in case of a medical acci- gram of reevaluation examinations for physi- dent or dispute. cian qualification and requires physicians to • The current quota-based human resource take it every two years. The NHFPC’s Science management system links physicians’ pen- and Education Department, in charge of in- sions, employment benefits, professional service training, has issued a policy requiring title, and access to research and training physicians to take training that earns them a opportunities directly to the hospital certain number of points. However, these where they work, and physicians fear that quality assurance measures are not linked dual practice might affect their perfor- with the relicensing process, and therefore mance evaluation and opportunities for health professionals do not always comply, promotion and clinical research and train- limiting the impact of the measures. ing at their primary hospital of affiliation. Current labor regulations compound the limiting effects of the quota system on the Lack of Managerial and Decision- mobility of health workers. Under Chinese Making Autonomy in Hiring Health law, a medical practitioner’s license to practice Workers must specify the medical facility where the professional will work; the category (such as Since 2006, all technical professionals, mana- clinical medicine, traditional Chinese medi- gerial staff, and logistics support staff in health cine, dentistry, or public health); and the spe- facilities have been hired through an open cialty. The licensing regulation allows health recruitment system. A March 2000 policy professionals to practice only in the facility directive directed PSUs to reform their person- specified in the license. This strongly restricts nel systems, abolish tenure status for manage- the mobility of doctors, most of whom are ment and professional technical staff, and hire employees of public hospitals, and adds to the health workers to fill posts through open difficulties that PHCs and private hospitals recruitment.11 Health institutions established face in recruiting qualified health workers. an employment relationship with employees L e v er 6 : S trengthening the   H ealth   W or k force 245 BOX 7.1  History of the Physician Dual Practice Policy in China “Moonlighting” by physicians emerged significantly in In November 2014, six ministries and agencies the 1990s in China, without any regulations or policies promote jointly issued a new policy directive to further ­ to assure service quality and safety. In 1999, the Law dual practice. This has opened dual practice to clini- for Licensed Doctors of the People’s Republic of China cal, dental, and traditional Chinese medicine doc- went into effect, stipulating that a physician may work tors with a middle-level professional title (or above) at only the one site specified by his or her license to who have worked in their specialty for more than five practice medicine. This law controlled moonlighting years. In addition, physicians no longer need written by physicians but also greatly restricted their mobility approval from the first facility or approval from the and significantly affected the ability of private sector local authority. The policy also emphasizes that hos- and PHC facilities to recruit qualified doctors, most of pitals should not discriminate against dual-practice whom were employed at public hospitals. physicians in professional title, promotion, or access to Fully aware of this negative effect, the govern- clinical research opportunities. ment took steps to counter it. The first Physician Provinces have formulated local policies based on Dual Practice Policy was promulgated in Septem- national guidelines. For example, Guangdong prov- ber 2009 to start a pilot in selected areas that ince’s dual-practice policy does not limit the number of allowed physicians to register to practice in more facilities where a physician can practice, though it spec- than one hospital or clinic on the condition that ifies a certain area such as a county or city. Physicians the physician (a) obtain the agreement of the first do not need approval from their primary hospital but facility; (b) obtain approval from the local health need only file an application at the local health bureau administration authority; (c) sign a legal agree- when they apply for a license. Since the issuance of this ment regarding malpractice disputes and litiga- new, more favorable policy, some public hospital physi- tion with each health facility where the physician cians reportedly have opened independent clinics. works; and (d) have the title of associate chief phy- Nationwide, the number of physicians applying sician or higher. for dual practice has increased only slightly. The In July 2011, the Ministry of Health issued a new most important reason is that employment benefits policy that (a) expanded the pilot to more prefec- are determined by the headcount quota and are tures, and (b) changed the minimum professional attached to the health facility to which the physi- title from associate chief physician (senior level) to cian belongs. The quota system is the root cause of attending doctor (middle level).a restrictions on physicians’ mobility. a. “Notification on Expanding Pilot of Physicians’ Dual Practice,” Ministry of Health (2011, No. 95). through contracts that specify the responsibili- local quota establishment office and the ties, rights, and benefits of both parties. finance bureau have defined the staffing The major problem associated with quota of each public facility (including the recruiting health workers is health facilities’ number for each major category such as GPs lack of autonomy. Within the public service, and nurses), the facility’s human resources the government defines the number, struc- department, guided by the BoHRSS, formu- ture, and responsibility of all posts in all lates the post-setting plan and submits it to three major categories: technical, managerial, the BoHRSS for approval. and logistics. The Ministry of Human Under current practice, the local govern- Resources and Social Security (MoHRSS) ment defines quotas and, in some cases, limits and local Bureau of Human Resources and the issuance of new quotas. Thus, health facil- Social Security (BoHRSS) set and manage the ities may not be able to hire the new staff they posts in PSUs and are responsible for policy need. Additionally, “open recruitment” is development and guidance, macro control, managed by the local BoHRSS rather than by supervision, and final clearance. Once the the health facilit y. T he standardized 246 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A qualification examinations, recruitment pro- Professional Title Evaluation cedures, and evaluation standards formulated Professional titles are designed to represent by the BoHRSS are also often unsuited to the the technical capacity of professional health health sector’s special characteristics. For workers. There are three levels of profes- example, the examination questions and eval- sional title for physicians in China: at the uation standards do not reflect the functions junior level are resident doctors, at the middle and skills of PHC workers and GPs, putting level are attending doctors, and at the senior them at a disadvantage and sometimes mak- level are associate chief physicians and chief ing it impossible to hire them. Moreover, a physicians. Advancement of professional title minimum education of three years of college is is an important component of career develop- required, whereas most THC staff only have ment; it is linked to salary and often to pen- training equivalent to secondary school. And sion and other benefits. the requirement of at least three applicants for China’s evaluation system for professional each position is difficult to meet for THC titles was established in 1955. For junior-level positions. and middle-level health professionals, there is a This post management system creates national syllabus, and a national technical rigidity in the management of the health qualification examination is held at a fixed time workforce. First, the widely reported quota once a year across the country with national shortages imply shortages of official posts in evaluation criteria. Evaluation for senior-level health facilities. The shortages affect PHC titles consists of an exam and individual assess- facilities more than hospitals because PHC ment at the provincial level. The provincial facilities rely more on government funding. evaluation committee is set up by the provincial Second, it restricts the highest-level post and health and human resources bureaus. the number of posts at PHC facilities—which The professional title evaluation system constrains the career prospects of PHC health has three major issues: it overemphasizes professionals. Third, health facilities lack publication; it does not reflect differences autonomy to define what and how many across positions, specialties, and levels of posts they would need to meet the service health care providers; and professional asso- needs of their local communities. Any ciations have only limited involvement. changes to the approved post-setting plan In 2015, the MoHRSS and NHFPC jointly need approval from the BoHRSS. issued an executive order providing guidelines There have been some recent movements to for improving the professional title evaluation reform human resource management policies for the PHC workforce.13 This policy order in China. In February 2014, the State Council aims to build a professional title evaluation sys- issued a new policy directive, “Personnel tem that corresponds to the features of medical Management Regulations for Public Service professionals. Following the policies of differen- Units,” and formally launched the Post tiated functions and the tiered care system, the Management System in Public Institutions.12 requirements for dissertation writing and for- The new regulations state that public institu- eign languages were softened as reference rather tions should set posts according to duties, than compulsory conditions. The policy order tasks, and needs, following relevant national places more emphasis on clinical skills and regulations. Each post should have a specific requires each province to make context-specific title, clearly defined responsibilities, evaluation implementation plans to improve the profes- standards, and qualification criteria. One goal sional title evaluation system. of this reform is to shift from identity-based (headcount quota) to post-based human resources management. Under the new policy, Recommendations for Human the government defines the category and grade Resources Reform of posts, and then the health facility is respon- sible for recruitment, evaluation, salary set- An adequate, well-functioning health work- ting, and training for the posts. force is critical to implement and maintain L e v er 6 : S trengthening the   H ealth   W or k force 247 the PCIC model. Human resources for health in the course of implementing a PHC model are a key component of health systems and of service delivery and to identify lessons that play a central role in delivering quality care at could be applied in the Chinese context. affordable prices to the population. Issues T he i nter n at ion a l rev ie w covered related to availability, distribution, and per- Australia; Brazil; Canada; Norway; Taiwan, formance of health workers pose big chal- China; Turkey; and the United Kingdom. lenges, and the extant literature is rich in Each has made significant efforts to expand country experiences with different ways of PHC coverage and to change how their health addressing these concerns. workforces are educated, deployed, man- This section outlines a set of recommenda- aged, and regulated in support of the tions to address the current health workforce expansion. challenges in China and the changes needed The four core action areas and corre- to implement PCIC. The recommendations sponding implementation strategies include are based on a review of international experi- the following (table 7.3): ence in strengthening the health workforce to implement or expand PHC as well as of 1. Build a strong enabling environment for China’s ongoing initiatives to reform its PHC workforce development to imple- health workforce. The review sought to ment PCIC. understand how selected countries and 2. Improve workforce composition and com- regions have adapted their health workforces petency for PHC service delivery. TABLE 7.3  Four core action areas and implementation strategies to strengthen the health workforce Core action areas Implementation strategies Strong enabling 1.  Establish general practice as a specialty, with equivalent status to other medical specialties. environment for Introduce a gatekeeping mechanism to direct patients to primary care providers as first development of point of contact, and mandate this arrangement once the PCIC system is well established. PHC workforce to Introduce career development prospects to develop and incentivize PHC workforce, including implement PCIC separate career pathways for GPs, nurses, mid-level workers, and community health workers. Raise compensation of PHC workers commensurate with other prestige specialties to increase recruitment, retention, and motivation. Balanced workforce 2.  Scale up the standardized training for resident doctors and GPs. composition and Accelerate ongoing successful efforts to increase supply of GPs and nurses. competency for PHC Reform the curriculum to upgrade medical training and build new skills and competencies service delivery required for PCIC. Improve on-the-job training to strengthen competency of current workforce and build new PHC competencies. Set up alternative cadres of health workers (such as clinical assistants, assistant doctors, clinical officers, and community health workers) to strengthen PHC delivery. Compensation system 3.  Increase basic wages of health workers, linking the exact level of increase to general labor with strong incentives for market trends in China to keep the health profession attractive. good performance Increase the percentage of basic salary vis-à-vis performance bonuses in physicians’ total income package. Increase the subsidy, and introduce or increase nonfinancial incentives, for health workers in rural and remote areas. Revise the system of incentives by linking income with performance assessment built on comprehensive performance indicators rather than revenue generation. Headcount quota system 4.  Give managers autonomy in human resources issues, including post-based recruitment, reform to enable a more salaries, deployment, evaluation, and training. flexible health labor Delink physicians’ licenses to practice from the facilities of employment. market and efficient Delink health workers’ employment benefits from the quota as well as from health facilities. health workforce Work toward abolishing the headcount quota system. management Note: GP = general practitioner; PCIC = people-centered integrated care; PHC = primary health care. 248 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A 3. Reform the compensation system to pro- change would play a key role in the coordina- v id e s t ron g i n c e nt ive s for go o d tion and integration of services to implement performance. PCIC. A gatekeeping mechanism could be 4. Reform the headcount quota system to introduced to ensure that patients are first enable a more flexible health labor market directed to PHC providers, and this arrange- and ef ficient health workforce ment could be mandated once the PCIC sys- management. tem is well established. International experience shows that large- Strategy 3: Introduce career development scale, rapid maintenance or expansion of prospects to develop and incentivize the PHC coverage is possible if there is political PHC workforce commitment and adequate investment. PHC-specific career paths need to be intro- Increasing the availability and accessibility of duced for GPs, nurses, mid-level workers, PHC personnel is a priority, but it is equally and community health workers to better important to ensure that health workers pro- develop and incentivize the PHC workforce. duce high-quality and responsive services. General practice must be established as a spe- This requires that health workers acquire a cialty with equivalent status to other medical professional culture in which evaluation and specialties and with the same strong attri- self-assessment are regarded as powerful butes of well-regulated standards of practice. tools to improve quality. This acculturation China’s current pilots of separate accredita- process starts during basic professional tion for rural assistant physicians and of a education and is maintained throughout a ­ separate professional title-promotion system career, ensuring continuing professional for PHC are good examples; these should be development. Table 7.4 summarizes the main evaluated and scaled up across the country. characteristics of the health systems of some of the countries covered by the review. Strategy 4: Raise compensation of PHC workers to be commensurate with that of other prestige specialties Core Action Area 1: Strong Enabling China has significantly increased its govern- Environment for the Development of ment funding of PHC through its subsidy for PHC Workforce a package of essential public health services, Strategy 1: Establish general practice as a but raising the compensation of PHC workers specialty with status equivalent to other to the level of other specialties would require medical specialties significant additional funding. The govern- It is imperative to improve the status of PHC ment could consider directing its annual to address the challenges in the PHC work- incremental health financing toward PHC. force. It is important to build consensus Most countries include high coverage of pre- among government, health providers, and the ventive and PHC services in their health general public that PHC is as important as insurance benefit package, usually with a hospitals and that the facilities at different fixed copayment or 80 percent reimburse- levels of the system simply have different ment rate. roles and functions in providing a continuum China is revising its health insurance poli- of care to citizens. cies to expand coverage of outpatient and PHC services, financing the expansion with Strategy 2: Direct patients to PHC provid- part of the annual increase in the insurance ers as the first point of contact premium and government subsidy. (In 2015, A consensus about the importance of PHC the per capita subsidy rose from RMB 320 to implies a change in the service delivery model RMB 380 and the individual contribution to ensure that PHC is the patient’s first point rose from RMB 90 to RMB 120.) This would of contact with the health system. Such a incentivize the use of outpatient services at 249 TABLE 7.4  Health system characteristics in selected countries GPs’ compensation Dominant mode of Dominant mode of specialist Gatekeeping by compared with Country Health system PHC provision services provision GPs or PHC specialists’ Australia Tax-financed universal health insurance Private group practices Private group practices, public Yes: GP referral required for the Lower supplemented by a small, compulsory, tax- hospitals, or private hospitals specialist to get reimbursed based health insurance levy Brazil Tax-based health system with supplementary Public health centers Private hospitals Yes Higher private health insurance; decentralized, with states, and municipalities having autonomy for service delivery Canada Universal health system providing tax- Private group practices Public hospitals or private Yes 20 percent lower financed hospital and physician services for all or private solo practices group practices on average residents; highly decentralized with provinces and territories responsible for planning and delivering health services as well as providing three-fourths of total health financing Turkey Social health insurance schemes Private group practices Public hospitals or private Yes Close or solo units hospitals United Kingdom Highly centralized, tax-financed national Private group practices Public hospitals Yes Close health service Note: GP = general practitioner; PHC = primary health care. 250 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A the PHC level, reduce reliance on hospitals, The community health agents are recruited and increase financing for PHC facilities. from the neighborhoods where they are deployed, trained for three months, and Core Action Area 2: Balanced Workforce employed by municipal health authorities. Composition and Competency for PHC They are responsible for a wide range of PHC Service Delivery services: chronic disease management, triage, child development, and public health (screen- Implementing people-centered health care in ing, immunizations, and so forth). Wherever China requires rebalancing the workforce the FHS is not fully implemented, PACS is a and induction of PHC workers with the transition model. Brazil has 234,767 commu- needed skills and competencies. The skill mix nity health agents in rural and urban periph- and scope of practice will need to be eral areas around the country (Johnson and reviewed, and education programs will be others 2013). needed to strengthen health workers’ cultural In 15 years (1998–2013), Brazil raised the and psychosocial competencies, communica- number of its family health teams 6.9-fold tion skills, and capacity to work in multidis- and the population covered 7.8-fold to reach ciplinary teams and to support patients in more than 60 percent of the total population managing their health better. Quality assur- (figure 7.8). Registration with a FHS care ance mechanisms such as accreditation of team is determined by whether a person lives programs and educational institutions and within the FHS team catchment area, not by certification of health professionals will need individual choice. In heavily populated areas, to be implemented or strengthened. there may be more than one FHS team per health facility; each team is assigned a spe- Strategy 1: Set up alternative cadres of cific territory and list of families to serve. health workers to strengthen PHC delivery FHS expansion proceeded unevenly across One trend across many countries is that of Brazil; coverage is higher in small municipali- introducing cadres of nonclinician physi- ties than in large ones, but it now reaches cians, clinical assistants, assistant doctors, more than 90 percent of Brazil’s 5,565 and clinical officers to help expand access to municipalities (Gragnolati, Lindelow, and basic PHC services. International experience Couttolenc 2013). shows that these cadres can be as efficient China may like to explore the possibility and cost-effective as traditional cadres in of producing and integrating alternative delivering health services within the limits of cadres of health workers, especially CHWs. their training when they work in supportive, The village doctors, for instance, could play supervised, multidisciplinary teams. a similar role if they could be better inte- Community health workers (CHWs), grated into the system and provided with sometimes hired from local communities, have improved training, compensation, and been widely used to provide outreach and pre- supervision. ventive services. Brazil, for instance, made CHWs a key part of its PHC service delivery Strategy 2: Accelerate successful efforts to model (Johnson and others 2013). Its health increase supply of GPs and nurses sector reform focused on PHC expansion, rap- Another trend across several high-income idly deploying its Family Health Strategy countries (Australia, Canada, Germany, (FHS) and Community Health Agents the Netherlands, and the United Kingdom) Program (Programa de Agentes Comunitários is that of delivering team-based PHC de Saude [PACS]). The FHS uses multiprofes- through the inclusion of more nurse practi- sional health teams composed of a physician, tioners, registered nurses, and other health nurse, nurse assistant, and four to six CHWs. staff to work alongside physicians (Freund PACS became important after the reorganiza- and others 2015). Australia, Canada, and tion of health service delivery around the FHS. the United Kingdom have added incentives L e v er 6 : S trengthening the   H ealth   W or k force 251 FIGURE 7.8  Community health agents’ coverage in Brazil’s Family Health Strategy, 1998–2014 70 60 50 Share of population, percent 40 30 20 10 0 19 8 19 8 19 9 20 9 20 0 20 0 20 1 20 1 20 1 20 2 20 2 20 3 20 3 20 3 20 4 20 4 20 5 20 5 20 6 20 6 20 6 20 7 20 7 20 8 20 8 20 8 20 9 20 9 20 0 20 0 20 1 20 1 20 1 20 2 20 2 20 3 20 3 13 9 9 9 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 19 ESF coverage ACS coverage Source: DATASUS database, Brazil Ministry of Health, http://datasus.saude.gov.br/. Note: Gaps indicate missing or unavailable data. ACS = community health agent (agente comunitário de saúde); ESF = Family Health Strategy (Estratégia de Saúde da Família). in their GP reimbursement schemes to Furthermore, GPs in the United Kingdom employ nurses to deliver PHC services. increasingly work in multipartner practices. A Most high-income countries give greater typical practice team consists of 5–6 GPs, 1 recognition to family medicine and nurs- nurse practitioner, 2–3 nurses, and 6–10 i ng, a nd most have ma ny fa m i ly administrative staff (Roland, Guthrie, and physicians. Thomé 2012). Teams may also include district In the United Kingdom, strategies to nurses, health visitors, midwives, community improve the accessibility and quality of PHC psychiatric nurses, and allied health profes- services have included expanding nurses’ sionals and social workers. The patient’s first scope of practice. Its National Health point of contact is usually a practice of self- Service Plan of 2000 introduced “new work- employed GPs because the NHS requires peo- ing practices” as a major step toward ple to register with a GP or a GP practice. GP advanced-level nursing practices. Debate on practices are responsible for referring patients expanding the functions of nurses went on to specialist services in hospitals or to commu- for more than a decade, and in April 2012 a nity-based professionals. Financial incentives new law came into effect allowing the more are offered to GPs based on 75 quality-of-care than 20,000 nurses who have undertaken a indicators, many of which relate to the care of specialist degree-level course and hold a sep- patients with chronic conditions (Roland, arate registered qualification to prescribe Guthrie, and Thomé 2012). There is no from the same list of medicines as doctors restriction on ­public-private multiple practice, within their specialty and competence. The though consultants (as specialists are known) Health and Social Care Act of 2012 pro- must inform employers of their private prac- motes integrated, personalized, and proac- tice commitments and obtain NHS authoriza- tive care by better coordinating the National tion to use NHS facilities or staff for private Health Service’s (NHS) hospital- and com- purposes. munity-based health services, including Canada has increased its federal and PHC and social care. p rovincial public investments in PHC ­ 252 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A since 2003. To increase the number of multi- depending on regional socioeconomic char- disciplinary teams providing primary care, acteristics. Failure to complete mandatory provinces committed to providing at least service can result in prohibition on practic- 50 percent of their population with 24/7 ing medicine in Turkey (Aran and Rokx access to multidisciplinary PHC teams by 2014). The effect on retention of profes- 2011. Each province designed its own model. sionals after completion of the compulsory Progress has been slow and unequal across service is not known. provinces, although nurse practitioners have become better integrated into the primary Strategy 3: Scale up standardized training care sector (Freund and others 2015). An for resident doctors and GPs estimated three-quarters of Canada’s family Because health workers from rural back- physicians now work within multiprofes- grounds are more likely to practice in rural sional practices (Marchildon 2013). areas after completing their studies, one strat- Worldwide, numerous policies have been egy is to decentralize professional schools, adopted to improve the geographical distribu- especially in remote and rural areas, and to tion of physicians and nurses. Recruitment and recruit local students, as has been done in the retention of qualified health workers to work in far north of Norway. In another approach, rural and deprived areas is a challenge even for Taiwan, China, has developed programs in high-income countries such as Canada and the offshore islands. And many Chinese provinces United States. Financial incentives—increased have implemented special targeted training salaries and benefits—have been used with programs to recruit and train medical students some success, but they are recognized as only to work in PHC and rural areas. In 2015, for part of the solution to attract and retain health example, Jiangxi province started recruiting workers in rural and remote areas. Examples 200 high school graduates each year for three of efforts to address geographical disparities in years to receive three-year college degrees in health care include the following: clinical medicine with no tuition fees. The pro- vincial government provides a financial sub- • In Brazil, the challenge of scaling up access sidy (RMB 5,000 per year) to the trainees. to medical services in rural and remote After graduation, these students will work as regions remains daunting. Practice in public health specialists in rural THCs. Brazil’s urban areas remains more attrac- From a practical perspective, policy solu- tive, partly because multiple jobs (highly tions may need to differentiate according to prevalent in Brazil) are only possible in local context and provide special modalities urban areas, where private insurance and for rural deprived areas. For example, care delivery schemes coexist with Brazil’s China’s current “5+3”-year standard training publicly funded health system (Sistema (five years of medical university training fol- Único de Saúde) and offer attractive com- lowed by three years of resident training) for plementary income opportunities. GPs and other physicians would not be viable • In Taiwan, China, a recently developed, for rural THCs in deprived areas because detailed strategy includes mobile clinics doctors who are qualified at this level will and communication technologies to pro- quickly move to cities and work in urban hos- vide access to quality services in remote pitals. A “3+2”-year training model might be regions. more realistic for recruiting and retaining • In Turkey, a compulsory service law was health workers for THCs (three years of reintroduced in 2005 to address persistent c ollege-degree medical school training ­ shortages of hospital specialists and GPs in followed by two years of resident training). ­ certain regions. It requires recent graduates An approach called “integrated human from public medical schools to practice in resource management” has been used in the public health sector for one to two some provinces of China, such as Shaanxi, years, and recent graduates from medical to recruit and retain qualified health specialty training to serve two to four years ­ professionals for rural THCs that cannot L e v er 6 : S trengthening the   H ealth   W or k force 253 recruit certified (assistant) physicians. This In general, the official pay of health workers approach allows the county hospitals to in China is not attractive, in particular at the recruit doctors as county hospital employ- grassroots level and in the rural areas. The ees and rotate them to work in THCs or health workers’ income relies heavily on village clinics. the revenues they can generate for the hospi- tal as reflected in their salary structure. On Strategy 4: Improve on-the-job training to average, the basic salary accounted for an strengthen and build PHC competencies of average of 23 percent of total compensation, current workforce allowances for 20 percent, and performance Health professional education and training for 57 percent.14 should be reformed to improve the competen- The structure is even more skewed in cies of the current workforce and build new urban hospitals. An NHFPC national salary PHC competencies. Measures for this pur- survey (implemented by the National Health pose would include the following: Development Research Center) in secondary and tertiary urban hospitals found that for • Introducing new education concentrations health workers in public hospitals, basic sal- for new types of cadres such as clinician ary accounts for only 13–14 percent of the assistants and CHWs total salary; allowances and subsidies for • Reforming medical and nursing curricula 14 percent; and performance-based pay and toward PCH competencies such as com- bonuses, which are linked to hospital service munication skills and patient-centered ser- income, for 74 percent. vices (Hou and others 2014) Although a combination of fixed payment • Revising clinical training to include more with variable performance-based payments is exposure to community-level and primary desirable, China may like to revise its com- care and to encourage multidisciplinary pensation system to reduce reliance on ser- team work vice revenue-based bonuses and to increase • Developing cohorts of trainers who can base salary and hardship allowances. support the development of new competen- On average, this percentage is below cies and scaling-up of standardized resident 30 percent—far below the international norm training for GPs and other physicians of more than 50 percent. • Linking in-service training with recertifi- cation of health professionals Strategy 2: Increase basic wages in line with labor market trends to keep health profes- Strategy 5: Reform the curriculum to sion attractive upgrade medical training and build new Raising PHC workers’ compensation to levels skills required for PCIC similar to those in other prestigious special- All countries have used curriculum reforms ties would make a PHC career more attrac- and continuing education to equip profes- tive and competitive. Almost all countries sionals with PHC competencies. In Brazil, where PHC is a priority have significantly new recruits must do induction training in improved PHC workers’ compensation. PHC; in Turkey, all must undergo special In countries with strong PHC, as in training when family medicine centers and Canada and the United Kingdom, GPs tend community centers are established. to earn more than specialists (reflecting their more comprehensive role). Increasing GPs’ relative earnings also affects the medium- Core Action Area 3: Compensation and long-term supply of GPs. Countries such System Reform to Provide Strong as Brazil, Turkey, and the United Kingdom Incentives for Good Performance have removed or reversed the differential Strategy 1: Increase the share of basic salary between compensation of hospital specialists relative to performance bonuses in health and PHC doctors (figure 7.9). In China, the workers’ total income packages differential remains large. 254 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 7.9  Ratio of hospital specialist pay to primary GP pay, selected countries and years 4.5 4.0 3.5 Ratio of specialist to GP pay 3.0 2.5 2.0 1.5 1.0 0.5 0 Brazil Canada Canada China China Turkey Turkey Turkey United United 2012 2001 2010 2013a 2013b 2008c 2008d 2010 Kingdom Kingdom 2002e 2010 Sources: McPake and others 2015. Note: GP = general practitioner. a. “China 2013a” compares public hospital in an average and primary health institution. b. “China 2013b” compares national-level hospital “professional and technical posts” with primary health institution “health institutional in-service staff.” c. “Turkey 2008c” shows the relative pay of hospital specialist to primary GP (within the country’s Family Medicine Program). d. “Turkey 2008d” shows the relative pay of hospital specialist to primary GP (outside the family medicine program), better reflecting the historical divergence (that is, before the 2005 launch of the country’s Family Medicine Program, which assigns each Turkish citizen to a state- employed family physician). e. “United Kingdom 2002e” data are from DH (2010): consultant (specialist) net earnings are as estimated from graph, p. 23. The size of the increase needs to be linked • In Canada, family physicians earn twice to general labor market trends in China. as much. Figure 7.10 shows how different categories of • In Canada and the United Kingdom, health workers are paid relative to their coun- nurses earn about 50 percent of those lev- try’s income distribution. These data are not els, placing them between the fourth and directly comparable because they are not dis- fifth quintiles of their populations’ average aggregated by type of health professional. per capita income. However, they show health workers’ place within the income distribution of several As for China, pay is significantly worse countries, as follows: for workers in the country’s PHC institu- tions, who earn less than 30 percent of • In the United Kingdom, the 10 percent of the top decile’s average income—putting population at the top of the income distri- them a little above the population’s aver- bution has an average per capita annual age income (between the third and fourth income of US$111,252. A hospital consul- quartiles of the per capita income distri- tant (specialist) earns US$164,345, bution). However, hospital workers over- approximately 1.5 times the average all earn 1 to 1.5 times the top decile’s income of the highest income decile. average income. • In Brazil, family physicians are paid 4.5 times the average income of the highest Strategy 3: Revise the incentive system by income decile. linking income to comprehensive performance • In Turkey and the United Kingdom, assessment rather than revenue generation doctors in general earn around the average ­ Countries like Australia, Canada, and the income level of the top decile. United Kingdom include incentives within L e v er 6 : S trengthening the   H ealth   W or k force 255 FIGURE 7.10  Ratio of health professionals’ pay to average income per capita of 10th richest decile, China and selected countries 5.0 Ratio of pay to top decile’s average per capita income 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 Ho ital ns, p lic s m GP Ho ysi ns, p es Fa es, blic Ge inst tal sp ph au ses Ho pec n sp urs ily ns l n s, e ph te Pu al n ts Sp l ho on l pe l y ( pi t se ala GP PH ho s n- l GP Se l co y GP sp ho l sp am GP oy nt l s ita ial ita Ho M ita ar os alis ve e ita rse at ia s ily iva i m lta sp cia ub bl urs Ho N m ia C spi ra ti l p cia iar sp iali Al ysic Ho ital ing nur ita sp ec sp sp nu riv TC osp wi nd ed ed P) ur pu no ta ne itu Fa ysic ci il u m pr ita ysi xil FM id s a ri lf- ns h ph pl ic F it ls H S s ily ita e m Fa h Ho im ur ln Pr ita sp Ho Brazil, 2012 Canada, 2010 China, 2010 Turkey, 2010 United Kingdom, 2010 Source: World Bank data. Note: FMP = Family Medical Program (Turkey); GP = general practitioner; PHC = primary health care; TCM = traditional Chinese medicine. reimbursement schemes for GPs to encourage cervical screening, asthma care, and indige- them to employ nurses to deliver primary nous health (Cashin and others 2014). care. Another trend seen in many European In China, the comprehensive compensa- countries is that of contracting GPs as entre- tion reforms in Sanming prefecture of Fujian preneurs, with remuneration topped up province offer a successful example (box 7.2). through various pay-for-performance incen- In recent years, remuneration systems have tives (Kringos and others 2013). This has become quite complex globally, especially as resulted in a surge of practices run as part- countries experiment with innovative pay- nerships of several physicians or by private ment methods to find new ways of incentiv- companies. izing health workers. As a result of this con- Similar experiments are being carried out tinuous trial process, countries typically in Australia where, because of inherent weak- adopt a combination of payment methods, nesses in the fee-for-service payment scheme including salary, fee-for-service, capitation, for GPs, the government introduced a performance bonuses, and so on. Practice Incentive Payment program in 1998. For example, in countries where the com- This pay-for-performance scheme provides pensation method was primarily fee-for-­ incentives around three areas: quality of care, service, elements such as salaries (Canada), capacity strengthening, and support in rural capitation fees (Belgium and France), areas. The quality-of-care component pro- ­ p erformance (France), and integrated fees vides incentive payments for diabetes care, (Belgium and Denmark) are being introduced 256 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A as additional payments. In countries where provinces with low population density GPs were traditionally salaried, capitation and (Wranik and Durier-Copp 2010). fee-for-service are being added (Finland and At the same time, the new payment meth- Sweden). To increase efficiency and quality of ods are raising costs and putting pressures on services, the United Kingdom introduced a the financial capacity of the country’s health pay-for-performance scheme (box 7.3). system. Physician incomes in Canada have Fee-for-service payment has traditionally increased substantially in the past decade (to been the predominant mode of remunera- four and a half times that of an average salary tion for most physicians in Canada, but in Canada), aided by the collective bargaining alternate remuneration methods have been model that has put pressure on provinces to introduced over the past 10 years. In 2013, continually increase compensation. Public the largest category of physician remunera- support has guided the relative strength of the tion was a mixed method of payment, and different parties in the collective bargaining the proportion of physicians being paid pre- process in Canada over the years (Ontario, dominantly by fee-for-service fell from for example, has been able to freeze remuner- 51 percent in 2004 to 38 percent in 2013. ation for doctors due to the shift in public Family physicians have a higher rate of support), but doctors have generally been able blended payments (46 percent) than special- to successfully negotiate higher wages at ists (37 percent).15 Blended payments in times when the public felt that doctor short- Canada have been associated with some ages created long waiting times. positive effects on preventive care, collabo- There is also a trend of general practices ration, and recruitment and retention in being run as partnerships of several GPs or BOX 7.2  Sanming’s comprehensive public hospital compensation reform In 2013, Sanming prefecture (Fujian province) intro- • Salary weights that reflect the capacity of dif- duced comprehensive public hospital compensation ferent types of hospitals to generate labor- reform with three major components: based service income • New governance structure: Hospital directors are • The director’s performance assessment directly appointed and paid by the government. results. They are given management autonomy over daily • Salaries comprising a basic allowance plus perfor- operations. The government assesses hospital mance incentives: The salary of each hospital staff directors’ performance according to the extent to member consists of the following: which the hospital has achieved annual perfor- • Basic allowance defined by technical grade, mance targets and tasks set by the government. seniority, and managerial position • New criteria for hospital salary budgets: The gov- • Payment for services provided, such as number ernment ended reliance on revenue generated from of outpatient services performed and number lab tests, physical exams, and drug sales (fully of inpatients discharged—the workload being implementing a zero-markup policy on drug sales). adjusted considering the service complexity as It now defines the total amount of funds that can defined by diagnosis-related groups be used for salary for each of the 22 hospitals in the • Awards or sanctions based on patient satis- prefecture, using three determining criteria: faction rates, malpractice, additional working • The hospital’s labor-based medical service rev- hours, and emergency medical aid. enue, excluding income from lab tests, physical exams, and medical supplies Source: Ying 2014. L e v er 6 : S trengthening the   H ealth   W or k force 257 BOX 7.3  GP pay-for-performance incentives in the United Kingdom The United Kingdom comprehensively reformed the of the new contract were payments for essential remuneration of general practitioners (GPs) in 2004. ­ s ervices (global sum), enhanced services, out-of- Contracts and payments went from being indepen- hours care, and the QOF (Boyle 2011). dent GP-based to practice-based and from largely GP incomes rose 58 percent between 2002–03 capitation-based to a significant proportion of pay- (before the reform) and 2005–06 (after the reform), for-­performance (Doran and Roland 2010; Krone- and they are now closer to the incomes of hospi- man and others 2013). The main objectives of the tal specialists, especially for nonsalaried partners reform were to improve the quality of care; help in general practice. In addition, the number of GPs recruit and retain GPs; improve GPs’ job satisfaction, (particularly salaried GPs) increased in line with pay, and working conditions; improve the unequal population growth, and vacancy rates fell (Doran distribution of personnel; and increase primary health and Roland 2010). The number of nurses employed care (PHC) productivity (Doran and Roland 2010). in primary care also increased as a result of the pay- Before the reform, GP contracts included a basic for-­performance incentive linked with chronic disease allowance (with a gradient by seniority) supplemented management. (The use of nurses has been associated by allowances based on the number and characteris- with increased quality of care in chronic disease man- tics of patients on the GP’s list. The reform introduced agement [Freund and others 2015].) Performance a pay-for-performance component for quality require- against the QOF’s quality indicators improved in the ments that are outlined in the Quality and Outcomes first three years (Doran and Roland 2010). Framework (QOF). The QOF has four domains: One potential perverse incentive was that care for nonincentivized diseases and activities might • Clinical standards, emphasizing noncommunica- be neglected, and the evidence on this has been ble diseases such as cancer and hypertension mixed. There remain concerns that quantifiable • Organizational standards, including informa- aspects of care may be prioritized over nonquantifi- tion, communication, patient education, and so able aspects, that external incentives may replace forth internal motivation, and that the incentives might • Additional services standards, such as cervical undermine professionalism (Doran and Roland screening, child health, maternity and contracep- 2010). The United Kingdom started a scheme in tive services) 2008 in which an element of GP pay was based on • Patient experience standards. patient satisfaction surveys, but this proved prob- QOF revenues were expected to increase prac- lematic and was withdrawn in 2011 (Roland, Guth- tice revenue by 15–20 percent. The key components rie, and Thomé 2012). by private providers. Table 7.5 provides an Commonly used options include housing, overview of how primary care and specialist job opportunities for spouses, education care are organized in selected countries, and allowances for children, and opportunities for how physicians are paid. further training (such as in-service training and scholarships for postgraduate studies). Strategy 3: Increase the subsidy, and intro- Changes in health workforce compensation duce or increase nonfinancial incentives, also require changes in the approach to setting for health workers in rural and remote areas compensation in China, which can vary from China will also need nonfinancial incentives to a rigid bureaucratic process to a more flexible attract and retain PHC workers, especially in and negotiated process between employers rural and remote areas. International experi- and employees. The former approach often ences suggest that financial incentives alone requires executive and legislative action to often cannot provide sufficient motivation, and change compensation practices. The latter nonfinancial incentives can help meet the needs approach depends on the extent to which and expectations of health professionals. employers are required to bargain with 258 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 7.5  Physician payment and predominant service provision mode, selected countries Primary care Outpatient Inpatient physician specialist specialist Predominant mode of Predominant mode of Country payment payment payment primary care provision specialist services provision Australia FFS FFS Salary Private group practices Private group practices or public hospitals Brazil Salary Salary Salary Public health centers Private hospitals Canada FFS FFS FFS Private group practices or Public hospitals or private private solo practices group practices Turkey Capitation PFP and salary PFP and salary Private group or solo units Public hospitals United Salary, Salary Salary Private group practices Public hospitals Kingdom capitation, FFS Sources: McPake and others 2015; Paris, Devaux, and Wei 2010. Note: FFS = fee-for-service; PFP = pay for performance. unions, the availability of arbitration proce- • Coordination: the extent to which wage dures to solve disputes, and whether workers setting is coordinated across sectors or have the right to strike as part of the compen- companies, the level of national govern- sation-setting process (Gregory and Borland ment involvement, and the frequency of 1999). wage setting. The structure of unions and employers’ organizations inf luences the degree of The impetus to change the approach to collectivization in the compensation-setting wage setting can come from a range of fac- process; these organizations can be defined by tors. In OECD countries, labor market con- geographical scope or by the type of workers cerns (such as shortages or a geographic (nurses, doctors, and so forth) or activities that maldistribution of health workers) have they cover. Trade unions and professional asso- spurred overall pay increases or called into ciations are recognized actors for representing question the responsiveness of a national health workers, and agreements usually cover standardized approach to varied and local- the entire health workforce in specific occupa- ized shortages in occupations, geographic tions or sectors, regardless of membership regions, or specialties. High-cost central status (Buchan, Kumar, and Schoenstein 2014). urban areas as well as remote rural areas are Buchan, Kumar, and Schoenstein (2014) often identified as meriting “more” than the reviewed wage-setting mechanisms for hospi- national average pay rates to offset recruit- tal staff in eight OECD countries (Canada, ment and retention challenges; and certain France, G ermany, New Z ealand, the specialties that are relatively difficult to Netherlands, Norway, Portugal, and the recruit to may also be identified as “deserv- United Kingdom). Their study looked at three ing” above-standard rates of pay. features of wage setting (table 7.6): Other pressures have included pay equity issues, particularly given the high numbers of • Collective bargaining: the extent to which women workers in nonmedical professions; wages are determined collectively, nor- structural changes in pay systems (for exam- mally in agreement with professional asso- ple, to increase flexibility or to delink pay in ciations and trade unions the health sector from pay in the broader • Centralization: the level or levels at which public sector or service); attempts to improve wages are set, as an indicator of the extent organizational productivity and the quality to which the wage-setting process is cen- of care; and improving the international com- tralized or localized petitiveness of pay. L e v er 6 : S trengthening the   H ealth   W or k force 259 TABLE 7.6  Characteristics of compensation setting for hospital staff, selected OECD countries Type of characteristic Findings Collective • All countries recognize trade unions and professional associations as representing the bargaining hospital workforce. (collectivization) –– The level of union membership (coverage) varies significantly across countries, but collective agreements normally cover all the workforce in designated occupations or sectors, irrespective of their membership status. • In all countries, employers are involved directly in negotiating and achieving collective agreements. –– In Canada, New Zealand, the Netherlands, Norway, and the United Kingdom, employers are represented in national wage setting by some type of employers’ association that has a specialist wage-setting capacity. Centralization of • Most countries have a core national or sectorwide model. wage setting –– In France, New Zealand, the Netherlands, Norway, Portugal, and the United Kingdom, the primary focus is at the national level, either across the whole health sector or in subsectors or specialties within health. • Although national agreements are at the heart of policy, France, Norway, and the United Kingdom (to an extent) provide scope for “top-up” wage setting at the local level. –– New Zealand also has some separate regional or local collective agreements, which are in part a legacy of a previous decentralized wage-setting model. • Of the eight countries, Canada has the greatest focus at the province level. • Germany has a mixed pattern between national and regional approaches as well as a trend toward fragmented wage setting. Coordination • All countries reported some degree of coordination of wage setting across the health or and government hospital sector. involvement –– Coordination is based on national health sector (or subsector) collective frameworks (France, the Netherlands, Norway, Portugal, and the United Kingdom), or occurs at the province level (Canada) or across central and local governments (Germany). • Cross-sectoral coordination within the broader public sector was reported in some countries (for example, New Zealand). • Norway uses a broader cross-sectoral coordinated approach based on “front runner” industries setting the benchmark for wage setting. • Reflecting the high level of public provision, most countries reported direct or indirect government involvement in wage setting. –– In Portugal and the United Kingdom, the government is the main funder or employer of the hospital workforce. • France, Portugal, and the United Kingdom have an annual wage-setting cycle; Norway has a biannual process; in the other countries, the wage-setting cycle varies between 18 months and 3 years. Source: Buchan, Kumar, and Schoenstein 2014. Note: OECD = Organisation for Economic Co-operation and Development. Eight OECD countries were studied: Canada, France, Germany, New Zealand, the Netherlands, Norway, Portugal, and the United Kingdom. Core Action Area 4: Headcount Quota financing policy. As noted earlier in this System Reform for a More Flexible chapter, the Chinese government is aware of Health Labor Market and Efficient this issue and is taking action to reform the Health Workforce Management system. The reform would require at least four sets of related actions. The headcount quota system leads to ineffi- ciencies in the management of the Chinese Strategy 1: Give managers autonomy on health workforce and needs to be improved human resources issues to be consistent with broad health sector First, health facility managers would need reform trends including increasing hospital to be given the necessary autonomy on autonomy, increasing health labor market human resources issues and be left to man- mobility, and performance- or results-based age their staffs on the basis of the post 260 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A rather than quota. This approach would Ireland, Israel, Italy, and Spain), central or end differences between workers with and subnational governments are responsible for without a quota post. recruitment of staff for public hospitals. In Each health worker would have a stan- Australia and Portugal, recruitment is done dardized labor contract with the health facil- by public hospitals but requires approval ity that would describe the responsibilities, from the central or subnational government scope, and accountability of the post. The (Buchan 2015). government would continue to define the cat- As for wage setting, in more than half (19) egories and grades of common posts but of the OECD member countries, salaried would grant health facility managers the medical staff negotiate work contracts autonomy to decide their workforce composi- directly with the public hospital. This pattern tion in term of posts, grades, technical quali- probably reflects specific administrative fications, and professional titles. arrangements unique to these countries and Health facility managers would be respon- need not imply that salaries are set by public sible for post-based recruitment, post-based hospitals themselves. However, it is worth deployment, post-based evaluation, post-based noting that the three countries with hospital salary setting, and post-based training. They autonomy over pay levels (Poland, Sweden, would be able to fire a worker who fails to per- and the United States) also have contracts at form according to the contract. In turn, man- the hospital level (Buchan 2015). agers would be evaluated by the government Table 7.7 provides an overview of respon- and held accountable for the performance of sibility for hospital recruitment and remuner- their staff and that of the health facility. ation in the same five countries reviewed for Many OECD countries practice central- comparison throughout this chapter. ized wage setting simultaneously with decen- tralized recruitment. Fourteen OECD coun- Strategy 2: Delink physicians’ licenses to tries set medical staff remuneration in public practice from the facility of employment hospitals at a national or subnational level To increase the mobility of health workers, while giving responsibility to hospital manag- China may wish to consider delinking a physi- ers for recruitment (Buchan, Kumar, and cian’s license to practice from the facility of Schoenstein 2014). In five countries (Greece, employment. The country’s 2009 Physician TABLE 7.7  Hospital staff recruitment and remuneration in five countries Recruitment of Remuneration level of Recruitment of other Remuneration level of other Country medical staff medical staff health professionals health professionals Australia Hospital managers Hospital managers have Hospital managers Hospital managers have have complete autonomy within state-level have complete autonomy within state- autonomy negotiated pay scales autonomy level negotiated pay scales Brazil Decentralized, with Decentralized, with state Decentralized, with Decentralized, with state and municipal and municipal governments state and municipal state and municipal governments having having autonomy governments having governments having autonomy autonomy autonomy Canada Hospitals must Pay scale set or negotiated Central or subnational Pay scale set or negotiated negotiate with local at national level government decides at national level authorities Turkey Central or subnational Pay scale set or negotiated Central or subnational Pay scale set or negotiated government decides at national level government decides at national level United Hospital managers have Pay scale set or negotiated Hospital managers have Pay scale set or negotiated Kingdom complete autonomy at national level complete autonomy at national level Sources: McPake and others 2015; Paris, Devaux, and Wei 2010. L e v er 6 : S trengthening the   H ealth   W or k force 261 Dual Practice Policy has already paved the As also noted earlier, if a health worker way for this transition. For example, leaves the facility, he or she loses all associated Guangdong province no longer limits the employment benefits, including his or her pen- number of facilities where physicians can work sion. Ending the quota system and delinking as long as they can reach agreement with each benefits from the health facility where individ- facility. Physicians are only required to file a uals are employed, thus making the benefits record with the local health bureau to indicate portable, would remove a large penalty for the health facilities where they work. However, mobility. The government has already started this is not the practice for the whole country, pension reforms to link pensions with PSUs so and in Guangdong it only applies to physicians that the workers outside of the headcount at the middle level and above. quota could qualify for the employment pen- Although multiple practice can increase sion pooling scheme. mobility and retention, it can negatively affect the quantity and quality of care pro- Strategy 4: Work toward abolishing the vided in the public sector. Especially among headcount quota system physicians, the higher earnings and better Reforming and eventually abolishing the working conditions associated with private headcount quota system is a critical step to practice (as in Brazil) can make it difficult to enable efficient health workforce management retain workers in the public sector. Many and for doctors to evolve from being “hospital countries that allow multiple practice also property” to becoming independent practitio- have a problem with doctors channeling ners. Abolishing the headcount quota system patients from the public facility to their pri- would require complementary reforms as pre- vate practice instead. Turkey passed a law in conditions or building blocks—reforms in the 2010 prohibiting multiple practice for all pension system, the post-management system, doctors (except university-based doctors, the professional-title-promotion system, the who can engage in multiple practice pro- government’s fiscal inputs, and the manage- vided they fully meet their daily commit- ment of opportunities for clinical research and ments in the public sector) and raising public in-service training. Such reforms take time to sector salaries. implement well, though many local pilots have Some countries prohibit dual practice begun. A platform would need to be estab- (whereby doctors combine part-time private lished to replace the PSU in performing some practice with a public sector job), while oth- of the human resource management roles. ers regulate or restrict it with different inten- Professional associations could take on sities and regulatory instruments. Ultimately, many of these roles. In many countries, they the success of each approach depends on the play important roles in governance. Their institutional context, resources, and the gov- most common function is to set standards ernment’s ability to enforce regulations. and assure quality through accreditation and certification. They also are well posi- Strategy 3: Delink health workers’ benefits tioned to manage professional title promo- from the quota and from health facilities tion and continuing education. In many China might also consider delinking the countries, professional associations repre- employment benefits of health workers from sent their members in developing legislation the quota as well as from health facilities, a and compensation-setting processes as well process that has already started with the as in advocacy for patients’ rights. delinking of pensions as part of recent However, these roles may conflict with reforms. This would relieve an important broader health system or social objectives— constraint on mobility in China’s health sec- as happened, for example, in Canada when tor. Currently, as noted earlier, the employ- professional associations pressed the govern- ment benefits associated with a post are ment to increase compensation by more than determined by the headcount quota and the consumer price index, or in Brazil, when attached to the health facility. the medical association blocked reforms to 262 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A expand nurses’ scope of practice. As China reforms in other areas and cannot be dealt considers a more engaged role for profes- with as if they are separate. For example, sional associations in the governance of its reforms in financing and payment systems health workforce, it should also be aware of are preconditions for changing the com- potential negative impacts. pensation system of health professionals. Salary increases will need increased fund- ing for personnel costs. The coherence and An Implementation Road Map full alignment of reform policies in a com- The core action areas and associated strate- prehensive package of reforms is critically gies address the main elements of the health important for success. workforce: production, recruitment, deploy- • There is no “one size fits all” solution. ment, management and regulation, perfor- Policy designs as well as reform implemen- mance evaluation, and compensation. tation will need to adjust and adapt to Implementing these strategies would entail local contexts. Rural and urban regions both short- and long-term steps. Some need may need different reform models at dif- immediate action because they set the stage ferent times because their needs in relation for others. Others, like improving quality, to the health workforce differ; policy solu- will take longer to implement. tions will need to be matched to the needs Several principles are useful to keep in in different situations in the country. mind in implementing human resource policy • The transformation of a health workforce reforms: takes time. Good sequencing and coordi- nation of actions will contribute to suc- • The scale of the human resources chal- cessful reform outcomes. Implementation lenges and the vital importance of over- will require a long-term vision plus an coming this bottleneck require high-level implementation pathway defining short-, attention and commitment. Human medium-, and longer-term action plans. resources for health should be on the agenda of groups leading the health sys- Table 7.8 summarizes the specific actions, tem reform and be an integral part of the designates the responsible agencies to lead overall health reform action plan. implementation, and notes the complementary • Human resources issues are at the root of measures that may be necessary to support the challenges faced by health care reform implementation and maintenance of the human in China. They are interwoven with resource reform actions and desired outcomes. 263 TABLE 7.8  Road map for implementation of health workforce reforms Recommendation Short-term Medium-term Responsible Complementary (core action area) actions actions agencies reforms needed 1: Build a strong • Establish independent system of professional • Establish general practice as a • State • Ensure full functioning of PHC by enabling titles and career development prospects for specialty (family medicine) with Council putting in place a gatekeeping environment for the PHC workforce (particularly for GPs) equivalent status to other medical • MoF mechanism to ensure PHC is the development of the • Introduce PHC-specific career development paths specialties • MoHRSS patient’s first point of contact PHC workforce to to develop and incentivize PHC workforce • Enhance compensation system • NHFPC • Shift health care financing toward PHC implement PCIC • Include separate career pathways for GPs, nurses, (monetary and nonmonetary) for • Revise health insurance policy to mid-level workers, and CHWs. PHC workforce relative to other expand coverage for PHC specialties to make PHC careers more • Target government incremental health attractive and competitive sector funding to PHC 2a: Improve workforce • Continue to expand training for GPs and nurses • Produce and integrate new and • NHFPC • Raise status of the PHC workforce (see composition (skill mix) • (Re)train village doctors (to upgrade their scope alternative cadres of health workers, • MoHRSS Recommendation 1) for PHC service delivery of practice) such as clinical assistants, assistant • Introduce payment mechanisms that • Develop specific policy package of qualification doctors, clinical officers, and CHWs incentivize the use of multidisciplinary and promotion for rural PHC workforce to • Increase salary for specialties in teams improve recruitment and retention of PHC shortage, such as pediatricians and • Establish medical alliances among workers psychiatrists, to attract medical hospitals and PHC facilities that enable • Adopt regulatory measures to mandate rural or students rotation of health workers within an PHC service (for example, in return for training, alliance requiring a period of service after training is completed) • Target recruitment of students from rural areas to increase recruitment to rural and remote settings • Integrate recruitment: county hospitals recruit doctors as county hospital employees and rotate them to work in THCs or village clinics (Table continued next page) 264 TABLE 7.8  Road map for implementation of health workforce reforms  Continued Recommendation Responsible Short-term actions Medium-term actions Complementary reforms needed (core action area) agencies 2b: Improve workforce • Scale up standardized training for resident • In vocational training schools and • MoE • Introduce quality assurance competencies for PHC doctors and GPs medical universities, introduce new • NHFPC mechanism for health professionals’ service delivery • Reform curriculum to upgrade medical training medical education programs and • MoHRSS education through accreditation of to add PCIC skills and competencies (such as courses for new workforce types training schools communication skills) and education and training such as clinical assistants, assistant for multidisciplinary team practice doctors, clinical officers, and CHWs • Improve current in-service training to improve competency in current workforce and to build new PHC competencies • Link in-service training requirements with relicensing and re-certification 3: Reform compensation • Increase basic compensation of health workers • Rationalize compensation structure • State • Adjust pricing schemes to increase system to provide to change current incentive structure (based on (increase percentage of basic salary Council charges for labor-based services, such strong incentives for quantity of services) relative to performance bonuses) (including as doctor consultations, surgeries, and good performance • Increase subsidy for rural and remote area MoHRSS, nursing services health workers MoF, • Increase percentage of hospital • Introduce or increase nonfinancial incentives NHFPC, expenses allocated for personnel (such as housing) to attract and retain health NDRC) costs to bring China’s average (below workers in rural and remote areas 30 percent) closer to international practice (50 percent) • Integrate fragmented financing, especially financing for PHC, to increase leverage for payment and purchasing • Reform payment system to link payment with improved quality of services rather than with revenue generation • Reform payment system to incentivize new PCIC service delivery approach (Table continued next page) 265 TABLE 7.8  Road map for implementation of health workforce reforms  Continued Recommendation Responsible Short-term actions Medium-term actions Complementary reforms needed (core action area) agencies 4: Reform headcount • Grant autonomy to health facility managers • Delink health workers’ employment • State • Change the way the government quota system to on HR management (including post setting, benefits from the quota and from Council provides its subsidy to health enable a more flexible recruitment, deployment, performance health facilities through pension (including facilities—moving away from paying health labor market evaluation) by moving from quota-based to post- reform MoHRSS, for a fixed headcount of health and efficient health based management • At urban and rural county level, MoF, workers to paying for tasks and workforce management • Apply realistic recruitment policies for recruiting gradually reform quota system NHFPC) performance, or link the subsidy to per health workers for rural PHC centers to transform the physician from capita demand-side financing • Delink the practice license from a specific health “hospital property” to “individual • Reform government pension system facility, to increase mobility of health workers practitioner” to delink the pension from PSUs and • Establish performance evaluation system to • Enhance the function and roles of merge it with social pension pooling increase accountability of health facility managers professional associations schemes • Define an appropriate regulatory policy for health • Institute other complementary reforms worker multiple practice to delink practitioners from the quota system, including professional title promotion, opportunity for clinical research, and continuous training— establishing a platform to replace the PSU in performing some HR management roles • Define the private sector’s role in provision and financing of health services (to define best policy related to health worker multiple practice) Note: CHW = community health worker; GP = general practitioner; HR = human resources; MoE = Ministry of Education; MoF = Ministry of Finance; MoHRSS = Ministry of Human Resources and Social Security; NDRC = National Development and Reform Commission; NHFPC = National Health and Family Planning Commission; PCIC = people-centered integrated care; PHC = primary health care; PSU = public service unit. 266 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Annex 7A  Supplementary Tables TABLE 7A.1  China’s health workforce: Classifications, numbers, and percentages of total, 2012 Primary Secondary Tertiary classification classification classification Description Health professional n.a. n.a. Includes licensed physician; registered nurse; pharmacist 6,675,549 and assistant pharmacist; medical laboratory technician and 73.2% inspector; medical imaging technician; health supervisor; intern doctor (including pharmaceutical intern, student nurse, and intern technician); and other health professionals. Health professionals working in management (hospital director and associate director, party secretary, and so on) are not included. This group of health personnel usually receives higher medical education. n.a. Physician n.a. Includes licensed physicians and assistant licensed physicians. 2,616,064 28.7% n.a. n.a. Licensed Includes staff whose title is “licensed physician” on their physician certificates as medical practitioners and who work in 2,138,836 medicine, prevention, and health care; excludes those working 23.5% in management. This group has a bachelor’s degree or higher, majoring in medicine, from a college or university. There are four categories of licensed physicians: clinical, traditional Chinese medicine, stomatology, and public health. n.a. n.a. Assistant Includes staff whose title is “assistant licensed physician” on licensed their certificates as medical practitioners and who work in physician medicine, prevention, and health care; excludes those working 477,228 in management. They graduate from colleges, universities, 5.2% or junior colleges with a medical vocational degree. Assistant licensed physicians can be divided into four categories: clinical, traditional Chinese medicine, stomatology, and public health. n.a. Registered nurse n.a. Includes health professionals registered as practicing nurses 2,496,599 and with a Practicing Nurse Certificate, engaging in nursing 27.4% according to the statutory nursing regulation: “to protect life, relieve patients’ pain and improve health.” n.a. Pharmacist, n.a. Includes professionals responsible for providing knowledge of including assistant drugs and pharmaceutical services: chief pharmacist, associate pharmacist 377,398 chief pharmacist, pharmacist-in-charge, pharmacist, and 4.1% assistant pharmacist. Apothecary is not included. n.a. Lab technician n.a. Includes medical laboratory technicians and medical imaging 249,255 technicians: chief technician, associate chief technician, 2.7% technician-in-charge, and technician. Village doctor n.a. n.a. Originally called “barefoot doctor”; includes people working in 1,094,419 village clinics who have a Village Doctor certificate. Those who 12% work in village clinics without the certificate are called “health workers.” Other technician n.a. n.a. Includes nonhealth personnel who engage in repairing 319,117 medical equipment, health education, scientific research, 3.5% teaching, and other technical work. (Table continued next page) L e v er 6 : S trengthening the   H ealth   W or k force 267 TABLE 7A.1  China’s health workforce: Classifications, numbers, and percentages of total, 2012  Continued Primary Secondary Tertiary classification classification classification Description Manager n.a. n.a. Includes staff in charge of health care management, disease 372,997 control, health supervision, medical research and teaching, 4.1% and so on, especially those engaged in administration. Supportive worker n.a. n.a. Includes skilled or unskilled staff engaged in operation, 653,623 maintenance, logistic support, and so on. Skilled workers 7.2% include inspector, toll collector, registrar, and so on. Excludes two kinds of staff: (a) those classified as “other technicians” (laboratory technician, technician, and research assistant), and (b) those classified as “managers” (finance person, accountant, and statistician. Note: n.a. = not applicable. TABLE 7A.2  Urban and rural distribution of China’s health workforce, 2003 and 2013 2003 2013 Classification Urban Rural Ratio Urban Rural Ratio Total health workers 3,515,780 1,759,006 2.00 4,488,500 5,291,983 0.85 All health professionals 2,828,419 1,478,052 1.91 3,680,276 3,520,302 1.05 Certified physicians and assistant 1,216,003 651,954 1.87 1,360,118 1,434,636 0.95 physicians Certified physicians 1,026,607 459,422 2.23 1,261,432 1,024,362 1.23 Certified nurses 928,367 337,592 2.75 1,603,913 1,179,208 1.36 Per 1,000 population Health workers 4.88 2.26 2.16 9.18 3.64 2.52 Certified physicians and assistant 2.13 1.04 2.05 3.39 1.48 2.29 physicians Certified nurses 1.59 0.50 3.18 4.00 1.22 3.28 Sources: NHFPC 2004, 2014. TABLE 7A.3  Average salaries in China, by occupational category, 2005–12 Nominal RMB Occupation 2005 2006 2007 2008 2009 2010 2011 2012 2012 rank Agriculture, forestry, 8,207 9,269 10,847 12,560 14,356 16,717 19,469 22,687 19 animal husbandry, fishery Mining 20,449 24,125 28,185 34,233 38,038 44,196 52,230 56,946 5 Manufacturing 15,934 18,225 21,144 24,404 26,810 30,916 36,665 41,650 14 Utilities 24,750 28,424 33,470 38,515 41,869 47,309 52,723 58,202 4 Construction 14,112 16,164 18,482 21,223 24,161 27,529 32,103 36,483 15 Transport, warehouse, 20,911 24,111 27,903 32,041 35,315 40,466 47,078 53,391 7 postal services Information technology 38,799 43,435 47,700 54,906 58,154 64,436 70,918 80,510 2 Wholesale and retail 15,256 17,960 21,074 25,818 29,139 33,635 40,654 46,340 12 (Table continued next page) 268 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 7A.3  Average salaries in China, by occupational category, 2005–12  Continued Nominal RMB Occupation 2005 2006 2007 2008 2009 2010 2011 2012 2012 rank Lodging and restaurants 13,876 15,236 17,046 19,321 20,860 23,382 27,486 31,267 18 Finance 29,229 35,495 44,011 53,897 60,398 70,146 81,109 89,743 1 Real estate 20,253 22,238 26,085 30,118 32,242 35,870 42,837 46,764 11 Rental and leasing 21,233 24,510 27,807 32,915 35,494 39,566 46,976 53,162 8 Science and technology 27,155 31,644 38,432 45,512 50,143 56,376 64,252 69,254 3 service Irrigation, environment, 14,322 15,630 18,383 21,103 23,159 25,544 28,868 32,343 17 and public facilities management Community services 15,747 18,030 20,370 22,858 25,172 28,206 33,169 35,135 16 Education 18,259 20,918 25,908 29,831 34,543 38,968 43,194 47,734 10 Health and social 20,808 23,590 27,892 32,185 35,662 40,232 46,206 52,564 9 protection Culture, sports, and 22,670 25,847 30,430 34,158 37,755 41,428 47,878 53,558 6 entertainment Public administration and 20,234 22,546 27,731 32,296 35,326 38,242 42,062 46,074 13 social organization Total 18,200 20,856 24,721 28,898 32,244 36,539 41,799 46,769 n.a. Source: China Labor Statistics Yearbook, 2006–13. Note: n.a. = not applicable. 269 TABLE 7A.4  Governance framework for health human resources management in China Governing bodies Public Professional Health institute Major issues or Subject Task or area Law or regulation Central govt. Local govt. service unit association (hospital or clinic) challenges Medical College or n.a. MoE, in n.a. n.a. n.a. n.a. Mismatch between education and university consultation with educational training medical NHFPC investments and education labor market demand Postgraduate Guidelines for Standardized NHFPC n.a. n.a. n.a. n.a. None training; resident GP Training, NHFPC, 1999 doctor or GP standardized training In-service Guideline and Planning NHFPC n.a. n.a. Develop n.a. None training for GP In-Service Training, training NHFPC, 1999 courses In-Service Training and Requirement for THC curricula Health Professionals, NHFPC, 2004 In-Service Training Requirements for Village Doctors, NHFPC, 2004 (Table continued next page) 270 TABLE 7A.4  Governance framework for health human resources management in China  Continued Governing bodies Public Professional Health institute Major issues or Subject Task or area Law or regulation Central govt. Local govt. service unit association (hospital or clinic) challenges Entry Establishing Physicians: Chinese Law of NHFPC formulates Licensing: health National n.a. n.a. Standardized qualification Licensing Medical Practice, exam procedures authority at county Medical qualification standards, 1999 and higher levels Exam Center exam does not qualification implements differentiate across examinations, the national positions, specialties, and certification exam and levels of health practice care providers licensing Nurse Management, MoH, Market entry, Licensing: health n.a. n.a. n.a. Pharmacist market 1993 NHFPC authority at county entry qualifications Nurse Regulations, State and higher levels not standardized Council, 2008 because of different management agency Guideline on Qualification MoHRSS and SFDA Licensing: FDA at n.a. n.a. n.a. of Pharmacist and National city and higher Qualification Exam, levels MoHRSS and SFDA, 1999 Village Doctor NHFPC formulates Licensing is with n.a. n.a. n.a. Management Regulation, management rules county health State Council bureau Quota n.a. Guideline on Quota-Setting Central post Each province, n.a. n.a. n.a. Quota standards Standards for General establishment prefecture, and cannot keep pace Hospitals, 1978 office, MoF, and county sets local with changing Guideline on Quota-Setting NHFPC jointly standards in health service Standards for Traditional set standard for compliance with needs or progress Chinese Medicine Hospitals, headcount quota national standards in positions and 1986 functions of health Guideline on Quota- professionals Setting Standards for MHC Huge differences in Institutions, 1986 pay and benefits for Guideline on Quota-Setting health professionals Standards for Urban with and without Community Health Centers, quota 2006 Guideline on Quota-Setting Standards for Township Health Centers, 2011 Guideline on Quota-Setting Standards for CDCs, 2014 (Table continued next page) TABLE 7A.4  Governance framework for health human resources management in China  Continued 271 Governing bodies Public Professional Health institute Major issues or Subject Task or area Law or regulation Central govt. Local govt. service unit association (hospital or clinic) challenges Post n.a. Guideline on Post-Setting MoHRSS and Local HR and n.a. n.a. Health institutes Health facilities have management Management in Health MoH formulate health bureaus and facilities no autonomy on Service Institutes and guidelines on set number and have no post setting Facilities, Ministry of post-setting structure of posts autonomy on Limited promotion Personnel and MoH, 2007 management in for every public post setting space for PHC health health service health institution workers institutes and facilities Recruitment Open Guideline on Deepening The national Local HR bureau n.a. n.a. Some institutes Health facilities have competitive HR Management Reform level formulates implements participate in no autonomy in recruitment in Health-Service Units, the guideline required the process, recruitment Central Party HR Ministry, and recruitment recruitment some do not Rigid recruitment Ministry of Personnel, MoF, procedures procedures and MoH, 2000 Professional Professional Guideline on Strengthening MoHRSS and Provincial n.a. n.a. In provinces Professional title title evaluation title setting and Health Professional Titles NHFPC formulate bureau of HR that separate evaluation allows professional title Evaluation, Ministry of management responsible for professional title limited space for evaluation Personnel and MoH, 2000 procedures setting evaluation evaluation and career development HR central of procedure recruitment, for PHC workers NHFPC organizes for high-level health facilities the national exam professional titles; have autonomy for middle- and health bureaus over recruiting junior-level implement In provinces professional titles that integrate professional title evaluation and recruitment, health facilities have no autonomy to recruit (Table continued next page) 272 TABLE 7A.4  Governance framework for health human resources management in China  Continued Governing bodies Public Professional Health institute Major issues or Subject Task or area Law or regulation Central govt. Local govt. service unit association (hospital or clinic) challenges Performance Includes Guideline on Implementing NHFPC Local health n.a. n.a. Health facilities Assessment indicator evaluation government Performance-Based formulates the bureau formulates formulate own system needs evaluation Evaluation in Health Service implementation local plan implementation improvement of health Institutions, 2010 plan for accordingly plans Third-party facilities and Implementation Plan performance evaluation system health facility of Performance-Based evaluation yet to be established evaluation of Evaluation in Community management HMIS system yet to staff Health Centers, 2011 be incorporated to Guideline on Implementing support monitoring Performance-Based and evaluation Evaluation in Township Feedback and use Health Centers and Village of evaluation results Clinics, 2011 need enhancement Guideline on Implementing to incentivize Performance-Based performance Evaluation in County Maternal and Child-Health Institutions, 2010 Guideline on Implementing Performance-Based Evaluation in Emergency Rooms, 2010 Guideline on Implementing Performance Based Evaluation in Health Promotion Institutions, 2011 (Table continued next page) 273 TABLE 7A.4  Governance framework for health human resources management in China  Continued Governing bodies Public Professional Health institute Major issues or Subject Task or area Law or regulation Central govt. Local govt. service unit association (hospital or clinic) challenges Compensation n.a. Guideline on Internal MoHRSS formulates Local bureau of HR n.a. n.a. n.a. Fixed wage level Income Distribution, wage standards for formulates local cannot provide Reform of Health Service public institutions, implementation incentives for good Institutes and Facilities, including public plan based on performance MoH, 2002 hospitals central guideline Salary structure Guideline on Implementing and sets the total needs adjustment; Internal Income wage ceiling a large portion is Distribution, Reform Action for each public from revenue-based Plan of Health Service institution bonus Institutes and Facilities, Wage gaps too Ministry of Personnel, MoF, narrow between and MoH, 2006 different positions; Guideline on Implementing disparity significant Performance-Based Salary between rural and in Public Health Institutes urban income and Primary Health Care Facilities, Ministry of Personnel, MoF, and MoH, 2009 Note: CDCs = centers for disease control and prevention; FDA=State Food and Drug Administration; GP = general practitioner; HMIS = health management information system; HR = human resources; MCH  = maternal and child care institution; MoE = Ministry of Education; MoF = Ministry of Finance; MoH = Ministry of Health; NHFPC = National Health and Family Planning Commission; PHC = primary health care; SFDA = State Food and Drug Administration; THC = township health center. 274 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 7A.5  Headcount quota formulation standards for Chinese health institutions Issue Scope of application date Policy directives Standards Urban general 1978 Draft Principles on The state estimates staff quotas for public hospitals hospitals, university Organization and according to the number of beds. In accordance with teaching hospitals, Quota System of national requirements, hospitals with fewer than 300 beds county hospitals, General Hospitals are calculated to have a staff-bed ratio of 1 to 1.3–1.4; those county specialty (trial) with 300–500 beds, 1 to 1.4–1.5; and those with more than hospitals, and 500 beds, 1 to 1.6–1.7. outpatient clinics Of the total staff, administrative management are 8–10 percent; support workers are 20 percent; and health care technical personnel are 70–72 percent. Among the latter, doctors and TCM doctors account for 25 percent; nursing staff, 50 percent; pharmacy staff, 8 percent; inspection staff, 4.6 percent; radiation inspection personnel, 4.4 percent; and other staff, 8 percent. Traditional Chinese 1986 Formulation Standards The bed-staff ratio is 1 to 1.3–1.7. Management personnel, medicine hospitals on Organizational other technical staff, and workers account for 28–30 percent, Structure and of whom management personnel are 6–8 percent and other Manning Quota of technical staff are, 2 percent. Medical technical staff should National Traditional account for 70–72 percent of the total. TCM personnel should Chinese Medicine be more than 70 percent of all medical staff. Hospitals (trial) Township health 2011 Guidance on Quota In principle, the number of staff is equivalent to 1 percent of centers Formulation Standards the total target population; the actual number is set according for Township Health to population, traffic conditions, and financial capacity. Centers Professional and technical personnel account for not less than 90 percent of the total quota, and public health personnel for not less than 25 percent of professional and technical staff. The priority is to ensure the number of GPs. Management work, if possible, should be done by medical staff, and managerial positions should be set according to the actual situation. Community health 2006 Guidance on Quota Community health service centers are staffed with two centers Formulation Standards to three GPs and one public health physician per 10,000 for Community Health residents; TCM practitioners are provided in each community Centers health service center within the total staff quota. The GP–nurse quota should be 1 to 1, and other personnel should be no more than 5 percent of the total. The actual quota can be set according to the tasks, responsibilities, population served, service radius, and other factors. Centers serving a population of 50,000 residents or more can reduce the standards appropriately in setting the quota. Centers for disease 2014 Guidance on The staff quota for CDCs, based on the unit of provinces control and Quota Formulation (autonomous regions and municipalities), is allocated in prevention Standards for Centers accordance with their control, classification of approval, for Disease Control and coordination roles. In principle, the proportion is 1.75 and Prevention persons per 10,000 residents; provinces (autonomous regions and municipalities) of more than 500,000 square kilometers with population density fewer than 25 persons per square kilometer can set the proportion at up to 3 persons per 10,000 residents. The provincial-level CDCs exercise overall control and arrangements and dynamic adjustment of staff quotas. In the personnel structure, professional and technical personnel account for not less than 85 percent of the total, and health technical personnel for not less than 70 percent. Comprehensive management work should be done, if possible, by the professional and technical personnel, and logistics services will be gradually socialized. (Table continued next page) L e v er 6 : S trengthening the   H ealth   W or k force 275 TABLE 7A.5  Headcount quota formulation standards for Chinese health institutions  Continued Issue Scope of application date Policy directives Standards Sanitation 2010 Advice on In accordance with the principle of consistent responsibilities, inspection Implementing accountability, and appropriate personnel to perform duties institutions Supervising and safeguard work implementation, and considering the Responsibilities and factors of population, workload, service scope, and economic Strengthening Food level, 1–1.5 health supervisors should be allocated per area Safety and Sanitary with 10,000 residents. Inspection Maternal and child 1986 Quota Standards for The staff quota in MCH institutions at or above the county care stations Maternal and Child level should be set at 1 per 10,000 total population. In regions Care Stations (trial) with vast land area and fewer people as well as inconvenient transport, as well as in large cities, the ratio is 1 per 5,000; for provinces with dense populations, it is 1 per 15,000. Medical technical personnel should account for 75–80 percent of the total personnel in MCH hospitals and 80–85 percent in MCH centers. Managerial positions can be set according to the actual situation and scale: above the prefecture level, at 2–4 for MCH hospitals and 1–3 for MCH centers. Village clinics 2014 Management According to the target population, rural health services Guidelines on Village status, expected demand, and geographical conditions, in Clinics (trial) principle at least 1 village doctor per 1,000 population should be provided. Specific standards will be formulated by the provincial health administrative departments. Note: CDCs = centers for disease control and prevention; GP = general practitioner; MCH = maternal and child care institution; TCM = traditional Chinese medicine. Notes  6. There were 17,243 incidents of violence against medical staff in 2010 (Hou and others 1. “Guidance of the National Health and Family 2014). Planning Commission and Seven Other   7. The headcount quota system was created in Departments on the Establishment of a 1956 after the Working Committee on Standardized Residency Training System,” Headcount of the State Council and the NHFPC (Guowei kejia fa 2013, No. 56). Ministry of Health issued a joint policy direc- 2. “Guiding Opinions of the State Council on tive, “Principles of Headcount Management for the Establishment of a General Practitioner Hospitals and Outpatient Clinics.” Under the System in China, July 2, 2011,” State Council quota system, the Chinese government estab- (2011). lishes the management system for all public 3. This reflects a broader picture of the health institutions and defines employee-headcount workforce in China: only 28.6 percent have a standards for various PSUs (including hospitals, university or higher degree (more than five THCs, CHCs, centers for disease control and years’ medical education), and 38.8 percent prevention, sanitary inspection stations, and so have three years’ junior college or even less forth). Based on these national standards, prov- education. inces formulate provincial standards, taking 4. Compensation data from “National Health local conditions into consideration. Financial Annual Report 2012,” NHFPC,   8. “Guideline on General Hospital Organization Beijing. Quota Formulation (78),” Ministry of Health 5. Survey data from “Survey of China’s Highest- (No. 1689). Scoring Students in National University   9. “Provisional Regulations for Professional Entrance Exam, 2013,” Ai Rui Shen Research Certificate System,” Ministry of Personnel Institute, affiliated with the Ai Rui Shen (1995). China University Alumni Alliance Network 10. “Notification on Pilot of Physicians’ Dual (cuaa.net). Practice,” Ministry of Health (2009, No. 86). 276 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A 11. “Deepening Personnel Reform in Public DH (Department of Health). 2010. Equity and Medical Service Units” (2000). Excellence: Liberating the NHS. London: 12. “Personnel Management Regulations for HMSO. Public Service Units,” State Council (2014). Doran, Tim, and Martin Roland. 2010. “Lessons 13. The executive order improved “Guideline on from Major Initiatives to Improve Primary Care Strengthening Health Professional Titles in the United Kingdom.” Health Affairs 29 (5): Evaluation,” Ministry of Personnel and MoH 1023–29. doi:10.1377/hlthaff.2010.0069. (2000), which is further described in annex Freund, Tobias, Christine Everett, Peter Griffiths, 7A, table 7A.4. Catherine Hudon, Lucio Naccarella, Miranda 14. Compensation data from “National Health Laurant, and Lincoln Chen. 2015. “Skill Mix, Financial Annual Report 2012,” NHFPC, Roles and Remuneration in the Primary Care Beijing. Work forc e: W ho A re t he H e a lt hc a re 15. “2013 National Physicians Survey” (data set) Professionals in the Primary Care Teams Across of the College of Family Physicians of Canada, the World?” International Journal of Nursing the Canadian Medical Association, and the Studies 52 (3): 727–43. Royal College of Physicians and Surgeons of Gragnolati, M., M. Lindelow, and B. Couttolenc. Canada (accessed July 19, 2018), http:// 2013. Twenty Years of Health System Reform nationalphysiciansurvey.ca/surveys/2013​ in Brazil: An Assessment of the Sistema Único -survey/. de Saúde. Directions in Development Series. Washington, DC: World Bank. Gregory, R. G., and J. Borland. 1999. “Recent References D evelopment s i n P ubl ic S e c tor L abor Aran, Meltem, and Claudia Rokx, eds. 2014. Markets.” In Handbook of Labor Economics, “Turkey on the Way of Universal Health vol. 3, edited by Orley Ashenfelter and David Coverage Through the Health Transformation Card, 3573–3630. Amsterdam: Elsevier. Program (2003–2013).” Health, Nutrition, Hou, Jianlin, Catherine Michaud, Li Zhihui, Zhe and Population Discussion Paper No. 91326, Dong, Baozhi Sun, Junhua Zhang, Depin Cao, World Bank, Washington, DC. and others. 2014. “Transformation of the Boyle, Seán. 2011. “United Kingdom (England): Education of Health Professionals in China: Health System Review 2011.” Health Systems Progress and Challenges.” The Lancet 384 in Transition 13 (1): 1–486. European Health (9945): 819–27. Observatory on Health Systems and Policies, Johnson, C. D., J. Noyes, A. Haines, K. Thomas, C. World Health Organization Regional Office Stockport, and A. N. Ribas. 2013. “Learning for Europe, Copenhagen. from the Brazilian Community Health Worker Buchan, J. 2015. “Health Sector Wages in Model in North Wales.” Globalization and Context.” Background report commissioned Health 9 (1): 25. by the World Bank, Washington, DC. K ringos , Dion ne, Wien ke B oerma, Yan n Buchan, J., A. Kumar, and M. Schoenstein. 2014. Bourgueil, Thomas Cartier, Toni Dedeu, “Wage Setting in the Hospital Sector.” Health Toralf Hasvold, Allen Hutchinson, and others. Working Paper No. 77, Organisation for 2013. “The Strength of Primary Care in Economic Co-operation and Development Europe: An International Comparative Study.” (OECD), Paris. British Journal of General Practice 63 (616): Cashin, Cheryl, Y-Ling Chi, Peter Smith, Michael e742–50. Borowitz, and Sarah Thomson, eds. 2014. Kroneman, Madelon, Pascal Meeus, Dionne Sofia Paying for Performance in Health Care: Kringos, Wim Groot, and Jouke van der Zee. Implications for Health System Performance 2013. “I nter national Developments i n and Accountability. Berkshire, U.K.: World Revenues and Incomes of General Practitioners Health Organization (acting as the host orga- from 2000 to 2010.” BMC Health Services nization for, and secretariat of, the European Research 13 (1): 436. Observatory on Health Systems and Policies). Liu, X. 2015. “Health Worker Labor Market Daermmich, A. 2013. “The Political Economy of Situation Analysis in China.” Background Healthcare Reform in China: Negotiating report commissioned by the World Bank, Public and Private.” SpringerPlus 2 (1): 448. Washington, DC. L e v er 6 : S trengthening the   H ealth   W or k force 277 Marchildon, Gregory. 2013. “Canada: Health Qin, X., L. Li, and C. R. Hsieh. 2013. “Too System Review.” Health Systems in Transition Few Doctors or Too Low Wages? Labor 15 (1): 1–179. Supply of Health Care Professionals in McPake, Barbara, and others. 2015. “Wage China.” China Economic Review 24 (1): Setting in Hospital and Primary Care: Case 150–64. Studies of Five OECD Countries.” Unpublished Roland, M., B. Guthrie, and D. Thomé. 2012. background report commissioned by the World “Pri ma r y Medical Ca re in the United Bank, Washington, DC. Kingdom.” Journal of the American Board of N H DRC ( Nat ion a l H e a lt h D evelopment Family Medicine 25 (1): S6–S11. doi:10.3122/ Research Center). 2014a. “Establishing in jabfm.2012.02.110200. Public Hospitals the Compensation System Waldmeir, P. 2013. “China’s Doctors Not Part which Aligns with the Sector Features.” Study of S o c ie t y ’s E l it e .” F i n a n c i a l T i m e s , report, NHDRC, National Health and Family October 6. Planning Commission, Beijing. Woodhead, M. 2014. “How Much Does the ———. 2014b. “Monitoring and Evaluation Average Chinese Doctor Earn?” Chin a Report on Talent Planning, 2013–14.” Report, Medical News, March 31. N H DRC , National Health and Family Wranik, D. W., and M. Durier-Copp. 2010. Planning Commission, Beijing. “Physician Remuneration Methods for Family NHFPC (National Health and Family Planning Physicians in Canada: Expected Outcomes and Commission). 2004. China Health and Family Lessons Learned.” Health Care Analysis 18 (1): Planning Statistical Yearbook 2004. Bejing: 35–59. Peking Union Medical College Press Ying, Yachen. 2014. “Achieve Win-Win-Win ———. 2014. China Health and Family Planning Situation by Jointly Reforming the Drug Sector, Statistical Yearbook 2014. Beijing: Peking Health Service Sector and Health Insurance Union Medical College Press. Sector: Field Study Report on Public Hospital ———. 2015. “Report on the Development of Reform in Sanming.” August. Unpublished case Standardized Residency Training in China study. Beijing: Health Development Research (2014).” Report, NHFPC, Beijing. Center (NHFOC). Paris, V., M. Devaux, and L. Wei. 2010. “Health Yip, W. C., W. Hsiao, Q. Meng, W. Chen, and Systems Institutional Characteristics: A Survey X. Sun. 2010. “Realignment of Incentives for of 29 OECD Countries.” Health Working Paper Health-Care Providers in China.” The Lancet No. 50, OECD, Paris. 375 (9720): 1120–30. 8 Lever 7: Strengthening Private Sector Engagement in Health Service Delivery Introduction and together they accounted for 19.4 percent of all hospital beds, 14.7 percent of admis- The health care system in China has moved sions, and 12 percent of outpatient visits in from an exclusively state-run system to one 2015. The number of private primary care that is decentralized and open to private sec- facilities also has grown considerably recently, tor investment and service provision. The and is about equal to the number of public pri- foundations for private participation in the mary care facilities. production, financing, and delivery of health Limited in size but rapidly growing in goods and services were laid during the early market share, private investment is set to days of liberalization of the economy in the transform the health market in China. On 1970s. Over time, the government has the one hand, the private sector offers alter- relaxed the rules for private investment in natives. On the other hand, the development health care and explored ways to nudge it of a health care delivery system in which pro- closer to the emerging vision of the future of viders, whether public or private, have strong health care in China. Reforms since 2000— incentives to generate revenues and operating and especially in the 12th Five-Year Plan surpluses is raising ethical, legal, economic, (State Council 2011) and the State Council and political issues. Despite central policies policy directives issued in 20151—affirmed pushing for a greater role for the private sec- the role of private capital in developing tor in health care, many local governments China’s health care system and of further continue to focus their service planning and encouraging private participation in the public financing on public service providers. health sector. Whether guided by prospects of more and The rapid rise of the private sector in health better health care or by concerns related to care poses many opportunities and challenges high levels of profit making, the continuing for the government, investors, and people of development of private health care enterprise China. There are now more than 10,000 pri- in China is being watched closely by all vate hospitals in China, constituting stakeholders. 52.6 ­percent of all hospitals in the country. This chapter examines the private health Most are small (with fewer than 100 beds), sector in China and proposes a way forward 279 280 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A to strengthen private engagement in health grew 20 times and manpower 5 times com- care delivery as it relates to the government’s pared with 1949 levels (Liu 1994). objectives of reforming health service deliv- ery and improving value for money. Phase 2: Opening of Health to the The rest of the chapter is organized as Private Sector, 1979–92 follows: Remarkable economic growth and rising • “Evolution of Policies on the Private standards of living after 1978 spurred Health Sector” briefly reviews the changes increased demand for health care, which out- in government policy toward the private paced the government’s ability to expand sector’s engagement and investment in the health services. To alleviate the increasing Chinese health care system. shortages in health services, the government • “Scope and Growth of China’s Private decided to relegalize private medical practice Health Sector” describes the private sec- in 1980. This laid the foundation for the con- tor’s current size, scope, and recent version of state- and collective-owned medi- growth. cal institutions to private ownership. • “Key Challenges” discusses the challenges By 1985, private investment in health care China faces in dealing with private enter- was legally allowed. During this period, the prise in health care. Cooperative Medical System collapsed, and • “Recommendations for Strengthening many “barefoot doctors” started private prac- Private Sector Engagement” draws upon tice with fee-for-service payment (Liu and oth- experiences from within China and ers 2006; Ramesh, Wu, and He 2014).2 Organisation for Economic Co-operation and Development (OECD) countries to offer a series of actionable recommenda- Phase 3: Experimentation in the Private tions for strengthening private sector par- Sector, 1993–99 ticipation and engagement in health care. As economic reforms took hold in the 1990s and pressures to deliver more and better Evolution of Policies on the health care intensified, the State Council and Ministry of Health published several policies, Private Health Sector rules, and laws encouraging cautious devel- Policies related to the development of the pri- opment of nonstate health organizations to vate sector in health care in China have supplement the government system. New evolved over time, and can be categorized ownership categories were created for health into five phases, as discussed in detail below institutions, including state, collective, pri- (table 8.1). vate, and Chinese-foreign joint ownership. By the end of the 1990s, almost all rural ambulatory care was delivered by private Phase 1: Socialization of Medicine, providers, either in private practice or con- 1949–78 tracted by the village health post (Liu and Before the People’s Republic of China (PRC) others 2006). Some regions set up small and was founded, health services were delivered medium-size private hospitals, but the policy primarily by the private sector. During this and planning environment still did not period, China moved to socialized medicine, encourage large-scale or organized growth of and in 1963, private practice became illegal. private hospitals (Gu and Zhang 2006). Over the next three decades, the govern- ment developed the largest public health Phase 4: Gradual Growth, 2000–08 institutional network and health workforce in the world (Huang and others 2009). The During the 2000s, a series of important number of health facilities in the public sector p olicies opened up the health sector to ­ L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 281 TABLE 8.1  Evolution of policies on the private health sector in China, 1949–present Phase Policies, opinions, regulations Impact on private sector Phase 1: • 1951 MoH Policy adjusting relationships • Rationalized health services and practices Socialization between public and private health sectors between sectors during 1950s of medicine, • 1951 MoH Policy implementing cooperatives • Eventually closed all private practice by mid- 1949–78 between public and private sectors 1960s • 1963 MoH interim management measures for medical practitioners Phase 2: • 1978 3rd Plenary Session of 11th Central • Opening up of Chinese economy including the Opening up Committee health sector to private enterprises to the private • 1980 MoH regulations permitting individual • Policies lay foundation for private practice, sector, 1979–92 private practice private ownership of state- or collective-owned • 1985 State Council and MoH circular on policy medical institutions, and foreign investment issues in health care reform in health • 1989 MoH/MoFTEC rules on establishing • Reforms stress improving efficiency and quality international hospitals and clinics in China and of health services through public-private on governing competition Phase 3: • 1992 South Inspection Speech of senior leader • Chinese leadership and 14th National Congress Experimentation Deng Xiaoping and the 14th National Congress set the direction for market opening in the private • 1994 State Council document regulating • State Council and MoH respond by cautiously sector, 1993–99 administration of health institutions opening up health sector to nonstate forms of • 1998 State Council creation of medical health ownerships insurance for urban workers • Other health policies reference nonstate • 1999 MoH guidance on regional health medical institutions planning • However, government stresses dominance • 1999 MoH opinions on urban community of the public sector in health, describing the health services private sector as “supplemental” Phase 4: • 2000 State Council guidance on health system • In early part of decade, many policy reforms Gradual growth, reforms in cities and towns further open up the health sector to nonstate 2000–08 • Notice of health affairs planning in 11th Five- actors, offering unprecedented opportunities Year Plan and of issuing planning on medical • Private sector experiences some setbacks and health affairs "12th Year" • Private sector growth is limited, less than expected Phase 5: • 2009 Central Committee and State Council • 2008–09 reforms have significant impact on Renewed publish opinions on deepening health system the private sector opportunities, reform • Reforms reinforce dominance of public sector 2009–15 • 2010 MoH notice guiding pilot reforms of in all aspects of the health system but also public hospitals allow for increased private sector role in • 2010 State Council opinions encouraging financing and delivery of health services further development of private medical • In 2012, first targets for private sector set by institutions the State Council: 20 percent of all beds and • 2012 State Council document on planning outpatient volume by 2015 medical and health sectors in 12th Five-Year • Article 40 envisions nonprofit institutions as the Plan central component of a new, more diversified • 2013 State Council issues several opinions health system promoting development of health services • Draft NHFPC Five-Year Plan specifies key areas industry for private sector growth beyond hospitals • 2013 Opinions accelerating development of and opens the door for contracting the private private ownership of health care sector to deliver services on behalf of the • 2015 Main tasks of deepening medical reform government • Private sector policies accelerate, laying the foundation to redirect private sector growth and explicitly reiterating “vigorous” support for a mixed ownership health system to achieve health for all Note: MoFTEC = Ministry of Foreign Trade and Economic Co-operation; MoH = Ministry of Health; NHFPC = National Health and Family Planning Commission. 282 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A nonstate actors, offering unprecedented • In 2014, the Ministry of Health (MoH) opportunities, including for Chinese-foreign outlined specific areas for private sector jointly owned medical institutions. The growth, including premium services, niche ­ private-for-profit (PFP) and private-not-for- services in rehabilitation and geriatrics, profit (PNFP) categories came to be recog- diagnostic and laboratory services, and nized, and all private health facilities were pharmacy chains. It also supported required to register as either of the two (Gu expanding the number of joint ventures. and Zhang 2006). This opened the door to the contracting of Several PFP and PNFP hospitals opened the private sector for delivery of publicly across China during this period, which also financed health services, and to the trans- saw many hospitals converting from public to fer of business and management skills to private ownership status. At the same time, public facilities. some private hospitals that were too small to be economically viable converted to public Recent policies related to the role of the ownership. private sector in health are more specific, des- ignating subsectors in which private sector expansion is encouraged. For example, the Phase 5: Renewed Opportunities for the “Planning Layout of National Medical and Private Sector, 2009–15 Health Services System (2015–2020)” explic- To further promote the growth of the private itly notes that nonpublic hospitals could sector in the health market, the 2009 health deliver basic medical services, compete with reforms and 2010 public hospital reforms public hospitals, offer premium services to created additional “space” for the private sec- meet nonbasic needs, provide niche services tor in the financing and delivery of health ser- on rehabilitation and geriatric nursing, and vices. This encouragement has continued so on.3 The guideline encourages the nonpub- with a series of new policies, opinions, and lic sector to run health facilities, including regulations that have come out each year specialized traditional Chinese medicine since, including the following: (TCM) hospitals; rehabilitation hospitals; nursing homes (stations); and medical facili- • The 2010 State Council opinion was a ties specializing in oral diseases, geriatrics, landmark policy that prompted all prov- and noncommunicable diseases. inces and metropolitan centers to publish In 2015, the policy statement enshrined in supporting documents to grow the private the “Main Tasks of Deepening Medical health sector in ways that reflected their Reform 2015” laid the foundation for redi- local conditions. recting private health sector growth and In 2012, the State Council, in the Twelfth • explicitly reiterated commitment to a mixed- Five-Year Plan, enunciated a specific target ownership health system to achieve health for for the share of the private sector in the all.4 This “vigorous” support called for a health sector at 20 percent of all beds and greater private sector role in areas beyond con- outpatient service volume by 2015 (State ventional medical service delivery, including Council 2011). fitness and prevention. The 2015 policy state- • In 2013, Article No. 40 reinforced health ment also moved away from national targets reform goals set in 2009 and encouraged and envisioned a mixed health system in additional future development of health which “public medical institutions lead joint service delivery through the private sector. development with nonpublic medical institu- It envisioned the development of nonprofit tions” and suggested a complementary role institutions as an important part of a more with “nonprofit medical institutions as the diversified health system, to be established main body and for-profit medical institutions with “social” capital and market-driven as the supplement.” Key themes running development. through the 2015 policy statement include L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 283 “creating a level playing field,” “loosening categories commonly used internationally: restrictions over market access,” “relaxing public (as reported in official Chinese statistics price control over medical services provided also), private not-for-profit, and private for- by nonpublic medical institutions,” and profit, as described in box 8.1. strengthening “multi-site practice and markets for human resources and talent in health.” Private Sector Activity in Hospital Services The number of private sector hospitals in Scope and Growth of China’s China, especially PNFP hospitals, has grown Private Health Sector rapidly in recent years. Between 2005 and Government statistics show the current status 2012, the total number of hospitals grew by and development of key segments of the pri- 24 percent, from 18,644 to 23,170. During vate sector since 2005, when the “private” cat- this period, the number of public sector hos- egory was first established. To enable interna- pitals fell by 14 percent; PFP hospitals tional comparisons, providers and facility increased by 116 percent (from 2,971 to ownership are reported using the three 6,403); and PN FP hospitals increased BOX 8.1  Hospital ownership categories in China: Public, PNFP, and PFP Standard international reporting has three catego- To enable comparisons using the three ownership ries: public, private not-for-profit (PNFP), and private categories commonly used internationally, we used the for-profit (PFP). “Not-for-profit organizations” are ­ Chinese statistics to create a category of PNFP (miny- those established for purposes other than generating ing feiyinglixing) by removing government-owned not- profit and in which none of the organization’s income for-profits from all not-for-profits (fei yinglixing 非营 is distributed to owners or investors. Not-for-profits 利性); and a category of PFP (min­ ying yinglixing) by include private not-for-profit organizations as well as removing the PNFPs from the official statistics on non- government-owned nonprofit organizations such as state providers. The assumption is that public providers public hospitals and government-owned community such as public hospitals are registered as not-for-profit health centers (CHCs). organizations according to management form—that Chinese statistics report two categories of regis- is, there are a negligible number of public for-profits, tration status (按登记注册类型分): regardless of whether public hospitals may behave sim- ilarly to for-profit organizations. • Public (公立) government-owned providers: pub- The three-way ownership definition used in this lic hospitals and other providers, including both book is consistent with China’s own definition of pub- state-owned (central and local governments) and lic hospital (our “public” category exactly matches collective-owned providers the official statistics of gongli) and with interna- • Private providers: “people-run” (minying 民营) or tional standards differentiating not-for-profits from nongovernment (feigongli 非公立), including joint for-profits within the private category (fei gongli). ventures, cooperatives, purely private providers, It is difficult to develop a comprehensive picture of and hospitals funded from sources in Hong Kong China’s overall health sector beyond hospitals, in SAR, China; Macao SAR, China; Taiwan, China; part because grassroots providers (jiceng yiliao jigou), or foreign countries. including village clinics, do not have consistently Statistics also report a separate category of profit reported statistics on ownership or management. status or “management form” ( 按分类管理分 ): Even in 2009, the profit status for many grassroots for-profit (yinglixing 营利性 ) and not-­ f or-profit providers was listed as “unknown.” (fei yinglixing 非营利性 ), which may be either China also usually reports statistics (for example, ­ government-owned or private. for hospitals or ambulatory clinics) using a different (Box continued next page) 284 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A BOX 8.1  Hospital ownership categories in China: Public, PNFP, and PFP (continued) three-way ownership categorization, by “manage- collect data that allowed consistent categorization of ment work unit” (zhuban danwei 主办单位): private providers as PFP or PNFP. China may wish to retain China-specific cat- • Government-managed (zhengfu ban 政府办) egories that use the management work unit and • Managed by society (shehui ban 社会办) r egistration to define “government-managed ­ • Managed by individuals (siren ban 私人办). ­public”  ( 政府办公立医疗机构 ), “society-managed Although these categories provide additional public” (社会办公立医疗机构), and “society-man- information about management form, they are not aged private” (社会办民营医疗机构), which include legal definitions of ownership, residual control rights, provider ­organizations managed “by society” and or legal authority. Providers managed by “society” “by individuals.” But these ownership categories (or by individuals) may be for-profit or not-for-profit. are not the international standard, and China could The statistics on the “government-managed” category enable meaningful international comparisons by include substantially fewer facilities than the numbers also reporting according to international standard registered as public; those managed by society or by categories, just as China reports and compares individuals do not only include private providers. It health expenditures internationally according to the is not known how consistently contracted-out facili- World Health Organization (WHO) and Organisa- ties are categorized in different parts of China, that is, tion for Economic Co-operation and Development if a government-owned hospital is contracted out to (OECD) international standard categories (public be managed by a PFP firm, whether it is classified as financing and private financing) with their respective government-managed (zhengfu ban 政府办) or man- subcategories, while retaining China’s own three aged by society (shehui ban 社会办). During fieldwork categories of health financing (government, social interviews, some localities reported that they did not health insurance, and out-of-pocket). Sources: Morris and others 2014; Rudd 2011; Turner and others 2016. by 1,681 percent (from 190 to 3,383) 9 out of 10 outpatient and inpatient visits to ­(figure  8.1).5 During the same period, the per- hospitals still took place in the public sector. centage of all hospitals that were in the public Private hospitals are generally smaller than sector fell from 83 percent to 58 percent, and public hospitals in China, as is reflected in their share of inpatient admissions fell from their shares of hospital beds. In 2012, 86 per- 96 percent to 89 percent. cent of all hospitals beds were in the public PFP hospitals increased from 16 percent to sector, with the private sector accounting for 28 percent of the total, and their share of the remaining 14 percent (figure 8.4). admissions rose from 3 percent to 6 percent. Almost all private hospitals (96 percent) PNFP hospitals increased from under 1 per- have fewer than 100 beds, while only cent to 15 percent of the total, though their 60 ­percent of public hospitals are that small share of admissions grew from 1 percent to figure 8.5). Because capital investment and (­ just 5 percent, reflecting their small size management requirements are less demand- f igure 8.2). These trends make it unlikely (­ ing in small and medium-size hospitals, the that the target set by the government in 2012 private sector expands more easily in that size of a 20 percent private sector share of hospi- facility (Hou and Coyne 2008). Almost all tal beds and outpatient services, with an hospitals of more than 500 beds are operated emphasis on PNFP ownership, would have by the public sector. been met by 2015.6 About 30 percent of private sector hospi- Growth and shares of inpatient admis- tals are specialty hospitals, where higher mar- sions are similar to outpatient visits gins for advanced technologies produce huge figure 8.3). By 2010, the private health sector (­ profits, compared with only 13 percent of had taken root firmly in China, even though public sector hospitals (Hou and Coyne 2008; L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 285 FIGURE 8.1  Number of Chinese hospitals, by ownership type, 2005–12 18,000 16,000 15,483 14,309 13,850 14,000 13,384 12,000 Hospitals, number 10,000 8,000 6,403 6,000 5,096 4,038 4,000 2,971 3,383 1,972 2,000 1,341 190 0 2005 2008 2010 2012 Public Private not-for-pro t (PNFP) Private for-pro t (PFP) Source: World Bank analysis of data from China Health Statistical Yearbook 2013, Ministry of Health, Beijing. Note: For complete definition of the PNFP and PFP categories, see box 8.1 of the chapter text. Tang and others 2014). Almost all new pri- FIGURE 8.2  Market share of inpatient admissions in China, by vate hospitals are specialty hospitals because hospital ownership type, 2005 and 2012 it is easier for a specialty hospital to become a. 2005, percent b. 2012, percent eligible for insurance reimbursements than for 1 3 6 a general hospital (Yip and Hsiao 2014). 5 Much of the private sector expansion has been in urban private hospitals that deliver high-end services, such as cosmetic surgery, VIP services,7 and “checkups” that are not integrated with chronic disease case manage- ment (Yip and Hsiao 2014). Key informant interviews indicated that difficult market conditions and regulatory policies have driven investment and growth in nonessential service areas where fewer obstacles exist. At 96 89 the same time, there are also important cases Public PFP PNFP of private companies providing specialty or Source: World Bank analysis of data from China Health Statistical Yearbook 2013, Ministry of essential services (not elective services) to the Health, Beijing. general public under the social health insur- Note: PFP = public for-profit; PNFP = public not-for-profit. For complete definition of the PNFP and PFP categories, see box 8.1 of the chapter text. ance schemes. Indeed, although systematic data are lacking, existing studies do not sup- Nationally, a disproportionate percentage port the common perception that private pro- of private hospitals—41.5 percent—are in viders only serve the wealthiest patients in eastern China (table 8.2). This is also true of China. Liu and others (2006) report that other types of hospitals. Western China has there is nothing in the literature, nor in any the next largest share (32.5 percent) of pri- data, that supports the notion that the private vate hospitals, and Middle China has the sector serves primarily the better-off people; least (26 percent). in fact, the available data suggest that the pri- There is no correlation between level of vate sector extensively serves low- to middle- economic development in provinces, mea- income groups as well. sured by income per capita, and the size of 286 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 8.3  Trends in hospital use in China, by ownership type, 2005–12 a. Hospital admissions 228,866 250,000 205,254 187,381 176,890 200,000 164,911 Admissions, tens of thousands 152,650 138,677 132,003 150,000 100,000 50,000 12,827 12,469 10,955 9,674 9,419 5,560 8,593 7,867 7,686 7,163 6,711 6,395 5,389 3,207 1,948 1,091 0 2005 2006 2007 2008 2009 2010 2011 2012 b. Outpatient visits 11,331 12,000 9,708 10,000 Outpatient visits, tens of thousands 8,724 7,810 8,000 6,873 6,079 5,270 6,000 4,900 4,000 2,000 730 666 549 499 441 361 358 315 294 259 222 206 170 129 80 30 0 2005 2006 2007 2008 2009 2010 2011 2012 Public Private not-for-private (PNFP) Private for pro t (PFP) Source: World Bank analysis of data from China Health Statistical Yearbook 2013, Ministry of Health, Beijing. Note: For complete definition of the PNFP and PFP categories, see box 8.1 of the chapter text. the private health sector, measured by per- Private Sector Activity in Primary centage of outpatient visits (figure 8.6).8 The Health Care Services share of private visits is low (6 percent) in the The number of primary health care (PHC) rich provinces (such as Shanghai) and in poor facilities in the private sector—using the provinces (such as Gansu) alike; it is also Chinese definition of grassroots providers equally high in the well-off province of (jiceng yiliao jigou), which includes all non- Jiangsu (19 percent) as well as in the low- hospital facilities such as village clinics; income province of Guizhou. L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 287 FIGURE 8.4  Number of beds in Chinese hospitals, by ownership type, 2005–12 400 357.9 350 301.4 300 261 Beds, tens of thousands 250 230.1 200 150 100 50 26.6 31.7 12.6 10 17.1 15 22.4 1.3 0 2005 2008 2010 2012 Public PNFP PFP Source: World Bank analysis of data from China Health Statistical Yearbook 2013, Ministry of Health, Beijing. Note: PFP = public for-profit; PNFP = public not-for-profit. For complete definition of the PNFP and PFP categories, see box 8.1 of the chapter text. FIGURE 8.5  Composition of Chinese public and private hospitals, by number of beds, 2012 a. Public hospitals b. Private hospitals Percent Percent 8 1 2 1 10 9 26 5 7 25 61 15 11 19 0–49 beds 300–399 0–49 beds 300–399 60 96 50–99 400–499 <0.5 50–99 400–499 percent percent 100–199 500–799 100–199 500–799 200–299 800+ 200–299 800+ <0.5 Source: China Health Statistical Yearbook 2013, Ministry of Health Beijing. township health centers (THCs); clinics (zhen- of PFP PHC ­ f acilities increased by only suo, menzhenbu, yiwushi); community health percent, from 202,537 in 2008 to 220,642 9 ­ stations; and community health centers—has in 2012. also grown. The number of PNFP PHC facili- By 2012, PFP and PNFP facilities each ties increased the most, from 15,204 in 2008 accounted for 24 percent of PHC facilities, to 216,614 in 2012 (figure 8.7). The number and the combined private sector share 288 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 8.2  Regional distribution of private hospitals in China, 2012 Regional share (%) Hospital type Total number East Middle West General hospital 6,047 39.8 22.4 37.7 TCM hospital 571 48.0 32.4 19.6 Combination 187 40.1 26.7 33.2 Ethnic 23 8.7 30.4 60.0 Specialty 2,905 43.1 32.4 24.8 Nursing 53 90.6 3.8 5.7 Total 9,786 41.5 26.0 32.5 Source: Ministry of Health service statistics. Note: TCM = traditional Chinese medicine. FIGURE 8.6  Correlation between private hospital share of outpatient visits and per capita income across provinces, 2012 100,000 90,000 Shanghai 80,000 70,000 Jiangsu GDP per capita, RMB 60,000 50,000 40,000 30,000 Anhui 20,000 Gansu Guizhou 10,000 0 5 10 15 20 Share of hospital outpatient visits to private hospitals, percent Beijing Hebei Inner Mongolia Jilin Shanghai Zhejiang Fujian Shandong Hubei Guangdong Hainan Sichuan Yunnan Shaanxi Qinghai Xinjiang Tianjin Shanxi Liaoning Heilongjiang Jiangsu Anhui Jiangxi Henan Hunan Guangxi Chongqing Guizhou Tibet Gansu Ningxia Sources: China Health Statistical Yearbook 2013, table 5-2-3, Ministry of Health, Beijing. (48 percent) approached the public sector informants in China noted that the private share (52 percent) in number of facilities but sector share of the supply chain has grown not in share of outpatient visits, which is not for other goods and services, such as diag- consistently reported. nostics, drug supplies, medical equipment, Recategorization reduced the percentages and dentistry services.9 of PFP and public PHC facilities from 32 The private sector employs a greater per- percent and 66 percent, respectively, in 2008 centage of all health workers in China than to 24 percent and 52 percent in 2012. Key its 10–11 percent share of admissions and L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 289 FIGURE 8.7  Growth in Chinese PHC facilities, by ownership type, 2005–12 600,000 485,113 475,544 460,927 415,870 500,000 Number of facilities 400,000 225,949 220,462 216,614 214,833 300,000 202,537 200,000 100,000 No data 15,204 PFP and PNFP 0 2005 2008 2010 2012 Public PNFP PFP Source: World Bank analysis of data from China Health Statistical Yearbook 2013, Ministry of Health, Beijing. Note: PHC = primary health care; PFP = private for-profit; PNFP = private not-for-profit. For complete definition of the PNFP and PFP categories, see box 8.1 of the chapter text. beds (table 8.3). Of the 9.1 million health Private sector expansion needs to be personnel in China in 2012, 18 percent strengthened to address key health sector worked in the private sector and 82 percent priorities—such as greater access to health ­ in the public sector (table 8.4). The percent- care in poorer regions — or to complement age is similar for all cadres from physi- government efforts in priority areas like reha- cians to administrative staff, except that bilitation, elderly care, and integrated 37 ­p ercent of village physicians are in the management of noncommunicable diseases. ­ private sector. Provinces seek to attract private capital to remote rural areas or new peri-urban areas not already well served by government pro- viders, whereas private capital demonstrates Key Challenges an inclination to stay in cities where medical Even though laws and regulations in China resources are already plentiful. This section encourage private capital investment in the discusses these issues in detail. health sector, private providers still face many challenges entering the health market Difficulty in Developing a Shared Vision at the local level. Despite the acceleration in of the Private Sector Role recent years in the pace and scope of policies promoting private health care production The central government has enacted a rich set and delivery, the country still lacks a unified of national policies related to private sector vision for the role of private providers in engagement, but the accelerated pace of pol- improving service delivery or contributing to icy development has added to uncertainty national health objectives. And no consensus and wide variability in interpretation and has yet formed across government agencies action by provinces (Brixi and others 2013). on whether the private sector should be com- Some provinces and municipalities have hesi- plementary, supplementary, or integral to the tated, while others have taken decisive steps. public delivery system. Quantity targets have (Annex 8A, table 8A.1, summarizes major spurred private sector growth in ways not relevant national polices, grouped by 14 top- consistent with national health objectives. ics, showing examples of the regional 290 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 8.3  Health employees in China, by sector, 2010–12 Tens of thousands Hospitals Primary health carea Year Public Private Private (%) Public Private Private (%) 2010 377.0 45.8 10.8 235.0 93.2 28.4 2011 398.1 54.6 12.1 241.8 95.7 28.4 2012 428.2 65.5 13.3 248.3 95.4 27.8 Source: Ha 2014. a. “Primary health care” comprises grassroots providers using the Chinese definition (jiceng yiliao jigou), which includes all nonhospital facilities such as village clinics; township health centers; clinics (zhensuo, menzhenbu, yiwushi); community health stations; and community health centers. TABLE 8.4  Health employees in China, by type and sector, 2012 Sector Health cadre Total number (tens of thousands) Public (%) Private (%) Physician 261.6 81.5 18.5 Nurse 249.7 86.3 13.7 Pharmacist 37.7 86.4 13.6 Technician 36.4 88.2 11.8 Other 81.5 88.0 12.0 Village physician 109.4 62.7 37.3 Other technicians 31.9 88.4 11.7 Administrative 37.3 85.1 15.0 Logistics 65.4 84.6 15.4 Total 9.1 million 82.2 17.8 Source: Ha 2014. differences in their interpretation and imple- adopted the national goal, while some others, mentation.) The inconsistency is a challenge like Hunan, have set even more ambitious but also an opportunity to learn from varied targets by suggesting that by 2020, the two experiences across China. indicators should exceed 25 percent of the National policies support private invest- total.11 Changsha went further, aiming to ments in the health sector in China. For exceed 30 percent by 2020. Likewise, example, the “Opinions of the State Council Shenzhen also aims for the two indicators to on Comprehensively Scaling up Reform of exceed 30 percent by 2015 and to have two County-Level Public Hospitals” calls for to three private medical institutions that meet local authorities to “implement policies that the standards for tertiary hospitals. support and guide nongovernment invest- Overall, about two-thirds of provinces ment in hospitals.”10 and mu nicipalities encou rage private However, not all provincial governments s ector hospitals, and more than half ­ follow up with the same sense of urgency or ­ e specially encourage “high-end” (that scale. One example of differences across is, modern and high-tech) and specialty provinces is in their implementation of the hospitals. Many refer to “hospitals with target of a 20 percent private sector share of certain scale and level” and a few, such as hospital beds and inpatient admissions by Hubei, set targets for the number of beds 2015. Some provinces, such as Yunnan, have by type of hospital. L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 291 One approach to increasing the private are not well aligned with people-centered sector share in hospitals has been to convert integrated care (PCIC) and that do not com- ownership of public hospitals. For example, plement government efforts. The incentives the Zhejiang provincial government adopted created by the current fee-for-service pay- mixed state-private ownership as a transition ment system have driven both private and stage to promote private investment (“social public provider investment into nonessential capital”) while exploring ways to assure ser- areas that generate revenue (such as VIP ser- vice quality and to handle state assets. vices, overprescribing, cosmetic surgery, and However, some provinces proceeded with “checkups” not integrated with chronic dis- ownership conversions without strong mech- ease management) rather than into areas that anisms in place to prevent loss of state assets. improve health and strengthen PCIC. One example is Ku n m ing Child ren’s Special private sector incentives and con- Hospital, for which the city intended to build straints differ considerably across health a new campus with joint contributions from authorities. For example, Yunnan set up a the hospital and the government. The hospi- special fund of RMB 20 million annually in tal was taken over by CR Pharmaceutical, 2010 (increased to RMB 40 million per year which, despite state ownership, is publicly in 2014) to encourage private sector expan- listed, creating the risk of state assets being sion and to develop exemplary institutions. transferred to private investors. Yunnan offers special support to private There is much less policy clarity about the firms, including medical institutions, to role of the private sector in primary care and attract talent, such as an RMB 150,000– community services, which reflects differing 300,000 housing subsidy for key personnel. interpretations of whether the private sector Some provinces and municipalities limit should be an integral part of the health sector development of public hospitals to facilitate or have a more limited supplementary role. entry by nongovernmental institutions. For Almost all provinces encourage the private example, Shaanxi province regulations do sector to set up medical services in geo- not allow new public medical institutions, graphic areas with inadequate public ser- and they seek to shut down VIP medical ser- vices, especially new towns, rural and remote vices in public hospitals and leave this market areas, and peri-urban areas. However, pri- segment for private medical institutions. vate (and public) providers prefer to stay in Chongqing’s health administration strictly urban areas where medical resources are controls VIP medical services allowed in its already abundant. public hospitals. Shenzhen in Guangdong Many provinces encourage private invest- province spends its own budget to directly ment in specific medical areas where existing subsidize PFP medical institutions. health care resources are lacking or weak, Stakeholders have been wary of policy such as rehabilitation, nursing, geriatrics, and changes that would shift the political context chronic disease management. A few provinces of private sector engagement and reverse pre- encourage private sector delivery of basic ser- vious policies. Many local initiatives were vices at CHCs and THCs. A few stress tradi- designed to be able to reverse course without tional Chinese medicine or Hui and Tibetan much difficultly, so markets and investors medicine. Anhui province stands out in having could not be confident of long-term support a comprehensive vision for the private health that would make important complementary sector role that follows practices in OECD investments profitable, such as investing in countries. Anhui encourages the private sector human resource training. Policy uncertainty to provide basic medical services in rural areas has encouraged a short-term investment and to establish high-end specialty medical focus, which fueled suspicion that private institutions in urban areas. engagement pursued short-run profits and The overall result is that a lot of private was ill-suited to long-term development goals sector development has been in services that in the health sector. 292 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Shortcomings in the Regulatory Compliance with regulations is much Framework Overseeing Private Sector more likely, and hence enforcement much Development easier and more feasible, if those being regu- lated agree that the regulations serve a useful To grow, the private sector requires a well- and desirable purpose. The modern and more functioning governmental stewardship successful approach to regulation is to ­ mechanism—one that has the capacity of approach it in a collaborative way rather than monitoring (and shutting down, as necessary) in a coercive way. However, there is limited facilities seen to be endangering patient safety capacity in China to engage the private sector or defrauding social health insurance. in policy discussions, and there are almost no Regulatory frameworks for accountability and direct interactions between policy makers quality assurance, however, exhibit wide local and the private health sector. variations and are not uniformly strong. It is widely believed that private providers are more likely than their public counterparts to engage Difficult Market Entry in false advertising, overtreatment, or fraudu- lent billing practices, and unsurprisingly, the Public and private sector stakeholders inter- private health sector in China does not have a viewed in 2014 and early 2015 agreed that good reputation with health consumers. Even market conditions have been a significant though some private sector providers have barrier to private sector growth in the health overcome this perception and established a sector. The “Measures to Promote the reputation of higher quality of services than Growth of Nonpublic Medical Institutions” public hospitals (for example, United Family include detailed policies to relax entry barri- Hospital in Beijing and Shanghai), and some ers, expand financing channels, promote have achieved high operational efficiency (such resource flows, and improve the regulatory as the Aier Eye Hospital Group and Wuhan environment for private sector develop- Asia Heart Hospital),12 this general impres- ment.13 However, private health sector sion is unlikely to change soon, given limited growth still faces constraints, especially rela- government capacity to monitor and sanction tive to other sectors. low-quality or unqualified providers. It is not easy to open a private health facil- Qinghai is an example of the serious diffi- ity. There are multiple agencies to deal with, culties that localities with limited resources several reports to file, and many payments to and experience face in monitoring health ser- make (Zhang 2006). If foreign investors are vices. Qinghai has only two quality supervi- involved, a new facility requires approvals sion staff to monitor approximately 300,000 from the local health authority, the National medical personnel, and information systems Health and Family Planning Commission are incomplete. The province relies on bor- (NHFPC), the Ministry of Commerce, the rowed personnel from the health and family N at ion a l D e velopm e nt a nd R e for m planning commission and other related orga- Commission, and environmental protection nizations in the area to carry out special agencies; a business license from the State checks. Even in relatively rich Fujian and General Bureau of Industry and Commerce; Zhejiang provinces, regulatory frameworks and registration with the State General for accountability and quality assurance are Bureau of Tax (Glucksman and Lipson 2010). not sufficient and have wide local variation. The process can take more than a year. Localities that aggressively promote private Another barrier to expansion is the prohi- sector development, such as Suqian city in bition against consolidating finances across Jiangsu, have recognized the need for rigor- affiliates that allow the private business to ous regulatory policies to ensure that inter- offset tax liability. Each municipality wants ests of patients are protected (Chen 2015; Liu its share of taxes, and so each subsidiary is 2015; Zhou 2015). considered a separate, individual cost center. L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 293 In addition, the numbers and types of hospitals, and are making the best doctors medical equipment are highly regulated by more mobile and easier to recruit (as further local and provincial authorities in their discussed in chapter 7). Provincial govern- annual capital plans. Consistent implementa- ments have begun to experiment with multi- tion of the “Measures to Promote the Growth site license policies, but implementation var- of Nonpublic Medical Institutions” and the ies widely. Guangdong and Fujian, for “Planning Layout of National Medical and example, have adopted a pioneering set of Health Services System (2015–2020)”14 could reforms, while Qinghai (a poorer province in ease these barriers to expansion of qualified western China with low population density private providers. and shortages of health personnel) has yet to implement the new multisite practice policy, and its private health care industry continues Uneven Implementation of Reforms to face human resource shortages. Allowing Doctors to Practice at Multiple In practice, the multiple-institute—or Facilities “dual practice”— policy has not been enough Government policies and practices tend to put to create a functioning labor market for the private health care industry at a disadvan- health workers and is not working well for tage relative to the public sector and affect its either the private or the public sector. It has ability to compete fairly in the marketplace. not freed up as many staff as the private sec- One huge problem until recently was access to tor had hoped. Although the policy allows human resources, with physicians responding doctors to work in multiple institutes without to the requirement of registering and working approval, in practice public hospital adminis- in only one facility by opting to work in public trators make it difficult. As an owner of one hospitals, which offered them a known and private hospital in Beijing noted, “Dual prac- stable career track. Professional recognition, tice is an interim step, but it is not revolution- career development, salary compensation, and ary. Doctors are still tied to public hospitals.” pension benefits were all linked to the physi- A public hospital administrator explained the cian’s employment contract with a specific administrative challenge that dual practice (usually public) health facility. creates for the public sector: “It is difficult to Some hospitals have devised successful plan when a percentage of your most experi- human resources strategies to counteract enced staff are off-site each week.” This is a these disincentives. One PNFP 1,000-bed particularly intractable policy reform to tumor tertiary specialty hospital with a joint- implement. stock ownership system has explored reforms Out-of-date professional and malpractice to attract and retain talent, including a retire- liability also constrains labor movement ment compensation fund to match the com- between the sectors. Few private insurance pensation levels of public hospitals as well as companies offer limited liability insurance. a special fund for managers and physicians to Physicians and other health workers are study and practice in other hospitals. A large reluctant to move the private sector where tertiary general hospital in Zhejiang that there is no safety net against malpractice. achieved a top 10 national ranking after merging two public hospitals and converting Uneven Implementation of Reforms to PNFP status retained authorized staff- Restricting Social Health Insurance quota status for more than 90 percent of the Reimbursement to Private Hospitals hospital employees. Its employees take the same bureau examination as public hospital Private hospitals face reimbursement restric- employees, which has aided recruitment and tions from social health insurance (SHI), retention of physicians. which gives preferential treatment to public The latest reforms allow doctors to prac- facilities (Liu, Guan, and Gao 2013). In cities tice at multiple facilities, including private where private enterprises are eligible, limited 294 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A insurance funds are first directed toward to be ownership-neutral, but effective differ- public facilities; they go to private enterprises ences remain. For example, Yunnan’s reim- only if there is money left over. The latest bursement rate is tied to the class of a medi- reforms are changing this, with more and cal institution as rated by the local social more private hospitals being considered for security department, which can be lower inclusion in public health insurance networks than the rating by the local health depart- on the same terms as public hospitals. Private ment, resulting in lower reimbursements to facilities that are approved for SHI reim- some private providers. (See box 8.2 for a dis- bursement report dramatic patient increases, cussion of Yunnan Kidney Hospital, an including patients traveling from afar for spe- example of the difficult market conditions cialized health care. More than any other that even a PNFP medical institution catering policy initiative, SHI reimbursement reform to an underserved population has faced.) is creating pressure on public hospitals and The increasing adoption of budget caps or has started to enable the private sector to global budgeting experiments also implicitly compete successfully for patients. favors incumbent market participants, usu- Even so, private sector representatives ally public hospitals, over newer entrants. have complained that SHI reimbursement The global budgets are usually negotiated levels are too low. In principle, approval for directly between local health insurance man- SHI “dingdian” provider status is supposed agement offices and individual hospitals, BOX 8.2  Difficult market conditions for Yunnan Kidney Disease Hospital Founded in 1997, Yunnan Kidney Disease Hospi- • Tax burden: The local government charges enter- tal is a nonprofit specialty hospital in Kunming, the prise income tax at a rate of 25 percent of any capital city of Yunnan province. At the end of 2014, retained earnings from all medical institutions the hospital had 140 employees, including 42 physi- (including nonprofit ones). This is not conventional cians. With 160 beds and about 100 hemodialysis practice in most other provinces. machines, the hospital provides care for more than • Staffing: The hospital recruited more than 100 doc- 400 renal patients who need long-term treatment. A tors over the years, but only 32 chose to stay. Most, new campus built in 2015 added 600 beds, a kidney medical center, cardiology center, and women and especially young doctors, left for public hospitals children medical center. The hospital often admits after acquiring professional credentials. The hospi- patients with severe complications who are declined tal has been able to attract doctors from public hos- by public hospitals. It sometimes reduces or waives pitals because it gives substantial authority to fees for the poor who cannot afford to pay. doctors, including the authority to reduce and even The chairperson explained the hospital’s difficulties waive medical charges for low-income patients. with finance, tax, staffing, and insurance as follows: • Insurance reimbursement: In Yunnan, the reim- bursement rate is tied to a medical institution’s • Finance: The Guarantee Law of 1995 states that grading by the local social security depart- medical facilities belonging to a nonprofit hospital ment, which may differ from the rating system that operates for public objectives shall not be used used by the local health department. For exam- as guaranties or collateral. This makes it very difficult ple, some hospitals rated Grade-3 by the local to get bank mortgage loans. Private nonprofit hospi- health department are only rated Grade-2 by the tals have to turn to private equity (PE) or venture local social security department and are reim- capital (VC) firms for funding to expand. However, bursed at a higher rate. Regular delays in insur- once PE or VC firms control nonprofit medical insti- ance payments adds to the hospital’s financial tutions, they will no longer be considered nonprofit. pressure. Sources: Morris and others 2014; Rudd 2011; Turner and others 2016. L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 295 based on average hospital revenue in several quality and efficiency, these arrangements of previous years. This policy is not supportive convenience can undermine potential gains of private hospitals with their short history of to society from fair competition, especially if development, especially recent entrants with they promote market power and even monop- limited hospital revenue in the start-up phase. olies. This highlights the importance of devel- The current price structure gives incen- oping institutions in China to promote and tives to both public and private providers to regulate well-functioning health markets that overtreat and overprescribe. This underscores promote coordination and integration. the importance of changing payment and purchasing arrangements to promote social value from both public and private providers, Recommendations for as highlighted in recent policies reinforcing Strengthening Private Sector calls to “stop the mechanism of subsidizing Engagement health facilities with income from medicines” China has taken important steps to formulate (county hospital reform) and instructions to policies that enable the private sector to “reduce prices for examination, treatment, deepen and expand its role in the health sys- and tests with large medical equipment” but tem. However, much remains to be done to set reasonable price increases for medical ser- nudge private sector entities to deliver high- vices to “reflect the technical value of the quality, effective health services that improve staff, particularly treatment, surgery, nurs- the lives and health of China’s population. ing, beds, TCM, and other services; and The experiences of OECD countries in their establish a dynamic price adjustment mecha- struggles to reconcile expectations, policies, nism based on changes of cost and revenue ideologies, and actions may also be useful to composition.” These proposed changes to the China’s own reform process. price structure will enhance the role of SHI Three core action areas and corresponding in channeling both private sector growth and implementation strategies include the follow- existing public sector providers toward pro- ing (as summarized in table 8.5): viding better social value in health. Some private investors have entered into 1. A clear, shared vision of the private sec- public-private partnership (PPP) with public tor’s potential contribution to health sys- hospitals as a strategy to work around mar- tem goals ket difficulties, especially to gain access to 2. Regulatory and enforcement capacity to qualified physicians and specialists and to steer health services production and deliv- market their services to high- and middle- ery toward social goals income consumers. In exchange for an annual 3. A level playing field across public and pri- service fee to the public hospital and direct vate providers to promote active private payment to public doctors, a private hospital sector engagement. can use the public hospital’s name and staff. In a few cases, the public hospital is a minor- Core Action Area 1: A Clear, Shared ity shareholder in the newly created private Vision of the Private Sector’s facility. Contribution to Health System Goals There are opposing views on whether both partners benefit from these PPP arrange- Strategy 1: Identify how the private sector can ments. Almost no evaluation has been done contribute most effectively to China’s vision of the impact on patients’ well-being or on for health sector development, and publicly efficiency and costs. There are numerous endorse and articulate this shared vision anecdotes about private parties not fulfilling A clear articulation of the role of private contractual agreements and necessitating enterprise in China’s health care system is government repurchase. Instead of being important to send an unambiguous message broadly beneficial partnerships that promote to the industry and to allay any ethical or 296 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 8.5  Three core action areas and implementation strategies to strengthen private sector engagement in health service delivery Core action areas Implementation strategies 1: A clear, shared vision on • Identify areas where the private sector can contribute most effectively to China’s private sector’s contribution to broad vision for the health sector’s development; endorse the shared vision and health system goals articulate it publicly and communicate it widely • Move away from quantity targets for private sector market share and instead use a combination of supportive policies and regulatory structures to align private sector entities with health system goals • Formalize the engagement process by drafting guidelines for provincial leadership groups to implement according to local conditions 2: Regulatory and enforcement • Conduct a systematic review of existing regulations to harmonize and eliminate capacity to steer health service out-of-date and inconsistent regulations delivery toward social goals • Review the current institutional framework and empower it with skills and resources needed to govern a mixed health system with both public and private participants • Adopt policies and regulatory measures to guide private sector engagement and minimize risks associated with growth of poor-quality private providers • Implement guidelines for key regulatory functions • Strengthen regulatory capacity at different levels of the government by training provincial and municipal governments in indirect management of mixed (public and private) health systems, tools of government, and the new regulations and implementation guidelines • Allocate sufficient resources for enforcement 3: Level playing field across • Issue clear guidance on private sector planning, entry requirements, surplus use, and public and private providers community service requirements • Identify and remove access barriers related to health professionals, land use, equipment purchasing, and professional title appraisal • Introduce equal contracting standards and payment principles for public and private providers ideological concerns that may be lingering in financing (averaging 75 percent) than in ser- any sections of the government or society. vice delivery (averaging 35 percent as mea- This vision should be widely communicated sured by the share of inpatient beds and of to all stakeholders and publicly endorsed. licensed medical professionals in 15 OECD Central to this articulation are clear state- countries) (Rothgang and others 2010). ments on Private ownership is usually highest in pri- mary care and smallest in hospital services, • Whether the private sector is seen as with outpatient specialist services tending to becoming an integral part of China’s fall in between (table 8.6). Private sector ser- health system; vice delivery is often done by independently • Which forms of organization are preferred licensed physicians who contract with the (for-profit versus not-for-profit); and government or the social insurance system. • Which areas private participation is most Self-employed doctors or private organiza- sought in (such as outpatient or inpatient tions under contract with public or social care, specialist tertiary care, VIP services, insurance organizations deliver primary care pharmacies, rehabilitation therapies, elder- in most OECD countries; the general practi- care, diagnostics and laboratory services, tioners (GPs) in the United Kingdom are and prevention). a well-know n example. As table 8.6 shows, only 10 OECD countries—Chile, In most OECD countries, the government F i n l a n d , H u n g a r y, I c e l a n d , I s r a e l , plays a much larger role in health care Mexico, Portugal, Slovenia, Spain, and L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 297 TABLE 8.6  Shares of predominant mode of service provision in OECD countries, by subsector, early 2010s Primary care Outpatient specialist Hospital acute care Country Predominant form Share (%) Predominant form Share (%) Predominant form Share (%) Australia Private 89 Private Public 70 Austria Private 80+ Private Public 73 Belgium Private 75 Private 80 NFP 66 Canada Private 52 Public Public 100 Chile Public 30 Public 40 Public Czech Republic Private 90 Private 60 Public 91 Denmark Private Public Public 97 Estonia Private 77 Public 84 Public Finland Public 88 Public 90 Public 89 France Private 65 Private 90 Public 66 Germany Private 76 Private 76 Public 49 Greece Private 60 Private 60 Public 69 Hungary Public 100 Public 61 — Iceland Public 95 Private 60 Public 100 Ireland Private Public Public 88 Israel Public Public — Italy Private 65 Public Public 82 Japana Private Private NFP 74 Korea, Rep. Private Private NFP 65 Luxembourg Private Private Public 68 Mexico Public 78 Public Public 65 Netherlands Private 54 Private 50 NFP 100 New Zealand Private 52 Public Public 81 Norway Private Public 70 Public 99 Poland Private 76 Private Public 95 Portugal Public 100 Public Public 86 Slovak Republic Private 98 Private Public 60 Slovenia Public 67 Public 65 — Spain Public 97 Public 88 Public 74 Sweden Public Public 17 Public 98 Switzerland Private Private 60 Public 83 Turkey Public Public   Public 90 United Kingdom Private 100 Public 95 Public 96 United States Private   Private 30–50 NFP 60 Sources: Hospital acute care data from Paris, Devaux, and Wei 2010. Primary care and outpatient specialist service data (as of April 2014) from Question 27, OECD Health System Characteristics (HSC) survey 2012 (http://www.oecd.org/els/health-systems/characteristics.htm) and Secretariat estimates. Note: — = not available; NFP = not-for-profit; OECD = Organisation for Economic Co-operation and Development. Shading of table cells designates the following: = private; = public; and = not-for-profit. a. Japan “private clinic” data from the OECD HSC 2008 questionnaire. 298 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Turkey—provide primary care mainly countries always involve considerable effort through public clinics and salaried health to build and strengthen policies, processes, personnel.15 Any private primary care provi- and regulatory structures for “indirectly” sion is segmented: that is, it takes place out- governing health care service provision. side the core network and is paid for pri- Likewise, OECD countries also offer many vately, often out-of-pocket. examples that can inform the preferred sub- OECD countries offer many examples of sector concentration of private health service different relative sizes of for-profit and not- providers. In most OECD countries, private for-profit private enterprises in the health service provision plays a strong role in health sector (figure 8.8). Denmark, Iceland, and ­ care delivery—more so in certain subsectors New Zealand limit the core hospital network such as primary care than in others such as to government-owned hospitals.16 Most hospital services. Private providers deliver a countries have relatively little private owner- large share of services in outpatient care, ship of hospitals. Canada and the Netherlands where services are delivered by independently permit only not-for-profit hospitals in the licensed physicians who contract with the gov- private sector. In the few countries with ­ ernment or the social insurance system. This relatively large private hospital sectors, not- ­ subsector (and others such as retail pharma- for-profit hospitals dominate in Belgium, cies, laboratory services, and so on) is charac- Japan, the Republic of Korea, and the United terized by well-established quality criteria, States; the private for-profit sector plays an which makes it readily contractible and open important role only in Germany, Greece, to competition. France, Mexico, and the Slovak Republic. These characteristics do not apply to all inpatient services, and, accordingly, the share Strategy 2: Move away from quantity tar- of private provision is comparatively lower in gets, instead combining policies and regula- the hospital subsector in OECD countries. tory structures to align private sector entities Outpatient specialist services tend to fall in with health system goals between, with more public ownership than in No OECD country has used quantitative tar- primary care services and with ­ p olicies gets to expand the private sector. Policy ini- deployed that constrain operation more than tiatives that expand private activity in OECD in primary care but less than in hospitals. FIGURE 8.8  Composition of hospital beds in OECD countries, by ownership type, 2012 or latest data Share of hospital beds, percent 100 80 60 40 20 0 Slo nd Hu ia De ary Cz Es y ep a lic Tu l ey d CD Au l Gr a Au ce a ain ly Ge ce ite any s p. No k ae ga Ko tate a ar h R ni i ali lan Ita n str Re ub ee an rw rk la OE Isr ve ec to Sp rtu ng nm str Un rm Ice Po Fr a, S Po re d Public hospitals Private not-for-pro t hospitals Private for-pro t hospitals Source: OECD Health Statistics 2013 database 2013. Note: OECD = Organisation for Economic Co-operation and Development. L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 299 Differing historical trajectories are one private enterprise to focus. This clarity will reason for the different shares of not-for- help the capital markets as well as subna- profit and for-profit hospitals. The other rea- tional governments, both of which can then son is the perceived policy trade-off between develop appropriate supervisory and regula- mobilizing capital (easiest with corporate for- tory mechanisms to guide the private sector profit entry) and aligning incentives, because in ways that best complement the existing corporate for-profit entities may be more pre- public system of health production and disposed to opportunistic behavior to delivery. increase profits unless a rigorous regulation framework is in place and enforced (Deber Core Action Area 2: Regulatory and 2002). Unlike in primary care where private Enforcement Capacity to Steer Health provision is well accepted, profit-oriented Service Delivery Toward Social Goals hospitals are viewed with concern—perhaps because hospital services are inherently The private sector can potentially make a harder to measure and therefore harder to strong contribution to helping achieve purchase wisely and regulate (Preker, national health goals, but this requires a con- Harding, and Travis 2000).17 ducive policy and regulatory environment. Indeed, regulations and incentives have a Strategy 3: Formalize the engagement pro- powerful influence on all health providers, cess by drafting guidelines for provincial whether public or private. leadership groups An expanded and integrated private sector Strategy 1: Conduct a systematic review to role will need sound regulations to ensure harmonize and eliminate outdated and patient safety, assure quality, and promote inconsistent regulations efficiency to align private sector development Private provision is widespread in OECD with social value in health. As noted earlier, health systems, but providers do not operate regulations are easier to enforce if they are in totally free markets. Expanded private well accepted by those being regulated. activity in OECD countries has invariably Acceptance is more likely if the private sector come with considerable effort to build and is consulted and input solicited and if regula- strengthen policies and processes for “indi- tions are informed by a good understanding rectly” governing health care service of the perspective of nonstate actors. provision. Dialogue that enables the government and Governments use several kinds of policy private sector to understand and appreciate tools to manage mixed health systems by each other’s perspectives can build trust and influencing service providers to achieve foster collaboration. health care goals of access, financial protec- The State Council might consider drafting tion, efficiency, and cost containment guidelines for dialogue between provincial (table 8.7). China can draw on considerable leadership groups and the private health international experience in designing and ­ sector. The guidelines could explain the ratio- implementing effective regulatory mecha- nale for dialogue and engagement; engage- nisms to oversee and guide the provision of ment principles (such as mutual respect, health services, whether delivered by the pub- accountability, and results reporting); and lic sector or by private enterprises. mechanisms (forums, meetings, and work- shops) that could be implemented according Strategy 2: Review the institutional frame- to local conditions, including the range of rel- work and empower it with the skills to ­govern evant stakeholders. a mixed health system with both public and It is important that China decide and state private participants its preferences for select forms and subsectors In New South Wales, Australia, the Ministry in the health sector where it would like of Health is the regulatory authority for 300 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 8.7  Tools of health care governance or indirect policy Tool Description Health sector example Social Rules to influence individual and organizational All medical doctors must belong to a professional regulation behavior, with an administrative apparatus to association, and follow its guidelines for high- sanction noncompliance quality care. Economic Rules on prices, output, or entry and exit of firms Medical insurance reimbursement rates are set for regulation into markets, with laws and judicial processes to specific services, with maximum markup margins implement and enforce them for pharmaceuticals. Public Information to a target audience to influence Websites give public access to data on hospital information knowledge and positively alter behavior quality measures. Contracting Enlistment of private organizations to deliver Government contracts with a private provider for services on behalf of government agency to an diagnostic services in a defined geographical area. identified group via a formal agreement Accreditation A recognition of quality and other characteristics, Health facilities that meet a set of clearly specified based on well-specified criteria—usually required criteria (for staffing, quality standards, services, size, for social funding eligibility opening hours, and so on) receive accreditation, which is reviewed periodically. Taxation Tax laws that encourage or discourage certain NFP hospitals are exempt from certain taxes behaviors by individuals or organizations by if a percentage of their patients are of low diminishing or increasing tax obligations and socioeconomic status. Import duties or sales taxes affecting costs are waived or reduced on items with strong social benefits (such as vaccines). Source: Adapted from Salamon 2002. Note: NFP = not-for-profit; OECD = Organisation for Economic Co-operation and Development. privately owned and operated private health Canada and the Netherlands take a facilities across the state. Guided by the Private “­s egmented” approach, in which the core Health Facilities Act of 2007, regulation hospital provider network includes only pub- focuses on maintaining appropriate and con- lic and private nonprofit hospitals. There is sistent standards of health care and profes- an assumed degree of alignment between sional practice in private health facilities as national health goals and those of managers well as on planning for and providing compre- of public and nonprofit hospitals who are not hensive, balanced, and coordinated health ser- under pressure to generate surplus revenues. vices throughout the state. The legislation also The policy tools used to guide core networks sets requirements for licensing, including min- like these are sometimes referred to as a form imum standards for safe, appropriate, and of trust-based governance, and they are likely high-quality health care for patients in private to be ill suited to for-profit organizations health facilities. (Costain 2000; Deber, Topp, and Zakas Regulating hospital services poses a quite 2004). For-profit private hospitals are not eli- different set of challenges than regulating gible for social insurance reimbursement in outpatient services. There are two different Canada and the Netherlands. They are lightly approaches to managing mixed hospital net- regulated and exempt from regulations that works: an “integrated” approach where all seek to ensure equal access and financial sus- hospitals operate under the same regulations, tainability, because they do not have contrac- and a “segmented” approach in which differ- tual relationships with funding bodies (Busse ent regulations apply depending on owner- and others 2002). ship or whether organizations operate In France, Germany, and Switzerland, the ­ for-profit or not-for-profit. Different owner- core provider network takes an “integrated” ship and organizational structures in OECD approach, consisting of public, nonprofit, and countries and their regulatory systems offer for-profit hospitals that all operate under the useful illustrations for China to consider. same governance regime. Hospitals are L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 301 relatively independent, and corporate (for- compliance issues. Contracts typically profit) hospitals deliver a substantial share of include both case-based reimbursement rates services. These countries use mechanisms for and agreed minimum and maximum service managing capacity expansion (services and volumes.18 Rates and volumes are agreed infrastructure) that work with providers of all upon and updated in negotiations between ownership types (Ettelt and others 2008). This representatives of the purchaser organiza- approach allows health agencies to ensure ade- tions and the hospitals (Busse and others quate and equitable access, and it gives private 2002, 136–41), supported by agencies capa- hospitals a degree of certainty about expected ble of sophisticated cost analyses of case- volume of demand. based reimbursement levels.19 Box 8.3 These integrated systems have well-­ describes the indirect policy tools used in established institutional contracting pro- Germany’s fully integrated mixed hospital cesses that specify providers’ obligations and market. This governance regime was devel- constitute the main mechanism for resolving ­ oped and refined over many years. BOX 8.3  Use of indirect policy tools in Germany’s hospital market Germany has around 2,000 general hospitals, with • Mandatory quality reporting to the Federal Qual- about equal numbers of private for-profit, private ity Assurance Agency; and not-for-profit (owned by churches or other charitable • Social insurance reimbursement contracts that bodies), and public (owned by the subnational gov- require providers to have quality management sys- ernments). Even public hospitals are largely autono- tems in place and to collect and submit standard- mous of government in their day-to-day operations. ized quality information. Hospitals get most of their funding from (non- Hospitals are accredited by the Cooperation for profit) social health insurance organizations and are Transparency and Quality in Healthcare (­ established paid per case. (There is also a small private insur- by the Association of Social Health Insurers), the ance sector.) Providers are reimbursed at agreed- Chamber of Physicians, the Hospital Federation, upon rates, which are set to reflect the costs of and the Nursing Council. Accreditation is voluntary, treatment. Social health insurance organizations’ but hospitals have strong incentives to earn accredi- representative body and an association represent- tation because patients and referring physicians can ing all ­participating hospitals meet periodically to choose among hospitals. review reimbursement rates and permitted extra Policy makers also use economic regulations. charges (such as copayments; supplements for supe- The Federal Cartel Office, a key economic regula- rior accommodation) as well as to negotiate changes. tory agency, reviews proposed mergers and sales of Patients choose hospitals; funds follow those choices. hospitals and is responsible for curbing potential Prices are fixed, so hospitals compete on qual- monopoly power and excessive prices. Regional hos- ity. Those with reputations for higher-quality health pital planning bodies also estimate demand for hos- care are likely to be more successful at attract- pital services and approve new hospital facilities or ing patients and thus earn more income. However, extensions, irrespective of ownership, for inclusion patients typically have little information or exper- in the socially funded network. tise to assess quality. So, even strong competition The outcomes from the governance regime are for patients does not ensure high-quality services. not perfect. Prices are higher than in some other This problem is offset (to some degree) by the use of markets in the region. There is a tendency for pro- social regulation policy tools including viders to oversupply well-reimbursed services. But • Extensive and effective professional self-regulation overall, the system supplies a sufficient quantity of and a requirement that all physicians be members mostly high-quality medical services for an overall of their professional association; cost that the public is, on average, willing to pay. Source: Busse and Blümel 2014. 302 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Strategy 3: Adopt policies and regulatory of the policy tools used to regulate private pro- measures to guide private sector engagement viders and ensure good performance (box 8.4). and minimize risks associated with growth These illustrate the methods most frequently of poor-quality private providers used to regulate and manage p ­ rimary care China’s early initiatives experimenting with delivery in the OECD countries. hospital ownership conversions had some suc- Primary care practitioners in Denmark cesses, but many joint ventures and conversions must obtain a license to practice from the to full private ownership had such disappoint- Health and Medicines Authority, which is ing results that the government repurchased part of the Ministry of Health. They operate them or reasserted strong oversight (Chen and under contract with the regional govern- Zhang 2015). These experiences underline the ments that manage capacity and quality by importance of (a) payment incentives that regulating a list of approved practitioners nudge private providers to national goals, and who can receive public reimbursement. A (b) contractual and other institutional arrange- new practitioner cannot obtain authorization ments to regulate and monitor quality as well for public reimbursement unless a regional as to ensure accountability. There are also cau- government determines it needs an additional tionary examples of potential conflicts of inter- practice; otherwise, the practitioner can buy est when supplier firms invest in hospital a retiring doctor’s practice (and reimburse- ownership. ment authorization). Contracts are nationally The famous (or infamous) ownership standardized and define services to be pro- transformations that took place in Suqian vided, opening hours, required postgraduate city in Jiangsu after 2000 have provoked education, and payments, which are a combi- sharply conflicting opinions.20 Opponents of nation of capitation and fee-for-service. A set market-oriented approaches cite negative of social regulations also influence how pro- aspects: anecdotal evidence of supplier- fessionals practice. induced demand, overtreatment, higher med- All practitioners must follow practice ical cost increases, and compromised medical guidelines and quality standards, which are quality due to profit-driven behavior. set and continuously updated by their profes- Proponents of market-oriented approaches sional associations. They must also allow the argue that the privatizations brought vitality national Quality Unit of General Practice to and increased supply and diversity of health collect quality data and standardized user services through market competition. survey responses from their patients (elec- There has been no rigorous evaluation of tronic records automate this). The Quality the outcomes of Suqian’s reforms, and indeed Unit provides reports to each practice that few facts are available for many of the hun- track and compare their prescribing, use of dreds of ownership conversion cases and tests, consistency with treatment guidelines, other early private sector engagement efforts and patient outcomes with aggregated data in China. This is a missed opportunity for for other doctors in the area. This has valuable learning. Systematic and transparent enabled improved quality of care and patient study of the results of new policies are essen- outcomes, less hospitalization, and lower tial to be able to expand and replicate suc- total costs (Schroll and others 2012). cesses as well as to understand and redirect China’s own experience includes exam- or stop reforms that fail to achieve their ples of effective and successful contracting intended goals or that have unacceptable, for primary care services (box 8.5). A study unexpected negative effects. by Hou and others (2012) of more than 5,000 community health stations in 28 cit- Strategy 4: Implement guidelines for key ies in 2008 found that government-owned regulatory functions and private stations provided similar basic Denmark is a good example of how rigorous medical services, and private stations pro- and transparent data collection is a core part vided public health ser vices (such as L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 303 BOX 8.4  Use of indirect policy tools in primary care in Denmark Primary care and hospital services in Denmark are Practitioners’ Association, which, along with the free of charge as part of the universal, tax-funded College of General Practice, continuously develops health system. Primary care practitioners play a and updates guidelines and distributes them to all gatekeeping role: patients must be referred by a pri- primary care practitioners. The Quality Unit of mary care unit to obtain care from a specialist or General Practice, a joint body between the Associa- hospital. Practitioners are private and self-employed, tion of Regions and the General Practitioners’ Asso- and most work solo or in small group practices. ciation, coordinates quality development activities Patients can choose among nearby practices, and and establishes practice quality standards, which most patients register with a practitioner. Primary members must follow. Patient and practice data are care practitioners are contracted by regional authori- submitted to the Quality Unit as well as patients’ ties, using a nationally agreed-upon contract that responses to standardized user surveys about the is negotiated every two years between the General care they receive. The Quality Unit uses these data to Practitioners’ Association and the payers’ associa- identify unusual patterns of prescribing, diagnostic tion (Health Service Bargaining Committee). This tests, and noncompliance with guidelines and also kind of “institutional contracting” is common in to benchmark patient health outcomes, all with the health care. goal of identifying opportunities to improve quality Primary care reimbursement is sufficiently high of care and patient outcomes. that the average incomes of primary care practitioners A review of 31 European countries exploring are higher than those of many doctors working in the strength of various features of primary care hospitals. The regional government pays a capitation found Denmark to be particularly strong (Kringos for each registered patient, which provides about one- and others 2013). Danes are satisfied with primary third of income; the rest comes from ­ fee-for-service care services: 91 percent of Danish respondents to reimbursement. Patients can switch registration to a a Eurobarometer survey rated the quality of fam- different practice after three months. The “money fol- ily doctors as “good,” compared with the European lows the patient,” so practitioners’ incomes depend Union average of 84 percent (EC 2007). Of course, on their success in attracting and keeping patients. no system is perfect; a 2013 OECD review con- Because reimbursement rates are administratively cluded that primary health care in Denmark faces determined, and primary care is free of charge to challenges related to increasing public and political patients (except for small copayments for pharma- expectations around the continuity of care (OECD ceuticals), the only way practitioners can compete for 2013), and the balance between capitation and fee- patients is to offer better care. for-service is contentious. But overall, Denmark’s There are also social regulations that influence governance regime for primary care appears to work practice. All practitioners belong to the General well and align private actors with social goals. Source: Pedersen, Andersen, and Søndergaard 2012 if not otherwise indicated. establishing resident health records and including financial reporting, internal and distributing health education materials) external audits, and quality assurance when paid adequately to do so. mechanisms. This finding supports those of numerous OECD countries have encountered many studies and reviews in other countries that challenges in developing effective policy and ownership or organizational form has far regulatory structures to govern mixed-­ less effect on results than regulatory over- ownership health service delivery systems. It sight, contractual management, quality is a long-term process that requires constant monitoring, and effective incentives or dis- monitoring and improvement, and adjust- incentives. Effective harnessing of private ments in response to evolving conditions. entrepreneurship will be more scalable once China can avoid many of the pitfalls encoun- China strengthens its regulatory frame- tered in other countries as they opened up the works for governance and accountability, health sector. 304 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A BOX 8.5  Chinese experiences in purchasing community services from private and public providers Shandong province has experimented with pur- assessment mechanism links provider performance chasing community services since 2008 and has to government subsidies and involves account- established a new model of community health ser- ing firms, experts in community health care, local vice characterized by government leadership, social officials, and community representatives. Payment responsibility, market mechanisms, and purchasing includes an element of pay-for-performance in that of services from public and private providers. In the criteria for CHS performance scores include Weifang city, for example, the system involves pub- multiple dimensions of basic care, chronic disease lic bidding and competitive awarding of contracts management, community outreach, and commu- to public and private community health centers nity resident satisfaction. Organizations with poor (CHCs). performance receive informed criticism (and no per- The coverage areas of CHCs and community formance rewards). Organizations with two consec- health stations (CHSs) were defined scientifically at utive low rankings can be removed from the social the city level. Community health service centers are health insurance (SHI) community services network. set up in places with high concentrations of popula- A case study assessing Weifang’s experience tion without existing centers. Medical institutions in in contracting with private providers for urban the defined areas can bid to operate these centers. primary and preventive health services used data Large hospitals, enterprise hospitals, private medical from administrative records; a household survey of institutions, and other health resources are guided to more than 1,600 community residents in Weifang run community health care services by the municipal and in comparator City Y; and a provider survey finance bureau and health bureau. In the Municipal of more than 1,000 staff at CHSs in Weifang and Government Procurement Center open tender, bids City Y, supplemented by interviews with key infor- were won by 16 CHCs (of which 4 were private) and mants (Wang and others 2013). All CHSs in both 64 CHSs (of which 38 were private). Public commu- cities are nonprofit (public or private-not-for-profit nity health service organizations are run by public [PNFP]). Analysis revealed that government and hospitals or the Chinese Center for Disease Control private CHSs in Weifang did not differ statisti- and Prevention (China CDC); private community cally in their performance on contracted dimen- health service institutions are either purely private or sions, after controlling for size and other CHS were transferred from enterprise hospitals. characteristics. In contrast, comparison City Y To ensure a reasonable and data-based fee had lower performance and a large gap between schedule, a third-party professional organization public and private providers. This was not because estimated the costs of labor and transportation, the public providers served a more vulnerable mix communication, basic supplies, and public health of patients. In fact, residents in the communities services in accordance with the relevant standards. served by private CHSs were of lower socioeco- The payment and purchasing mechanisms were nomic status (more likely to be uninsured and to designed to be simple yet achieve accountability for report poor health) than residents in communities the multiple government objectives. The third-party served by a government-owned CHS. Source: Wang and others 2013. Achieving a regulatory environment that out-of-date or inconsistent regulations. The offers similar, predictable, and consistent research for this study found examples of conditions for public, PFP, and PNFP service underregulation in some areas and overregu- providers will be greatly helped by a system- lation in others as well as inconsistent inter- atic review of the implementation of the pretations and implementation of regulations “Measures to Promote the Growth of on PNFPs. Making both public and private Nonpublic Medical Institutions”21 and other health service organizations subject to a “sci- recent policies to harmonize and eliminate entific performance evaluation system”22 will L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 305 enhance accountability across the health sys- Estonia, Latvia, Lithuania, Poland, and tem. Indeed, many of the issues discussed in Romania.) earlier chapters that need to be resolved to enable China to develop a world-class health Strategy 6: Allocate sufficient resources for care system also undermine China’s ability to enforcement harness the private sector to help meet social New tasks may require learning new skills. goals in health. Provincial governments will need training to Many observers allege with considerable build capacity to monitor and improve qual- justification that public and private hospitals ity of all health care providers, using quality in China, regardless of profit status, act standards that are ownership-neutral. very much like for-profit hospitals. Under the Eligibility for SHI reimbursement is a strong current circumstances, public nonprofit and incentive to meet quality standards and can private not-for-profit hospitals might be con- apply equally to public and private providers. sidered illustrations of what Weisbrod (1988) It is likely that more resources will be needed referred to as “for-profits in disguise.” If by provincial and municipal governments to China wishes to foster nonprofit-oriented ensure the staff, transport, and training behavior among public and PNFP providers, needed to monitor compliance and enforce as stated in official policy documents, then regulations. incentive, policy, and regulatory structures Researchers have tried to study whether need to support that goal. OECD country competition has improved quality and regulations on ownership category (PFP ver- reduced cost in China (Liu and others 2009; sus PNFP); market size and antitrust enforce- Pan and others 2015). Consistent with the ment (preventing market dominance and/or international evidence, the available evidence monopolies under the pretense of integration from China shows that competition has the and coordination); and conflict of interest potential to improve quality and reduce cost (avoiding financial incentives for improper for specific services or cases, but the evidence behavior) could be useful examples. is mixed on broader systemwide effects (as discussed in box 8.7). This is partly Strategy 5: Strengthen regulatory capacity because the institutional context (such as by training provincial and municipal gov- payment incentives and regulatory enforce- ernments in indirect management of mixed ment capacity) matters greatly when assess- (public and private) health systems, tools of ing the impact of competition and ownership, government, and the new regulations and and these factors vary across China’s implementation guidelines provinces and over time. ­ In addition to regulations that apply to pro- Researchers and policy makers have been viders, regulations also need to enable and handicapped by the lack of systematic data support provincial governments to use indi- on quality of care and the extent of essential rect policy tools. A comprehensive review services provision, as well as a lack of mean- could identify gaps, duplication, and incon- ingful indicators of the case mix of patients sistencies in these regulations. It is also likely served by different providers. For example, that some new or additional institutions, Pan and others (2015) use emergency room skills, and resources will be needed to govern mortality as a hospital quality metric, but the a mixed health system. hospital’s case mix, and mortality after Reforms in Sweden (box 8.6) provide a patients leave the hospital, would be impor- good example of how governing a mixed tant variables to include in an actionable health system involved new tasks for existing quality evaluation. Similarly, Pan and others government agencies and the need for new (2015) use outpatient waiting times as a mea- central agencies. (Similar reforms were made sure of outpatient care “quality,” because no in Finland, Norway, and several East systematic data are available yet (or released European countries: the Czech Republic, for third-party assessment) in China to 306 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A BOX 8.6  Sweden’s “Choice” reforms: From government direct delivery to indirect governance of the primary care market Historically, Sweden’s primary care sector of pub- • The Swedish Competition Authority would evalu- licly owned health centers employing a multidisci- ate the competitive conditions of the primary care plinary workforce was not as strong as the hospital market, to ensure that public and private provid- sector. General practitioner (GP) visits per capita ers faced the same market conditions. and proportions of GPs to total physicians were • The Legal, Financial and Administrative Services low compared with Organisation for Economic Co- Agency (Kammarkollegiet) became responsible for operation and Development (OECD) peers. As in providing procurement support to the 21 county China, decision making in Sweden is decentralized; councils and developing a national website for 21 counties operate most hospitals and primary care tender documents to ensure transparent public facilities. Thus, the “Choice” health care reforms— procurement. begun in 2007 to improve access, responsiveness, • The National Board of Health and Welfare con- and quality of care while strengthening primary care tinued its previous role of supervising and moni- overall—had to consider the political ideologies of toring the quality of care and operations of all national and county governments. counties. It was left to counties to decide on the relative • The Swedish Association of Local Authorities importance of the different reform goals and to vary and Regions (an employer-interest organization) the reform design for the county accordingly. For offered legal advice, provided process support, instance, Stockholm county emphasized increas- and organized conferences for county councils ing access to care and therefore based a large share implementing the reforms. of provider payments (approximately 40 percent • The Swedish Agency for Health and Care Ser- of reimbursement) on the number of GP visits per vices Analysis (Vårdanalys), was established to capita. This was less important in all other coun- strengthen the position of patients and users by ties, where 80 percent or more of reimbursement analyzing health care and social care services from was through fixed capitation payments based on the the perspective of patients and citizens. number of people listed at each clinic. These central-level functions were key to provid- The Choice reforms started as a local initiative ing the framework within which the locally devel- in one county in 2007; two other counties volun- oped reforms operated. tarily implemented the reform later. Based on these Evaluation of the Choice reforms shows that experiences, in 2010 the central government issued access to primary care has improved. The number a new law requiring all counties to give “freedom of primary care providers increased 20 percent, of establishment” to private primary care providers contacts with GPs and other primary care provid- that fulfilled local market entry requirements, and ers per capita increased, and providers expanded provide public payment to people’s chosen providers. their phone and visit availability. Evaluations find no It was up to county councils to develop the details systemic differences between public and private pro- of the reform based on their context and objectives. viders in quality and efficiency of care. Most coun- Counties decided on eligibility criteria for market ties (except Stockholm) used capitation as the main entry, levels of public financing, whether patients basis for reimbursement, so the cost of the reform would register with a clinic actively or passively, and was predictable and cost neutral. Counties that reimbursement models. The national government set decided to strengthen primary care shifted spend- the direction but did not provide specific guidance ing from inpatient care to primary care. There has to county councils for developing their individual been heated debate, however, on whether the reform reform models. Many county councils wanted to affected equality in health care. Continuity of care learn from the pioneers’ experience, and learning does not seem to have improved in most counties. exchanges were facilitated. Comparisons of performance of counties is compli- New functions were established at the central cated by varying definitions of primary care tasks level, and various existing agencies were given new and compensation requirements. tasks: (Box continued next page) L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 307 BOX 8.6  Sweden’s “Choice” reforms: From government direct delivery to indirect governance of the primary care market (continued) County councils continue to tinker with and c ­ ompetition are met in all counties; however, in improve the Choice reforms. In Stockholm county, areas with very low population density, the choice for instance, the development department has of providers is limited by geographical distances. considered how to strengthen coordination of pri- The report also found that some counties have mary care, which reimbursement mechanisms have allowed public providers (owned by the counties) yielded best results, how to strengthen procedures to operate with a net loss, which was reprimanded for banning nonperforming providers from the since it threatens competitive neutrality. The report market, how to improve quality of care for patients recommends that county councils closely monitor with multiple and chronic diseases, and whether the financial situation of private providers to antici- competitive neutrality is being sustained. A 2015 pate and prevent adverse impacts on patients of report shows that the prerequisites for quality possible provider bankruptcy. Sources: Anell 2015; KKV 2014; Swedish NAO 2015; and interviews with Prof. Anders Anell and staff from the Stockholm county council in Stockholm, March 2015. BOX 8.7  Health care quality assurance and the role of fair competition Private sector advocates sometimes argue that quality (Bloom and others 2015; Cooper and others another layer of bureaucracy for quality assurance 2011; Gaynor, Moreno-Serra, and Propper 2013; Kes- is counterproductive because it deters innovation sler and McClellan 2000; Propper, Burgess, and Green and stifles entrepreneurial experimentation. Fair 2004). Yet other equally rigorous studies have found competition with public hospitals, they suggest, that competition can lead to quality problems (Capps will automatically raise quality standards. But this 2005; Gowrisankaran and Town 2003; Ho and Ham- is not consistent with evidence from China and ilton 2000; Propper, Burgess, and Gossage 2008). internationally. Competition with appropriate payment incentives Although public-private competition may bring (to control spending and reward quality without China benefits in innovation, years of rigorous underservicing) can contribute to social value, impact evaluations in Organisation for Economic Co-­ whereas competition under fee-for-service can lead operation and Development (OECD) countries show to a wasteful “medical arms race.” A key challenge that competition does not always improve efficiency is establishing purchasing and payment mechanisms and quality. Several studies in the United Kingdom and that align entrepreneurship with the goals of people- the United States have found that competition improves centered integrated care (PCIC) and social value. evaluate ambulatory care quality in terms of rigorous is that of the Helios hospital group process and outcomes of both private and in Germany (box 8.8), which has been public providers. Systematic evaluations will adapted by Switzerland as a national system. be extremely important to know whether reforms are having their intended effect, Core Action Area 3: Level Playing Field including whether private providers are help- across Public and Private Providers ing improve health outcomes. There are many useful examples of quality Leveling the playing field across public and assurance measures. One of the most private providers of health services promotes 308 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A BOX 8.8  Hospital quality measures at Germany’s Helios group Many health care providers in Europe use bench- examination of the hospital’s results promotes marking and related management tools to improve better performance, and they treat quality data the level and consistency of care. The Helios group for each clinic or department with the maximum systematically compares its performance with degree of transparency. results for other German hospitals, monitoring Helios also ensures that medical protocols are 30 diseases and procedures using 142 indicators appropriate and regularly updated. At monthly related to complications, death rates, and so on. All internal conferences, doctors from a selected spe- quality data are posted on the company intranet cialty present difficult cases they have encountered and provided to the health insurance organizations recently, and care protocols are reviewed in the light (Krankenkassen), which maintain quality manage- of specific cases four times a year. Twice each year, ment statistics and publish most of the data regu- the heads of specialty departments meet in groups of larly. The Helios managers have noted that close 25 to review the medical protocols. Source: Busse and Blümel 2014. active private sector engagement by helping that they are mobile and the labor market to create a competitive environment for cost- works. Many examples from within China effective service delivery. could be elaborated, especially from prov- inces such as Guangdong and Fujian, which Strategy 1: Issue clear guidance on private have been pioneers in this field. sector planning, entry requirements, surplus Additionally, China may like to continu- use, and community service requirements ously implement policies and regulations to Licensing a private health facility in China ensure the private sector of treatment similar remains variable and costly compared with to public institutions in such aspects as land public facilities. China may consider provid- use, equipment purchasing, and professional ing clearer guidance to provincial govern- title appraisal. ments on private sector planning, entry requirements, surplus use, and other commu- Strategy 3: Introduce equal contracting stan- nity service requirements as well as strictly dards and payment principles monitoring enforcement. Periodic reviews Finally, China may wish to ensure fair and and reforms of policies and regulations may even implementation across all regions of the be needed to ensure similar treatment of the recent reforms, lifting restrictions on reim- private sector and public institutions. bursing private facilities from SHI so that patients have access to more providers. Equal Strategy 2: Identify and remove access barri- contracting standards and payment principles ers related to health professionals for both public and private providers are nec- China may consider lifting the remaining essary to establish a level playing field, one in restrictions, in policy and practice, on allow- which both public and private sector health ing doctors to practice at multiple facilities so providers can grow. L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 309 Annex 8A  Supplementary Tables TABLE 8A.1  Overview and comparison of Chinese national and provincial private sector policies in health care service delivery Policy category National policy Regional interpretation and implementation Market entry Vision • There is lack of clarity in the national government’s • Without a shared vision, there is “space” for vision of the private sector role in health. provincial governments to develop their own vision. Ownership • Private sector policy focuses on the 20 percent • Interpretation of ownership types—particularly types target for hospital beds and outpatient service PNFP—varies by province. volume. • Provincial interpretation of the private sector • The PFP and PNFP categories are established in role falls into three categories: (a) deliver services business licensing and tax registration. in rural, peri-urban areas where there are weak • PNFPs are established and operated for the public’s public services; (b) establish high-end, large benefit rather than institutional profits. Any specialty hospitals; and (c) invest in specific balance or surplus in revenue after expenditures medical areas—rehabilitation, nursing, geriatrics, can only be spent on developing the organization chronic diseases—where the public sector is to improve quality, introduce advanced weak. Few mentioned basic medical services at technology, expand services, and so on. community and township levels. • PFP health care organizations are those whose • Anhui province’s vision is comprehensive: earnings are an economic return for investors. encourage the private sector to provide basic • Nongovernment services of a certain quality and medical services in rural areas and to establish high- scale are encouraged, with emphasis on shifting end specialty medical institutions in urban areas. from small to large medical groups and hospitals • To facilitate nongovernmental medical with high quality and advanced technology. institutions, some provinces limit the number • PNFPs are preferred because they are more of public hospitals and VIP services offered in aligned with the public interest. public hospitals. Public hospital • Policies propose reducing the number of public • Provinces want to convert underperforming conversion hospitals by encouraging private entities to public and state-owned hospitals. assume ownership of public hospitals in a • The reform process focuses on (a) evaluation variety of ways. and disposition of the property of the public • Initially, the government will pilot reforms of hospitals, and (b) resettlement of personnel. public hospitals in specific regions and in some • Government is concerned that national property hospitals run by state-owned enterprises (SOEs). ownership will contract and the workforce will • Hospital conversions should prioritize lose stability. prevention of national property loss and • All local governments published regulations safeguarding of employee resettlement. governing hospital conversions. • Experience has been mixed, with examples of both failure and success. Foreign capital • Foreign business actors can establish medical • Many provinces are gradually relaxing institutions in China through joint ventures or limitations on overseas investments. Some co-ownership with Chinese counterparts. are experimenting with wholly owned foreign • The percentage of foreign-owned capital is investments (WOFIs). expected to decline over time. • At present, most encourage overseas capital • Qualified foreign capital can eventually establish to establish high-end facilities and/or services wholly owned medical institutions in China. in areas where public resources are relatively • Disagreement remains on allowing WOFIs at the weaker. national level. • There are many gaps in policies related to • Overseas investors are allowed to sponsor both foreign investment. PNFP and PFP health care institutions in China. • Overseas investors are encouraged to establish health care institutions in the central and western regions of China. • Investors in Hong Kong SAR, China; Macao SAR, China; and Taiwan, China, receive preferential access in establishing foreign institutions according to relevant rules. (Table continued next page) 310 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 8A.1  Overview and comparison of Chinese national and provincial private sector policies in health care service delivery  Continued Policy category National policy Regional interpretation and implementation Land policy • Relevant departments are responsible for • Most provinces follow similar land use practices allocating land use for nonpublic medical in the private health sector. institutions and should align demand with • PNFP medical institutions cannot change the urban land plan. the use of land, but can access land through • PFP and PNFP medical institutions are subject transfers and approved sales or leases. to land policies and must be qualified licensed • PFP medical institutions can purchase land. operations. • Private medical institutions receive preferential • Private medical institutions are not allowed to treatment in Jiangsu if land is used for specific change the use of land without authorization. health services (such as nursing, geriatrics, and so on), and in Shaanxi if land is used for hospitals with 500 or more beds. Approval and • Regional health planning departments are • No two provinces issue licenses alike. There are licensing responsible for authorizing and licensing new still many unnecessary procedures and multiple PNFP institutions. “fees.” • Health departments are responsible for verifying • Multiple departments within health are that facility types, clinical specialties, bed involved, depending on the type of activity and volume, and so on are appropriate for service size. Approvals are required from multiple public capacity. agencies (Commerce, Trade, Fire, and so on). • Priority is given to private medical institutions • Key regulations governing market entry are that conform to local health priorities and plans. out of date, including the Regulations on • Once entry requirements are met, approval Administration of Medical Institutions (State should be timely. Council and MoH) and the Basic Standard of • There are no restrictions on nonpublic Medical Institutions (Trail). practice scope. • Most provinces are trying to simplify the • Foreign investors—through Chinese-foreign approval process. For example, some provinces joint ventures or cooperation—seek approval assign approval rights to different levels of from provincial health and commerce authority according to the investment scale or departments. number of beds. Others delegate approval right • Wholly foreign-owned medical institutions to the county and municipal levels. Some set seek approval from the MoH and Ministry of time limits to encourage faster turnaround. Commerce. Operations Taxes • PNFP medical institutions are entitled to • Policies and regulations governing taxes for the preferential tax policy according to private medical institutions are generous. national rules. • In many provinces, PNFP are exempt from value added tax (VAT) and other taxes (for example, land tax, building tax, and so on). • “Self-made” and “self-used” preparations are exempted from VAT, building tax, land use tax, and car and vessel tax. • In most provinces, PFP medical institutions engaged in improving medical conditions are exempted from the building tax and land use tax for the first three years of business. Some provinces extend this tax holiday beyond three years. (Table continued next page) L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 311 TABLE 8A.1  Overview and comparison of Chinese national and provincial private sector policies in health care service delivery  Continued Policy category National policy Regional interpretation and implementation Price • The prices of health services and medicine • Policies governing PNFP medical institutions, provided by PNFP medical institutions must health services, and medicine prices are comply with government pricing schemes. consistent across all the provinces because they • PFP medical institutions can set their prices of must conform to centrally set public health medical services independently. prices. • PNFP and public medical institutions are charged the same cost for electricity, water, and gas. • PFP medical institutions are free to set prices for health services and drugs. Most provinces require PFPs to offer open public access to pricing schemes. • Some provinces set price ceilings—Tianjin for specialty services and technologies, and Hunan for medicines. Medical • Eligible nonpublic medical institutions should • From the policy review, almost all local Insurance be incorporated into the designated hospital governments do not limit social health lists for the Urban Basic Medical Insurance insurance (SHI) provision to public facilities. Scheme for Employees, Retirees and Residents; The only exception is Shanghai. the New Rural Cooperative Medical Scheme • However, key informant interviews reveal this is (NRCMS); medical assistance; employment injury not the actual practice. insurance; and maternity insurance. • Reimbursement policies for eligible nonpublic medical institutions should be the same as for public institutions. • Local government authorities are prohibited from providing preferential approval of participation with insurance providers based on ownership. Large • Nonpublic medical institutions are allowed to • Similar to facility and business licensing, equipment equip facilities reasonably for practice scope, approval for large medical equipment is size, and target population. implemented in different ways across provinces. • Health departments approve the allocation • Although most provinces have relaxed the of large medical equipment based on the quotas for large medical equipment for regional or local large medical equipment nonpublic medical institutions on paper, in allocation plan. practice, provincial and local governments • Nonpublic medical institutions’ equipment continue to limit the number of approvals. needs should be given full consideration and integrated into the formulation and adjustment of the large medical equipment allocation plan. • Approval of large medical equipment is simultaneous with institutional and services license approval. • The health authority cannot limit the allocation of large medical equipment for qualified institutions that meet allocation quota. (Table continued next page) 312 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 8A.1  Overview and comparison of Chinese national and provincial private sector policies in health care service delivery  Continued Policy category National policy Regional interpretation and implementation Government • Health authorities are encouraged to contract See chapter 6 (Realigning Purchasing and purchasing eligible nonpublic medical institutions to Provider Incentives) provide health services on behalf of public health or government authorities. • Priorities for contracting include supporting health services in rural and remote areas. • Nonpublic medical institutions, especially community-level medical service centers and clinics, are encouraged to play an active role in the basic medical service system. • Nonpublic medical institutions shall receive compensation from the government in accordance with relevant regulations. • Nonpublic medical institutions should complement government initiatives during public health emergencies. Switching • PNFP medical institutions are not allowed to All provinces, except for Jiangsu province, legal status change to PFP status in principle unless there implement the policies consistently according is approval by the original administrative to terms outlined in State Council policies. authorities. • PFP medical institutions can apply to change to PNFP status by following procedures outlined in the law. After conversion, the institution must comply with applicable national price policy and tax policy. Personnel • Medical personnel sign labor contracts with • All provinces now encourage the “free flow” of nonpublic medical institutions according to the health personnel between public and private law and should be enrolled in social insurance. health facilities. Multisite • The National Health and Family Planning • Several provinces have policies prohibiting practice Commission (NHFPC), the National public health facilities from preventing multisite Development and Reform Commission (NDRC), practice. the Ministry of Human Resources and Social • Most provinces are moving to liberalize Security (MoHRSS), the State Administration of restrictions on multisite practice such as no Traditional Chinese Medicine, and the China longer requiring public hospital administrators’ Insurance Regulatory Commission issued the approval. policy opinions on doctor multisite licensed • Beijing is a pioneer in promoting multisite practice in January 2014, encouraging flow practice. of medical personnel between public and • However, in practice many provincial public nonpublic medical institutions. hospitals restrict labor movement. See chapter 7 (Strengthening the Health Workforce) for more details. Professional • Academic status, title evaluation, vocational skill status evaluation, or professional and vocational skill training of medical personnel should not be influenced by place of employment. • MoHRSS authority should include nonpublic institutions in professional training and other regular guidance according to their grade. (Table continued next page) L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 313 TABLE 8A.1  Overview and comparison of Chinese national and provincial private sector policies in health care service delivery  Continued Policy category National policy Regional interpretation and implementation Supervision • Nonpublic medical institutions are subject • All provinces stress the need to bring nonpublic of medical to State Council and MoH regulations on medical institutions into the quality system. institutions administration of medical institutions. • Provincial regulations emphasize quality and • Health authorities should include nonpublic client satisfaction as well as the importance of medical institutions in the medical quality accurate and regular reporting. control evaluation system. • However, many provincial government officials • Practice conditions of nonpublic medical interviewed explained that they do not institutions and associated personnel should be have sufficient staff or capacity to effectively checked, evaluated, and verified through daily monitor and supervise the growing number of supervision and management, site visits, and nonpublic medical providers. physician assessments. • Illegal medical practice and medical fraud is punishable by law. • Medical advertisements of nonpublic medical institutions are regulated, and false and illegal medical advertisements are forbidden. Note: MoH = Ministry of Health; MoHRSS = Ministry of Human Resources and Social Security; PFP = private for-profit; PNFP = private not-for-profit; WOFIs = wholly owned foreign investments. 314 TABLE 8A.2  Summary of policies on the social capital sponsoring medical institutions, by region Access policies Human resource policies Medical Price policies Tax Supervision insurance policies policies Encouraged Encouraged Encouraged Encouraging Permitting Simplifying Introducing Hiring Doctor Incorporating Non-profit For-profit Establishing in high- in remote/ in urban public overseas and high-end retired multisite into the hospital: medical supervisión tech and developing center hospital capital regulating talent staff licensed designated government institution: mechanism Regions characteristic regions districts ownership the approval practice hospital lists guidance exempt regions conversión procedures price / profit- from hospital: business market tax adjusted price Beijing ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Tianjin ¸ ¸ ¸ ¸ ¸ ¸ ¸ Hebei ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Shanxi ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Inner ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Mongolia Liaoning ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Jilin ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Heilongjiang ¸ ¸ Shanghai ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Jiangsu ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Zhejiang ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Anhui ¸ ¸ ¸ ¸ ¸ ¸ Fujian ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Jiangxi ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Shandong ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Henan ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ Hubei ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 315 Notes 14. “Planning Layout of National Medical and Health Services System (2015 –2020),”   1. “Planning Layout of National Medical and General Office of the State Council (Guo Ban Health Services System (2015 –2020)”; Fa 2015, No. 14) “O p i n io n s of t h e S t at e C ou n c i l o n 15. Budget allocations to public clinics are Comprehensively Scaling up Reform of t ypically based on the number of residents ­ C ou nt y- L evel P ubl ic Hospit a ls”; a nd served in a defined geographical area or catch- “Measures to Promote the Grow th of ment area. Nonpublic Medical Institutions,” General 16. OECD country hospital network data from Office of the State Council (Guo Ban Fa 2015, the OECD Health System Characteristics Nos. 14, 33, 45). Survey 2012, http://www.oecd.org/els/health​   2. Barefoot doctors are “farmers who received a -systems/characteristics.htm. short medical and paramedical training, to 17. “Measurability” refers to the precision with offer primary medical services in their rural which health care service inputs, processes, villages” (Yang and Wang 2017). outputs, and outcomes can be specified   3. “Planning Layout of National Medical and and measured (Preker, Harding, and Travis Health Services System (2015 –2020),” 2000, 782). General Office of the State Council (Guo Ban 18. In “cost and volume contracts” (widely used Fa 2015, No. 14). in the United Kingdom also), hospitals   4. “Main Tasks of Deepening Medical Reform,” receive an agreed-upon sum for a specified General Office of the State Council (Guo Ban baseline level of activity (number of cases, Fa 2015, No. 34). treatments), and beyond that level, funding   5. Hospital numbers and ownership data in the is per case, at a specified rate per case. The tables and graphs are from the Chinese Health baseline helps hospitals plan, and the maxi- Statistical Year Book 2013 (published mum volume helps authorities control through 2013 by the Ministry of Health) expenditure (see Duran and others 2005). unless otherwise indicated. The team ana- 19. For an overview of these mechanisms and lyzed the data using standard international processes, see the introduction to Busse and definitions (as explained in box 8.1) to deter- others 2011. mine the number of PFP and PNFP hospitals. 20. Suqian—a large city in Jiangsu, with very   6. Statistics for 2015 were not yet available at the li m ited gover n ment health assets and time of writing. resources to expand health services—from   7. “VIP service” refers to enhanced patient 2000, tried three different models to convert access to services for an extra charge. ownership of all township hospitals and most   8. The correlation coefficient between per capita urban hospitals: a transfer of net assets and income and private share is near zero (0.06) auction of intangible assets; a joint-stock and becomes negative when outliers Tianjin or cooperative system; and a merger and man- Jiangsu are excluded. d ator y ad m i n i s t rat ion . A mong m a ny   9. The review focused on private health services changes, the reformed hospitals gained deci- only. sion rights (including inputs, outputs, out- 10. “O p i n io n s of t h e S t at e C ou n c i l o n comes , and process /management) and Comprehensively Scaling up Reform of became residual claimants to hospital net County-Level Public Hospitals,” State Council revenues for all hospital operations, compet- (Guo Ban Ga 2015, No. 33a). ing on the market for patients. Hospitals 11. “Opinions on Promoting the Development of undertook some social functions without Health Services Industry,” Hunan Provincial government compensation (Chen 2015; Liu People’s Government ( Hunan Provincial 2015; Zhou 2015). Governmental Announcement 2012, No. 30). 21. “Measures to Promote the Grow th of 12. For a more detailed discussion of the Aier Eye Nonpublic Medical Institutions,” State Hospital Group and Wuhan Asia Heart Council General Office (Guo Ban Fa 2015, Hospital, see chapter 5. No. 45). 13. “Measures to Promote the Grow th of 22. As called for in the “Opinions on County Nonpublic Medical Institutions,” State Council Public Hospitals,” State Council General General Office (Guo Ban Fa 2015, No. 45). Office (Guo Ban Fa 2015, No. 33). 316 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A References Costain, D. 2000. “Regulating Quality & Price in Private UK Health Markets.” Oxford Policy A nell, A nders. 2015. “T he Public-Private Institute (OPI) Seminar Series “Issues in Pendulum—Patient Choice and Equity in Health Sector Regulation,” Oxford, U.K. Sweden.” New England Journal of Medicine Deber, R. 2002. “Delivering Health Care Services: 372 (1): 1–4. Public, Not-for-Profit, or Private.” Discussion Bloom, Nicholas, Carol Propper, Stephan Seiler, Paper No. 17, Commission on the Future of and John Van Reenen. 2015. “The Impact Of Health Care in Canada (Romanow Competition on M anagement Qualit y: Commission). Evidence from Public Hospitals.” Review of Deber, R., A. Topp, and D. Zakas. 2004. “Private Economic Studies 82 (2): 457–89. Delivery and Public Goals: Mechanisms for Brixi, H., Y. Mu, B. Targa, and D. Hipgrave. E n su r i ng t h at Ho spit a l s Me e t P ubl ic 2013. “Engaging Sub-national Governments in Objectives.” Unpublished background paper Addressing Health Equities: Challenges and for the World Bank, Washington, DC. Opportunities in China’s Health System Duran, Antonio, Igor Sheiman, Markus Scheider, Reform.” Health Policy and Planning 28 (8): and John Øvretveit. 2005. “Purchasers, 809–24. Providers, and Contracts.” In Purchasing to Busse, Reinhard, and Miriam Blümel. 2014. Improve Health Systems Performance, edited “Germany: Health System Review.” Health by Josep Figueras, Ray Robinson, and Elke Systems in Transition 16 (2): 1–296. Jakubowski, 187–214. London: European Busse, Reinhard, Alexander Geissler, Wilm Observatory on Health Care Systems and Quentin, and Miriam M. Wiley, eds. 2011. Policies. Diagnosis-Related Groups in Europe: Moving EC (European Commission). 2007. “Health and towards Transparency, Efficiency and Quality Long-Term Care in the European Union.” in Hospitals. Berkshire, U.K.: Open University Special Eurobarometer Report No. 283, EC, Press, McGraw-Hill International, for the World Brussels. Health Organization on behalf of the European Ettelt, Stefanie, Ellen Nolte, Sarah Thomson, and Observatory on Health Systems and Policies. Nicholas Mays. 2008. “Capacity Planning in Busse, Reinhard, Tom van der Grinten, and Per- Health Care: A Review of the International G u n n a r Sven s son. 20 02 . “Reg u lat i ng Experience.” Policy brief, World Health Entrepreneurial Behaviour in Hospitals: Organization on behalf of the European T he or y a nd P rac t ic e.” I n R eg u l a t i ng Observatory on Health Systems and Policies, Entrepreneurial Behaviour in European Copenhagen. Health-Care Systems, edited by Richard B. Gaynor, Martin, Rodrigo Moreno-Serra, and Saltman, Reinhard Busse, and Elias Mossialos, Carol Propper. 2013. “Death by Market 126–45. Buckingham, U.K.: Open University Power: Reform, Competition, and Patient Press for the World Health Organization. Outcomes in the National Health Capps, Cory Stephen. 2005. “The Quality Effects Service.”  Americ an Economic Journ al: of Hospital Mergers.” Economic Analysis Economic Policy 5 (4): 134–66. Group Discussion Paper 05- 6, Antitrust Glucksman, James, and Roberta Lipson. 2010. Division, U.S. Department of Justice. “Private Healthcare: A Tough Market to Chen, Qiulin, and Wei Zhang. 2015. “To Privatize Crack.” China Business Review 37 (1): 30–34. or Not to Privatize: The Political Economy of Gowrisankaran, Gautam, and Robert J. Town. Hospital Ownership Conversion in China.” 2003. “Competition, Payers, and Hospital Working paper, Renmin University of China Quality.” Health Services Research 38 (6 Pt 1): and Chinese Academy of Social Sciences. 1403–22. Chen, X. J. 2015. “Chen Xingjie: Falling Qiu He Gu, Edward, and Jianjun Zhang. 2006. “Health Did the Right Thing to Wipe Out All Public Care Regime Change in Urban China: Hospitals in Suqian.” Caijing.com article. Unmanaged Marketization and Reluctant Cooper, Zack, Stephen Gibbons, Simon Jones, Privatization.” Pacific Affairs 79 (1): 49–71. and Alistair McGuire. 2011. “Does Hospital Ho, Vivian, and Barton H. Hamilton. 2000. Competition Save Lives? Evidence from the “Hospital Mergers and Acquisitions: Does English NHS Patient Choice Reforms.” The Market Consolidation Harm Patients?” Journal Economic Journal 121 (554): F228–F260. of Health Economics 19 (5): 767–91. L e v er 7 : S trengthening P ri v ate S ector E ngagement in H ealth S er v ice D eli v ery 317 Hou, Wanli, Hong Fan, Jing Xu, Fang Wang, Yun The Role of Nongovernment Providers.” Chai, Hancheng Xu, Yongbin Li, Liqun Liu, Health Policy 77 (2): 212–20. Bin Wang, and Zuxun Lu. 2012. “Service OECD (Organisation for Economic Co-operation Functions of Private Community Health and Development). 2013. OECD Reviews of Stations in China: A Comparison Analysis with Health Care Quality: Denmark 2013. Raising Government-Sponsored Community Health Standards. Paris: OECD. Stations.” Journal of Huazhong University of Pan, Jay, Xuezheng Qin, Qian Li, Joseph P. Science and Technology 32 (2): 159–66. Messina, and Paul L. Delamater. 2015. “Does Hou, X., and J. Coyne. 2008. “The Emergence of Hospital Competition Improve Health Care Proprietary Medical Facilities in China.” Del iver y i n Ch i na?”  C hin a E c onomic Health Policy 88 (1): 141–51. Review 33 (C): 179–99. Huang, Cunrui, Haocai Liang, Cordia Chu, Paris, V., M. Devaux, and L. Wei. 2010. “Health Shannon Rutherford, and Qingshan Geng. Systems Institutional Characteristics: A Survey 2009. “The Emerging Role of Private Health of 29 OECD Countries.” Organisation for Care Provision in China: A Critical Analysis of Economic Co-operation and Development the Current Health System.” Asia Health Health Working Paper No. 50, OECD, Paris. Policy Program Working Paper No. 10, doi:10.1787/5kmfxfq9qbnr-en. Stanford University, Stanford, CA. Pedersen Kjeld Møller, John Sahl Andersen, and Kessler, Daniel P., and Mark B. McClellan. 2000. Jens Søndergaard. 2012. “General Practice “Is Hospital Competition Socially Wasteful?” and Primary Health Care in Denmark.” Quarterly Journal of Economics 115 (2): Journal of the American Board of Family 577–615. Medicine. 25 (Suppl 1): S34–S38. doi:10.3122/ KKV (Konkurrensverket [Swedish Competition jabfm.2012.02.110216J. Authority]). 2014. “Etablering och konkurrens Preker, A. S., A. Harding, and P. Travis. 2000. bland vårdcentraler – om kvalitetsdriven “‘Make or buy’ Decisions in the Production of kon k u r re n s o ch ekonom i sk a v i l l kor” Health Care Goods and Services: New Insights [Establishment and Competition among from I nstitutional E conomics and Health Centers: About Qualit y-Driven Organizational Theory.” Bulletin of the World Competition and Economic Conditions]. Health Organization 78 (6): 779–90. Report No. 2014:2, KKV, Stockholm. Propper, Carol, Simon Burgess, and Denise Kringos, D., W. Boerma, Y. Bourgueil, T. Cartier, Gossage. 2008. “Competition and Quality: T. Dedeu, T. Hasvold, A. Hutchinson, and oth- Evidence from the NHS Internal Market ers. 2013. “The Strength of Primary Care in 1991–9.” The Economic Journal 118 (525): Europe: An International Comparative Study.” 138–70. British Journal of General Practice 63 (616): Propper, Carol, Simon Burgess, and Katherine e742–50. Green. 2004. “Does Competition between Liu, G. L. 1994. “Privatization of the Medical Hospitals Improve the Quality of Care? Market in Socialist China: A Historical Hospital Death Rates and the NHS Internal Approach.” Health Policy 27 (2): 157–74. Market.” Journal of Public Economics 88 (7): Liu, G., H. Guan, and C. Gao. 2013. “Analysis on 1247–72. Status Quo of Private Care in China.” China Ramesh, M., Xun Wu, and Alex Jingwei He. Journal of Health Policy 6 (9): 41–46. 2014. “Health Governance and Healthcare Liu, Gordon G., Lin Li, Xiaohui Hou, Judy Xu, Reforms in China.” Health Policy and and Daniel Hyslop. 2009. “The Role of For- Planning 29 (6): 663–72. doi:10.1093/heapol/ Profit Hospitals in Medical Expenditures: czs109. Evidence from Aggregate Data in China.” Rothgang, H ., M. Cacace, L . Frisina, S. China Economic Review 20 (4): 625–33. Grimmeisen, A. Schmid, and C. Wendt. 2010. Liu, W. 2015. “Why Did Suqian Rebuild the The State and Health Care: Comparing Public Hospital Ten Years after Selling It?” OECD Countries. Basingstoke, U.K.: Palgrave Southern Weekend, infzm.com, March 5. Macmillan. http://www.infzm.com/content/108036. Salamon, Lester M., ed. 2002. The Tools of Liu, Yuanli, Peter Berman, Winnie Yip, Haocai Gover nme nt: A Guide to the Ne w Liang, Qingyue Meng, Jiangbin Qu, and Governance. New York: Oxford University Zhonghe Li. 2006. “Health Care in China: Press. 318 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Schroll, Henrik, René dePont Christensen, Janus Wang, Yan, Karen Eggleston, Zhenjie Yu, and Laust Thomsen, Morten Andersen, Søren Qiong Zhang. 2013. “Contracting with Friborg, and Jens Søndergaard. 2012. “The Private Providers for Primary Care Services: Danish Model for Improvement of Diabetes Ev idenc e f rom Urba n C h i na.” He alt h Care in General Practice: Impact of Automated Economics Review 3: 1–20. doi:10.1186​ Collection and Feedback of Patient Data.” /2191-1991-3-1. International Journal of Family Medicine 2012: Weisbrod, B. 1988. The Nonprofit Economy. Article ID 208123. doi:10.1155/2012/208123. Cambridge, MA: Harvard University Press. State Council. 2011. “The 12th Five-Year Plan for Yang, Le, and Hongman Wang. 2017. “Medical Economic and Social Development of the Education: What about the Barefoot Doctors?” People’s Republic of China (2011–2015).” The Lancet 390 (10104): 1736. Translated by the Compilation and Translation Yip, Wi n n ie , a nd Wi l l ia m H siao. 2014. Bureau, Central Committee of the Communist “Harnessing the Privatization of China’s Party of China. Beijing: Central Compilation Fragmented Health-Care Delivery.” The & Translation Press. Lancet 384 (9945): 805–818. doi:10.1016/ Sw e d i s h N AO ( N a t i o n a l A u d i t O f f i c e S0140-6736(14)61120-X. [Riksrevisionen]). 2015. “Primärvårdens Zhang, Mengzhong. 2006. “The Development s tyrning – efter behov eller efterfrågan?” ­ of Private - O w ned Hospitals in China: [Primary Care Management: As Needed or C a n   H e a l t h Po l i c i e s b e Tr a n s f e r r e d Demand?]. Review Report No. 2014:22, from Other Countries in China?” Network Swedish NAO, Stockholm. of Asia-Pacific Schools and Institutes of Tang, C., Y. Zhang, L. Chen, and Y. Lin. 2014. Public Administration and Governance, “The Growth of Private Hospitals and Their New Delhi. Health Workforce In China: A Comparison Zhou, Qiren. 2015. “Zhou Qiren Talks about with Public Hospitals.” Health Policy and Suqian Health-Care Reform.” 22ccom.net Planning 29 (1): 30–41. article. 9 Lever 8: Modernizing Health Service Planning to Guide Investment Introduction system increasingly top-heavy and have con- tributed to further escalating costs. To promote people-centered integrated care As mentioned in chapter 1, patients tend (PCIC) and ensure that the health system to go directly to hospitals even for outpatient places more emphasis on people’s needs, capi- care, and there is no gatekeeping at lower tal investment decisions must reinforce the levels. Since 2005, bed-per-population ratios strengthening of primary health care (PHC) have increased by 56 percent and admission so that the population can obtain access to rates have more than doubled, reaching levels affordable health care at the right place and higher than in most middle-income countries at the right time. This implies shifting the and approaching Organisation for Economic focus of capital investments from tertiary to Co-operation and Development (OECD) primary health care and, within tertiary care, to deepening the delivery system rather than averages. expanding it further. This is at the funda- The health sector in China is growing rap- mental basis of service planning, which is the idly. Industry analysts predict it will exceed substance of this chapter. US$1 trillion and constitute over 7 percent of More than half of all first contacts with the country’s gross domestic product (GDP) the health care delivery system in China for by 2020, which would triple 2010 levels and an illness occur in hospitals, which account make it the second largest health care mar- for 54 percent of the country’s health spend- ket in the world, behind the United States ing, according to the China Health ­Statistical (EIU 2015; Le Deu and others 2012). Annual Yearbook 2013. The number of inpatient capital investment in the health sector will discharges has grown at a rate of 12 percent potentially reach US$50 billion within the per year, and in keeping with this trend, hos- same time frame. The question of value for pital revenue has grown at an annual rate money with these resources—important even of 23.6 percent (2011–13) and is expected at existing levels—will become fundamental, to exceed RMB 4 trillion by 2017. Fueling especially as the country progresses toward this growth are the huge capital investments its commitment of affordable, equitable, and in the hospital sector, which have made the effective health care for all by 2020. 319 320 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A This chapter examines capital planning allocations, as a result of which more than strategies in China and in selected OECD 54 percent of total public health spending countries, and proposes a framework to is directed to hospitals. In comparison, lev- introduce modern service planning tech- els in OECD countries are predominantly niques in the capital investment planning below 50 percent (and averaging 38 percent). process. The next section examines the chal- Furthermore, the ratio of beds to population lenges in China’s current capital investment in most provinces has already exceeded the planning practices. Later the chapter draws average among OECD countries. upon experiences from within China and In the past decade, the number of hospitals OECD countries to offer a series of action- increased by 50 percent. During this period, able recommendations for aligning capital the number of public hospitals has decreased, investments with the service needs of the while the number of private hospitals has population served by the health system. increased, and the number of beds nation- wide has doubled. Although these levels of investments in hospitals may have been nec- Capital Investment Challenges in essary to satisfy unmet demand and growing China’s Health Sector population needs, continued expansion can Capital investment refers to the acquisition of have serious fiscal implications for the health capital assets or fixed assets such as land, clin- sector in the near future. ics, hospitals, or equipment that is expected to Addressing these problems calls for a shift be productive over many years. Two key chal- in how capital investment is planned in Chi- lenges characterize China’s current capital na’s health sector. The traditional input-based investment planning (CIP) model: (a) a lack planning system—in which decisions are not of investment planning, which contributes based on actual demand but are driven by to super scaling of investments, particularly high-level macro standards—has to give way at the hospital level; and (b) a focus on con- to an approach that considers the changing struction to expand network capacity rather epidemiological and demographic profiles than to deepen the existing infrastructure’s and emphasizes effective regionalization and capacity to better meet the population’s health integration of care with new technologies needs. With disproportionate expansion of (box 9.1). This people-centered service plan- hospital infrastructure in urban areas, the net ning approach—in which production and result is a hospital-centric system character- delivery of services are based on population ized by large, well-endowed urban hospitals needs—prioritizes public investments accord- and relatively few or poorly endowed rural ing to the burden of disease, where people ambulatory facilities. live, the kind of daily care people need, well- ness, and so on. Under this people-centered approach, CIP A Call for People-Centered Capital identifies and exploits all funding opportu- Investment Planning nities (including insurance and direct public Capital spending in the health sector among budgetary funding) to guide the develop- provinces in China accounts for 5–10 percent ment of facilities of the future and ensure that of total public spending on health. In com- excess capacity is not created that further parison, the average capital spending in exacerbates inefficiency and capital misal- OECD countries is 7 percent of total public location. It offers the opportunity to remake spending on health,1 while among European the health provider network—its design, cul- countries, capital spending varies between ture, and practices—to better meet the needs 2 percent and 6 percent of total public health of patients and families and the aspirations spending (Rechel and others 2010).2 of those that provide them with health care. However, each yuan invested in capital Consideration of the private sector’s role in also determines future recurrent expenditure meeting the population’s service needs is also L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 321 BOX 9.1  Distinguishing features of an effective health service planning approach •• Needs-based planning linked to specific health •• Increased proportion of outpatient care, including challenges primary health care, day surgeries, and day hospitals •• Long-term perspective by using demographic, •• Increase in general hospitals with fewer mono- epidemiological, and urban development plans ­ profile facilities •• Balance in real demand and supply •• Use of spatial analysis with geographic informa- •• Integrated networks that deliver services required tion system (GIS) to ensure access by catchment populations •• Integrated perspectives in terms of buildings, peo- •• Capital expenditure (CAPEX) allocations to ple, and technology provinces that correct for equity and level of •• Use of private sector as partner in reaching health deprivation goals. critical to reducing the public sector’s capital government’s efforts to reduce the pressure on requirements and optimizing utilization of prices. Further, the high level of fragmenta- existing capacity. tion and lack of transparency and account- The need to develop a CIP model driven by ability limit the effectiveness of these subsidies service planning based on population needs is as policy instruments. Public subsidies for well understood in China, and several efforts capital investments are not being fully used as have already been undertaken to improve a top-down mechanism to develop a rational, resource allocation and investment planning. patient-centered network capable of respond- Since the 1990s, regional health planning ing to the population’s changing health needs has been conducted as part of health policy while delivering value for money. reforms to improve performance of the health A key challenge, therefore, relates to coor- sector. In 1997, the National ­ Development dination and compliance with the national and Reform Commission (NDRC), ­ Ministry guidelines and standards at the provincial level of Health, and Ministry of Finance jointly to ensure that capital investments are used to issued “Guidance of Implementation of shape a people-centered provider network that Regional Health Planning,” which provided delivers the right care, at the right place, and at details on the concepts, contents, methods, the right time. Although the National Health procedures, and implementation of regional and Family Planning Commission (NHFPC)3 health planning, and demonstrated recogni- leads on setting broad planning goals, and the tion of the need for capital planning to be NDRC examines and approves the project, driven by population health needs. Local much of the investment is based on bottom- governments were expected to plan and up goals from the provinces and cities with- ­ project health care delivery according to out consideration of the service needs or the these guidelines. existing installed (public and private) capacity However, despite the efforts of the before approvals are issued. national government agencies, regional and An initial step in the right direction was local health planning has still not adopted the 2015 issuance of policy guidelines that an efficient and integrated service approach, aim to rationalize capital investments—­ and capital planning strategies continue to specifying functions and roles of health facili- favor larger hospitals. A significant share ties, staffing standards, vertical integration of all hospital investments is funded from across tiers, and horizontal integration across debt financing, off-balance sheet operations, types of care and aim to rationalize capital or land swaps, which also compromise the investments.4 However, capital investment 322 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A needs to be further integrated into regional growth rate (10 percent) over the four-year service planning and ensure that private sec- period remained significantly higher than tor capacity is considered within the targets that of other types of health facilities, while for 2020. the average annual growth rates of other facility types, including PHC facilities, became negative over time. Health Investment Planning Challenges: From 2012 to 2014, the local government A Survey of Three Provinces spent RMB 1.01 billion on capital investment An analysis of capital investment decisions projects related to prefecture-level facilities, in three provincial administrative regions— while only spending RMB 369.96 million on Sichuan province, Hubei province, and the capital investment projects related to major Tianjin municipality, which vary across disease prevention and control. Furthermore, demographic, economic development, pub- in 2012, Sichuan hospitals (including general lic resources, and health indicators—reveals or comprehensive, traditional Chinese medi- many fundamental challenges in the invest- cine, and special and national hospitals) rep- ment models being employed in the health resented RMB 40.79 billion of fixed assets, sector in China. whereas health care institutions at the basic Sichuan province. The main challenges level (including hospitals, township hospitals, for the western province of Sichuan include community health centers, and outpatient control of hospital expansion and an uneven departments) represented RMB 9.6 billion in distribution of health personnel. Figure 9.1 fixed assets. depicts the results of hospital-centric CIP in Figure 9.2 shows the heavy capital invest- Sichuan. Although hospital growth reached ment expenditures on hospitals in Sichuan a plateau in 2010–12, the average annual province relative to other types of health facilities over a four-year period: 2009–12. FIGURE 9.1  Growth rates of health facilities, Sichuan province, The distribution of health care person- 2008–09 to 2011–12 nel in Sichuan is likewise unbalanced: there are roughly twice as many health person- 12 nel per 1,000 residents in city or urban settings than in county and rural settings Percentage change in number of facilities 10 (table 9.1). To provide higher-quality care, officials in ­ S ichuan province must assess 8 how to encourage a more even distribution of health professionals according to service 6 need. Hubei province. The tendency to favor 4 larger hospitals is also apparent in the east- ern province of Hubei. Figure 9.3 shows the 2 growth rate of health facilities from 2008 to 2013. Hospitals were the only facility type to 0 maintain a steady increase after 2009. If these 2008–09 2009–10 2010–11 2011–12 trends continue, capital investment in Hubei Hospitals will remain hospital-centric, and the health Specialized public health institutions system will inadvertently attract patients who Basic level health care institutions should otherwise be seeking care at primary Other health care institutions facilities. Tianjin municipality. In the northeastern Source: Sanigest Internacional. municipality of Tianjin, health expenditure Note: “Hospitals” include general or comprehensive, traditional Chinese medicine, and special and national hospitals. “Basic-level health care institutions” include hospitals, consistently increased from 2000 to 2013, township hospitals, community health centers, and outpatient departments. and far more money went to city hospitals L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 323 FIGURE 9.2  Distribution of total capital expenditures in health, Sichuan province, 2009–12 1,000 900 800 700 RMB, millions 600 500 400 300 200 100 0 2009 2010 2011 2012 City hospitals County hospitals Community health centers Health care centers Source: Sanigest Internacional. TABLE 9.1  Distribution of health personnel, Sichuan province, by location type, 2008–12 Workers per 1,000 residents Personnel type 2008 2009 2010 2011 2012 Licensed (assistant) doctors City 2.18 2.31 2.44 2.60 2.73 County 1.06 1.09 1.11 1.17 1.23 Health technical personnel City 4.85 5.32 5.78 6.29 6.82 County 2.08 2.22 2.30 2.49 2.74 Licensed (assistant) doctors Urban 2.02 2.15 2.32 2.45 2.58 Rural 1.13 1.32 1.35 1.43 1.50 Health technical personnel Urban 4.79 5.29 5.79 6.26 6.82 Rural 2.51 2.68 2.80 3.05 3.33 Source: Sichuan Health Yearbook 2012, Sichuan Bureau of Health, Chengdu. than to any other type of health facility with government officials in charge of health (­figure 9.4). care CIP in each of the three provinces. These Survey methodology. A standardized ques- government officials represented the most rel- tionnaire was prepared to evaluate the CIP evant government departments: the Health process in the three provinces, benchmarking and Family Planning Commission (HFPC), existing practices against best practice for a the Development and Reform Commis- service-led planning model. Interviews using sion, the Finance Department, the Human a structured questionnaire were completed Resources and Social Security Department, 324 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A and the Construction Department. Hospi- addressed the following key components: tal administrators were also interviewed by knowledge of CIP, financing CIP (ability to employing the same questionnaire during site assess value for money of investments), and visits to medical facilities. The case studies project identification and evaluation. FIGURE 9.3  Growth rate of health facilities, Hubei province, Limited knowledge of capital investment 2008–09 to 2012–13 and planning techniques There was significant unanimity among the 10 three provinces’ responses to many of the interview questions regarding their knowl- edge of CIP. Decision makers acknowledged Percentage change in number of facilities 5 the importance and necessity of needs-driven CIP. The HFPC officials explained that all 0 infrastructure projects were developed in line with strategic health objectives. –5 Population health care needs were stated as the main consideration for CIP, but the inter- viewed officials gave different measures. For –10 example, in Sichuan province, population size and providers’ service radius were given as the –15 primary measure to define health care needs. In the Tianjin municipality, disease pattern –20 and incidence and services utilization (for 2008–09 2009–10 2010–11 2011–12 2012–13 example, number of visits, types of services, Hospitals medical expenses, and so on) were stated as Outpatient departments the main measures of health care needs. Specialized prevention centers In each of the three CIP components examined, the respondents stated that rates Source: Sanigest Internacional. FIGURE 9.4  Distribution of health capital expenditures, Tianjin municipality, 2000–13 400 350 300 250 RMB, millions 200 150 100 50 0 2000 2005 2010 2013 City hospitals County hospitals Community health centers Health care centers Other hospitals Source: Sanigest Internacional. L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 325 of service utilization and disease prevalence The data also indicate that in the three were the primary factors behind CIP by provinces surveyed, the practical configura- Health Bureau planners. Furthermore, only tion of health resources has been driven by the respondents from Tianjin stated that medical facilities rather than by population health objectives have a “very high” influence needs. For example, in a feasibility study on capital investment decisions; in Hubei, report of a county-level hospital in Renshou the influence was stated to be “medium” and that applied for relocation and expansion, the in Sichuan, “high.” In either case, however, number of beds, construction planning, and only population density is documented as the selection of location were initiated according key factor of consideration for capital invest- to facility needs (table 9.3). There is no indi- ments, and beds per population are used cation that the CIP considered the optimal as the key indicator for configuring health location(s) for the population to be served. sources (table 9.2). Both of these are the tra- Facility-based planning is not unique to ditional standards for CIP and bear little or Renshou; regional or local facilities often devi- no relation to service needs. ate from national standards. This indicates TABLE 9.2  Capital investment planning for health facilities, beds, and health personnel in studied Chinese provinces Personnel (per 1,000 Province Health facilities (population: health facility) Beds residents) Sichuan · 30,000–100,000 pop.: 1 community health center Per 1,000 residents: · Health technical · 300,000–500,000 pop.: 1 county-level hospital · Hospitals: 4.8 (of which public hospitals should personnel: 7 and 1 county-level traditional Chinese be more than 3.3) · Licensed (assistant) medicine hospital · Grassroots health facilities: 1.2 doctors: 2.5 · County with population over 500,000: permissible · Nurses: more than 2.8 to increase number of public hospitals, which · GPs: 2.5 should achieve 2nd-level standards · County with population over 800,000: hospitals should achieve 3rd-level standards · 1–2 million pop.: 1–2 city hospitals (2nd level); service radius of each should reach 50 kilometers Tianjin · 2,000–3,000 pop.: 1 community health For comprehensive or general hospitals: — station (village clinics) with service radius of · 3rd-level hospitals: 1,000–1,500 1.5 kilometers · 2rd-level hospitals: 500–800 · 15,000–25,000 pop.: 1 community health For traditional Chinese medicine hospitals: station (city) · 3rd-level hospitals: 800–1,500 · 50,000 pop.: 1 community health center · 2nd-level hospitals: 500–800 · 300,000–500,000 pop.: 1 city-level hospital Grassroots medical facilities (per 100,000 residents): (2nd level) · Community health centers or township hospitals: 15–30 (maximum should be lower than 50) Hubei · More than 3,000 pop.: 2 village clinics Per 1,000 residents: In general, for · 5,000–10,000 pop.: 1 community health station · In general: 4.0–4.5 (Wuhan city: 7.0–7.5) licensed doctors: · 30,000–100,000 pop.: 1 community health center · Township hospitals: 0.6–1.2 1.95–2.15 in urban area · Grassroots health facilities: 0.3–0.6 · More than 100,000 pop.: 2 community health Taking residents as standard for county-level centers hospitals: · More than 800,000 pop.: 1–2 county hospitals in · Below 100,000 population: 100–150 rural areas · 100,000–300,000 population: 200–300 · 300,000–500,000 population: 300–500 · 500,000–800,000 population: 400–600 · 800,000-1 million population: 500–800 · More than 1 million population: 800–1,000 Sources: Adapted from Configuration Standards of Health Resources in Sichuan 2008–20 and Sichuan Health Care Service Planning 2015–20; Tianjin Medical Facilities Layout Planning 2014–20; and Configuration Standards of Health Resources in Hubei 2011–15. Note: — = not available; GP = general practitioner. 326 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 9.3  Feasibility study results on relocation and expansion of Renshou County People’s Hospital, 2009 Financing estimation (RMB, tens of Reasons for relocation and expansion Project planning thousands) · Increased outpatient and inpatient needs · Total number of beds needed: 2–4 beds per 1,000 · Total investment: 16,451 by population (estimated by rate of pop. × total population · Infrastructure investment (three utilization) · Bed gap: total number of beds needed—existing years): 16,451 · Acute shortage of beds number of beds · Floating fund: 0 · Department scattered in different places · Planned number of beds for the hospital: current · Capital sources: 16,451 · Vulnerability of some buildings to number of beds (498) + bed gap × 30%) · Funds for postdisaster weather-related risks (earthquakes) · Bed dimensions: 88 square meters per bed reconstruction: 8,111 · Shortage of land for hospital expansion (following national guideline) · Asset replacement: 4,650 · Department setting restricted by limited · Construction area: 800 (number of beds) × 2,417.64 [1st year] + 2,232.36 [2nd year] space 88 = 70,400 square meters (following national + 0 [3rd year] guideline) · Loans (financial discount · Selection of location (factors considered): convenient guaranteed by government): 3,690 transportation, safety, nice environment Source: Adapted from feasibility study report of Renshou County People’s Hospital 2009. that CIP in China is often driven by medical Project identification and evaluation facility demand rather than population health Project identification and evaluation are care needs. important components of the CIP process. As in the other components of the survey, Absence of clear procedures to assess value the results were remarkably similar across for money of investments the three provinces. In each case, invest- Financing is a crucial part of capital invest- ments are prioritized based upon criteria set ment planning. Without proper financial by government policy and feasibility. Each management and planning, capital invest- official stated that existing infrastructure ment projects tend to lack direction and have is always examined in each province before a high probability of failure. All three prov- approving new projects, but there was no inces studied demonstrated an absence of mention as to how this was done. It is inter- clear management and economic principles esting to note that each province indicated to assess the potential profitability and sus- there does not exist a means to monitor, tainability of long-term investments or to evaluate, and report on infrastructure effec- determine the value for money of competing tiveness, efficiency, and sustainability. The investment projects. Although the govern- respondents also indicated that an informa- ment is moving to establish three-year bud- tion management system is not in place to geting, the NDRC investment approval pro- support monitoring and evaluation, nor are cess does not yet evaluate the sustainability of any cost-accounting tools and norms used in the investments based on projected cash flow CIP decision processes. Furthermore, each and operating expenditure or value for money respondent indicated there is no risk man- in terms of efficiency and affordability. agement process in place for capital invest- Responses from each province indicate ment projects. that all three levels contributed to the capital A lack of these components is cause for costs of health care, yet it was not specified concern. Although management and evalu- which carried a greater burden of the cost ation processes exist, if they do not employ for each province. Each official indicated the best, most up-to-date tools, they may that each facility bore the brunt of long-term not be reliable. It is reasonable to suggest costs. It was not specified how much of a bur- that efficiency issues in China’s health care den these costs are on facilities or whether capital investment may be alleviated with they are deterrents to investing in capital the proper use of management and evalua- improvement projects. tion tools. L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 327 Mismatch in procedures for administrative and epidemiological transitions associated reporting and planning clearances with an aging population, advances in medi- In China, the principle of administrative- cal technologies and pharmaceuticals, ris- affiliated management is employed in the ing public expectations, persistent health planning process, and hospitals (including inequalities, and so on. The challenge for those at the provincial, city, community, and these countries, as well as for China, is to county levels) are administratively linked reconcile health needs and expectations with with their corresponding level of govern- available resources. Several OECD countries ment. Each level of the government develops have made or are making this transition, and its own capital investment plan, while the their experiences offer important lessons provincial government makes the final deci- for China. sion in the overall planning. For example, Another major challenge that China needs Tianjin Medical University General Hospital to keep in mind concerns the lengthy time is directly affiliated with the Ministry of Edu- periods involved in planning, financing, cation instead of with the Tianjin municipal construction, and operation of new health government; however, this Level 3 hospital facilities. The interval between concept and develops its capital investment plan under commissioning of major hospitals can range the administration of the Development and from 5 to 10 years, while several more years Reform Commission of Tianjin. This creates are needed to construct the hospital (Rechel, confusion, especially because common infor- Erskine, and others 2009). This has impli- mation is not shared across different types cations both for hospital sustainability and and levels of governments involved, and proj- for responsiveness of health care delivery to ect identification and evaluation suffer in this population needs. The long time period from process. commissioning to operation can mean that Excessive capital investment in hospitals, many hospitals, when beginning to operate, particularly in urban areas, continues in Sich- do not meet the current (or future) health uan, Hubei, and Tianjin. Unless there are needs of their population. principles to guide the development of facili- Meanwhile, population needs are con- ties of the future, there is a real danger that stantly shifting. Health care demand is highly capital investment planning will simply per- sensitive to variations in the hospital’s catch- petuate the status quo, or worse yet, create ment population, including demographic excess capacity that will exacerbate the exist- changes and migration. The dynamic con- ing inefficiencies and capital misallocations. text of hospitals makes demand difficult to Planning clearances should consider the predict, both in terms of quantity and type private sector’s capacity and planned invest- of use. Furthermore, medical technologies ments in each province to ensure that the have advanced rapidly since the 1970s, with overall targets are achieved based on service a far-reaching impact on demand for clinical planning needs and population-based needs. services (Rechel, Erskine, and others 2009). Ensuring that hospitals created today can retain their relevance and value in the future Recommendations to Modernize is a profound challenge. Although providing health care goes beyond the physical asset, it Service Planning is the starting point in the delivery of sustain- China is not alone in its efforts to modify its able and high-quality clinical services at the capital investment strategy from one driven right place and the right time. This means by macro standards to one determined by ser- that the design of hospitals should be suffi- vice planning based on real population needs. ciently flexible to meet new requirements. OECD countries, although diverse, face a The following sections—each discussing number of common challenges when it comes a core action area for modernizing CIP for to capital investment for health: demographic health service delivery—provide illustrative 328 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A case studies from OECD countries from simultaneously providing levers for economic which lessons can be drawn for China. The control. Care pathways are likely to have case studies provide a variety of perspec- greatest impact on health capital investment tives on how the challenges outlined above when (a) they are applied across care settings, are being met in various contexts. They also not only to hospitals, and (b) they are backed demonstrate a variety of different approaches by appropriate systems of resource allocation to adopting a service-based capital invest- (Hindle, Dowdeswell, and Yasbeck 2004). ment strategy—from needs-based planning In addition, the case studies highlight the for care of the elderly, persons with disabili- need for comprehensive systems of capacity ties, and stroke patients to reform of regional planning and for the use of new measures of capital planning. hospital capacity that go beyond bed num- Although the case studies are diverse, bers. Bed numbers to measure hospital capac- common themes can be identified on the ity, although obsolete, are still used by many response of health services to population countries (both OECD and non-OECD). needs. For example, there is a clear trend Other countries are seeking measurements toward using systematized “care pathways” derived from systemized care pathways, or as a means of characterizing the provision at least more closely linked to actual capacity of health care services, including their links rather than bed numbers. However, this is a and integration with capital investment. Care methodology that is still in its infancy, and pathways aim to describe health care services more work is needed to develop a reliable and for specified disease syndromes and, ideally, robust characterization of hospital capac- encapsulate measurable inputs and outcomes. ity other than that based on bed ­ numbers. They provide a possible basis for translating Figure 9.5 shows how best practice in OECD ­ demographic and epidemiological trends into countries is linking the service planning with concepts that can be used for planning health the estimated investment cost requirements. capital investment. Furthermore, they offer The OECD case studies demonstrate the a means of engaging with clinicians while need for linking the operation of hospitals FIGURE 9.5  Translating health services to costs in OECD countries Service analysis Functional analysis Area analysis Cost analysis Demographic Number of Service p.a. Area and policy data units Demographic Functional BMs Area BM Cost BM service data (services/units) (m2/unit) ($/m2) Service plan Functional plan Area plan Cost plan (Oos so on) (beds, theaters, (m2) ($) and so on) Source: Interpretation of OECD 2008; Rechel, Erskine, and others 2009; Rechel, Wright, and others 2009. Note: BM = benchmark; m2 = square meters; OECD = Organisation for Economic Co-operation and Development. L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 329 with flexible financing models. The time peri- 1. A shift from the traditional input-based ods for renewing medical technologies and planning toward people-centered plan- buildings are becoming shorter, and issues of ning of capital investments based upon the life-cycle effectiveness and economic sus- region-specific epidemiological and tainability of hospitals are being recognized demographic profiles as more important. Those hospital projects 2. Engagement with all relevant stakehold- that have sought to design more adaptable ers and local communities in the planning buildings and services have also tended to process turn to more adaptable capital financing 3. Empowerment and enabling of regions models. and provinces to develop their own capi- Five core action areas and corresponding tal investment plans implementation strategies that can strengthen 4. Introduction of a Certificate of Need pro- capital investments in the health sector in gram to evaluate and approve new capital support of PCIC include the following (as investments in the health sector summarized in table 9.4): 5. Prioritization of community health projects. TABLE 9.4  Five core action areas and implementation strategies to strengthen capital investments Core action area Implementation strategies 1: Shift from traditional input- · Develop a regulatory framework in which capital investment in health is focused on improvement and based planning toward people- value centered planning · Adopt the service planning approach to capital investments, and require all future investments to be guided by an assessment of population needs · Develop a capacity planning tool that estimates financial and physical resource needs for the country’s hospital system by province, medical specialty, and level · Prepare province-level strategic plans that include 5–10 year perspectives on investment needs for infrastructure, equipment, technology, and human resource development · Integrate capital planning into a medium-term expenditure framework, and bring together planning and budgeting including consideration of private sector capacity · Create an enabling legal framework to support the new planning and governance arrangements, and support enforcement and compliance arrangements to ensure execution 2: Engagement with all · Identify different stakeholder groups and prominent community and private sector leaders, and relevant stakeholders and local formulate an engagement strategy for each stakeholder type communities in the planning · Conduct consultation sessions according to the strategy process · Require rigorous evaluation and public disclosure of all capital projects, including self-funded capital projects, financed through philanthropy or other in-kind contributions · Publish benchmark spending per bed by level of care and average bed size across provinces to ensure that standards are met 3: Empowerment and enabling · Establish provincial commissions on health investment and capital development of regions and provinces to · Prepare province-level strategic plans (master plans) that include 5–10 year perspectives on investment develop their own capital needs for infrastructure, equipment, technology, and human resource development to ensure investment plans consistency with the population’s evolving health needs · Include private capital investment in the establishment of regional health accounts that include total capital expenditures 4: Introduction of a Certificate · Require feasibility studies for all capital investments to be based on population health needs of Need program to evaluate · Require feasibility studies to demonstrate that the proposed capital investment is necessary to meet the and approve new capital identified and targeted need, considering the public and private supply in each region investments · Require all applications for new capital investments to be supported by a Certificate of Need as developed in the feasibility study 5: Prioritization of community · Earmark a percentage of provincial and city capital budgets for community projects health projects · Identify high-priority communities, and formulate multiyear community capital investment plans within the context of the new budgetary frameworks 330 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Core Action Area 1: Shift from approach reveal interesting differences. For Traditional Input-Based Planning example, the needs-based model predicted Toward People-Centered Planning a slightly increased bed capacity for the Netherlands as a whole in 2009, because China is a very large country and has a of the demographic of elderly citizens in diverse demographic and epidemiological the country. Because it takes into account the profile. An investment planning method that unequal demographic distribution of people is based on region-specific population needs in the municipalities, it is more accurate than instead of country-level averages will better the demand-based, linear approach. meet the health objectives of the population. A good example of CIP using the Hori- Several specific strategies will help reverse zon approach is in The Hague. In 2009, the the current planning logic and allow popula- needs-based model predicted that 4.3 beds tion needs to determine service planning. (per 1,000 population) would be needed for The needs-based Horizon method used care profiles 4 and 5, whereas the demand- to plan capital investment for care of the based model predicted 3.5 beds (Nauta, elderly in the Netherlands is an example of Perenboom, and Galindo Garre 2009). The CIP focused on meeting the health needs of needs-based model took into account the dif- the population while simultaneously care- ferences between The Hague and the national fully planning long-term capital investments population and calculated the need for the (box 9.2). different care profiles, giving The Hague a Comparisons between the Horizon needs- more accurate representation of capital needs based approach and the old demand-based for elderly care. BOX 9.2  Horizon’s three-step CIP model for eldercare in the Netherlands Like many countries, the Netherlands is faced sources, and patterns are distinguished using latent with the problems associated with an aging popu- class analysis. Care profiles developed from the analy- lation. To better plan for infrastructure for the ses indicate prevalent health concerns for the elderly. A elderly, Dutch health officials transitioned from lin- random population survey is then carried out to check ear, demand-based estimations to the needs-based whether the profile is reflective of the entire popula- H orizon method, which has given more accurate ­ tion. This survey is carried out yearly, ensuring that the estimations of the population, allowing for more data are updated and reflect the most current health efficient capital investment planning (CIP) centered needs of the elderly population. The number of persons on population health needs. belonging to a certain profile for a set geographical CIP for eldercare in the Netherlands has tra- area is predicted using demographics and predictions ditionally used a demand-based method, which about future demographic trends. calculates demand using the percentage of citizens Step 2 in the process is to determine the care above the age of 75. By 1998, it had become obvious needed for each profile, as each care profile states that the approach was proving to be insufficient, and a general condition of a surveyed group. This step the Netherlands moved to a needs-based approach. is relatively short, because the profiles are broken Called Horizon, this three-step approach uses mea- down and precategorized. sures of actual physical and mental disabilities to Step 3 involves ascertaining the most appropriate help plan capital investment projects (Nauta, Peren- setting of care, given the type of care needed. This boom, and Galindo Garre 2009). step assesses the needs of each profile and examines In Step 1, questionnaires and surveys are issued to the best option for the setting of care. The analy- capture personal health status, physical abilities, well- sis conducted in this step is crucial for CIP because being, and ability to cope with daily routines. Infor- it informs the plan of the care needs of the elderly mation about care issues is gathered from multiple population. Source: Nauta, Perenboom, and Galindo Garre 2009. L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 331 The needs-based Horizon model is flex- •• Conduct consultation sessions according ible and allows for greater long-term predic- to the strategy. tions. For example, the linear approach of •• Require rigorous evaluation and public the demand-based model suggests that the disclosure of all capital projects, includ- number of beds in The Hague be raised to ing self-funded capital projects, financed around 4.0 beds for 2019. The needs-based through philanthropy or other in-kind model, however, predicts that the need for contributions. beds will decrease to 3.9 by 2019. The needs- •• Publish benchmark spending per bed by based model allows for future-friendly capi- level of care and average bed size across tal investments, and it can be translated to provinces to ensure that standards are met. different sectors of health care. Horizon has proven to be a good model New South Wales (NSW), a state on the in the Netherlands for long-term, needs- east coast of Australia, has begun to imple- based capital investments for care of the ment a new capital investment method to bet- elderly. It shows that making needs-based ter meet the needs of its population of people projections is possible and, in fact, may give with disabilities. Known as the Sector Plan- more accurate estimations for capital invest- ning Framework, its flexible approach can ment projects. be modified to fit any population subgroup. China is faced with an aging popula- Among its key features, it places local com- tion, so such an approach could be used to munities, including people with disabilities, help direct CIP toward meeting the needs their families, and caretakers at the center of the elderly population in China. A plan- of the planning process and as joint parties ning method such as Horizon would allow in the planning process. It also helps the state Chinese health officials to plan capital deliver on its commitments to local commu- investments based upon a location’s unique nities in ways that best suit each commu- demographics instead of generalizing health nity. It recognizes that each community has needs across the vast country. This will allow unique health needs and that capital invest- for personalized health capital investments ments cannot be made in a “one size fits all” designed to meet specific health needs; as a manner if they are to meet all the disparate result, the health needs of each unique region health needs. will be met more readily by the capital, a step NSW is currently transitioning to this in the right direction toward improving pop- new approach (figure 9.6). Previously, capi- ulation health. tal planning for disability services was a centrally driven, program-based planning method that focused on service outputs and Core Action Area 2: Engagement with defined service models, driven by agency- All Relevant Stakeholders and Local based priorities (NSW Government 2011). Communities in the Planning Process This new Sector Planning Framework Involving all relevant stakeholders in the approach will contribute to a service system planning process, especially the target popu- that does the following; lation and the private sector, allows for capi- tal investment decisions to be made in ways •• Is people-centered, with the population that simultaneously meet health needs as of people with disabilities, their families, well as policy requirements. Key action steps and caretakers being the focus of deci- include the following: sion making and allowing them to provide input about the support they receive, who •• Identify different stakeholder groups and provides the support, and how and when prominent community and private sec- they receive it
 tor leaders, and formulate an engagement •• Is based on a lifespan approach that strategy for each stakeholder type empowers people with disabilities to be 332 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 9.6  Delivering a new approach to health sector planning Further, it will feed into the broader Fam- in New South Wales, Australia ily and Consumer Services (FACS) planning framework and ensure that locally based ADHC’s vision, statewide plans, strategies are reflected in the wider FACS and and corporate priorities in the NSW government’s plans to support the population of people with disabilities Current approach Future approach (NSW Government 2011). • Service and • Responses that are This approach allows for open dialogue program based person centered among all different levels of planning. Robust • Centrally directed and responsive to research and strong community involvement (top down) individual needs allow for investment plans to incorporate • Focused on • Guided by local or resources or funding community priorities projects that best fit the health needs of any allocation (bottom up) given population. Further, it helps tailor • Supported by service • Concerned with capital investments to the unique needs of agreements investment in social individual communities, contributing to the • Risk minimizing capital • Shared responsibility development of service-based investment and collaboration decisions. NSW has recognized that this • Strengths based method of planning is not limited to capital planning only for people with disabilities; it is an approach that can be modified for any Local sector plans given population. NSW’s Sector Planning Framework offers many attractive options for China. It offers a Source: NSW Government 2011. Note: ADHC = Ageing, Disability, and Home Care. way for China to incorporate each planning level into the investment planning process, actively involved in planning and design- allowing for capital investment decisions ing their own support arrangements that meet health needs and policy require- •• Focuses on maximizing the experiences and ments. The Sector Planning Framework is opportunities of people with disabilities; 
 designed to achieve coordination and align- •• Promotes economic participation and ment in the priorities among governments, employment opportunities agencies, providers, and communities, and •• Actively develops the capacity and social it builds cross-agency and public-private capital of the nongovernmental organiza- partnerships to enable easy integration into tion (NGO) sector and local communities.
 future systems. China may wish to employ the flexibility of this approach to address a In terms of logistics, the Sector Planning variety of different health concerns, while Framework will focus on long-term invest- not having to reinvent the process every ments of 5-10 years. It will involve planning time. based upon robust evidence, research, and data. This will help better meet the popula- Core Action Area 3: Empowerment tion’s needs while also embedding benefits and Enabling of Regions and Provinces realization into the planning cycle. It will to Develop their Own Capital integrate planning at various levels: state- Investment Plans wide, regional or local, and organizational. This will allow for a tailored approach for Empowering subnational levels in China to each community that still communicates well develop their own capital investment plans with NSW’s overall goals. It will strengthen require several key actions. The “Planning planning at the organizational level as well, Layout of National Medical and Health providing tools to support planning at the S ervices System (2015–2020)” provides an ­ local level with the help of organizations. incipient framework for this planning and L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 333 ensures its implementation will be a step in The SROS is the most important tool in the right direction.5 China may wish to fur- France’s regional capital investment and ther study. health care delivery planning. It focuses on One such example is the capital invest- hospital planning and on expensive treatment ment framework in France, where the health and technology provided in hospital settings. sector investment planning is based on Since its implementation in 2003, the SROS population needs and is executed through in each region has replaced the “national Regional Strategic Health Plans (Schéma medical map,” which was the quantitative Régional d’Organisation Sanitaire , or planning tool used by the Ministry of Health SROSs). SROSs set the overall strategic to divide each region into health care sectors goals for health care delivery; define priori- and defined norms for bed-population ratios ties, objectives, and targets; and determine for major disciplines within a geographical quantitative targets and the distribution of area (EOHSP, n.d.; Ettelt and others 2008). health care facilities within a region. SROSs In contrast to previous national planning are developed by regional health agencies practices, the purpose of the SROS is to bet- in consultation with stakeholders, includ- ter tailor health care delivery to the needs of ing the Ministry of Health, health insur- the local population. ance funds, hospital federations, health Related to capital investment planning, care professionals, and patient representa- SROSs determine capacity by specifying the tives (EOHSP, n.d.; Ettelt and others 2008). number of facilities in each region and subre- The Ministry of Health plays a coordinat- gion for each area of care (including g ­ eneral ing role and generates a catalogue of health medicine, surgery, maternity care, accident services based on an assessment of national and emergency care, neonatal care, radio- needs and priorities, which the regions therapy, cardiologic intensive care, and psy- incorporate in their own plans (Ettelt and chiatric care, as well as expensive technical others 2008). equipment such as magnetic resonance imag- ­ The regional health agencies are gener- ing scanners). They also define the volumes ally responsible for planning services and for certain types of service and benchmark for authorizing hospitals to deliver services them for comparison. “Service volumes” refer within the social health insurance system. to units such as numbers of patients, sites, They also oversee changes to the existing days (length of stay), procedures performed, hospital infrastructure, including restruc- and admissions, and they are expressed turing and mergers. The only exceptions in numbers of services or rates and show are new hospital developments (both private changes relative to previous volumes. The and public) and comprehensive emergency objective of planning on the basis of service centers, which have to be authorized by volumes rather than on bed-­ population ratios the Ministry of Health. Strategic planning is to limit oversupply, which is a persistent requires regional agencies to assess popula- problem in some cities (Paris) and regions tion health care needs on the basis of regional (south of France) (Ettelt and others 2008). health care utilization data and relevant Since the reform of national health plan- demographic data (such as on mortality and ning in 2003, trends in the French health care morbidity). Data for each region are analyzed system have moved toward increased effi- and compared with those for other regions to ciency. For example, the percentage of outpa- identify demand and supply. Expert estimates tient care has risen from 48 percent in 2001 of future trends in demand and technologi- to 53.4 percent in 2006, mostly through an cal change—largely based on epidemiological increase of day cases in acute care (from 30.9 data and trends observed in other countries percent to 39.2 percent) and in the follow-up (mainly the United States)—are taken into and rehabilitation sectors (60.3 percent to consideration for these assessments (Ettelt 67.4 percent). The average length of stay in and others 2008). acute care decreased from 5.7 days in 2001 334 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A to 5.4 days in 2006. However, the utilization The Certificate of Need (CON) program rate in the acute care sector (73.4 percent in is used extensively in the United States to 2005) is relatively low compared with that of evaluate and approve new capital investment neighboring countries, indicating continued projects. In 1974, the federal Health Planning overcapacity (EOHSP, n.d.). Resources Development Act mandated that The French experience—and in particu- all 50 states evaluate CONs before allowing lar how the SROSs are developed—offer the continuation of any health capital invest- lessons to China on how to involve all rel- ment projects, such as building expansions, evant stakeholders in the planning process. and ordering new high-technology devices. More important, however, are the lessons the The goal was to restrain facility costs and French experience offers in how to transition allow for a more coordinated planning of away from a bed-population ratio method health services and construction. Many states of determining health service configuration. established CON programs to receive federal As discussed earlier in the Chinese case stud- funding. Even though the Health Planning ies, CIP is still largely based upon population Resources Development Act, along with its projections and bed numbers. The same was funding, was cut in 1987, 36 states still main- true in France before the reforms in 2003; tain some form of a CON program, and the this method often led to oversupply of health remaining 14 states that do not have CON services, especially in urban centers like Paris programs have other mechanisms in place to and in southern France. China could learn regulate costs and duplication of services. from France by adopting the service volume method, whereby health service configura- Maine CON program tion and capacity are calculated based upon Each state in the United States has developed volumes of service (that is, type of service, its own unique approach to the program. numbers of patients, sites, lengths of stay, Many states have recognized the importance procedures performed, and admissions), of population health needs in CIP and rely which are then benchmarked with current on the analysis of population health needs levels. to implement capital investment projects. For example, in 2002, Maine passed a Cer- tificate of Need Act in an attempt to decrease Core Action Area 4: Introduction of a unnecessary construction and modification Certificate of Need Program to Evaluate of health facilities and duplication of health and Approve New Capital Investments services and hence to decrease costs and pro- in the Health Sector vide higher-quality health care. China already has a system of requiring fea- The core principles of Maine’s CON pro- sibility reports for all capital investments. gram are as follows (Ashcroft and Maine However, these feasibility reports use norms State Legislature 2011): set according to macro standards governing the size and scope of the intended service. •• Supporting effective health planning Feasibility studies are essential for gov- •• Supporting the provision of quality health ernments to evaluate and approve new care in a manner that ensures access to capital investment projects. The earlier cost-effective services discussion of a feasibility study for pos- •• Supporting reasonable choice in health sible relocation and expansion of the Ren- care services while avoiding excessive shou County People’s Hospital noted that duplication the construction planning and selection of •• Ensuring that state funds are used pru- location were determined according to facil- dently in the provision of health care ity needs rather than population needs. The services 
 Certificate of Need (CON) program pres- •• Ensuring public participation in the pro- ents a possible solution. cess of determining the array, distribution, L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 335 quantity, quality, and cost of these health •• There is a public need for the proposed care services 
 services. 
 •• Improving the availability of health care •• The proposed services are consistent with services throughout the state 
 the orderly and economic development of •• Supporting the development and availabil- health facilities and health resources for ity of health care services regardless of the the state. 
 consumer’s ability to pay 
 •• The proposed services are consistent with •• Seeking a balance, to the extent a balance the State Health Plan. 
 assists in achieving the purposes of this •• The proposed services ensure high-quality law, between competition and regulation outcomes and do not negatively affect the in the provision of health care 
 quality of care delivered by existing ser- •• Promoting the development of primary vice providers. 
 and secondary preventive health services •• The proposed initiative does not result in inappropriate increases in service utilization. 
 The Certificate of Need Unit (CONU) is •• The proposed project can be funded composed of a manager, three financial ana- within the capital investment fund (CIF). lysts, and administrative support. To distrib- ute the workload throughout the year, the Notably, the Maine CON program CONU processes applications for different requires the existence of public hearings for a projects on a staggered timeline. For a proj- proposed service before a capital investment ect to be considered, a CON application must project can begin. Applicants must prove be submitted to the CONU. After receiving their proposed capital investment is geared the application, a public information meet- toward meeting some public need. Further- ing is scheduled, and a public notice is issued. more, for each application, the CONU solic- This part of the process is essential for the its comments on the impact of each project CON program’s success. It allows for open on the health of Maine citizens from both the feedback from the public regarding projects, Maine Quality Forum and the Maine CDC allowing for citizens to express their needs (figure 9.7). and concerns. Any Maine citizen has the Maine’s CON program requires in-depth opportunity to provide a public testimony evidence of the potential worth of a capi- about a potential health capital investment tal investment project. Population health project (Ashcroft and Maine State Legisla- is considered to be of crucial importance, ture 2011). and because of the CON process, no capital The CONU considers the public testimo- investment project can be approved without nies, along with input from organizations detailed consideration of population health such as the Maine Quality Forum, the Maine needs from the applicant, the CDC, and the Center for Disease Control & Prevention Maine Quality Forum. This thorough exami- (CDC), and the Bureau of Insurance. By con- nation is less likely to favor larger hospitals in sulting with the Maine CDC, the CONU can a disproportionate way, allowing for efficient make informed decisions based upon popula- CIP to occur at all levels of health care to tion health needs. Throughout this process, best meet population health needs (Ashcroft several important factors are considered in and Maine State Legislature 2011). CON determinations (Ashcroft and Maine Similarly, China can consider the Maine State Legislature 2011): example in designing a CON program to approve CIP that is based on population •• The applicant is fit, willing, and able to health and public need. provide the proposed services at the proper standard of care. Kentucky CON program •• The proposed services are economically The state of Kentucky has also used the CON feasible. program to incentivize development of a full 336 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 9.7  Certificate of Need application and approval process The program also incentivizes the attain- in Maine, United States ment of robust quality indicators. This is especially important, given that Kentucky Informal consult currently has a poor health profile. Health officials wish to incentivize greater access Letter of intent to care for Medicaid members, the newly insured, and any remaining uninsured Technical assistance meeting citizens—a particularly important priority ­ for poor rural communities, where chronic diseases are rampant and the rates of unin- Application sured high. The CON program will also focus on Public notice achieving greater price transparency, espe- Record cially as health care in the United States open for Public informational meeting transitions from a fee-for-service model to comments a value-based purchasing framework. In addition, to modernize the health systems No public hearing 30 calendar days in Kentucky, CON will favor the adoption of new health technologies geared toward 30 calendar days Public hearing request the prevention and treatment of chronic dis- eases. Finally, the program encourages fur- 30 calendar days ther modernization by being more reflective Record closes of modern health care trends and population health needs. All the aforementioned points CONU analysis were brought forth in October 2014 as core principles for CON modernization.6 Record Preliminary analysis open for Michigan CON program comments 10 business days Another example from the United States that Record closes may be relevant for China is from Michigan. The CON program in for Michigan’s health CONU analysis sector has evolved over the years to include more services and to move away from a Briefing memo hospital-centric system. Michigan also intro- duced requirements to ensure that capital Commissioner review projects comply with standards, which has proved to be a challenge in China. Decision letter Initially, Michigan’s CON program was based solely on the costs of capital investment Source: “2010 Report, Certificate of Need Act.” ©Maine Department of Health and Human projects; as a result, primary attention was Services (Maine DHHS). Reproduced, with permission, from Maine DHHS; further permission required for reuse. given to hospitals, because they accounted Note: CONU = Certificate of Need Unit. for most health spending. The program ini- tially only covered hospital capital investment continuum of care. The program works to projects. However, as time passed, health promote and support providers and facilities policy officials realized that a CON program that seek to develop a robust continuum of based solely on costs could have a distorting care alone or in partnership with others, in effect on health care, adversely affecting both part owing to the evolving payment struc- quality and access. For example, the least tures and the ever-changing environment of costly location in which to start a new ser- the health care sector in the United States. vice might not be one that improves access or L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 337 might not have enough of the health profes- •• Proof that consumers make up the major- sionals needed to meet demand. The Michi- ity of a nonprofit hospital’s governing gan CON program evolved accordingly into body a more patient-centric, needs-based process, •• Proof that the hospital has the financial although considerable attention is still paid to capacity to both fund the construction and bed numbers (CRC of Michigan 2005). operate the facility following completion Michigan’s Act 256 of 1972 states a hos- •• Proof that the project complies with pital facility shall not be constructed, con- local and regional rules, regulations, and verted, added to, or modernized without first standards obtaining a CON that documents a demon- •• Other factors that contribute to the orderly strated need for the proposed project. This development of quality health care. need-based component has remained a part of the CON program. This act also placed Public Act 368 of 1978 further amended the responsibility for CON on the Depart- Michigan’s CON program, extending its ment of Public Health, and a commission coverage to nonhospital facilities (including was established to oversee the program. This nursing homes) and to certain clinical ser- helps to protect the state’s health-related CIP vices. This allowed Michigan to further con- from any biased political agendas. Accord- sider population health needs in its CIP as it ing to this act, a capital investment proposal moved away from the hospital-centric model. must contain the following (CRC of Michi- Ten years later, Public Acts 331 and 332 gan 2005)—good examples of the potential extended the CON program to include more criteria for capital investment projects in clinical services and defined seven covered China: medical equipment categories in state law. It also continued authorizations for regional •• The patterns and level of utilization, avail- health planning agencies under state law. Act ability, and adequacy of existing facilities, 332 allowed the Department of Public Health institutions, programs, and services in the to require the submission of data and statis- immediate community and region tics as a part of a CON application, and it •• The degree to which residents and physi- established in law the obligation to moni- cians in a community are provided access tor CON projects after approval to confirm to the hospital applying for the Certificate alignment with the approved project and of Need with population health needs (CRC of Michi- •• The availability and adequacy of services gan 2005). such as preadmission, ambulatory, or Michigan’s CON program is notably more home care services that may serve as alter- complicated than Maine’s, which highlights natives to hospital care the diversity of the CON programs in the •• The economies and service improvements United States. Each program is designed to that could be achieved from consolida- fit the state’s unique health policy procedures tion of highly specialized services or from and allows each state to create a program shared central services such as laboratory, that best suits its health population needs and radiology, and the like governmental processes. A proposed proj- •• The economies and service improvements ect in Michigan must meet the following six that could be achieved from affiliation or requirements (CRC of Michigan 2005): contractual arrangements between hospi- tals and others •• Meet an unmet need. •• The availability of personnel to fulfill the •• Include alternatives that have been con- services to be offered sidered and the reasons why the proposed •• Proof that the hospital does not discrimi- particular approach is best (if there are nate in activities including employment, no alternatives, the application must state room assignment, and training why). 338 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A •• Show that the proposed service is the least A program akin to the CON program costly. presents a possible solution to this. ­ •• Be delivered in compliance with oper- Each state’s CON program in the United ating standards and quality assurance States holds important lessons for China: standards; include a description of how the proposed project will assure appro- •• In Kentucky, the state recognized the dire priate utilization; indicate how proj- state of its population health and sought ect effectiveness will be measured; and to refocus its CON program on developing show that the applicant has complied, capital focused on the health needs. China both currently and historically, with fed- can consider doing the same by developing eral and state licensing and certification CON program-like processes to respond requirements. to health needs. By focusing on the health •• Demonstrate, if the project relates to a needs of a population, capital investment facility, that the facility where the pro- projects can help provide the best, most posed service will be delivered is viable by accessible care, improving overall popula- meeting one of six requirements. tion health. Kentucky has a long way to •• If a nonprofit applicant, show that con- go in terms of improving its population sumers make up a majority of the board. health, but it is starting by redirecting its CIP, providing a positive example of the Such requirements could prove useful for importance of service-based CIP. China, especially for provinces that have •• In Maine, applicants for capital invest- not fully complied with federal mandates. ment projects must prove that their pro- These requirements help base CIP on realistic posed capital investment is geared toward health needs and ensure that capital invest- meeting some public need. This is impor- ment projects contribute to the betterment of tant, because it helps reduce duplicated population health. services and helps direct capital invest- Michigan’s CON program is an example ment to areas of the state that need it of how such a program can evolve over time most. Also, public hearings are important and show that CON programs that are not features of the CON program process. hospital centric are the most successful at In these hearings, citizens can voice their providing capital investment guidance that is needs and opinions regarding potential aimed at improving the entire health sector. capital investments. This increases com- Michigan’s CON program is another exam- munication with health officials and the ple of how to orchestrate a CON program public and further provides for a people- to better plan capital investment structures centered CIP. Like Maine, China could around existing and unmet needs. develop a CON program-like process that relies heavily on actual population health U.S. Certificate of Need programs as a data to make informed decisions on capi- model for China tal investment. The CON programs as practiced in the •• In Michigan, the state realized that a United States hold a lot of promise for China, CON program based solely on costs where facility needs are often paramount in could have a distorting effect on health determining hospital expansion. A close look care, adversely affecting both quality and at the feasibility study of possible reloca- access—an important lesson for China. tion and expansion of the Renshou County The CON program evolved over the years People’s Hospital in Meishan city, Sichuan to include more services and move away province, for instance, reveals that construc- from hospital-centric CIP. Several require- tion planning and selection of location for the ments also ensure that capital projects Renshou hospital were determined according comply with standards; China has expe- to facility needs rather than population needs. rienced difficulty with compliance, so L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 339 such requirements could help its practices centers at population centers throughout the to better reflect actual policy. Michigan’s country (box 9.3). CON program shows how CON pro- Meeting the second component (concen- grams that are not hospital-centric are tration of complex services) required greater the most successful at providing capital centralization—from local general hospi- investment guidance aimed at improving tals to acute centers or to regional centers of the entire health sector. Michigan’s CON excellence—of those services that, because of program is another example of how to their complexity, required specialized skills orchestrate a CON program to better plan and expertise that could not easily or afford- capital investment structures around exist- ably be replicated in local hospitals. A key cri- ing and unmet needs. terion in the process of determining the final locations of those hospitals being designated as “acute” was that patients should have a Core Action Area 5: Prioritization of maximum travel time of one hour from any- Community Health Projects where in Northern Ireland to an acute facility As in China, capital investment in Northern offering full accident and emergency services. Ireland was once hospital centric and was A primary objective of this new model of largely focused on the acute sector. Begin- care is to improve accessibility of the public ning in 2007, Northern Ireland started to high-quality, timely services. The specific to redirect its capital investments toward location of individual facilities was deter- community-level facilities. The new model mined by a number of key factors, includ- sought to create an integrated continuum ing the core principles within the regional of facilities—from home care to primary, health strategy, urban or rural setting, community, subacute (step-down), and size of the local population, epidemiology, acute facilities—all supported by structured travel times and distances, critical mass networks. for staff, critical mass for specialist equip- The underlying strategy had two main ment, state and ­ location of current facili- components: enhanced services within the ties, improved accessibility, reduced waiting community and concentration of complex times and reduced hospital admissions, and services. Regarding the first component, affordability. Northern Ireland carried out a compre- Additionally, Northern Ireland has hensive regionwide planning exercise and attempted to incorporate flexible design prin- decided to develop 42 new community health ciples into its new configuration (­ figure 9.8). BOX 9.3  Physical redesign of Northern Ireland’s health system model Five elements defined the physical redesign of the 3. Reduction in the number of general hospitals health system in Northern Ireland: providing the full range of acute services from 18 to 10 1. Reduction of the number of Health and Social Care 4. Redevelopment of seven of the remaining nine Trusts (service provider organizations) from 17 hospitals as new nonacute step-down facilities to 5, according to geographic need, each providing with a focus on their local communities and the a full continuum of health and social care services ability to provide a wider range of intermediate to its local population care services 2. Designation or development of regional centers as 5. Creation of 42 new, one-stop community health the sole providers of a range of tertiary services centers (without bed accommodation) with the key that will benefit from centralization objective of preventing unnecessary hospitalization. Source: Rechel, Erskine, and others 2009. 340 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 9.8  The integrated health services model in Northern Ireland Local hospital 4 150,000–300,000 Acute Local hospital hospital Acute hospital 3 CHC Non-health agencies Local hospital 1.7 million Regional centre LHC 5 100,000 + LHC 2 CHC CHC LHC 1 Acute 20,000–70,000 hospital Individual LHC homes LHC 2,000–10,000 Other community LHC LHC facilities Level 5 Regional centre Level 4 Acute hospital Level 3 Local hospital Level 2 Community health centre Level 1 Local health centre Source: Department of Health, Social Services and Public Safety (DHSSPS), Belfast, Northern Ireland. Note: CHC = community health center; LHC = local health center. These principles included phased construc- The colocation of Level 1 and Level 2 tion to transition from existing to new facili- facilities has been encouraged within the ties; insertion of “soft” spaces (for example, model, particularly in areas of high popula- office space or educational accommodation tion density, where travel distances are more that can be relatively easily relocated) beside likely to be acceptable for access to general complex areas (such as those for critical care practitioners. Where sites for Level 3 or Level or imaging) that are likely to expand in the 4 facilities are already located at natural future and would be very expensive to move; population centers with good access to pub- and standardization (Rechel, Erskine, and lic transport, there are potential benefits in others 2009). colocating Level 1 and Level 2 facilities while L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 341 ensuring the retention of their separate iden- undergo such a physical transition and move tities and organizational structures. Where away from a hospital-centric system. Citizens such colocation is proposed, the resultant of Northern Ireland now have greater access arrangement has come to be referred to as “a to both community facilities and acute facili- health village.” The typical range of services ties, both of which have been designed to intended to be provided at each level is out- improve population health. lined in table 9.5. The focus on specific geographic needs The example of Northern Ireland shows offers an important lesson for China, which that it is possible for a health system to could greatly benefit from investing more TABLE 9.5  Services under Northern Ireland’s integrated health services model, by level Level Services Level 1: Local health centers · General practices · Construction cost range: £1–5 million · Noncomplex diagnostic testing · Level 1 facilities frequently incorporated · Basic treatments and nursing care into Level 2 facilities · A limited range of therapies Level 2: Community health centers · After-hours GP service · Construction cost range: £5–15 million · Outpatient clinics · Minor procedures · Noncomplex imaging and diagnostics · Children’s services · Physiotherapy · Speech therapy · Podiatry · Dental services · Social services · Mental health services · Multidisciplinary outreach teams · Voluntary sector · Community facilities · Pharmacy Level 3: Local hospitals · Urgent care center (as opposed to full accident and emergency care) · Construction cost range: £40–70 million · Ambulatory care center · Full diagnostics including radiological services · Day procedures or day surgery unit (Level 3 facilities can be designated to act as “protected elective centers”) · Step-down, rehabilitation, and GP beds · Mental health unit · Support services Level 4: Acute hospitals Full range of standard acute hospital services, including the following: · Construction cost range: £200–300 · Specialist-led accident and emergency care million · Critical care department · Acute medical and surgical departments · Pediatrics · Outpatient department · Radiology Level 5: Regional centers of excellence Specialized care, including the following: · Construction cost varies · Cancer treatment services · Generally, but not always, colocated with · Orthopedic services a Level 4 acute hospital · Cardiac surgery · Neurosurgery Source: Adapted from Rechel, Erskine, and others 2009. Note: GP = general practitioner. 342 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A in community health capital projects and 6. For more information on Kentucky’s CON increasing access to quality care. Recogniz- program, see the Cabinet for Health and ing the unique health and capital invest- Family Services of Kentucky website: https:// ­ ment needs at the community level, North- chfs.ky.gov/agencies/os/oig/dcn/Pages/cn.aspx. ern Ireland is focusing its capital investment on community-level facilities, transitioning References away from hospital-centric models. Further, Northern Ireland has dedicated some capital Ashcroft, Beth, and Maine State Legislature. investment toward creating flexible facili- 2011. “Certificate of Need: Process Appears Clear, Consistent and Transparent, 2011.” ties, which increases efficiency in the long Paper 26, Office of Program Evaluation and term and enables the health system to better Government Accountability, Augusta, ME. respond to future population health needs CRC of Michigan (Citizens Research Council of without needing to invest in new capital or Michigan). 2005. “The Michigan Certificate completely redesign facilities to meet unfore- of Need Program.” Report 338, CRC of Mich- seen needs. China may wish to explore this igan, Livonia, MI. flexible design. EIU (Economist Intelligence Unit). 2015. “China Healthcare Industry Report.” http://www.eiu​ .com (requires subscription). EOHSP (European Observatory on Health Sys- Notes tems and Policies). n.d. “Health Systems in 1. OEC D health spending data are from Transition (HiT) Profile of France.” Country the 2015 OECD. Stat Health Statistics health system report, EOHSP, World Health Dat abase, ht t ps: //stats.oecd.org /i ndex​ Organization (WHO) European Centre for .aspx?DataSetCode=HEALTH_STAT. Health Policy, Brussels. 2. National accounts data provide an idea of the Ettelt, Stefanie, Ellen Nolte, Sarah Thomson, and type of assets and capital spending. Although Nicholas Mays. 2008. “Capacity Planning in capital spending can fluctuate from year to Health Care: A Review of International Expe- year, in OECD countries overall, there is an rience.” Policy brief, World Health Organiza- even split between spending on construc- tion on behalf of the European Observatory on tion (that is, building of hospitals and other Health Systems and Policies, Copenhagen. health care facilities) and spending on equip- Hindle, D., B. Dowdeswell, and A.-M. Yasbeck. ment (medical machinery, ambulances, and 2004. “Report of a Survey of Clinical Path- information and communication technology ways and Strategic Asset Planning in 17 EU [ICT] equipment). Together they account Countries.” Report, Netherlands Board for for 85 percent of capital expenditure. The Hospital Facilities, Utrecht. remaining 15 percent is accounted for by Le Deu, Franck, Rajesh Parekh, Fangning intellectual property products—the result of Zhang, Gaobo Zhou. 2012. “Healthcare in research, development, or innovation. China: Entering Uncharted Waters.” McKin- 3. The NHFPC was created in 2013 from the sey Insights compendium, McKinsey & Co., former Ministry of Health and National Shanghai. Population and Family Planning Commis- Nauta, J., R. Perenboom, and F. Galindo Garre. sion. In March 2018, however, its functions 2009. “A New Horizon for Planning Ser- were integrated into a new agency called the vices and Health Care Infrastructure for the National Health Commission (“China to Elderly.” Proceedings, HaCIRIC (Health and Set Up National Health Commission,” Xin- Care Infrastructure Research and Innova- hua, March 13, http://www.xinhuanet.com​ tion Centre) International Conference 2009: /english/2018-03/13/c_137035722.htm). 44–53. 4. “Planning Layout of National Medical and NSW Government (New South Wales Govern- Health Services System (2015–2020)” (Guo ment). 2011. “Sector Planning Framework: Ban Fa 2015, No. 14). Policy Statement [Version 1.0].” Ageing, 5. “Planning Layout of National Medical and Disability, and Home Care (ADHC) policy Health Services System (2015–2020)” (Guo document, NSW Department of Family and Ban Fa 2015, No. 14). Community Services (FACS), Sydney. L e v er 8 : M oderni z ing H ealth S er v ice P lanning to G u ide I n v estment 343 OECD (Organisation for Economic Co-operation Rechel, B., S. Wright, B. Dowdeswell, and M. and Development). 2008. “OECD Annual McKee. 2010. “Even in Tough Times: Invest- Report 2008.” OECD, Paris. ing in Hospitals of the Future.” Euro Observer Rechel, B., J. Erskine, B. Dowdeswell, S. Wright, 12 (1): 1–3. and M. McKee. 2009. Capital Investment for Rechel, Bernd, Stephen Wright, Nigel Edwards, Health: Case Studies from Europe. Observa- Barrie Dowdeswell, and Martin McKee. 2009. tory Studies Series No. 18. Copenhagen: World Investing in Hospitals of the Future. Observa- Health Organization, on behalf of the Euro- tory Studies Series No. 16. Copenhagen: World pean Conservatory on Health Systems and Health Organization, on behalf of the European Policies. Conservatory on Health Systems and Policies. 10 Strengthening the Implementation of Health Service Delivery Reform Introduction up the recommended reforms described in earlier chapters. The next phase of China’s health care system • “A Four-Part Actionable Implementation development will center on comprehensive Framework” presents an operational imple- improvement in the value of care across all mentation framework that focuses on four levels of the system. Previous chapters have implementation systems: (a) macro imple- discussed the details of what must be changed mentation and influence, (b) coordination in each of the eight reform levers. Drawing and support, (c) service delivery and learn- on lessons from national and international ing, and (d) monitoring and evaluation. cases, they have also provided specific strate- • “Moving Forward: Effective, Sustainable gies to guide the implementation of each core Local Implementation” proposes, for action area. This chapter addresses the cen- tral challenge of how to implement these each of the four implementation systems important changes and focus on creating an listed above, strategies that are specific enabling organizational environment; it also and relevant to China. The organizational discusses the tools needed to operationalize platforms for frontline service delivery and sustain the core action areas and imple- improvement and learning are particularly mentation strategies suggested in the previ- important. For example, it is unlikely that ous chapters. Putting this environment in changes in payment incentives will be place is a precondition for effective imple- enough to enable low-performing organi- mentation and thus a critical first step. zations to transform themselves (Cutler Without it, progress may be elusive. 2014); improvement will also require a The rest of the chapter is organized as support system that builds capacity and follows: creates a facilitative climate to foster orga- nizational (and individual) change.1 • “Implementation Challenges” reviews • “Toward a Sequential Plan for Full-Scale barriers to implementation in China’s Reform Implementation in China” con- institutional and organizational environ- tains recommendations on sequencing and ment and discusses specifics of an imple- reaching full-scale health service delivery mentation model for spreading and scaling reform. 345 346 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Implementation Challenges principles and general guidelines to stimu- late local innovation, while allowing flexi- There is consensus that China has suffi- bility in applying the principles to local ciently robust policies in place for health ­ c onditions. Innovations are usually tried sector reform, but most observers acknowl- through pilot activities, which tend to be edge that the country has had difficulty sanctioned by the central government. translating these policies into scalable and As observed in a number of case studies sustained actions that improve service deliv- reviewed in this report, successful innova- ery. Typical of its development strategy in tions have indeed occurred, but few have other sectors, China has promoted health been scaled up. reform implementation mainly through Some policy makers suggest that innova- pilot projects. Experimenting with small- tions and reform implementation tend to be scale pilots operated by local governments “personalized,” responding to the prefer- has been effective in promoting and expand- ences of local leaders, and therefore are dif- ing economic reforms (Heilmann 2008), but ficult to replicate. This may relate to the lack it needs to make further progress in expand- of evidence-based analysis and feedback on ing health reforms. 2 This need has become reform progress and problems. Few innova- particularly evident in efforts to address tions have been evaluated using rigorous deep-rooted and complex issues related to methods, especially since all pilots were provider incentives, private sector engage- implemented under local contexts, and the ment, public hospital reform, and rebalanc- background of different localities nationwide ing service delivery. varies greatly. Part of the problem is the difficulty of The State Council’s Health Reform shifting away from direct facility manage- Leading Group is responsible for policy for- ment by government agencies to an arm’s- mation and oversight, but various central length or indirect approach to governance, in government agencies monitor how these which government agencies steer the health ­ policies are implemented, with each agency system through a combination of incentives, focusing on specific aspects of reform regulation, and other checks and balances (such as pricing, insurance, drug standards, (Meessen and Bloom 2007). Institutional human resources, medical services) aligned fragmentation, diffuse leadership, and vested with their respective mandates. Supervisory interests make this transition even more reports tend to be based on short fact-­ challenging. Under these conditions, even gathering site visits, which are often con- effective pilots cannot be maintained or ducted separately by representatives of dif- scaled up. ferent agencies. I n addition, the Moving forward to implement the recom- independence of any assessment can be mendations related to the eight reform levers questioned because central-level depart- will depend on careful management of imple- ments are not totally separate from their mentation impediments at three levels of the decentralized counterparts in provincial and system: central government, provincial and local governments. China has yet to system- local governments, and frontline service pro- atically put in place independent mecha- viders. The impediments at each of these lev- nisms for gathering information and evalu- els is taken up in turn. ating reforms. These conditions suggest that the central government may need to provide implementation-­ oriented guidance, consoli- Central Government: Dispersed date and strengthen implementation over- Oversight and Monitoring of Reform sight, and introduce systems to scrupulously Implementation monitor and v ­ alidate progress and assess Typical of China’s governance style, central implementation from a more systemic, “big government policy directives consist of picture” perspective. S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 347 Provincial and Local Governments: Frontline Service Providers: Weak Fragmented Coordination and Organizational Mechanisms for Leadership Providers to Lead and Share Learning about Health Care System Reform and Given the dispersion of roles over many Improvement institutions, health reform has not been pri- ­ oritized at some local levels. Resilient mecha- Health care improvement occurs on the nisms for holding local government leaders front lines: in households, village clinics, accountable for health reform implementa- community and township health centers, tion have yet to be put in place. Incentives for and hospital wards. Transformational value local officials to plan and implement health is seldom created by a single clinician or reforms are generally weak compared with, facility; it is more often generated by a for example, the incentives to promote eco- group of providers who cooperate with each nomic growth and development (Huang other and are collectively responsible for 2009; Ramesh, Wu, and He 2013). Local patient care. Reliable implementation of leaders’ performance and promotion are not policy reform at the facility level does not determined by the progress on health reform. happen by accident or by chance: deliberate Under these conditions, local officials are jus- and focused plans to ensure implementation tifiably reluctant to take on complex issues must be created and then executed. This has such as the profit-making interests of public been amply demonstrated internationally: hospitals. g o o d e x a m p l e s i n c lu d e t h e U n i t e d Putting in place new models of health ser- Kingdom’s Primary Care Collaborative vice delivery will require the strengthening of (described in annex 10A), the U.S. Veterans broad system coordination, particularly to Health Administration’s Patient Aligned overcome institutional fragmentation—both Care Teams (discussed in chapter 2), and hori zontal (across many gover n ment t he U. S . C e nt er s for M e d ic a re a nd ­ departments) and vertical (across the munici- Medicaid’s recent Partnership for Patients pal, county, and district levels of govern- (discussed in chapter 3). ment). Sustainable and scalable reform imple- International experience demonstrates mentation is compromised under the current that the proposed shift in organizational situation in which each department and goals from treatment delivery to outcomes agency tends to act to defend its own inter- improvement requires fundamental changes ests. Decisions on complex issues are often in organizational culture. “Naming and made through interagency bargaining, which blaming” denunciations to motivate changes weakens reform implementation (Huang in provider practices are insufficient to 2009; Qian 2015). Patchwork administrative encourage creation of a value-oriented deliv- actions negotiated among diverse government ery system. Instead, the evidence supports departments (with divergent interests) to the use of health-systems improvement address elements of the reform may be effec- methods, including performance reporting, tive in the short term but are not sustainable data transparency, and, perhaps most unless the government builds and institution- importantly, systematic application of spe- alizes its coordination capacity and creates cific learning models that allow institutions the organizational arrangements to make to make changes and learn from their impact them operational (He 2011). I n sum, (Garside 1998; Greene, Reid, and Larson effective, scalable, and sustainable implemen- 2012; Schouten and others 2008). Facilitated tation will require putting in place incentives collaborative approaches that allow peer and accountability mechanisms that will institutions to learn from one another’s suc- drive local leaders and government depart- cesses and failures in a fear-free environ- ments to coordinate and enforce health ment can rapidly accelerate implementation reforms. of policy reforms. 348 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A A Four-Part Actionable implementation, and ultimately to sustained Implementation Framework institutionalization of successful practices. Despite the strong evidence base support- Implementation consists of the set of activi- ing these frameworks, some caution is war- ties, processes, and interventions used to put ranted. Some components have stronger policies, reforms, and evidence into practice. empirical support than others. Also, imple- High-quality implementation is associated mentation is inherently intertwined with the with obtaining desired impacts (Aarons and contexts where it occurs: general, “one size others 2009; Durlak and DuPré 2008; fits all” solutions do not exist, and adapta- Meyers and others 2012; Wilson, Lipsey, tions tailored to local contexts will invari- and Derzon 2003). For example, in a review ably take place. The implementation steps of 483 studies in five meta-analyses and and organizational platforms proposed 59 additional studies, Durlak and DuPré below, and their sequencing and timing, will (2008) found a significant association vary according to local capacity, the sup- between the level of implementation and the porting environment, and other starting achievement of program outcomes. High- conditions. quality implementation increases the statisti- Bridging the gap between policies and cal probability of better program perfor- practice requires capacity, resources, account- mance and can lead to better benefits for ability, and a commitment to collaboration, program participants. evaluation, and learning. Drawing on the lit- There is a well-recognized gap between erature on implementation guidelines, the the health gains that could be achieved and discussion below proposes a simplified but those that are actually being realized around actionable implementation framework con- the world (WHO 2007). Part of this gap sisting of four systems adapted broadly to the results from shortcomings in implementation Chinese context. It is important to note that and putting knowledge into practice. In other these systems overlap and that further adap- words, evidence-based health technologies tations probably will be required for specific and service models are not reliably imple- situations. mented in many contexts. As the editors of the journal Implementation Science wrote in The Macro Implementation and their inaugural issue, “Uneven uptake of Influence System research findings—and thus inappropriate care—occurs across settings, specialties, and This system involves establishing the external countries” (Eccles and Mittman 2006, 1). “influence factors” that would create a cli- Drawing on a large body of literature, the mate that facilitates effective and sustained science supporting implementation has implementation (Fixsen and others 2005, 59). advanced considerably during the past two Implementation does not occur in a vacuum; decades. A number of actionable frameworks it occurs within an institutional, political, have emerged to assist planners, implement- and financial environment that establishes ers, and communities in their implementation leadership and advocacy (for a focus on efforts (Aarons, Hurlburt, and Horwitz implementation practices), sets goals and per- 2011; Damschroder and others 2009; Durlak formance targets, and scrutinizes the quality and DuPré 2008; Fixsen and others 2005; of implementation practices and their Meyers, Durlak, and Wandersman 2012; impacts. This enabling environment is essen- Meyers and others 2012; Peters, Tran, and tial for successful implementation of transfor- Adam 2013; Wandersman, Chien, and Katz mative reforms, which entail a need for new 2012; Wandersman and others 2008). These models and learning to overcome well-­ frameworks provide evidence-based guidance established routines and embedded interests. on the critical phases, action steps, and Research shows that a facilitating macro envi- ­ c omponents that contribute to effective ronment is associated with better outcomes S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 349 and with fidelity of implementation—that is, • Ensuring that reform has adequate admin- the degree to which implementation is aligned istrative support with intended expectations, design, and plans • Conducting on-site monitoring of imple- (Fixsen and others 2005; Meyers, Durlak, mentation activities, including document- and Wandersman 2012; Meyers and others ing adaptations of original plans and 2012). designs. Greater attention to and scrutiny of imple- mentation practices by senior policy makers The coordination and support system and leaders is critical to the process of service requires an organizational structure near the delivery reform. Specific considerations frontline implementation to carry out these include the following: functions and oversee the implementation process. Some countries have set up special • Creating clear accountabilities for imple- implementation task forces to encourage the mentation performance involvement and commitment of key stake- • Demonstrating leaders’ commitment to holders, monitor and control institutional the implementation process and political pressures, and guide frontline • Specifying expected implementation mile- providers through the complex change pro- stones and outcomes cesses (Dewan and others 2003). As described • Building a monitoring and feedback sys- below, China may consider establishing a tem to learn from implementation experi- fully empowered “leading group” or steering ences and adjust policies and guidelines committee at the provincial or local govern- • Mobilizing resources to support imple- mental levels to perform these functions. mentation processes • Arranging independent evaluations. The Delivery and Learning System One strategy for fostering an enabling The delivery and learning system is the main environment (as further described below) is locus of implementation and where many ser- to strengthen the central government’s vice delivery reforms and care improvement oversight and monitoring role in reform ­ solutions are designed and executed. It is at the implementation. front lines of service delivery: health care orga- nizations (for example, hospitals, township health centers, and community health centers); The Coordination and Support System networked groups of health care organiza- The coordination and support system aims to tions; and communities. It involves individual create the capacity and an enabling environ- behavioral and broader organizational change ment for effective implementation of frontline but also making the “culture of the organiza- reform. Key functions of the coordination tion” open to change (Garside 1998, S8). This and support system include the following: system is where evidence is put into practice and where implementers learn from their own • Coordinating and ensuring the commit- experience and customize and tailor experi- ment of key local stakeholders ence from elsewhere to their own situation. • A r ra ng i ng t ra i n i ng a nd te ch n ic a l Operationally, it involves creating assistance organizational arrangements for problem ­ • Developing and adapting implementation solving, practitioner-to-practitioner coach- plans and timelines ing and collaboration, and shared and con- • Communicating reform activities and tinuous learning. Transformation Learning expectations to communities, health care Collaboratives (TLCs) are proposed as the organizations, and health workers organizational building blocks for a delivery • Making frontline providers accountable and learning system in China, as described for implementation progress and results later in this chapter. 350 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A A learning system includes a set of methods useful for frontline practitioners to make including observation, experimentation, and system-level improvements, the information feedback that is applied to build knowledge will need to be returned to the front line for and enable the achievement of a patient-­ those practitioners’ interpretation and use. centered result. In the United States, the Systems may already be in place to aggregate National Science Foundation has supported local data, create distribution tables, and multidisciplinary, cross-sector workshops to benchmark values that can be fed back to explore the potential for a new “science of service delivery units for their reflection. learning systems” (Etheredge 2014; Friedman Guidance on how to improve the accuracy, and others 2014). A variety of technical mod- completeness, and timeliness of data submis- els for generating new insights (innovation), sion, processing, and feedback can be found developing supporting evidence (research), in the international literature (Mate and and ensuring reliable implementation of ­others 2009). impactful findings (dissemination and imple- In addition, it is highly recommended that mentation) have been described (IHI 2003; implementation be accompanied by a combi- Rogers 2010). Recognizing the need and nation of effective health services research importance for a health care system to be able (HSR) and impact evaluations to allow rigor- to learn rapidly from its own efforts to ous measurement of intended and unintended improve patient care, the U.S. Institute of effects and outcomes. Though these activities Medicine established a Committee on the are more methodologically demanding than Learning Health Care System in America monitoring, HSR and impact evaluations can (Yong, Olsen, and McGinnis 2010).3 provide valuable information for understand- ing a reform’s effects and can provide practi- cal guidance to key stakeholders, including The Monitoring and Evaluation System policy makers, about the progress of reform Monitoring and evaluating the effectiveness implementation. One additional focus—­ of implementation and the impact of reform combining both monitoring and impact is a critical but often overlooked component evaluation—is understanding why implemen- of the implementation process. Evidence tation was successful or not (Berwick, Nolan, needs to be gathered to learn from implemen- and Whittington 2008). In China, putting in tation and contribute to evidence-based place a robust monitoring and evaluation sys- adjustments and future policy making. tem to accompany reform implementation Careful monitoring can detect whether will require the close attention of the central implementation is aligned with stated objec- government in coordination with provincial tives, whether it is on track (or going off and local governments. track), and whether the implemented reforms match those that were intended. Moving Forward: Effective, To that end, it requires careful measure- ment, which in turn must respond to the Sustainable Local Implementation information needs of the various stakehold- Numerous experiments are under way in ers. Good measurement will require priori- China to operationalize the health reform tizing the establishment and strengthening policies, but for the reforms to be successful of high-quality national and subnational and brought to scale, they need to be deep, information platforms. Previous attempts to comprehensive, and implemented in a coordi- improve the quality of common data reposi- nated and deliberate manner. In building a tories for health care improvement have better health care delivery system for China, established a basis on which primary care a major challenge is reaching full scale: being data can be captured and analyzed for the able to test and spread reforms to health care purposes of making ongoing improvements. delivery systems in every municipality, For these data to become operationally county, township, and village. S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 351 Following the four-part implementation support to make them work. The main strat- framework presented in the previous sec- egies for each system are the following (as tion, this section suggests strategies and cor- further outlined in table 10.1): responding actions that China can consider to facilitate robust reform implementation • Macro implementation and influence sys- and scaling-up. These strategies are the criti- tem: Establish strong central government cal elements for planning, prioritizing, and oversight linked to national policy imple- sequencing interventions necessary to build mentation and monitoring guidelines a 21st-century health system. All will need • Coordin ation and suppor t syste m: strong, persistent central government Establish fully empowered coordination TABLE 10.1  Responsibility levels, strategies, and actions to scale up health service delivery reform, by implementation system type Actionable implementation systems System and level of responsibility Main strategies and responsibilities Actions Macro implementation Establish the external “influence factors” that Establish strong central government oversight, linked to national and influence system would create a facilitative climate for effective policy implementation and monitoring guidelines, responsible for (central level) and sustained implementation the following activities: Focus greater attention and scrutiny on • Create clear accountabilities for implementation performance, implementation practices by senior policy demonstrating leaders’ commitment to the implementation makers and leaders process, by specifying expected implementation milestones and outcomes • Build a monitoring and feedback system to learn from implementation experiences to adjust policies and guidelines • Mobilize resources to support implementation processes • Arrange for independent evaluations Coordination and Create capacity and an enabling environment Institute coordination and leadership mechanisms at the provincial support system for effective reform implementation by and local governmental levels (such as “leading groups” or steering (provincial and resolving overlapping accountabilities and committees) that build capacity and foster accountability for local levels) interagency complexities effective reform implementation as follows: • Coordinate and ensure buy-in of key local stakeholders • Arrange training and technical assistance • Develop and adapt implementation plans and timelines • Communicate reform activities and expectations • Conduct on-site monitoring of implementation activities, including documenting adaptations to original plans and designs Delivery and learning Set up a robust learning system—a • Develop local Transformation Learning Collaborative models to system (frontline set of methods including observation, foster frontline reform implementation and care improvement providers) experimentation, and feedback—that • Create an organizational arrangement for problem solving, frontline workers can use as they implement practitioner-to-practitioner collaboration and coaching, and service delivery reforms and help design and shared and continuous learning execute care improvement solutions, enabling them to build and share knowledge during implementation Monitoring and Monitor and evaluate the effectiveness of • Ensure strong and independent monitoring with robust data evaluation system reform implementation and impact—a critical feedback to the front line where care decisions are made (central level in but often overlooked component of the • Conduct rigorous impact evaluations to understand progress partnership with implementation process against overall health reform objectives provincial level) 352 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A and leadership mechanisms at the provin- reform planning and execution by provincial cial and local governmental levels that and local governments. Drawing on the core build capacity and foster accountability action areas presented in chapters 2–9, many for effective reform implementation activities could be covered by such guidelines • Delivery and learning system: Create local (table 10.2). The guidelines could describe Transformation Learning Collaboratives verifiable (and measurable) tasks or interme- at the network and facility levels that fos- diate outcomes related to reform implementa- ter frontline reform implementation and tion, and thus foster greater integrity of care improvement reform implementation at local levels. They • Monitoring and evaluation system: would not constitute an implementation plan Ensure strong and independent monitor- or a generalized blueprint but would need to ing and impact evaluation. be operational in nature, specifying categori- cally “what to do.” In turn, provincial and local governments would need to have full Macro Implementation and Influence authority to decide on “how to do it” — System: Establish Strong Central developing, executing, and sequencing imple- Government Oversight mentation plans based on local conditions. The central government might consider tak- Given the large number of government ing a more hands-on lead in guiding and institutions involved in the health sector, the monitoring the implementation of the decentralized nature of implementation, and reforms, including the eight levers. China the well-known difficulties in aligning insti- could consider assigning this mandate to the tutional positions, China may want to con- State Council, which would mean expanding sider assigning an official with a rank higher the authority, roles, and functions of the State than Minister to head the State Council Council Health Reform Leading Group cur- Health Reform Office (SCHRO). While con- rently responsible for health-reform policy troversial, an appointment with this very making (figure 10.1). high rank may be needed to influence institu- Guidelines for policy implementation and tional stakeholders and provincial governors. monitoring could be prepared to orient China might also consider granting SCHRO FIGURE 10.1  Proposed oversight, coordination, and management for implementation and scaling up of health service delivery reform Level Organizational arrangements Tasks and responsibilities • Policy implementation framework • Task agreements Central State Council Health Reform Leading Group • Monitoring system • Promotion system • Sequenced master implementation plan Provincial leading groups • TLC selection and oversight Provincial • Task agreements or local Local government leading groups • Support for institutional and nancial reforms Frontline TLC leaders • Implementation of service service delivery reforms and care delivery Transformation Learning Collaboratives (TLCs) improvement S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 353 TABLE 10.2  Sample monitoring guidelines for implementation of China’s value-driven future health system Reform component Key elements, selected core action areas Service delivery system A tiered health care delivery 1. Strengthened primary care is the first point of contact and the gatekeeper for patients’ use of the health care system based on people- delivery system, and is responsible for providing continuous and comprehensive care. (Primary care includes centered integrated care m-health outreach to communities, social services, and homes through use of community health workers (Lever 1) virtually connected to general practitioners and specialists.) 2. Within each care network, well-organized multidisciplinary teams of clinical and nonclinical personnel provide full cycle of care to patients. (People enroll with care teams and are stratified by risks and conditions. Teams assume joint accountability for treatment, prevention, and patient engagement.) 3. Vertical integration of care is provided at hospitals, primary care facilities, and communities by establishing multidisciplinary teams, evidence-based integrated clinical pathways and referral systems (such as for postdischarge care), and individualized care plans for patients with chronic conditions. 4. Horizontal integration of individual preventive and curative care services takes place at the primary care level. (Centers for disease control emphasize public health; individual preventive care is transferred to primary care.) 5. Information and communication technologies support provider-to-provider integration and empower frontline health workers. 6. A Transformation Learning Collaborative leadership team (separate from hospital management) forms and operates networks. Quality of care and patient 1. National authority assesses, regulates, and oversees quality of care in all institutions. engagement (Levers 2, 3) 2. Evidence-based health literacy campaigns are used to encourage healthy behaviors. 3. Patients’ self-management of chronic conditions is part of care plans. 4. Information on provider quality is publicly disclosed. Hospital reform and service 1. Public hospitals are granted more autonomy in management but within a strong regulatory and accountability integration (Levers 1, 4) framework that ensures accountability for supporting care integration, reducing costs and unnecessary care, and shifting low-complexity care to lower levels. 2. Managerial capacity building is supported by putting a professionalization plan in place for hospital management. 3. Tertiary hospitals provide highly complex care while supporting secondary hospitals and primary care facilities with technical assistance, research, and workforce development. 4. Secondary hospitals provide essential specialty care and are closely linked to primary care, providing technical support, supervision, and training. Professional medical staff are shared with primary care through formation of multidisciplinary care teams. Financial and institutional environment Incentives in purchasing 1. Strategic purchasing of health services is based on quality and efficiency criteria. and provider payments 2. Health providers’ income is delinked from service volume. (Lever 5) 3. Provider payment systems gradually shift from paying individual facilities to paying integrated care networks (for example, using capitation) and to paying for a package of services (for example, bundled payments for treating groups of patients with certain conditions). Strengthening of human 1. Standardized scientific professional development and education is in place for all health care professionals, resources (Lever 6) including physicians, nurses, and pharmacists. 2. Professional standing and sufficient income for primary health care providers are ensured to keep the health profession attractive. 3. Physician compensation and hospital-based quota systems are reformed to enable a more flexible labor market and efficient workforce management. 4. New and alternative cadres of workers are produced and integrated into the health workforce to strengthen primary health care delivery. Private sector engagement 1. Regulations support a level playing field whereby high-quality private providers can deliver cost-effective (Lever 7) services and compete on the same terms as the public sector. 2. Social health insurers purchase from private providers the services for which they are licensed and that meet quality standards. Service and capital planning 1. A new, people-centered planning model is based on province-specific population health needs and (Lever 8) demographic profiles. 2. Capital investment planning integrates all public financial resources. 3. An integrated capital planning process incorporates private provider participation and capacity. 354 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A sufficient authority and institutional indepen- current institutional framework. It could be dence to influence how resources are allo- considered an interim organizational cated and how provincial and local leaders arrangement, in part to mitigate the potential are assessed in terms of reform implementa- adverse effects of institutional fragmentation tion. SCHRO staffing would need to be on reform implementation—it would not strengthened to perform this proposed institutionalize interagency coordination. expanded role. SCHRO should also consider A longer-term solution would involve institu- establishing strong accountability mecha- tional consolidation as part of a much nisms to enforce reform implementation at broader reform to streamline the govern- the provincial and local levels, such as “task ment’s administration systems and organiza- agreements” with provincial and local gov- tional structures (box 10.1). ernments (as further discussed below). In Sanming (as mentioned in chapter 5), concerted and coordinated actions led by a leading group at the prefecture level and but- Coordination and Support System: tressed by exceptionally strong political sup- Establish Coordination and port enabled a successful series of deep Organizational Mechanisms for reforms. The Sanming experience suggests Provincial and Local Accountability and that the leading-group arrangement can Frontline Reform Implementation effectively coordinate decision making across Strengthening of accountability arrange- multiple government departments for plan- ments, particularly at the provincial and local ning and implementing complex reforms, at levels, is another essential ingredient to facili- least in the short term. Reformers have yet to tate effective implementation of reforms. Any put in place an institutionalized platform for accountability arrangement should be suffi- coordinating stakeholders that would formal- ciently powerful to align institutional inter- ize accountability mechanisms and incentives ests and leverage government priorities when for sustained reform implementation. dealing with providers and vested interests. As it considers the organizational struc- One solution would be to form empow- tures, distribution of responsibilities, and ered leading groups or steering committees at coordination of functions across agencies for the provincial level, led by high-level leaders health system governance, China may wish to (governors or party chiefs). Leading groups review the experiences in Organisation for could also be formed at local government lev- Economic Co-operation and Development els (county, municipality, and prefecture), (OECD) countries. Both international and depending on the context. Such groups Chinese experience suggest that implementing already exist in China and could be enabled health reform is a long-term endeavor, is tech- to oversee reform implementation and sup- nically and politically complex, and requires port frontline execution. The leading groups numerous adjustments while it is in progress. would need strong, active leadership by high- Desired outcomes may take time to material- level officials, broad political support, and ize because of many intervening factors, and full empowerment (and accountability) to unintended negative consequences can occur. implement reform within their jurisdictions. In a country as large as China, flexibility is The proposed leading groups could consist of also required to allow for the wide variation representatives from the various government in starting conditions and local contexts. agencies involved in the health sector, but The leading-group arrangement can be they should also include private sector repre- strengthened to support longer-term imple- sentatives and community leaders. mentation in several ways: An advantage of the proposed leading- group arrangement is that it is a well-known • The proposed provincial leading groups mechanism for interagency coordination and could be made accountable to central gov- has been applied successfully within China’s ernment through intergovernmental S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 355 BOX 10.1  Government administrative reforms and international experience Organizational restructuring has been a major social ­i nsurance system in which the state retains ­ feature of China’s administrative reforms for several ­ regulatory functions but delegates financing to social decades (Saich 2015; Xue and Liou 2012). Policies insurance agencies and service delivery to public and have called for streamlining administrative functions (increasingly) private providers. China may want to to promote coordination and reduce the overlapping explore the institutional governance arrangements of authorities and responsibilities. More recently, of health systems based on social insurance financ- these reforms are seen as part of a broader process to ing such as in Austria, Germany, the Republic of transform government functions to enable deepen- Korea, and the Netherlands. ing of economic, social, and other sectoral reforms; In the OECD, all agencies involved in health strengthen regulations; and delegate government system governance are generally under the jurisdic- power (Li 2015). Making government agencies tion of a single governing institution responsible for more effective through streamlining functions and policy making, strategies, and regulations. Over the “building a unified supervision platform” (Li 2015) past two decades, OECD countries have enacted is also considered critical to improving reform over- governance reforms that have added national agen- sight and implementation. Whether these reforms cies (that is, for quality oversight, assessment, and will lead to institutional consolidation or creation improvement as well as for performance and regu- of an institutionalized platform for an interagency latory monitoring) while at the same time consoli- coordination in the health sector remains an open dating overlapping functions and responsibilities question. across different levels of government, including the China may consider examining organizational consolidation of social insurance funds (Jakubowski structures, distribution of responsibilities, and coor- and Saltman 2013). These reforms aimed to exert dination of functions across agencies for health greater central influence. Similarly, in part to system governance among the Organisation for address coordination, cost containment, and equity Economic Co-operation and Development (OECD) concerns, national governments have strengthened countries. Most OECD countries have an array of their decision-making power along with that of the agencies—including central line ministries, self-­ corresponding lead health organization, inducing governing bodies, professional associations, affiliated the recentralization of functions. These centralizing institutes, independent commissions, and regional trends have been noted in different systems, includ- health authorities—that constitute the governance ing those based on the tax-funded National Health configurations of the health sector. a Institutional Service in the United Kingdom and the social insur- configurations depend on the following (Jakubowski ance system in Germany. and Saltman 2013; Mossialos and others 2015): However, in countries with strongly decentral- ized systems, greater central-level authority does not • The type of system (that is, a tax-financed national always result in greater policy or policy implemen- health system or social insurance system) tation integrity. Moreover, international experience • The extent of decentralization suggests that stronger government authority should • The degree of state involvement in three core not mean, for example, government interference in health system functions: regulation, financing, and operating social insurance systems. Clear division service delivery. of roles and authorities between government health institutions and social insurance agencies combined Over the past two decades, China has been with well-defined accountability to align the lat- migrating from a tax-funded national health ser- ter with government health policies and priorities vice in which the state plays a dominant role in are critical to coherent decision-making structures regulation, financing, and service delivery to a (Savedoff and Gottret 2008). a. For more details about the institutional configurations of the health sectors in OECD countries, see Mossialos and others 2015 and the Health in Transition (HiT) Series of the World Health Organization’s (WHO) European Observatory on Health Systems and Policies: http://www.euro.who.int/en/ about-us/partners/observatory​ /publications/health-system-reviews-hits/. For similar information about Asian countries, see the online resources of the Asia Pacific Observatory on Health Systems and Policies: http://www.wpro.who.int/asia_pacific_observatory/hits/series/chn/en/. 356 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A performance contracts or task agreements, specialists and hospitals, measuring and signed with SCHRO, that specify imple- tracking outcomes, and more. Although mentation benchmarks, anticipated results these changes can and should be driven by of the reforms, and, ultimately, population- national and provincial leadership, imple- health indicators. These agreements could menting them at local sites will require assis- be assessed and revised annually or twice a tance for local learning, problem solving, year. SCHRO could consider rewards and and adaptation. sanctions related to performance. To achieve better outcomes at lower costs, • A subset of these implementation-­ providers in China need to learn new ways to performance measures should be incorpo- deliver care. To support this learning process, rated into the career-promotion system as public and private providers can come together yardsticks to measure provincial and local to form associations committed to implement- g overnment performance in reform ­ ing the people-centered integrated approach implementation. and the corresponding financial and institu- • As suggested above, performance in tional reforms. If these associations are prop- implementing agreed-upon reforms should ­ erly organized and led, participating providers be vigorously monitored by SCHRO and will benefit from not having to reinvent their independently verified by SCHRO in part- care alone and separately; they can learn nership with academic institutions. together and help each other. Associations or National and regional workshops could be groups of providers can be organized in either held to review and compare performance urban or rural settings and be made account- across provinces. These reviews would able for on-the-ground implementation of identify some higher performers, whose reforms, under the oversight of the provincial efforts could be more carefully examined leading group and aligned with the policy to learn the contextually relevant ingredi- implementation framework to be developed by ents for success that may be replicable the State Council Health Reform Leading by others. Group. These associations could help move the care systems more quickly toward a new cu lt u re of coord i nated , cooperat ive , Delivery and Learning System: Create ­outcome-­oriented care. Transformation Learning Collaboratives Drawing from international experience, we to Implement, Sustain, and Scale Up propose that Transformation Learning Frontline Reforms Collaboratives—partnerships of groups of The shift to focus on improving outcomes facilities within a county, district, or municipal- rather than just delivering treatments—that ity—should be established to implement, man- is, on value rather than procedures—will age, and sustain reforms on the front line.4 The require fundamental changes in organiza- driving vision behind the TLC concept is to tional culture. Health care organizations assist and guide local care facilities such as vil- (whether networks, hospitals, community lage clinics, township health centers, commu- health centers, or township health centers) nity health centers, and county and district hos- would greatly benefit from adopting contin- pitals to implement and scale up the reformed uou s l e a r n i n g a nd probl e m - s olv i n g service delivery model and close the gap approaches to hasten the successful imple- between “knowing” and “doing.” Provincial mentation of reforms. The service delivery (and local) leading groups can select the facility reforms recommended in earlier chapters alliances or networks, hospitals, and primary include a number of important changes at care facilities to participate in TLCs. care facilities throughout China: using evi- This approach for shared joint learning dence-based care protocols, extending among all parties in a geographic area has e-health innovations, integrating care, fol- been tried and tested all over the world, lowing clear guidelines for referral to including in Brazil, Chile, Germany, Portugal, S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 357 BOX 10.2  Evidence supporting the use of TLC methodology The Transformation Learning Collaborative (TLC) patients. Project teams tested 4,400 changes methodology has led to impressive results in several between September 1999 and August 2000, large health care systems in a number of countries, involving about 1,000 patients. Sixty-five percent and it has been adopted and locally improved by of the projects showed at least a 50 percent reduc- many organizations. These sample results are repre- tion in the time to first treatment (Griffith and sentative of hundreds more like them: Turner 2004). • The Boston-based nonprofit Partners in Health • The U.S. Veterans Health Administration (VHA) (PIH) adapted the “Breakthrough Series” model used a collaborative learning approach to reduce to improve care for people in low- and middle- waiting times in primary care clinics by 53 percent, income nations. In Peru, where 9 out of 10 ­ people from 60 days to 28 days. As the largest integrated with tuberculosis die, PIH’s patients have an delivery system in the United States, caring for more 80 percent cure rate. The program’s success per- than 6 million patients, the VHA continued to work suaded the World Health Organization (WHO) to to spread its “advanced access” modela to health add medicines for this disease to the WHO list of services across its entire system. From July 2002 to essential drugs (Shin and others 2004). October 2003, the total number of veterans waiting • Nash Health Care Systems in North Carolina in decreased from more than 300,000 to fewer than the United States reduced the average number of 50,000 (IHI 2004; Schall and others 2004). days on a ventilator by 34 percent and the average • The United Kingdom’s National Health Service length of stay by 25 percent for ventilator patients. (NHS) launched its National Primary Care Col- Cases of ventilator-associated pneumonia dropped laborative in 2000. The Collaborative is now by more than 50 percent during the collaborative. perhaps the world’s largest health care improve- Patients in the protocol group averaged more than ment project. Encompassing nearly 2,000 prac- US$35,000 savings in hospital charges, compared tices nationwide and covering almost 18.2 million with patients in the baseline group (IHI 2003). patients, it has helped to reduce the waiting time • The Singapore Healthcare Improvement Net- for an appointment with a general practitioner by work (SHINe) has used collaborative methods to an average of 60 percent (Oldham 2004). improve patient safety in all acute-care medical • The United Kingdom’s NHS Modernization institutions in the country, including hospitals that Agency formed the Cancer Services Collabora- serve the behavioral health and long-term care tive in 1999 to improve access and care for c ­ ancer needs of the population.b a. The “advanced access” model is also known as open access or same-day scheduling, in which a sizable share of the day’s appointments are reserved for patients desiring a same-day appointment (IOM 2015). b. For more information, see the SHINe website: http://shine.com.sg/. Singapore, Sweden, the United Kingdom, and use to help their members make improve- the United States. Box 10.2 summarizes some ments; and (d) proposed sequencing of inter- of these experiences and their impact. Annex ventions within a TLC. 10A describes how the TLC methodology has been used in Scotland to improve patient Principles, structures, and managerial safety and more broadly in the United philosophy of TLCs ­ Kingdom to strengthen primary care. A TLC is a structure that supports shared The rest of this subsection discusses how learning and rapid change among a group of China might structure and operationalize providers or organizations. Instead of trying TLCs, including (a) the basic principles, to achieve results alone and separately, structure, and managerial philosophy under- ­ participants in a TLC have the opportunity lying TLCs; (b) the tiered management sys- to learn together, exchange ideas and les- tem to support TLCs; (c) processes that TLCs sons learned, and share information on 358 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A measurements and results to encourage county or an urban municipality. At the start that exchange. A TLC capitalizes on the idea of service delivery reform implementation, that “two heads are better than one” and each participating province would select the that “many heads are better than a few.” most natural administrative level for the The approach moves away from defining TLC: county, district, municipal, or prefec- performance indicators, identifying under- ture. TLCs would be formed and rolled out performers, and public “naming and sham- over time, and all health care organizations ing.” Naming and shaming can generate a (public or private) within the province would culture of fear—a situation that often leads be expected to join a TLC at some point. to incomplete and distorted data, corrodes In most provinces, some combination of the spirit of innovation, and undermines the TLC types would be needed. For example, in will to improve. In the TLC model, continu- a rural setting, a TLC could consist of a ous improvement for everyone is the goal, county hospital, township health centers, vil- and everyone (even the best performers) is lage clinics, and private providers. Urban recognized as having the capacit y to TLCs could consist of tertiary hospitals, dis- improve. Facility-level teams are encouraged trict hospitals, community health centers, to test and improve new systems without community health stations, and private pro- fear of failure. Data are scrutinized, not to viders. Other combinations of facilities are identify underperformers but rather to high- also possible. Figure 10.2 displays three pos- light, celebrate, and learn from those who sible examples of TLC partnering arrange- have outperformed the rest. Recognition ments: at the county or rural level, at the and celebration of performance, not the municipal or urban level, and at the prefec- instillation of fear, is the currency of the tural rural and urban level. TLCs and drives all parties to higher levels of performance. The TLC management system The TLC model is a structure for rapidly Depending on local conditions, TLCs can be disseminating better practices to all facilities formed and overseen by provincial leading in a geographic region, whether in a rural groups (PLGs) or by local leading groups FIGURE 10.2  The TLC model in three different arrangements a. County (rural) TLC b. Municipal (urban) TLC c. Prefectural (rural and urban) TLC County hospital Tertiary hospital Prefecture hospital District hospitals County hospital County hospital THCs CHCs THC THC THC THC THC VC VC CHS CHS VC VC VC VC VC VC VC VC VC VC VC VC CHS CHS VC VC VC VC VC VC VC VC VC VC Communities Communities Local communities Local communities Local communities Local communities Local communities Source: ©World Bank. Permission required for reuse. Note: CHC = community health center; CHSs = community health services; THC = township health center; TLC = Transformation Learning Collaborative; VC = village clinic. S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 359 (LLGs). The PLG or LLG would define the TLC activities is critical for the success of any number of participating facilities and the geo- individual TLC. This set of tasks would graphical scope of the TLC, appoint leaders, include building the capability and technical invite facilities and teams to participate, and skills of TLC members in how to manage sci- host its activities. It would appoint a TLC entific improvement of systems. To acquire management team consisting of trusted local these skills, TLCs should consider forming hospital and clinic leaders, assisted by a sys- technical partnerships with leading Chinese tems improvement adviser and program man- academic institutions that would contribute agement staff from the par ticipating technical know-how and confer some of their hospitals. reput at ion a l st reng t h on t he T LC s. Given the operational nature of TLCs, International partners and technical assis- PLGs should consider making TLC manage- tance could also be provided through the ment separate from government administra- Chinese academic institutions as needed. tive leadership. Strong communication and continuous data flow between the TLCs and TLC processes to make improvements PLGs would strengthen the PLGs’ gover- Each TLC would be organized as a time- nance and stewardship roles while allowing delimited (18- to 24-month) learning system. the TLCs sufficient room to innovate. The Before launching the TLC, the PLGs or LLGs PLGs should also ensure the active participa- would agree on the specific set of reform ini- tion of multiple providers and avoid hospital tiatives to be implemented and on a set of capture of TLC leadership. The leading measures to track the implementation prog- groups could sign task agreements with the ress of all the participating facilities and insti- TLC’s leadership. tutions. For example, one reform initiative It would be important that the PLGs work could involve the transition to team-based on the more macrolevel changes and improve- care, which would facilitate care for chronic ments in the institutional and financial diseases such as diabetes. All the participat- environment across the multiple TLCs (that ­ ing facilities could track their progress using is, those at the provincial level) to remove spe- agreed-upon “process measures” (such as the cific barriers that impede progress within sev- proportion of frontline staff on the clinical eral TLCs operating in a province. For exam- care teams, the proportion of patients ple, as TLC participants seek to spread the assigned to a clinical care team, the numbers changes that are needed to produce better of annual visits by patients assigned to a care care at lower cost, they will encounter barri- team, or the numbers of medicines pre- ers that make the reforms difficult. Removing scribed) and “outcome measures” (such as those barriers would require actions by senior the percentage of diabetic patients with gly- leaders and leading groups above the level of cosylated hemoglobin levels of less than the TLC participants. For example, the TLCs 8 percent). probably would need the PLGs’ support to Figure 10.3 illustrates how a collaborative deal with issues relating to changes in human could work—from formation of facility-level resources policies, supply-chain problems, teams to attend the TLC learning sessions to reorientation of incentives, capital planning action periods bet ween meetings, all and investment, and promoting engagement s upported by various knowledge-sharing ­ with the private sector. International experi- activities. ence demonstrates that a critical function of For mat ion of fac i l it y-level tea ms. senior leadership (in this case, provincial lead- Organizations participating in the TLC ership) is to remain in touch with the TLC would send facility-level teams to the TLC members and focus on solving upstream meetings. The facility-level teams would con- problems to allow the TLCs to progress. sist of three to five people from each facility, Establishing the appropriate managerial including operational leadership and key clin- capacities to guide, support, and operate ical staff. 360 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A FIGURE 10.3  Design of a Transformation Learning Collaborative Action Action period period T T T T T T Learning Learning Learning T session 1 session 2 session 3 T T T Supporting elements: Monthly teleconferences, site visits, point assessments, data-sharing architecture T Facility-level team composed of three to five people including the operational manager, clinical leaders, and data managers Source: ©World Bank. Permission required for reuse. Learning sessions. TLC teams from all Supporting activities. Teams would sub- participating facilities would meet face-to- mit monthly progress reports and be sup- face in learning sessions every four to six ported by conference calls, peer site visits, months to discuss successes, barriers, and and web-based discussions to enable them to challenges; share better practices; and share information and learn from national describe lessons learned as they collectively experts and other health care organizations. seek to implement the reforms in primary The aim is to build collaboration and support health care. Learning sessions are a mixture member organizations as they try out new of collective will-building; didactic training ideas, even at a distance. on specific improvement skills and technical Throughout the process, teams would issues that might be pertinent to the health use a methodology known as the Plan-Do- reform; and sharing of lessons and ideas by Study-Act (PDSA) cycle to iteratively test participating teams to overcome specific ideas for improving system performance (as implementation barriers. further described in annex 10B). 5 During Learning sessions conclude with each team an action period, for example, teams would establishing a plan for implementation dur- test different ways of implementing team- ing the action period that follows each learn- based care. Teams might try different ing session. For example, in a recent learning approaches to structuring their teams or session of a collaborative that focused on different communication strategies, such as improving access to primary care, partici- a daily morning meeting to review all pants spent one quarter of the time learning assigned patients. Scheduling might take technical skills in a plenary format, one quar- various forms. Teams might test an innova- ter working in their teams to plan future tive technology for grouping patients actions, and the rest of the time exchanging according to various characteristics and practical tips and tools with other participat- conditions to perceive patterns. Teams ing teams. would use a web-based data collection por- Action periods. Between the face-to-face tal to submit monthly progress reports on TLC learning session meetings, there would the agreed-upon measures. To continue the be action periods when the facility-based example mentioned above, such measures teams would test and implement interventions might include the percentage of diabetic in their local settings and collect and report patients with glycosylated hemoglobin of data to measure the interventions’ impact. less than 8 percent or with blood pressure S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 361 under control. These data would be avail- system capable of independently assessing able to the entire TLC community for all to and verifying implementation progress and see and review. reform impacts. This could be achieved in partnership with academic institutions. Reform sequencing and measurement Based on the proposed implementation within a TLC guidelines and existing monitoring systems, As TLCs begin to be rolled out in the selected SCHRO could develop implementation reform provinces, it would be important to benchmarks and other metrics to track think about the sequence of implementation. reform implementation. Table 10.3 contains (For an example of an implementation path- examples of value-oriented indicators cate- way and guidelines, see annex 10C.) TLCs gorized by the three overarching goals of might focus on one or more of the eight the reform effort: better care, better health, reform areas. It would be difficult to predict and lower cost. which reforms each TLC would select, Regardless of the specific pathway taken because the details of their circumstances through the available reform priorities, each would likely determine which reforms are reform would need a clear, universal mea- most important to the TLC leaders. A full surement framework to help guide TLC lead- menu of the reforms should be made avail- ers and the provincial leadership groups to able to the TLC leaders at the outset, and as understand the progress being made on the soon as a team starts work, its leaders should front lines. As a particular reform matures devise a master “reform pathway” in consul- within facilities, progress would need to be tation with representatives from the partici- measured and understood so that TLC lead- pating health care facilities. ers and PLGs can encourage the TLCs to move on to new areas of reform. More operationally, the PLGs could track Monitoring and Evaluation System: the TLCs’ progress and, together with the Ensure Strong and Independent central government, monitor data on selected Monitoring and Impact Evaluation indicators of utilization, cost, quality, and The State Council may consider establish- outcome (table 10.4). The tracking of prog- ing a strong monitoring and evaluation ress should be complemented by impact TABLE 10.3  Sample indicators for monitoring health reform implementation, by reform goal Reform goal Indicators Goal 1: Achieve better • Admission rates for complications for diabetes, hypertension, and chronic lung disease in care for individuals secondary and tertiary hospitals: aim for 20 percent reduction in two years • Number of patients whose first contact for an illness episode occurs in primary care: aim for a 20 percent increase in two years • Antibiotics prescriptions at primary care facilities and outpatient clinics: aim for a 25 percent reduction in two years Goal 2: Achieve better • Percentage of population ages 18–75 years with hypertension and whose blood pressure was health for populations adequately controlled (below 140/90): aim for 20 percent improvement in two years • Percentage of patients with diabetes with hemoglobin A1c below 8 percent: aim of 20 percent improvement in two years • Percentage of women ages 16–64 years who received one or more Pap tests to screen for cervical cancer: aim of 20 percent improvement in two years Goal 3: Achieve • Inpatient admissions per 1,000 population: aim of 15 percent reduction in two years affordable costs • Length of stay: aim of 20 percent reduction in stays at secondary and tertiary hospitals in two years • Total spending per insured: aim of health cost inflation similar to consumer price inflation, as indicated by quarterly reports of social insurance agencies 362 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 10.4  Scoring system for Transformation Learning Collaboratives TLC stage Indicators 1. Formation • TLC has been formed. • Aim for implementation has been set, and baseline measurement has begun. 2. Activity • TLC is meeting regularly. • Participating teams are beginning local implementation activities. 3. Testing • Changes are being tested, but no improvements are seen yet. • Data on measures are being reported consistently. 4. Process improvement • Improvements have been recorded in processes identified as critical to achieving collaborative aim. 5. Outcome improvement • Improvements have been recorded in outcomes related to the collaborative aim. Note: TLC = Transformation Learning Collaborative. Each TLC is graded 1–5 on this scale based on how the TLC is progressing. These data can be averaged at the desired level of aggregation for performance review by provincial and national authorities. FIGURE 10.4  Sequential plan for scaling up reform This section describes a “waved” sequence implementation that could be used to achieve provincewide spread of the eight reform levers (Barker, Reid, and Schall 2016). Build Test Go to Phases of TLCs could be rolled out gradually in Set up phases to all cou nties and distric ts. scalable unit scaling-up full scale scaling-up Depending on the local context and starting Leadership, communication, social networks, culture of Adoption conditions, there might be more than one urgency and persistence mechanisms TLC per jurisdiction (such as a large munici- pality or county). Four phases would be Learning systems, data systems, infrastructure for scaling-up, Support required to spread TLCs throughout a human capacity for scaling-up, capability for scaling-up, province:6 systems sustainability • Phase 1: Set up the TLC, including the Source: ©World Bank. Permission required for reuse. provincial and local preparatory steps for implementation of reforms • Phase 2: Develop the scalable unit—a pro- totyping phase. evaluations that use rigorous methodologies • Phase 3: Test scaling-up — a phase that to measure impact and allow comparison expands the core knowledge in a variety of across sites that are implementing similar settings that are likely to represent differ- reforms. ent contexts that will be encountered at full scale Toward a Sequential Plan for Full- • Phase 4: Go to full scale—a phase that Scale Reform Implementation in unfolds rapidly to enable a larger number of sites or divisions to adopt or replicate China the intervention. The province is the recommended unit of focus for implementing health reform in Figure 10.4 illustrates this sequential China over the next five to seven years. A implementation and also shows a series of well-designed and detailed plan is needed for adoption mechanisms and support systems scaling-up across a province—that is, for that facilitate the implementation sequence. ensuring that all facilities in a province par- “Adoption mechanisms” are the system ele- ticipate in a TLC and implement the reforms. ments needed to facilitate the implementation S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 363 TABLE 10.5  TLC sequential implementation plan, by phase, time interval, and jurisdiction Phase Time interval TLC rollout in counties and districts 1. Set up Month 0 0 counties 2. Develop scalable unit Month 3 1–2 “initial” C&Ds 3. Test of scaling-up Month 12 Wave 1: 10 C&Ds 4. Full-scale rollout Month 24 Wave 2: 10 additional C&Ds Month 36 Wave 3: 10 additional C&Ds Month 48 Wave 4: 10 additional C&Ds Month 60 Wave 5: 10 additional C&Ds Month 72 Wave 6: 10 additional C&Ds Note: C&Ds = counties and districts; TLC = Transformation Learning Collaborative. of change: leadership, communication, social core action areas to design a series of more- networks, and a culture of urgency and per- specific reform interventions (as shown in sistence. “Support systems” are the health- the example in box 10.3). It is essentially system building blocks that must be strength- a guide that identifies, sequences, and, ened to enable implementation: learning to the extent possible, synchronizes the systems, data systems, infrastructure, human reform actions, adapting them to the local resources, and technical capability. setting. It would focus on the core actions Table 10.5 shows a sample sequence for presented in chapters 2–9 but would be care- the rollout of TLC across counties and dis- fully sequenced, taking into account key tricts in a hypothetical province with about elements of hospital reform, the PCIC 60 counties and districts. model, and quality improvement, as well as structural changes to payment mechanisms, purchasing arrangements, team configura- Phase 1: Set Up the TLCs tions, and workforce development that In the set-up period, provincial leaders would would need to precede certain changes in begin to build the “how” of implementation, clinical processes. No master pathway will first examining the province’s administrative be perfect, and therefore its components structures to identify where the TLCs ought to should be flexible and iterative, allowing be created. Decisions would need to be made provincial leaders to work with TLC leaders on how many urban and rural TLCs would be to amend the master reform pathway over needed, which specific facilities would join time. This phase could be accomplished specific TLCs, and which TLCs would be quickly, within three months. launched first and which in subsequent years. More in-depth implementation pathways, PLGs or LLGs would be set up, and the TLC derived from specific core action areas of the management teams would be established to master pathway, would provide more detailed begin developing the province’s first TLC. specifications for what needs to be done. Governance arrangements, measurement Annex 10A presents an example of such an frameworks, task agreements, and pathways implementation pathway to guide and would be agreed upon between national, pro- sequence activities for implementing specific vincial, and local management teams. components of the proposed PCIC model. Provincial leaders would also examine Implementation pathways provide clear the full menu of reforms and SCHRO’s objectives, overall milestones, measurable implementation guidelines to derive a master outputs, and specific activities to achieve reform pathway particular to their local cir- them. Based on contextual requirements, cumstances. Such a master pathway would each province would choose its own imple- outline the key change concepts, goals, and mentation pathways. 364 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A BOX 10.3  Example of a master reform pathway: primary care enhancement for patients with complex needs This master pathway is designed to help primary Identify individuals who are good candidates for care teams redesign care to meet the needs of your enhanced care design includes these tasks: patients living with complex needs. It illustrates a • Identify individuals through a combination of master pathway that includes ideas that teams can approaches. test and implement; tips and guidance culled from • Use real-time identification. the experience of expert faculty and primary care teams from around the world; resources to support Revolutionize patient engagement includes these teams’ progress; and examples of care models. tasks: Several key change concepts are necessary to improve the health and cost outcomes of patients • Develop processes to recruit people into care. with complex needs (figure B10.3.1). Teams with • Adopt patient engagement strategies that existing care models and those just beginning to are tailored to and informed by your focus develop programs for the complex-needs population population. are encouraged to work through this design process Develop an enhanced care model to fit the needs and to refine and focus care interventions. assets of the focus population includes these tasks: Choose your BHLC (Better Health at Lower Cost) • Co-create individualized care plans to learn about population and learn about its assets and needs and prepare for care redesign. includes the following tasks: • Develop your transformed care model through • Identify your overall population. iterative testing. • Identify population segments and select the focus • Develop work processes to ensure consistent care population for intervention. delivery. FIGURE B10.3.1  Sample “Better Health at Lower Cost” reform pathway Identify Develop an Choose your Strengthen individuals who enhanced care BHLC population Revolutionize partnerships are good model to fit the and learn about patient within and candidates for needs and assets its assets and engagement outside of your your enhanced of the BHLC needs organization care design population Source: ©World Bank. Permission required for reuse. Note: BHLC = Better Health at Lower Cost. (Box continued next page) S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 365 BOX 10.3  Example of a master reform pathway: primary care enhancement for patients with complex needs (continued) Strengthen partnerships within and outside of your and effectively and to work at scale, execute these organization includes these tasks: strategies: • Strengthen community partnerships to meet the needs • Understand the data and how to use them and enhance the strengths of the focus population. • Cultivate ongoing investment to build a sustain- • Cultivate a healthy team in your enhanced care able program design. • Develop a learning system Finally, the data must be used effectively. Even • Work at scale the best models will fail if implementation is poorly • Scale up services to all individuals in the target accomplished. To help teams use data efficiently population. Phase 2: Develop the Scalable Unit Phase 3: Test of Scaling-Up The “scalable unit” is the smallest represen- This phase involves testing the set of inter- tative unit of the system targeted for full- ventions to be taken to scale. The successful scale implementation. The county, district, or strategies that aided implementation in the municipality would be the ideal scalable unit initial TLCs would need to be tested in a within the province. This is where the action broader range of settings before going to happens for implementation and where the full scale. International experience suggests TLC would be operationalized. that testing should take place in 10 addi- In each geographical area that is targeted tional TLCs in each of the selected reform in a province, at least one and preferably provinces starting in Year 2 of the reform more than one initial TLC would be set up in period. the first year. The purpose of these initial During this phase, all necessary infra- TLCs is to intensively test local ideas for best- structure required to support full-scale practice implementation. An important out- ­ i mplementation should be documented, come of this work would be a set of well-­ understood, and adjusted as needed, documented, context-sensitive strategies to including workforce development (for aid implementation of specific reforms that example, leadership, managerial, and front- could be further tested and refined.7 line capacity); information systems manage- The choice of facilities to participate in ment; and the supply chain. This phase is an this initial phase of implementation is of the important opportunity to build the confi- utmost importance. Research on change dence and will of leaders and frontline staff management and the diffusion of innovation to support the changes. As the work pro- suggests it is good practice to identify front- ceeds, new insights from the reform imple- runner innovators who have the will and mentation will lead to a more nuanced and motivation to make a change. Further, expe- mature set of context-­specific strategies and rience in China and internationally has ideas for change that can be used for full- shown that strong political commitment is scale implementation throughout the prov- needed to overcome entrenched interests in ince. This phase would last one year. the health sector, make the difficult choices involved, and bring about the relentless Phase 4: Go to Full Scale focus on execution that is needed to achieve results. This phase will last approximately This is a rapid deployment phase in which a nine months. tested set of reforms within each province, 366 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A now supported by a reliable data feedback and supported in the early stages so that system, can be rapidly adopted by frontline TLCs can be prepared to take on the role of staff throughout the province. While some mentoring subsequent TLC participants. In adaptation of the intervention to local envi- this way, successes are multiplied across the ronments may still be required, there would province and transformation is greatly be less emphasis on contextual adaptation accelerated. during this phase. Significant will, knowl- As shown earlier (in table 10.3), the sug- edge, experience, and infrastructural support gested plan for achieving provincewide imple- and capacity need to be in place before mov- mentation is to spread the reforms in succes- ing to this phase of scaling-up. sive annual waves of 10 counties and districts At this point, a series of waves of TLCs until the full province is covered. After the would be launched within each of the prov- first year (in which one or two initial TLCs inces selected for reform (as shown in are established in each province), the second table 10.5). Each wave of scaling-up would year would see TLCs launched in the next be informed by the knowledge gained from wave of jurisdictions (counties, districts, and the previous wave. The best-performing TLC municipalities) as a test of scaling-up. A year participants from early waves may coach new after that, the next round of 10 counties and TLC teams in subsequent waves. This devel- districts would be launched, and so on until opmental step needs to be explicitly described the full province is covered. Annex 10A  Case Examples of Transformation Learning Collaboratives (TLCs) Case 1: The Scottish Patient Safety Goals Programme The initial aim of the SPSP collaborative was to reduce in-hospital mortality by 15 percent Background over five years. To achieve this larger mor- Since 1999, Scotland has had its own tality objective, the collaborative’s leader- National Health Service, NHS Scotland, ship set specific goals, such as to reduce independent from the NHS in the rest of the “crash calls” by 30 percent and reduce United Kingdom. Scotland has 36 acute-care Staphylococcus aureus bacteraemia infec- hospitals, supervised by 14 local health tions by 30 percent. An additional goal was boards. NHS Scotland has a long history of to field-test the collaborative methodology focusing on value, safety, and innovation. In in the NHS for the first time. If the test 2004, it participated in the United Kingdom’s proved successful, the plan was to apply this Safer Patients Initiative, a large-scale methodology to primary care and postacute improvement program that tested organiza- care as well. tionwide ways to improve patient safety within hospitals. Through this program, one hospital, Ninewells Hospital, reduced patient Implementation harm by more than 60 percent in three years. The model for collaborative learning was Inspired by that success, in 2007 Scotland introduced in four clinical areas: general clin- lau nched the S cot tish Patient Safet y ical wards, critical care units, perioperative Programme (SPSP) —a Transformation theatres, and stewardship and use of medica- Learning Collaborative (TLC) to improve tions. To spread successful changes and share patient safety in all of the country’s hospitals. learning, teams at different hospitals came This collaborative proved so successful that it together to participate in face-to-face learn- has now been replicated by NHS Scotland in ing sessions where they learned from their primary care, long-term care, mental health, peers’ successes and challenges in trying to and maternity care.8 implement the same changes. S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 367 In preparation for this collaborative work, Between the in-person events were the Scotland engaged extensively in building action periods when teams returned to their motivation and knowledge at the leadership hospitals to test and implement changes. level and established crucial organizational During these times, clinicians from each of infrastructure for improvement. The national the four clinical areas participated in monthly leadership set the policy framework, and calls, reporting progress and setbacks and local health boards developed focused imple- discussing them with other teams. There were mentation plans and communicated them to also regional facilitators who became well the hospital staffs. Safety was deemed a pri- acquainted with the hospitals in their region, ority, as reiterated by leaders at all levels of conducted site visits, and provided support. the system in discussions at health board Frequent informal discussion also took place meetings and safety walk-arounds in hospi- throughout the collaborative. Leaders reached tals (during which leaders discuss safety out to chief executives, board chairs, medical issues with frontline staff). directors, nurse directors, and patient groups The local health boards were also asked to discuss planning and progress, and teams to appoint a program manager for patient spoke frequently to other teams when they safety, and later an executive leader, for had questions or encountered challenges. each of the four clinical areas. A leadership Clinical teams were asked to select a series team (including leaders from the Scottish of improvements from the SPSP collaborative’s government and Scottish nongovernmental toolkits, but implementation was decidedly a organizations [NGOs]) was formed specifi- local activity with local teams encouraged to cally for the program. The Scottish govern- make changes to the interventions as needed ment also set up the National Advisory to tailor them to their own environments and Board, chaired by the chief medical officer, experiences. For example, for one intervention to oversee the program, with representa- procedure, one organization set up a cart to tives from the N HS, government, and transport supplies to the bedside, while patient groups. another developed a procedure pack contain- To develop capability for implementation ing all the necessary supplies for one patient. and improvement, more than 200 clinicians What mattered was that each team decided on across the 36 hospitals received extensive an approach to implementation that worked instruction in the science of implementation for them, and adhered to it rigorously. and improvement. Board members and lead- Data management was another key ­ element. ers also received instruction in providing Clinicians were required to enter all patient- appropriate leadership to steward these care data into a web-based data ­portal that was efforts effectively. developed expressly for the SPSP. To minimize The structure of the SPSP included bian- the workload, data already collected by other nual nationwide meetings (learning sessions), programs—such as on infection rates—were during which the leadership team and the integrated automatically. local health board teams from the four clini- After an initial focus on acute care, the cal areas met to share their experiences in initiative gradually expanded to incorporate overcoming obstacles and solving problems. mental health, primary care, and pediatric Importantly, prominent national leaders such and maternal care. Each clinical community as the chief executive of NHS Scotland has its own collaborative learning structure. attended these sessions, sending the message In November 2014, delegates from all four that the collaborative was a national priority communities came together to share learning and that teams would be held accountable. In at a national conference. addition, regular events were held to convene 50–300 clinicians for one or two days to Outcomes build skills in measurement, testing changes, Through late 2012 and early 2013, SPSP and spreading innovation. achieved its aim of reducing hospital 368 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A mortality by 15 percent. Other specific out- Goals comes include (a) a 66 percent reduction in The goals of the NPDT were to create mea- ventilator-acquired pneumonia; (b) a 70 percent surable improvement in access to primary reduction in the central-line bloodstream care; improve the management of patients infection rate; (c) a 79 percent reduction in with established coronary heart disease; and Clostridium difficile (commonly called “C. build capacity- and demand-management diff ”) infections in patients 65 and older; systems between primary and secondary (d) a 15 percent decrease in the crash call care. All primary care practices participating rate; and (e) at least a 21 percent increase in in the collaborative were asked to work on all medication reconciliation. three areas at the same time. The broader goal was to develop a team that would create capacity and capability for Case 2: The Primary Care Collaborative, improvement in primary care group clinical United Kingdom (Prepared in practice. Government leaders were commit- Collaboration with Sir John Oldham) ted to sustainable, long-term change. Background In 1997, a new government was elected Implementation in the United Kingdom on a promise The initial collaborative structure began with to improve public services, in particular 20 PCGs or PCTs and was closely managed the N HS. The National Primary Care by the NPDT. Each PCG or PCT was invited Development Tea m (N PDT), a small to select five group practices to participate. national leadership group, was formed. The Thus, about 100 primary care group prac- NPDT was charged with improving primary tices participated in the first “wave” of the care throughout the country by reducing collaborative. waiting times and improving the effective- At the individual clinic level, each primary ness of service delivery. To spread change, care practice would choose a full-time project the team used the collaborative methodology manager, employed by the site but trained by described in this chapter through a project and accountable to the national collaborative called the Primary Care Collaborative. In management team. The project managers’ late 2000, the government launched the responsibilities were to coach the core prac- NHS Plan, a policy framework for reform- tices, ensure that data were submitted, gener- ing the health service and, along with it, the ate local awareness, and recruit new PCGs or Primary Care Collaborative as its “effector” PCTs. They received bimonthly training in or implementation plan. This policy frame- coaching and group skills, process flow and work envisaged primary care redesigned redesign, data management, communication around the patient; “seamless” service; and and spread techniques, and improvement sci- a focus on patient experience and outcomes ence. They also created a community among (Oldham 2004). themselves, sharing challenges and solutions. In the United Kingdom at the time, pri- They had a dedicated area on the Primary mary care doctors typically worked together Care Collaborative website to facilitate infor- in small group practices of three to four phy- mation sharing in this community. sicians. These small group practices were To expand the national learning collabor- organized into primary care groups (PCGs) ative, the initial five participating practices that were reorganized later into primary care would coach the remaining 15–20 colleague trusts (PCTs). At the start of the collabora- practices within each PCG or PCT. This led tive, there were 310 PCGs or PCTs, each to faster spread of better practice ideas for ­ c ontaining about 20–25 physician group policy implementation: there was an almost practices. These PCGs or PCTs were the scal- palpable unleashing of the power of peer-to- able units that joined the Primary Care peer coaching and knowledge building. The Collaborative. five initial participating practices linked the S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 369 remaining 15–20 practices within the same 1 million patients were engaged in this PCG or PCT to the collaborative formally, national transformative initiative, and there through their coaching and mentoring. was strong momentum for change. Typically, the collaborative effort would run This was good progress, but how to reach for two years. In Year 1, the initial five sites the remaining 3,000 practices covering would focus on learning and on making 30 million additional patients? changes to their own practices; in Year 2, Here the NPDT needed a different solu- they would seek out their colleagues and tion. It created regional nodes: small semiau- coach them on the better practices and imple- tonomous administrative units that were mentation ideas that they had learned. typically headquartered at the site of one of To expand this further, the N PDT the best performers from the nationally led launched a series of four subsequent collaborative. These nodes were called NPDT “waves,” each of which was its own collab- centers and were tasked by the NPDT to orative of 20 PCG or PCTs of 100 primary coordinate regional collaborative waves that care practices. Each proceeded along similar essentially duplicated the formula used for lines to the initial collaborative. Year 1 the national collaborative implementation focused on changes within the initial partici- just described. pating sites. Year 2 focused on coaching the remaining practices to adapt and adopt the Outcomes reforms. The launch of the waves was stag- What was achieved by all this implementa- gered with four to five months between the tion activity? Within three years, 4,900 start of each wave. Each collaborative wave practices were engaged in the nationwide was chaired by a knowledgeable and experi- initiative, covering 31 million people. By enced clinician who had credibility with par- Januar y 20 0 4, the collaborative had ticipants. The main role of the chairpersons achieved a 72 percent improvement in access was to lead the in-person learning sessions, to primary care doctors (that is, reduced chair meetings during and between the waiting time) and a fourfold reduction in learning sessions, and represent the collab- mortality due to cardiovascular disease in orative to a wider audience. the participating collaborative sites com- Throughout the two years of a wave, the pared with the rest of the United Kingdom. core practices, and the spread practices as Most importantly, it had established a they joined, supplied monthly data. By the robust, sustainable infrastructure for end of 2001, some 1,500 practices covering improvement and spread. Annex 10B  Plan-Do-Study-Act Cycles of Change The Plan-Do-Study-Act (PDSA) cycle guides the conduct of small-scale tests to improve individuals and organizations to test ideas system performance. for change systematically to determine whether the change can generate viable The PDSA Cycle improvement. PDSA cycles have emerged from a long tradition of hypothesis testing A PDSA cycle includes four key steps, as fur- and change management in both science and ther outlined below. Plan refers to a specific industry. Originally expressed by statistician planning phase. Do is a phase for trying the Walter Shewhart as the Plan-Do-Check-Act change and observing what happens. Study is cycle, it was later refined by W. Edwards the phase for analyzing the results of the trial. Deming to its current Plan-Do-Study-Act Act is the phase for devising next steps based form (Kiran 2016). This simple four-step on the analysis (keep the change, adjust the process has been used in a variety of health change, or discard the change and try some- care and non-health care settings to guide thing else). 370 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Step 1: Plan—Plan the test or observation, • Setting aims for improvement: The aim including a plan for collecting data  should be time-specific and measurable; it should also define the specific population • State the objective of the test. of patients or other system that will be • State the questions you want to answer, affected. Examples of an aim would be and make predictions about what will that all diabetic patients receive optimal happen and why. c a r e a c c o rd i n g to Wo rld H e a lt h • Develop a plan to test the change. (Who? Organization (WHO) guidelines within What? When? Where? What data need to the next year, or that all identified tuber- be collected?) culosis patients are managed according to W HO’s I ntegrated Management of Step 2: Do—Try out the test on a small scale  Adolescent and Adult Illness (IMAI) guidelines in the next 18 months. • Carry out the test.  • Establishing measures: Teams use quanti- • Document problems and unexpected tative measures to determine whether a observations.  specific change actually leads to an • Begin analysis of the data. improvement. For the aims cited above, measures might include the percentage of Step 3: Study—Set aside time to analyze the diabetic patients with glycosylated hemo- data and study the results  globin of less than 8 percent or the percent- age of identified tuberculosis patients who • Complete the analysis of the data.  convert to sputum- or culture-negativity. • Compare the data to your predictions.  • Selecting changes: Ideas for change may • Summarize and reflect on what was come from the insights of people who learned. work in the system, whether from change concepts or other creative thinking tech- Step 4: Act—Refine the change, based on niques or by borrowing from the experi- what was learned from the test  ence of others who have successfully improved. • Determine what modifications should be • Testing changes: The PDSA cycle is short- made.  hand for testing a change in the real work • Prepare a plan for the next PDSA cycle. setting—by planning it, trying it, observ- ing the results, and acting on what is learned. This is the scientific method Methods for PDSA Implementation adapted for action-oriented learning. The PDSA cycle can be implemented in any • Implementing changes: After testing a organization and by any team by using these change on a small scale, learning from methods: each test, and refining the change through several PDSA cycles, the team may imple- • Forming the team: Including the right ment the change on a broader scale—for ­ people on an implementation team is criti- example, for an entire pilot population or cal to a successful improvement effort. on an entire unit. Teams vary in size and composition. They • Spreading changes: After successful imple- might include some combination of pri- mentation of a change or package of mary care physicians, specialists, nurses, changes for a pilot population or an entire and other health care workers. Each orga- unit, the team can spread the changes to nization builds teams to suit its own needs other parts of the organization or in other and the improvement goal at hand. organizations. S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 371 Annex 10C  Example of a Detailed Implementation Pathway Lever 1: Shaping Tiered Health Care table 10C.1), using population empanelment Delivery System with People- Centered and risk stratification (Pathway 1.2 Integrated Care [table 10C.2] and Pathway 1.3 [table 10C.3]) The core action areas of Lever 1 selected for • Core Action Area 2: Functional multidis- the sample implementation pathway include ciplinary teams (Pathway 1.3, as shown in the following (further detailed in chapter 2): table 10C.3) • Core Action Area 7: Measurement, stan- • Core Action Area 1: Primary care as the first dards, and feedback (Pathway 1.4 point of contact (Pathway 1.1, as shown in [table 10C.4] and Pathway 1.5 [table 10C.5]) TABLE 10C.1  Pathway 1.1: Increase the use of primary care as first point of contact Pathway 1.1 rationale: Having primary health care perform gatekeeping functions limits specialty care access and can help systems reduce overuse of inappropriate care and move toward providing the right care at the right place at the right time. Milestone Outputs Activities Milestone 1: Protocol for Output 1: Means of concordant Activity 1: Assess community and provider perceptions of gatekeeping community-concordant gatekeeping is determined. functions, including severity of concerns about restricted choice. gatekeeping is developed Activity 2: Determine package of incentives and consequences for gatekeeping that encourages in concert with providers and insurance mechanism. Get input from a variety individuals to make their of local and provincial stakeholders, especially regarding successful and primary care provider problematic gatekeeping efforts in the past. the entry point into the health system Activity 3: Determine options for financial incentives to encourage PHC when they need care, utilization and discourage patients from bypassing to specialist care. while also maintaining Output 2: Patient education about Activity 1: Provide physicians with the knowledge necessary to build strong, elements of patient primary health care and services is long-lasting relationships with their patients. choice that promote trust. provided to increase community Activity 2: Reach out to local patient groups and the public around a campaign Timeline: Years 1–2 trust in the health care system. to encourage utilization of PHC services first. Milestone 2: Pilot program Output 1: Begin institutionalizing Activity 1: Depending on results of community input above (during Milestone 1 to test whether agreed- incentives and processes activities), develop metrics and measure capture systems to monitor Output 2. upon gatekeeping devised under Milestone 1. scheme is under way. Output 2: Measure uptake of Activity 1: Set indicators such as visit rates at PHC facilities, bypass rates for a set Timeline: Years 1–2 gatekeeping. of conditions that could be treated at PHC facilities, and patient and provider experience with the system. Activity 2: Set monitoring systems to measure indicators (see Pathway 1.5 [table 10C.5]). Output 3: Use continuous Activity 1: Establish feedback system for providers and patients to assess their learning, and feedback from perspectives on strong and weak points of the gatekeeping model (see providers and clients, to inform Pathway 1.5 [table 10C.5]). future iterations of the model. Milestone 3: Increase Output 1: Increase the hours Activity 1: Pilot offering weekend and evening hours at PHC facilities on a accessibility of primary that primary care services are limited basis. care services to patients available. Activity 2: Publicize campaign about after-hours care as an alternative to long using existing facilities wait times and high out-of-pocket costs for specialist or hospital visits. for care delivery. Timeline: Years 1–2 Activity 3: Provide financial incentives to PHC providers to offer more available times for patient visits. Activity 4: Measure the effect of these access policies on visit rates and use of specialist services in affiliated network providers. Output 2: Offer same-day visits Activity 1: Work with PHC providers in two counties to save at least 30 percent to patients with acute needs. of visit slots each day for acute or urgent walk-in patients. Activity 2: Measure the effect of this open-access plan on the volume of visits and nearby hospital utilization. Note: PHC = primary health care. 372 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 10C.2  Pathway 1.2: Stratify panels based on risk for poor outcomes and high utilization, and develop care plans for most at-risk population Pathway 1.2 rationale: Proactively targeting individuals at risk for poor outcomes and high utilization of health services enables provision of higher-intensity, coordinated care in a PHC setting. Milestone Outputs Activities Milestone 1: Empaneled Output 1: Algorithm(s) for risk Activity 1: Determine goals for risk stratification (such as to improve health or populations are stratified stratification are developed. reduce costs). by risk. Activity 2: Determine which algorithm for risk stratification fits best Timeline: Years 1–2 with local needs, including disease burden, stated goals, and available resources. Potential algorithms used elsewhere include Adjusted Clinical Groups, Hierarchical Condition Categories, Elder Risk Assessment, Chronic Comorbidity Count, Charlson Comorbidity Index, and Minnesota Health Care Home Tiering. Activity 3: Pilot the algorithm(s) using historical data and determine the threshold for designation as “high-risk.” Output 2: Risk stratification of Activity 1: Beginning with populations already empaneled to MDTs, use the patients is initiated. algorithm(s) to identify “high-risk” patients per panel. Activity 2: Determine how often to update the list(s) of “high-risk” empaneled patients (for example, every three months or six months). Activity 3: Assign high-risk patients for longitudinal care management by one member of the MDT. Milestone 2: Individualized Output 1: Care pathways for Activity 1: Select the highest-burden and/or highest-cost diseases for care plans for high-risk the highest-burden diseases targeting. patients are developed. are developed. Activity 2: Assemble MDTs of experts from across care levels (primary care Timeline: Years 1–2 and hospitals) for consultation. Activity 3: Review literature to determine the best evidence-based practices for treatment. Activity 4: Develop standardized care pathways that clarify team members’ roles and responsibilities at each level of care, establish explicit referral criteria, and establish postdischarge follow-up procedures. Output 2: Care pathways for Activity 1: Establish a monitoring framework for assessing care pathway a small number of selected processes and associated health outcomes (see Pathway 1.5 [table 10C.5]). diseases are piloted. Activity 2: Begin applying care pathways for selected disease(s). Activity 3: Institute regular review meetings with MDTs across care levels to assess bottlenecks to implementation, propose solutions, and set new coverage and completeness targets. Note: MDT = multidisciplinary team; PHC = primary health care. S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 373 TABLE 10C.3  Pathway 1.3: Form multidisciplinary teams and empanel population Rationale: Multidisciplinary team (MDT) formation and population empanelment promote the accountability of primary health care system and providers for population health outcomes; shift care from hospital-centric to people-centered integrated care; improve patient-provider relationships and trust; and alleviate the consequences of acute shortages of human resources for health by maximizing reach and efficacy through multidisciplinary teams. Milestones Outputs Activities Milestone 1: Care Output 1: Location- Activity 1: Assess the disease burden, demographic profile, and preferences of the provision is specific optimal local population. reorganized into composition of MDTs is Activity 2: Assess available human resources, including physicians, registered nurses, nurse multidisciplinary teams determined. practitioners, community health workers, and other health personnel such as mental health Timeline: Years 1–2 experts, pharmacists, nutritionists, and social workers. (Work to harmonize these efforts with the human resource reforms recommended in chapter 7.) Activity 3: Conduct a rapid inventory of skills and technical areas for which staff at lower- facility levels need support. Activity 4: Delineate each team member’s roles and responsibilities and designate a team leader. Activity 5: Develop technical assistance support plans that are appropriate for the local setting and meet the needs of lower-level staff. Plans should include details on the frequency of assistance provision, the topics to be covered, and the roles and responsibilities of staff at all levels. Potential activities include on-site mentoring at primary care facilities by higher-level staff (including outpatient services, inpatient rounds, case discussions, and lectures) or embedding primary care staff at higher-level facilities for extended training opportunities. Output 2: MDTs are Activity 1: Conduct outreach events (seminars, roundtable discussions, and so on) with formed and providing health care providers to explain the rationale for transitioning to team-based care, answer services. A culture questions, and address concerns. of cooperation is Activity 2: Group the care providers according to the criteria established in Output 1. fostered. Activity 3: Provide opportunity for teams to review their roles and responsibilities and adapt to local and team needs. Activity 4: Designate a care coordinator on each team. The exact training and professional status of this person can vary across regions; what is important is that the care coordinator fulfills a standardized function of tracking high-risk patients across care settings and proactively reaching out to them. Activity 5: Roll out the MDTs in a small pilot setting. Activity 6: Plan and implement regular team-based trainings to promote a collaborative culture with supportive supervision. Milestone 2 Output 1: The Activity 1: Weigh the options for empanelment, based on community feedback and (concurrently guidelines and available technical capacity. Options include purely geographic empanelment, with Milestone 1): protocol for choosing empanelment based on patient choice, or a combination of the two. Empanelment based The population is and implementing on patient choice and the combination approach require the technical capacity and real- empaneled to MDTs. different empanelment time communication to track patients if (or as) they switch between providers to ensure Timeline: Years 1–2 strategies are that gaps in care are minimized. prepared. Activity 2: Test different empanelment approaches in a small sample of diverse geographic areas. Activity 3: Develop patient registries and the information technology (IT) and human resource capacities needed to maintain and use them. Activity 4: Define the panel type and size based on these protocols. Output 2: Provider and Activity 1: Conduct training events for multidisciplinary provider teams to explain the community buy-in rationale for empanelment and to clarify provider roles in the new system. for empanelment Activity 2: Conduct community outreach events to provide information on the goals of is generated. empanelment and to explain what these changes mean for how patients access care. Empanelment is initiated. Activity 3: Register people in the panels and assign them to MDTs. Note: MDT = multidisciplinary team. 374 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 10C.4  Pathway 1.4: Review pilot and scale up Rationale: A systematic review of the progress made on pilots facilitates scaling-up. Milestone Outputs Activities Milestone 1: Spread the Output 1: Develop the algorithm(s) Activity 1: Determine the goals for risk stratification (for example, to rollout of risk stratification for risk stratification based on the improve health or to reduce costs) within the new geographic area. and care pathways. needs and goals of the new locale. Activity 2: Determine which algorithm for risk stratification fits best with Timeline: Years 3–4 local needs (including the disease burden) and available resources. Activity 3: Pilot the algorithm using historical data and determine the threshold for designation as “high-risk.” Output 2: The expanded pool of Activity 1: Within new geographic areas, use the algorithm to identify patients is risk stratified. “high-risk” patients per panel. Activity 2: Determine how often to update list of “high-risk” empaneled patients (for example, every three months or six months). Activity 3: Assign high-risk patients for longitudinal care management by one member of the MDT. Output 3: Care pathways are Activity 1: Establish a monitoring framework for assessing the care applied to newly risk-stratified pathway processes and associated health outcomes (see Pathway 1.5 patient panels. [table 10C.5]). Activity 2: Begin applying the care pathways for selected disease(s). Activity 3: Institute regular review meetings with MDTs across care levels to assess bottlenecks to implementation, propose solutions, and set new coverage and completeness targets. Output 4: Care pathways are revised Activity 1: Assemble local MDTs of experts from across care levels (primary based on the technical capacity of care and hospitals) for consultation. local providers and the population Activity 2: Update the care pathways based on new feedback from burden of disease. local experts. Milestone 2: Clinical pathways Output 1: The highest-burden or Activity 1: Assemble MDTs of experts from across care levels (both primary are developed for new highest-cost diseases not already care and hospitals) for consultation. diseases. covered by clinical pathways are Activity 2: Review literature to determine the best evidence-based Timeline: Years 3–4 identified. practices for treatment. Activity 3: Develop standardized care pathways that clarify the roles and responsibilities of team members at each level of care, establish explicit referral criteria, and establish postdischarge follow-up procedures. Milestone 3: Establish Output 1: Each clinical pathway Activity 1: Conduct listening sessions with local providers to determine process for regular review is assessed at least once every bottlenecks to implementation, technical problems with clinical of pathways. two years. pathways, or other perceived needs for change. Timeline: Years 3–4 Activity 2: Review new literature to ensure fidelity of disease-specific clinical pathways to the most recent evidence base. Activity 3: Make necessary updates to the pathway based on Activities 1 and 2. Activity 4: Review revised pathway(s) with local stakeholders from multidisciplinary care teams across care levels. (Table continued next page) S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 375 TABLE 10C.4  Pathway 1.4: Review pilot and scale up  Continued Milestone Outputs Activities Milestone 4: MDT capacity Output 1: Use of MDTs is expanded Activity 1: Determine location-specific optimal composition of the MDTs. is spread and deepened beyond pilot sites, following the Activity 2: Facilitate formation of teams and a culture of cooperation; through technical assistance process described in Milestone 1. continuous training is essential. between upper- and lower- level facilities. Output 2: Health workers staffing Activity 1: Identify appropriate facilities for scaling-up within the tiered Timeline: Years 3–4 community and township health care system, including county hospitals, THCs, and village clinics centers (THCs) receive technical Activity 2: Conduct an inventory of skills and technical areas for which assistance from specialists in staff at lower-facility levels need support. county and district hospitals. Activity 3: Develop technical assistance support plans that are appropriate for the local setting and meet needs of lower-level staff. Plans should include details on frequency of assistance provision, topics to be covered, and roles and responsibilities of staff at all levels. Potential activities include on-site mentoring at primary care facilities by higher-level staff (including outpatient services, inpatient rounds, case discussions, and/ or lectures) or embedding primary care staff at higher-level facilities for extended training opportunities. Activity 4: Implement technical assistance activities within defined groups of care teams. Activity 5: Continually assess patient perspectives, provider perspectives and knowledge, and health outcomes (see Pathway 1.5 [table 10C.5]). Milestone 5: Empanelment Output 1: Guidelines and protocols Activity 1: Assess available options, local technical capacity, and of populations to are developed or fine-tuned for community needs. multidisciplinary teams choosing and implementing Activity 2: Test different empanelment approaches. is scaled up to additional empanelment strategies for new localities. locales. Activity 3: Develop patient registries. Timeline: Years 3–4 Activity 4: Define panel type and size based on protocols. Activity 5: Register the panels to MDTs. Output 2: Provider and community Activity 1: Conduct a training event for multidisciplinary provider teams buy-in for empanelment is to explain the rationale for empanelment and clarify provider roles in the achieved. new system. Activity 2: Conduct community outreach events to provide information on the goals of empanelment and to explain what these changes mean for how patients access care. Note: MDT = multidisciplinary care team; THC = township health center. TABLE 10C.5  Pathway 1.5: Implement a continuous M&E plan to track reform progress and inform iterative improvements Rationale: Continuous monitoring and evaluation, implemented as an integral part of the reforms from the start, is critical to the success of reforms. Ongoing measurement helps to ensure that reforms are undertaken with fidelity to the planned process; enables the collection of feedback from critical reform stakeholders (providers, clients, and communities); allows for early detection and correction of unintended negative consequences; and provides up-to-date data with which to track outcomes. Milestones Outputs Activities Milestone 1: Conduct Output 1: The measurement Activity 1: Design an M&E framework and identify key processes, intermediate continuous M&E of system is prepared. outputs, and outcomes to track, including critical PCIC functions such as care multidisciplinary team rollout. continuity and coordination, patient feedback, and provider perspectives. Timeline (continuous with Activity 2: Assess the available data sources for tracking key indicators identified Pathway 1.1): Years 1–4 in the M&E plan. Activity 3: Develop new data collection tools to track indicators not captured by existing data sources, as needed. (Table continued next page) 376 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE 10C.5  Pathway 1.5: Implement a continuous M&E plan to track reform progress and inform iterative improvements  Continued Milestones Outputs Activities Output 2: Ongoing M&E Activity 1: Implement continuous monitoring of the identified indicators and quality improvement Activity 2: Iterate within the pilot site to address any identified problems or are conducted. improvement opportunities. Activity 3: Continue the measurement and improvement cycle for expanded rollout of MDTs in Years 3–4. Milestone 2: Conduct Output 1: The measurement Activity 1: Design an M&E framework and identify key processes, intermediate continuous M&E of system is prepared. outputs, and outcomes to track, including critical measures of key PCIC empanelment rollout. functions such as care continuity and coordination, patient feedback, and Timeline (continuous with provider perspectives. Key indicators to track include the following: Are the Pathway 1.1): Years 1–4 panel sizes within the guidelines and manageable for teams? Do the MDTs know the patients and communities they are responsible for? Activity 2: Assess the available data sources for tracking key indicators identified in the M&E plan. Activity 3: Develop new data collection tools to track indicators not captured by existing data sources, as needed. Output 2: Ongoing M&E Activity 1: Implement continuous monitoring of the identified indicators. and quality improvement is Activity 2: Iterate to address any identified problems or improvement conducted. opportunities. Activity 3: Continue the measurement and improvement cycle for expanded rollout of empanelment in Years 3–4. Milestone 3: Conduct Output 1: Assess Activity 1: Based on the criteria used to select a risk stratification algorithm, continuous M&E of risk appropriateness of risk identify key indicators to track to determine whether most at-risk patients stratification and clinical stratification strategy. are being identified. For example, if high risk for cardiovascular outcomes is a pathway rollout. criterion for risk stratification, indicators to track might include CHADS2 scores, Timeline (continuous with the ACC-AHA CVD risk calculator, and others. Pathway 1.2): Years 1–4 Activity 2: At regularly scheduled intervals, compare populations identified as high-risk with those not designated as high-risk on selected outcome measures such as admissions for MI, CVA, COPD, and PAD. Activity 3: If the risk stratification algorithm is not adequately capturing high-risk patients, change as needed. Output 2: The measurement Activity 1: Design an M&E framework, including identifying key processes, system is prepared. intermediate outputs, and outcomes to track. Activity 2: Assess available data sources for tracking key indicators identified in the M&E plan. Activity 3: Develop new data collection tools to track indicators not captured by existing data sources, as needed. Output 3: Ongoing M&E Activity 1: Building on the scheduled pathway review meetings, implement and quality improvement is continuous M&E of the pilot, including measures of key PCIC functions such as conducted. care continuity and coordination, patient feedback, and provider perspectives. Activity 2: Iterate within the pilot site to address any identified problems or improvement opportunities. Note: ACC-AHA CVD = the American College of Cardiology and American Heart Association cardiovascular disease risk prediction algorithm; CHADS2 = stroke risk assessment score in atrial fibrillation (standing for congestive heart failure, hypertension, age ≥ 75, diabetes mellitus, stroke); COPD = chronic obstructive pulmonary disease; CVA = cerebrovascular accident; MDT = multidisciplinary team; M&E = monitoring and evaluation; MI = myocardial infarction; PAD = peripheral artery disease; PCIC = people-centered integrated care; PHC = primary health care. S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 377 Notes the most successful approaches was to pay physicians a salary instead of paying them 1. Such an environment will also be needed for based on the volume of procedures performed. effective and sustained implementation and This had an immediate impact on C-section scaling-up. rates and became a key strategy that other 2. “Guidance on Comprehensively Scaling-Up organizations implemented also. Reform of County-Level Public Hospitals,” 8. For more information about the Scottish State Council General Office (Guo Ban Fa Patient Safety Programme, see https://ihub​ 2015, No. 33); “Guidance of the General .scot/spsp/about-us/. Office of the State Council on Promoting Multi-level Diagnosis and Treatment System,” State Council General Office (Guo Ban Fa References 2015, No. 70). Aarons, G. A., M. Hurlburt, and S. M. Horwitz. 3. For more information, see the Institute of 2011. “Advancing a Conceptual Model of Medicine website on the Roundtable on Value Evidence-Based Practice Implementation in & Science-Driven Health Care: http://iom​ Public Service Sectors.” Administration and .nationalacademies.org/Activities/Quality​ Policy in Mental Health and Mental Health /­VSRT.aspx. Service Research 38 (1): 4–23. 4. This subsection draws on the following Aarons, G. A., D. Summerfield, D. B. Hecht, sources for evidence: Brush and others 2009; J. F. Silovsky, and M. J. Chaffin. 2009. “The Franco and M arquez 2011; Hulscher, I mpac t of Evidence -B ased Prac tice Schouten, and Grol 2009; IHI 2003; Jones Implementation and Fidelity Monitoring on and Piterman 2008; Kritchevsky and others Staff Turnover: Evidence for a Protective 2008; and Schouten and others 2008. Effect.” Journal of Consulting and Clinical 5. The Plan-Do-Study-Act (PDSA) cycle guides Psychology 77 (2): 270–80. individuals and organizations to test ideas for Barker, Pierre M., A my Reid, and Marie change systematically to determine whether W. Schall. 2016. “A Framework for Scaling Up the change can generate a viable improvement. Health Interventions: Lessons from Large- PDSA cycles have emerged from a long tradi- Scale Improvement Initiatives in Africa.” tion of hypothesis testing and change manage- Implementation Science 11 (1): 1–11. ment in both science and industry. Briefly, the Berwick, Donald M., Thomas W. Nolan, and cycle works as follows: Teams thoroughly John Whittington. 2008. “The Triple Aim: plan to test the change, taking into account Care, Health, and Cost.” Health Affairs 27 (3): cultural and organizational characteristics. 759–69. They do the work to make the change in their Brush, John E. Jr., Edna Rensing, Frank Song, standard procedures, tracking their progress Sallie Cook, Janet Lynch, Leroy Thacker, Sarat using quantitative measures. They closely Gurram, Robert O. Bonow, Joani Brough, and study the results of their work for insight on C. Michael Valentine. 2009. “A Statewide how to do better. They then act to make the Collaborative Initiative to Improve the Quality successful changes permanent or to adjust the of Care for Patients with Acute Myocardial changes that need more work. This process Infarction and Heart Failure.” Circulation continues iteratively over time, and refinement 119 (12): 1609–15. is added with each cycle. For a more detailed Cutler, David. 2014. The Quality Cure: How description, see annex 10B. Focusing on Health Care Quality Can Save 6. Similar efforts to scale reforms using the same Your Life and Lower Spending Too. Berkeley approach have been used in the United and Los Angeles, CA: University of California Kingdom, for example, with great success, Press. leading to major improvements in waiting Damschroder, Laura J., David C. Aron, Rosalind times, cardiovascular care, and patient E. Keith, Susan R. Kirsh, Jeffery A. Alexander, satisfaction. and Julie C . Lowery. 20 09. “Fostering 7. For example, in a project seeking to reduce Implementation of Health Services Research cesarean sections in Brazil, teams tested vari- Findings into Practice: A Consolidated ous approaches to reduce financial incentives Framework for Advancing Implementation to perform this procedure. Ultimately, one of Science.” Implementation Science 4 (1): 50. 378 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A Dewan, N. A., D. Conley, D. Svendsen, S. P. He, Jingwei Alex. 2011. “Combating Healthcare Chon, J. R. Staup, and A. L. Miller. 2003. “A Cost Inflation with Concerted Administrative Qualit y I mprovement Process for Actions in a Chinese Province.” Public I mplementing the Texas A lgorithm for Administration and Development 31 (3): Schizophrenia in Ohio.” Psychiatric Services 214–28. doi:10.1002/pad.602. 54 (12): 1646–49. Hei l ma n n , S ebast ia n. 20 08. “Pol ic y D u rl a k , J . A ., a nd E . P. D u P ré. 2 0 0 8 . Experimentation in China’s Economic Rise.” “Implementation Matters: A Review of S t u di e s in C o mparative Int e r n ation al Research on the Influence of Implementation Developments 43 (1): 1–26. doi:10.1007/ on Program Outcomes and the Factors s12116-007-9014-4. Affecting Implementation.” American Journal Huang, Yanzhong. 2009. “An Institutional of Community Psychology 41 (3–4): 327–50. A nalysis of China’s Failed Healthcare doi:10.1007/s10464-008-9165-0. Reform.” In Socialist China, Capitalist China: Eccles, Martin P., and Brian S. Mittman. 2006. Social Tension and Political Adaptation under “Welcome to Implementation Science .” Economic Globalization, edited by Guoguang Implementation Science 1 (1): 1. Wu and Helen Landsowne, 75–86. New York: Etheredge, L. M. 2014. “Rapid Learning: A Routledge. Breakthrough Agenda.” Health Af fairs Hulscher, Marlies, Loes Schouten, and Richard (Millwood) 33 (9): 1155–62. Grol. 2009. Collaboratives. Quest for Quality Fixsen, Dean L., Sandra F. Naoom, Karen A. and Improved Performance (QQUIP) Series. Blasé, Robert M. Friedman, and Frances London: The Health Foundation. Wallace. 2005. “Implementation Research: A IHI (Institute for Healthcare Improvement). 2003. Synthesis of the Literature.” Monograph, “The Breakthrough Series: IHI’s Collaborative National Implementation Research Network Model for Achieving Breakthrough at the Louis de la Parte Florida Mental Health Improvement.” Innovation Series white paper, Institute, University of South Florida, Tampa. IHI, Boston. Fra nco, L . M ., a nd L . M a rque z . 2011. ———. 2004. “The Courage to Act on ‘What If “Effectiveness of Collaborative Improvement: . . . ’: 2004 Progress Report.” IHI, Boston. Evidence from 27 Applications in 12 Less- I O M ( I n s t i t u t e o f M e d i c i n e) . 2 0 1 5 b . Developed and Middle-Income Countries.” “Transforming Health Care Scheduling and BMJ Quality & Safety 20 (8): 658–65. Access: Getting to Now.” Washington, DC: Friedman, C., J. Rubin, J. Brown, M. Buntin, The National Academies Press. M. Corn, L. Etheredge, C. Gunter, and others. Jakubowski, E., and R. B. Saltman, eds. 2013. The 2014. “Toward a Science of Learning Systems: Changing National Role in Health System A Research Agenda for the High-Functioning Governance: A Case-Based Study of 11 Learning Health System.” Journal of American E u r o p e a n C o u n t r i e s a n d A u s t ra l i a . Medical Informatics Association 22 (1): Observatory Studies Series 29. Copenhagen: 43–50. doi:10.1136/amiajnl-2014-002977. World Health Organization (acting as the host Garside, Pam. 1998. “Organisational Context for organization for, and secretariat of, the Q u a l it y : L e s s on s f rom t h e Field s of European Observatory on Health Systems and Organisational Development and Change Policies). Management.” Quality in Health Care 7 Jones, Kay, and Leon Piterman. 2008. “The (suppl): S8–S15. Effectiveness of the Breakthrough Series Greene, S. M., R. J. Reid, and E. B. Larson. 2012. Methodology.” Australian Journal of Primary “Implementing the Learning Health System: Health 14 (1): 59–65. From Concept to Action.” Annals of Internal Kiran, D. R. 2016. Total Quality Management: Medicine 157 (3): 207–10. doi:10.​ 7 326/​ Key Concepts and Case Studies. Oxford, U.K., 0003-4819-157-3-201208070-00012. and Cambridge, MA: Butterworth-Heinemann. Griffith, Clive, and Jill Turner. 2004. “United doi:10.1016/B978-​0-12-811035-5.00003-9. Kingdom National Health Service, Cancer Kritchevsky, S. B., B. I. Braun, A. J. Bush, M. R. S e r v i c e s C ol l ab or at ive ‘ I mprove m e nt Bozikis, L. Kusek, J.-P. Burke, E. S. Wong, and Partnership,’ Redesign of Cancer Services: others. 2008. “The Effect of a Quality A National Approach.” European Journal of I mprovement Collaborative to I mprove Surgical Oncology 30 (Suppl 1): 1–86. A nti m icrobial Prophyla x is in Su rg ical S T R E N G T H E N I N G T H E I M P L E M E N TAT I O N O F H E A LT H S E R V I C E D E L I V E RY R E F O R M 379 Patients: A Randomized Trial.” Annals of Qian, Jiwei. 2015. “Reallocating Authority in the Internal Medicine 149 (7): 472–80. Chinese Health System: An Institutional Li, Keqiang. 2015. “Streamline Administration, Perspective.” Journal of Asian Public Policy, Delegate Power, Strengthen Regulation and 8 (1): 19–35. doi:10.1080/17516234.2014.​ Improve Services to Deepen Administrative 1003454. R efor m s a nd Tra n sfor m G over n m e nt Ramesh, M., Xun Wu, and Alex Jingwei He. Functions.” National teleconference address, 2013. “Health Governance and Healthcare May 12, State Council, the People’s Republic Reforms in China.” Health Policy and of China, B eijing. ht tp: //english.gov.cn​ Planning 29 (6): 663–72. doi:10.1093/heapol​ /­premier/speeches/2015/05/23/content​_ 28147​ /­czs109. 5113213526.htm. Rogers, E. M. 2010. Diffusion of Innovations. Mate, Kedar S., Brandon Bennett, Wendy New York: Simon and Schuster. Mphatswe, Pierre Barker, and Nigel Rollins. Saich, Tony. 2015. Governance and Politics of 2009. “Challenges for Routine Health System China. 4th ed. London: Palgrave. Dat a M a n agement i n a L a r ge P ubl ic Savedoff, William D., and Pablo Gottret. 2008. Programme to Prevent Mother-To-Child HIV Governing Mandatory Health Insurance: Transmission in South Africa.” PloS One 4 (5): Learning from Experience. Washington, DC: e5483. World Bank. doi:10.1596/978-0-8213-7548-8. Meessen, Bruno, and Gerald Bloom. 2007. S cha l l , M a r ie W., Ter renc e Du f f y, A n i l “Economic Transition, Institutional Changes Krishnamurthy, Odette Levesque, Prashant and the Health System: Some Lessons from Mehta, Mark Murray, Renee Parlier, Robert Rural China.” Journal of Economic Policy Petzel, and John Sanderson. 2004. “Improving Reform 10 (3): 209–31. Patient Access to the Veterans Health Meyers, Duncan C., Joseph A. Durlak, and Administration’s Primary Care and Specialty Abraham Wandersman. 2012. “The Quality Clinics.” The Joint Commission Journal on Implementation Framework: A Synthesis of Quality and Patient Safety 30 (8): 415–23. Critical Steps in the Implementation Process.” Schouten, Loes M. T., Marlies E. J. L. Hulscher, American Journal of Community Psychology 50 Jannes J. Evan Everdingen, Robbert Huijsman, (3–4): 462–80. doi:10.1007/s10464-​012-​9522-x. and Richard P. T. M. Grol. 2008. “Evidence Meyers, Duncan C., Jason Katz, Victoria Chien, for the Impact of Quality Improvement Abraham Wandersman, Jonathan P. Scaccia, Collaboratives: Systematic Review.” BMJ 336 a nd A n n i e Wr i g ht . 2 01 2 . “ P r ac t i c a l (7659): 1491–94. Implementation Science: Developing and Shin, Sonya, Jennifer Furin, Jaime Bayona, Kedar Piloting the Quality Implementation Tool.” Mate, Jim Yong Kim, and Paul Farmer. 2004. American Journal of Community Psychology “Community-Based Treatment of Multidrug- 50 (3–4): 481–96. Resistant Tuberculosis in Lima, Peru: 7 Years Mossialos, Elias, Martin Wenzl, Robin Osborn, of Experience.” Social Science & Medicine and Chloe Anderson, eds. 2015. “International 59 (7): 1529–39. Profiles of Health Care Systems, 2014: Wandersman, Abraham, Victoria H. Chien, and Australia, Canada, Denmark, England, France, Jason Katz. 2012. “Toward an Evidence-Based Germany, Italy, Japan, The Netherlands, New System for I nnovation Suppor t: Tools, Z ealand, Nor way, Singapore, Sweden, Training, Technical Assistance, Quality Switzerland, and the United States.” Annual A ssu ra nce / Qua l it y I mprovement for publication, The Commonwealth Fund, Implementing Innovations with Quality to New York. Achieve Desired Outcomes.” Americ an Oldham, John. 2004. Sic Evenit Ratio Ut Journal of Community Psychology 50 (3–4): Componitur: The Small Book about Large 445–59. System Change. Chichester, U.K.: Kingsham Wandersman, Abraham, Jennifer Louis Duffy, Press. Paul D. Flaspohler, Rita Noonan, Keri Lubell, Peters, David H., Nhan T. Tran, and Taghreed Lindsey Stillman, Morris Blachman, Richard Adam. 2013. Implementation Research in Dunville, and Janet Saul. 2008. “Bridging the Health: A Practical Guide. Geneva: Alliance Gap bet ween Prevention Research and for Health Policy and Systems Research, World Practice: The Interactive Systems Framework Health Organization. for Dissemination and Implementation.” 380 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A American Journal of Community Psychology Xue, Lan, and Kuotsai Tom Liou. 2012 . 41 (3–4): 171–81. “Government Reform in China: Concepts and W HO (World Health Organization). 2007. Reform Cases.” Review of Public Administration Everybody’s Business: Strengthening Health 32 (2): 115–33. doi:10.1177/0734371X12438242. Systems to Improve Health Outcomes: WHO’s Yong, Pierre L., LeighAnne Olsen, and J. Michael Framework for Action. Geneva: WHO. McGinnis. 2010. Value in Health Care: Wilson, S. J., M. W. Lipsey, and J. H. Derzon. Accounting for Cost, Quality, Safety, Outcomes, 2 0 03 . “ T h e E f f e c t s of S c ho ol - B a s e d and Innovation: Workshop Summary. Learning Intervention Programs on Aggressive Behavior: Healthcare System Series, Roundtable on Value A Meta-Analysis.” Journal of Consulting and & Science-Driven Health Care. Washington, Clinical Psychology 71 (1): 136–49. DC: National Academies Press. A Supplementary Tables TABLE A.1  Eight levers and recommended core actions for high-quality, value-based health service delivery in China Levers (strategic directions) Core action areas Lever 1: Shaping tiered 1. Primary health care as the first point of contact health care delivery 2. Multidisciplinary teams system in accordance with people-centered 3. Vertical integration, including new roles for hospitals integrated care (PCIC) 4. Horizontal integration model 5. Advanced information and communication technology (e-health) 6. Integrated clinical pathways and functional dual referral systems 7. Measurement, standards, and feedback 8. Accreditation and certification Lever 2: Improving 1. Organizational structure to lead the creation of an information base and development of quality of care (QoC) in strategies for quality improvement support of PCIC 2. Systematic QoC measurement and continuous use of resulting data to support quality improvements 3. Transformation of management practice to improve QoC in health facilities Lever 3: Engaging 1. Health literacy citizens in support of the 2. Self-management practices PCIC model 3. Shared decision making 4. Supportive macro environment for citizen patient engagement in health promotion and improvement Lever 4: Reforming 1. Strong accountability mechanisms for autonomous public hospitals to strengthen public hospital performance governance and 2. Incentives aligned with public objectives and accountabilities management 3. Sound organizational arrangements for public hospital governance 4. Gradual delegation of decision rights to hospitals 5. Managerial capacity building (Table continued next page) 381 382 H E A LT H Y C H I N A : D E E P E N I N G H E A LT H R E F O R M I N C H I N A TABLE A.1  Eight levers and recommended core actions for high-quality, value-based health service delivery in China  Continued Levers (strategic directions) Core action areas Lever 5: Realigning 1. Provider payment reforms in support of PCIC incentives in purchasing 2. Coherent, consistent incentives and stronger integration of care and provider payment 3. Rational distribution of services by facility level 4. Capacity building of insurance agencies to equip them to become strategic purchasers Lever 6: Strengthening 1. Strong enabling environment for development of primary health care (PHC) workforce to the health workforce implement PCIC 2. Balanced workforce composition and competency for PHC service delivery 3. Compensation system with strong incentives for good performance 4. Headcount quota system reform to enable a more flexible health labor market and efficient health workforce management Lever 7: Strengthening 1. A clear, shared vision of the private sector’s potential contribution to health system goals private sector 2. Regulatory and enforcement capacity to steer health service delivery toward social goals engagement in health service delivery 3. A level playing field across public and private providers Lever 8: Modernizing 1. Shift from traditional input-based planning toward people-centered planning health service planning 2. Engagement with all relevant stakeholders and local communities in the planning to guide investment process 3. Empowerment and enabling of regions and provinces to develop their own capital investment plans 4. Introduction of a Certificate of Need program to evaluate and approve new capital investments 5. Prioritization of community health projects TABLE A.2  Government policies in support of the eight levers for health service delivery reform in China Levers Government policy statements in support of each lever References 383 Lever 1: •• Adhere to the people-first principle, and attach primary importance to safeguarding the •• Opinions of the CPC Central Committee and the State Council on Shaping rights and interests of the people’s health. Deepening the Health Care System Reform (Zhong Fa 2009, No. 6) tiered health •• Adhere to the tenet of serving the people’s health with health care undertakings; regard •• The Notification on Health Sector “Twelfth Five-Year Plan” (Guo Fa care delivery safeguarding the people’s health as the center, and take the entitlement of basic health 2012, No. 57) system in care services to everyone as the fundamental aim and outcome. •• Suggestions of the CPC Committee on the 13th Five-Year Plan for accordance •• Emphasize the combination of prevention, treatment, and rehabilitation. Strengthen the National Economic and Social Development with people- prevention of chronic diseases. •• Guidance of the General Office of the State Council on Overall Pilot centered •• Make community health the “gatekeeper.” Strengthen the three-tiered health service net Reform of Urban Public Hospitals (Guo Ban Fa 2015, No. 38) integrated in rural areas. Improve the health service system based on the TCH. •• State Council General Office Opinions on the Full Implementation care (PCIC) •• Promote the construction of the health care information system. Take advantage of the of Comprehensive Reform in the County Level Public Hospital model network information technology to promote the cooperation between urban hospitals (Guo Ban Fa 2015, No. 33) and community health service institutions. •• Planning Layout of National Medical and Health Services System •• Establish a coordinated service system, and on the basis of enhancing grassroots services, (2015–2020) (Guo Ban Fa 2015, No. 14) optimize allocation of resources with the application of legal, social, administrative, •• Construction Planning of Grassroots Health Professionals Focusing and market tools to improve the quality of medical care and guide reasonable medical on the General Practitioners (Fa Gai She Hui 2010, No. 561) treatment. •• Guidance of the General Office of the State Council on •• Establish the multilevel diagnosis and treatment model characterized by initial diagnosis Promoting Multi-Level Diagnosis and Treatment System by grassroots institutions, a two-way referral system, separate treatments for urgent and (Guo Ban Fa 2015, No. 70) chronic disease, and close cooperation between hospitals at different levels. •• Guiding Opinions on Further Regulating Community Health Service •• Focusing on the cultivation of general practitioners, establish the system of basic medical Management and Improving Health Service Quality (Guo Wei Ji Ceng and health personnel training. Fa 2015, No. 93) Lever 2: •• Strengthen regulation of health care service behavior and quality, improve the health care •• Zhong Fa 2009, No. 6 Improving service standards and quality evaluation system, regulate the management system and •• Guo Fa 2012, No. 57 quality of work flows, quicken the formulation of the treatment protocols, and complete the health •• Guo Ban Fa 2015, No. 38 care (QoC) care service quality surveillance networks. •• Guo Ban Fa 2015, No. 33 in support of •• Enhance the management and control of medical quality. Clinical examination, diagnosis, •• Suggestions of the CPC Committee on the 13th Five-Year Plan for PCIC treatment, drug use, and the use of implant medical apparatus should be regulated. National Economic and Social Development Lever 3: •• Strengthen health promotion and education. Carry out health education, strengthen the •• Zhong Fa 2009, No. 6 Engaging dissemination of medical and health knowledge, advocate a healthy and civilized lifestyle, •• Guo Fa 2012, No. 57 citizens in promote rational nutrition among the public, and enhance the health awareness and •• Guo Ban Fa 2015, No. 38 support of the self-care ability of the people. •• Guo Ban Fa 2015, No. 33 PCIC model •• Build sound and harmonious relations between health care workers and patients. •• Guo Ban Fa 2015, No. 14 •• Investigate timely to irrational use of drugs, material, examination, and repetitive •• Suggestions of the CPC Committee on the 13th Five-Year Plan for examinations for economic benefit. National Economic and Social Development •• Promote the transparency of hospital information and establish a regular display system, •• Guo Wei Ji Ceng Fa 2015, No. 93 including financial situation, performance assessment, quality safety, price and inpatient cost, and so on. (Table continued next page) 384 TABLE A.2  Government policies in support of the eight levers for health service delivery reform in China  Continued Levers Government policy statements in support of each lever References Lever 4: •• Transform government functions and promote separation of functions of government •• Zhong Fa 2009, No. 6 Reforming agencies and public institutions as well as separation of administration and business •• Guo Fa 2012, No. 57 public operations. •• Guo Ban Fa 2015, No. 38 hospital •• Perfect the management mechanism of public hospitals, and complete the corporate •• Suggestions of the CPC Committee on the 13th Five-Year Plan for governance legal person management system. National Economic and Social Development and •• Promote innovation in modern hospital management, promote the professional management specialization of dean teams, and improve public hospital management. Implement the autonomous right of public hospitals, such as in personnel management, internal distribution, and operations management. Lever 5: •• Along with economic and social development, efforts should be made to uplift the •• Zhong Fa 2009, No. 6 Realigning fund-raising and pooling levels step by step, narrow the gap between different insurance •• Guo Fa 2012, No. 57 incentives in schemes, and eventually achieve the fundamental unity of those schemes. •• Guo Ban Fa 2015, No. 38 purchasing •• Explore the establishment of an integrated urban and rural health insurance scheme. •• Guo Ban Fa 2015, No. 33 and provider •• Implement the reform of the mode of health insurance payment. •• Opinions on Implementing the Control of Total Medical Insurance payment •• Utilize the fundamental function of health insurance, strengthen the budget for revenues Payment (Ren She Bu Fa 2012, No. 70) and expenditures of medical insurance fund, and establish various payment methods, •• Notification of Pilot of DRGs Reform (Fa Gai Jia Ge 2011, No. 674) in which payment according to the type of disease is the major form and other forms •• Opinions on Further Improving the Reform of Health Insurance like payment by person and payment by service unit may also be used. Promote the Payment (Ren She Bu Fa 2011, No. 63) diagnosis-related group (DRG) system. •• Guo Ban Fa 2015, No. 70 •• Establish effective, open, and fair negotiation mechanisms and risk-sharing mechanisms •• Suggestions of the CPC Committee on the 13th Five-Year Plan for between the insurance agencies and designated medical institutions. National Economic and Social Development •• Establish the restriction mechanism of medial expense growth; control the unreasonable growth. •• Implement the basic health insurance settlement directly as well as cost accounting and control. The various health insurances should regulate, control, supervise, and restrict the behavior of medical services and medical prices, effectively control medical costs, and regulate the medical service behavior of the working staff. (Table continued next page) TABLE A.2  Government policies in support of the eight levers for health service delivery reform in China  Continued 385 Levers Government policy statements in support of each lever References Lever 6: •• Promote the medical talent system and innovation of mechanisms. •• Zhong Fa 2009, No. 6 Strengthening •• Establish the reasonable incentives of income distributions, and improve the treatment •• Guo Fa 2012, No. 57 the health of medical staff. Establish a personnel system and salary system suitable for the medical •• Guo Ban Fa 2015, No. 38 workforce industry. The salary of the medical staff should not be linked to profit. •• Guo Ban Fa 2015, No. 33 •• Implement the system of comprehensive performance evaluation and post •• Guo Ban Fa 2015, No. 14 performance-based salary in line with service quality and workload, and effectively •• Several Opinions on Promoting and Regulating Doctors’ Multi-Sited mobilize the initiatives of health care workers. Practice (Guo Wei Yi Fa 2014, No. 86) •• Deepen the reform of the headcount quota system. In terms of headcount setting, •• Guo Ban Fa 2015, No. 70 income distribution, professional title evaluation, management and deployment, •• Suggestions of the CPC Committee on the 13th Five-Year Plan for and personnel inside or outside the authorized size should be considered as a whole, National Economic and Social Development and the reform of endowment insurance system should be carried out according to •• Guo Wei Ji Ceng Fa 2015, No. 93 national regulation. •• Adopt the employment system and post management system, and establish a flexible employment mechanism. Ensure the autonomous right of public hospitals in recruiting people. •• Promote registered physicians’ multisited practice. Lever 7: •• Encourage and guide social capital to sponsor health care undertakings. •• Zhong Fa 2009, No. 6 Strengthening •• Promote the development of nonpublic health care institutions, and form a health •• Guo Fa 2012, No. 57 private sector care system with multiple categories of investors and diversified investment modes. •• Guo Ban Fa 2015, No. 38 engagement Encourage social forces to invest in the medical industry through funding of new •• Guo Ban Fa 2015, No. 33 in health construction or participating in restructuring. •• Guo Ban Fa 2015, No. 14 service •• Encourage and promote the incentives of nonpublic hospitals. •• Several Policy Measures to Accelerate the Development of Medical delivery •• Further ease entry requirements. Institutions Sponsored by Social Force (Guo Ban Fa 2015, No. 45) •• Carry out the tax policy of nonpublic hospitals. •• Notification on Launching the Pilot of Establishing Wholly Foreign- •• Carry out the same policies with the public hospitals when the nonpublic hospitals are Owned Hospitals (Guo Wei Yi Han 2014, No. 244) designated medical institutions. •• State Council General Office Opinions on Further Encouraging and •• Improve classification management of medical institutions, and introduce the regulation Guiding the Social Capital to Hold a Medical Institution (Guo Ban Fa of nonprofit hospitals, such as the nature of business and the usage of surplus. 2010, No. 58) •• Suggestions of the CPC Committee on the 13th Five-Year Plan for National Economic and Social Development •• Guo Wei Ji Ceng Fa 2015, No. 93 (Table continued next page) 386 TABLE A.2  Government policies in support of the eight levers for health service delivery reform in China  Continued Levers Government policy statements in support of each lever References Lever 8: •• Strengthen regional health planning. •• Zhong Fa 2009, No .6 Modernizing •• Optimize medical resources allocation. •• Guo Fa 2012, No. 57 health service •• Plan resources in a differentiated manner at different levels. At the city level and below, •• Guo Ban Fa 2015, No. 38 planning basic medical services and public health resources will be planned according to size of •• Guo Ban Fa 2015, No. 33 to guide population and service radius; at the provincial level and above, resources will be planned •• Guo Ban Fa 2015, No. 45 investment according to needs and priorities in different regions. •• Guo Ban Fa 2015, No. 14 •• Instruct the health facilities to procure equipment in a rational manner according to •• Guo Ban Fa 2015, No. 70 their functions, skill competency, disciplinary development, and the health needs of the •• Suggestions of the CPC Committee on the 13th Five-Year Plan for general public and in the spirit of resource sharing. National Economic and Social Development •• The implementation condition of planning should be taken as the basis of the hospital construction, financial investment, performance assessment, medical insurance payment, personnel allocation, and beds arrangement. The constraint of planning should be enhanced, and the execution condition of the planning should be made public regularly. TABLE A.3  New policy guidelines on tiered health service delivery and recommended core actions 387 Policy guideline Levers supporting policy guideline Core actions supporting policy guideline 1. First diagnosis at the grassroots Shaping tiered health care delivery system in accordance with Primary health care as the first point of contact people-centered integrated care (PCIC) model (Lever 1) 2. Dual referral Shaping tiered health care delivery system in accordance with Integrated clinical pathways and functional dual referral systems PCIC model (Lever 1) 3. Interaction between the upper and Shaping tiered health care delivery system in accordance with the Vertical integration, including new roles for hospitals grassroots levels PCIC model (Lever 1) 4. Specify diagnosis and treatment Shaping tiered health care delivery system in accordance with the Vertical integration, including new roles for hospitals functions of medical institutions of PCIC model (Lever 1) different grades and categories 5. Enhance capability building of the Strengthening the health care workforce (Lever 6) Strong enabling environment for development of primary health care grassroots health care team workforce to implement PCIC Compensation system with strong incentives for good performance 6. Enhance grassroots capability in Shaping tiered health care delivery system in accordance with the Vertical integration, including new roles for hospitals health care PCIC model (Lever 1) Realigning incentives in purchasing and provider payment (Lever 5) Rational distribution of services by facility level Strengthening private sector engagement in health service A clear shared vision on the private sector’s potential contribution to delivery (Lever 7) health system goals 7. Consolidate sharing of regional Realigning incentives in purchasing and provider payment (Lever 5) Correct and realign incentives to reverse the current irrational medical resources distribution of service by level of facilities 8. Speed up health care informatization Shaping tiered health care delivery system in accordance with the Advanced information and communication technology (e-health) PCIC model (Lever 1) 9. Improve mechanism for reasonable Shaping tiered health care delivery system in accordance with the Vertical integration, including new roles for hospitals allocation of medical resources PCIC model (Lever 1) Realigning incentives in purchasing and provider payment (Lever 5) Rational distribution of services by facility level 10. Improve medical insurance payment Realigning incentives in purchasing and provider payment (Lever 5) Coherent, consistent incentives and stronger integration of care system reform Rational distribution of services by facility level 11. Establish and improve the profit Realigning incentives in purchasing and provider payment (Lever 5) Capacity building of insurance agencies to equip them to become distribution mechanism strategic purchasers 12. Structure a division of labor and Shaping tiered health care delivery system in accordance with the Vertical integration, including new roles for hospitals coordination mechanism for medical PCIC model (Lever 1) institutions Note: “New Policy Guidelines” refers to the “Guidance of the General Office of the State Council on Promoting Multi-Level Diagnosis and Treatment System” (Guo Ban Fa 2015, No. 70). Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, we leverage electronic publishing options and print- on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. We follow the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on Forest Stewardship Council (FSC)– certified paper, with nearly all containing 50–100 percent recycled content. The recy- cled fiber in our book paper is either unbleached or bleached using totally chlorine-free (TCF), processed chlorine–free (PCF), or enhanced elemental chlorine–free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://www​.worldbank.org/corporateresponsibility. ISBN 978-1-4648-1263-7 World Health Organization Western Pacific Region SKU 211263