73350 Good Practice in Health Financing Lessons from Reforms in Low and Middle Income Countries Version for Review June 6, 2007 Health, Nutrition and Population (HNP) Human Development Network The World Bank Table of Contents Acknowledgments .................................................................................................... 3 Executive Summary ................................................................................................. 5 Part 1 ......................................................................................................................... 9 1. Introduction .............................................................................................................................. 10 Background ...................................................................................................................... 10 Audience .......................................................................................................................... 14 2. Health Financing Functions ..................................................................................................... 15 3. Criteria for Defining 'Good Practice' and Choosing Country Cases ..................................... 18 4. Summaries of Country Cases ................................................................................................... 29 Chile ................................................. ................................................................................ 29 Colombia .............................................. ............................................................................ 32 Costa Rica ........................................................................................................................ 36 Estonia ............................................................................................................................. 40 Kyrgyz Republic ............................................................................................................... 43 Sri Lanka .......................................................................................................................... 46 Thailand ........................................................................................................................... 48 Tunisia ............................................... ............................................................................... 51 Viet Nam ................................................... ........................................................................ 54 5. Conclusion - Enabling Factors for Expanding Coverage ....................................................... 57 Appendix 1- Definitions ........................................................................................ 70 References ............................................................................................................... 73 Part 2 ....................................................................................................................... 87 Country Case Studies ..................................................................................................................... 87 Chile ................................................................................................................................. 88 Colombia ....................................................................................................................... 132 Costa Rica ...................................................................................................................... 187 Estonia ........................................................................................................................... 23 7 Kyrgyz Republic ........................................................................................................... 279 Sri Lanka ....................................................................................................................... 319 Thailand ........................................................................................................................ 360 Tunisia ........................................................................................................................... 390 Viet Nam ....................................................................................................................... 448 Good Practice in Health Financing, p. 3 ACKNOWLEDGMENTS Many individuals have contributed to making this report a reality. The country case studies have all been prepared by country-based experts who have first-hand knowledge of the health financing and health coverage arrangements - and the history of health policy reforms - in their respective countries. The work has been guided by a Steering Committee representing all six of the World Bank's regions. The Steering Committee in conjunction with the Health, Nutrition, and Population unit of the Human Development Network developed criteria for the definition of a 'good practice' and provided recommendations on choice of country cases as well as appropriate experts to prepare the cases. The HNP unit under the guidance of George Schieber, Pablo Gottret and Hugh Waters (Johns Hopkins University) with the assistance of Ajay Tandon, Adam Leive, Lisa Fleisher, Bjorn Ekman (Health Economics Program, Lund University, Sweden), and Axel Rahola (now an Advisor to the Minister of Budgeting, France) conceived and carried out this study. The peer reviewers are Ruth Levine (Director of Programs and Senior Fellow, Center for Global Development), John Langenbrunner (Senior Health Economist, MENA), and Christoph Kurowski (Senior Health Specialist, LCSHD). Toomas Palu provided extremely useful comments on the Estonia case study. Other key individuals contributing to this initiative include: Individual Study Authors: Chile - Ricardo D. Bitnin and Urcullo C. Gonzalo, Bitnin y Asociados. Colombia - Diana Pinto, Fedesarrollo. Costa Rica - James Cercone, Sanigest Health Solutions. Estonia - Triin Habicht, Head of Health Economics Department, Estonian Health Insurance Fund; and Jarno Habicht, Head of WHO Country Office, World Health Organization Regional Office for Europe. Kyrgyzstan - Melitta Jakab, WHO Health Policy Advisor, Kyrgyz Republic; and Elina Manjieva, WHO and Center for Health Systems Development, Kyrgyz Republic. Sri Lanka - Ravi P. Rannan-Eliya and Lankani Sikurajapathy, Institute for Health Policy, Colombo, Sri Lanka. Thailand - Suwit Wibulpolprasert and Suriwan Thaiprayoon, Ministry of Health, Thailand. Tunisia - Chokri Arfa, National Institute of Labor and Social Studies, Tunisia; and Hedi Achouri, Director General of Public Health Facilities, Ministry of Public Health, Tunisia. Vietnam - Bjorn Ekman, Health Economics Program, Lund University, Sweden; and Sarah Bales, Health Policy Unit (HPU), Ministry of Health, Viet Nam. Good Practice in Health Financing. p. 4 Steering Committee: Paul Gertler (Former Chief Economist, Human Development Network, Professor of Economics, Haas School of Business, University of Cali fomi a, Berkley). Robert Holzmann (Director, HDNSP). Cristian Baeza (Acting Director, HDNHE). Kei Kawabata (Sector Manager, HDNHE). Akiko Maeda (Sector Manager, MNSHD). Eduard Bos (Lead Population Specialist, HDNHE). Alexander Preker (Lead Health Economist, AFTH2). Peter Berman (Lead Health Economist, SASHD). Adam Wagstaff (Lead Economist, DECRG). Mead Over (Lead Economist, DECRG). Abdo Yazbeck (Lead Health Economist, WBIHD). Logan Brenzel (Senior Health Specialist, HDNHE). John Langenbrunner (Senior Health Economist, ECSHD). Paolo Belli (Senior Health Economist, SASHD). Mark Charles Dorfman (Senior Economist, HDNSP). Good Practice in Health Financing, p. 5 EXECUTIVE SUMMARY This report focuses on nine countries that have carried out, or are in the process of carrying out, health financing reforms that have either resulted in significant expansions of health care coverage for their populations or sought to maintain such coverage following a severe and prolonged political or economic shock (e.g., following the collapse of the former Soviet Union). In doing so, the report seeks to expand the evidence base on successful health financing reforms in low and middle-income countries. The countries chosen for the study are: Chile, Colombia, Costa Rica, Estonia, the Kyrgyz Republic, Sri Lanka, Thailand, Tunisia, and Vietnam. With health at the center of the global development policy stage on humanitarian as well as economic and health security grounds, the international community and developing countries are increasingly focused on scaling up health systems to meet the Millennium Development Goals (MDGs) as well as assuring sustainable financing over the long- term to perpetuate these gains. With the significant scaling up of aid and the realization by both donors and countries that money alone is not enough to achieve health gains, increasing attention has focused on the lack of good evidence about what works and what doesn't. In 2004, the Center for Global Development published a seminal report on "Millions Saved: Proven Successes in Global Health", which documented some 17 public health interventions of 'proven' effectiveness (Levine and Kinder 2004). A significant contribution of the report was an explicit specification of criteria for defining 'what works'. The report was highly acclaimed as an important synthesis of the global evidence base on public health interventions. Concomitantly, the global community as well as individual donors are increasingly focusing on 'results-based' aid and lending. These concerns about 'results' have triggered a major debate and actions with respect to the need for appropriate data for decision-making and monitoring and evaluation. Central to this exercise is the need for countries to implement effective evidence-based policies. Unfortunately seminal work in a number of areas including human resources for health, health systems and health financing has shown that the critical evidence bases are often lacking (Joint Learning Initiative 2004, Gottret and Schieber 2006, Center for Global Development 2006). This study attempts to partially fill this void by providing assessments of low and middle-income countries which have 'successfully' expanded health insurance coverage to their populations in order to improve their health status and provide financial Good Practice in Health Financing. p. 6 protection against catastrophic medical expenses. Coverage expansions are measured in terms of numbers of individuals covered and out of pocket payments as a percent of overall health spending. In addition to expansions in coverage, success is also assessed based on the the fact that the countries selected have done well in terms of achieving average or better than average health outcomes (e.g. infant mortality, life expectancy, and maternal mortality) for the amounts spent and for their income and educational levels. Nine country cases -- Chile, Colombia, Costa Rica, Estonia, the Kyrgyz Republic, Sri Lanka, Thailand, Tunisia, and Vietnam -- were selected by an expert steering committee representing all six World Bank regions. Each of the country cases is analyzed based on a standardized taxonomy which captures most of the key health and non-health sector specific factors that affect the performance of that country's health financing system. As the study relies only on existing ex post reform information, experiments with control groups or randomization are not possible, and thus no attempt is made to attribute causality. The selected countries have achieved remarkable health sector outcomes - in terms of life expectancy, infant and child mortality, and access to essential health services. For the most part spending levels are about average for their income levels; yet, virtually all the countries perform better or much better than average in terms of health outcomes when compared to comparable income countries. All nine countries have immunization coverage rates above 88 percent. All but the Kyrgyz Republic have a life expectancy at birth above 70 years, and infant mortality rates below 25 per 1,000 live births. Two of the countries - the Kyrgyz Republic and Vietnam - are low income countries (i.e., with 2005 per capita GNI below $875), the rest middle income ($876 - $10,725) with Estonia, Columbia, and Chile upper middle income (greater than $3,466). The majority do well in terms of providing financial protection defined in terms of relatively low out of pocket payment spending shares. However, four of the countries - not surprisingly the three poorest (Kyrgyz Republic, Vietnam, Sri Lanka) and Tunisia - perform less well with out of pocket payments above 40 percent of all health spending, although in the Sri Lanka case these payments disproportionately fall on the non-poor. The former Soviet Union (FSU) countries of Estonia and the Kyrgyz Republic, despite significant reductions in health-related infrastructure and other fixed costs, still have over twice as many hospital beds and physicians per capita when compared to the other countries. Trying to generalize' enabling conditions' for successful coverage expansions from these nine case studies is difficult, particularly as these cases are hardly representative of all global experience and are not a random sample . Nevertheless, these cases do represent countries that have undertaken serious reforms and based on our definitions of success appear to have done relatively well. The key question is: as health reforms are very country-specific, is it possible to find a common set of enabling conditions in these nine successes or is the conclusion simply that everything is country-specific and that one cannot generalize? It is clear that the paths taken by all these countries are very different and heavily contingent on the political economy and institutional arrangements in the individual countries. However, based on an analysis of expenditure, outcome, and demographic performance for these Good Practice in Health Financing. p. 7 countries against global trends and the detailed case studies of each of the countries, there would appear to be a number of general 'enabling' conditions common to almost all these countries. In addition, for several different smaller groupings of these countries, a number of other common factors are highlighted. General 'Enabling' Conditions In virtually all of the country cases several common characteristics are apparent: • robust growth during the bulk of the reform period • long-term government commitment • strong institutional and policy environment • high population literacy levels • rural focus • coverage expansions accompanied by carefully sequenced and significant delivery system (including human resources for health) and provider payment changes • flexibility and mid-course corrections • strong and explicit primary care focus • Strong fiscal commitment to cover the poor. From subsets of these nine cases other factors emerged including: • importance of good information systems and evidence-based decision-making • strong stakeholder support • efficiency gains and copayments used as financing mechanisms • consolidation of risk pools • limits to decentralization While it can be argued that one could have posited these results ex ante, these findings are based on a standardized definition of success in tenns of coverage expansions and financial protection in the context of average or below average spending countries which have achieved good health outcomes. In addition, development assistant for health has played a relatively minor role as a health financing source in most of these countries. The factors identified above make sense on conceptual grounds and would all be likely hypotheses for any rigorous evaluation of health financing change impacts. While these findings are useful, the more important contribution for the current evidence base is not these findings per se, but rather the need for the global community to do a much better job in evaluating and disseminating the global evidence base. To do this at least five important actions need to be taken. First, we need to rigorously define what we mean by successes in health financing. Second, we need to collect standardized and appropriate qualitative and quantitative information. Third, health Good Practice in Health Financing, p. 8 systems characteristics must be described in enough detail so that all the critical components and their interactions can be clearly identified and assessed. Fourth, rigorous evaluations need to be undertaken. Fifth, these evaluations - conducted in a policy relevant and user friendly manner - need to be disseminated to all stakeholders. To date, we in the global community have done a rather poor job in these areas. However, with the growing importance of this issue, it is incumbent on the global community to take charge and move this agenda forward. It is hoped that global stakeholders through various aid effectiveness fora as well as the G-8 and Paris Declaration Process will rise to the occasion and put this desperately needed work on track. Good Practice in Health Financing. p. 9 PARTl Good Practice in Health Financing, p. 10 1. INTRODUCTION Background A significant number of developing countries have recently undertaken ambitious health reforms to improve resource mobilization for healthcare in order to advance the goal of universal health care coverage and to improve financial protection against impoverishment due to catastrophic illness costs. These experiences - which are of great relevance to other low and middle-income countries about to undertake similar reforms - have not been consistently evaluated and documented, In order to partially address this gap, the World Bank's Health, Nutrition and Population (HNP) Unit in the Human Development Network has prepared this report, documenting experiences in financing significant expansions of health care coverage in low and middle-income countries, based on case studies of nine 'good practice' countries. This report focuses on countries that have 'successfully' carried out - or are in the process of carrying out - important health financing reforms that have resulted in expansions of health care coverage and improved financial protection. In doing so, the report seeks to expand the evidence base of 'good practices' in health financing reforms in low and middle-income countries. Contributing to this evidence base is particularly important now as health is at the center stage of global development policy on humanitarian as well as global economic and health security grounds, As a result, both the international community and developing countries themselves are increasingly focused on scaling up health systems to meet the MDGs as well as assuring sustainable long-term financing. With the significant scaling up of aid and the realization by both donors and countries that money alone is not a sufficient condition for development effectiveness, increasing attention has focused on the lack of good evidence about which policies work and which do not. In 2004, the Center for Global Development published a seminal report on "Proven Successes in Global Health", which documented some 17 public health interventions of 'proven' effectiveness (Kinder and Levine 2004). A significant contribution of the report was an explicit specification of criteria for defining 'what works' which included the important element of large scale implementation. The report was highly acclaimed as an important synthesis of the global evidence base on public health interventions. 1 Concomitantly, the global community and individual donors have increasingly focused on the need for results-based aid and lending. The World Bank's new HNP strategy openly decried the lack of available evidence to assess whether billions of dollars of HNP lending over the previous decade has led to significant results (World Bank 2007a). I There has been a recent update of the report which includes three additional cases. See Levine (2007). Good Practice in Health Financing. p. 11 These concerns about 'results' have triggered a major debate and actions with respect to the need for appropriate data and evidence for decision-making and serious monitoring and evaluation. The recent creation of the Health Metrics Network provides one example of global action to assure availability of appropriate health information. Central to this exercise is the need for countries to have knowledge of the global evidence base and to implement effective evidence-based policies. Unfortunately, seminal work in a number of areas induding human resources for health, health systems, and health financing has shown that the critical evidence bases are often lacking (Joint Learning Initiative 2004, Gottret and Schieber 2006, Center for Global Development 2006). For example, where is the comprehensive compilation/reference for successful health financing reforms in developing countries? The lack of such evidence-based policy collations to a large extent reflects the complexities of the health sector, country and donor priorities as well as the inherent difficulty of doing rigorous evaluations in the social sciences. Measuring health outcomes, other than sentinel events such as death, is difficult methodologically and complex and costly data-wise. Numerous health and non-health related factors affect an individual's health status. Individual behavior is an important determinant of health outcomes and difficult to measure and change (Schieber, Fleisher, and Gottret 2006). Figure I below highlights the many different confounding factors that interact in the health sector milieu. These complexities are exacerbated by the enormous number of interactive factors impacting most health policies. In the context of developing countries with insufficient funds to meet many basic and often competing needs, it is not surprising that rigorous policy evaluation has not been a high priority. Similarly, given these needs, donors have focused their efforts more on getting money out the door to assist countries, often in dealing with catastrophic and unpredictable situations such as the Tsunami or Avian flu. As a result, much of what passes for 'evidence' is in fact anecdotal and small scale: a box in a report indicating that in a particular health center or district a particular policy resulted in a particular outcome. Few rigorous evaluations are national in scope. 2 2A couple of notable exceptions are the national-level health policy evaluations in Cambodia and Mexico. In Cambodia, different modalities of the health system reform were randomly allocated across districts and evaluated (see Bhushan, Keller, and Schwartz 2002). In Mexico, the impact of conditional cash transfer programs was evaluated using a national phase-in controlled randomized design (see Gertler 2004). Good Practice in Health Financing, p, 12 Figure 1: Determinants of Health, Nutrition, and Population Outcomes Achieving Change in HNP ~ Behavior of Individuals/Households Income ./ "- Education / ~ Water Performance of Health System Health Status Outcomes Sanitation -Clinical Effectiveness 'Fertility ( 'Accessibilityand Equity 'Mortality -Quality and Consumer Satisfaction 'Morbidity 'Economic Efficiency 'Nutritional Status Macro- economic i Health Care System Environment Delivery Structure Institutional Capacity 'Facilities (public & private) 'Regulatory & Legal Framework 'Staff (public & private) -Expenditure & Finance ~ ~ 'Information, Education, & ·Planning & Budgeting Systems . / communication 'Client & Service Information/Accountability 'Incenfes Governance '--- Projects and Policy Advice Source: Wagstaff, Yazbeck, and Claeson, "Poverty Reduction Strategy Framework" In the health financing area, there have been remarkably few single or multi-country studies which define what is meant by a successful reform, describe the countries in question in a standardized way and with sufficient detail so that the reader can truly understand the interplay of the many key factors, determine causality based on rigorous evaluation methods, and provide generalizable lessons for other countries. This study attempts to partially fill this void. First, success is defined as countries significantly increasing the number of individuals covered, the depth of coverage in terms of the richness of the benefit package, and the level of financial protection. For countries which previously had universal coverage but have been faced with drastic medium term political or economic shocks (e.g., countries of the former Soviet Union), a successful reform is also one in which the country resurrects health sector revenues and expenditures back to a satisfactory pre-economic crisis base level. The thorny problems of defining and measuring these success factors are discussed below in detail. As discussed below, nine country cases were selected by an expert steering committee from all six World Bank regions based on expansions of the breadth and depth of coverage and attainment of good health outcomes at reasonable spending levels. The countries chosen are: Chile, Columbia, Costa Rica, Estonia, the Kyrgyz Republic, Sri Lanka, Thailand, Tunisia, and Vietnam. Each of the country cases is then described based on a standardized taxonomy which captures most of the key health and non-health sector specific factors that affect the performance of a health financing reform. As the Good Practice in Health Financing, p. 13 study relies only on existing ex post reform secondary data, one cannot do experiments with control groups or randomization, and thus no attempt is made to attribute causality. While success is measured in terms of increases in the breadth and depth of coverage and levels of financial protection, it is also implicitly based on the panel's normative judgment that these countries have done well not just in expanding coverage but also in terms of average or lower than average overall health spending levels and average or better than average health outcomes (infant mortality, life expectancy, and maternal mortality) for the amounts spent and their income and educational levels. The reform experiences and lessons to be learned from these 'good practice' cases are discussed in detail for each country. Each country chapter includes a description of important background information, including: the country's economic, social and political environments; descriptions and evaluations of key health coverage reforms; conditions for success; and lessons for other countries. In classifying countries by income level, this report uses current World Bank definitions which are based on 2005 estimates of Gross National Income (GNI) per capita, as follows (World Bank 2007): • Low-income = $875 per capita or less; • Lower middle-income = $876 to $3,465; • Upper middle-income = $3,466 to $10,725; • High-income = $10,726 or more. The report is organized into two parts. Part I describes key issues in expanding health care coverage in low and middle-income countries, defines what is meant by successful expansions of coverage and financial protection, discusses the criteria used to choose the country cases, and summarizes the findings from each of the nine country case studies. The concluding chapter of Part I draws the generalizable lessons from the country cases. The full country case studies are contained in separate chapters in Part II of the report. The country case studies are organized so as to systematically address a core set of issues. The following areas are addressed by each case; however the emphasis given to the different points may vary across cases depending on the nuances of that case: 1. The country's economic, social, and political environments. The implications of individual country's institutional structures cannot be understated, and should be carefully described as part of the analysis and lessons learned. 2. Descriptions, chronologies, and evaluations of reforms to expand health care coverage. The case studies consider reforms implemented over a long-term time frame (up to the past 30 years; and even earlier for Sri Lanka). 3. Overview of the health delivery system with an emphasis on delivery system configuration, provider payment issues, equity, and ease of access. 4. The depth of the benefits package. Good Practice in Health Financing, p. 14 5. Financial protection and access to services for chronic and catastrophic conditions. 6. Revenue sources. 7. Health indicators and outcomes, and their distribution (focus on equity). 8. Spending effectiveness. 9. The financial sustainability of the reforms. 10. Other key issues, potentially including: • How risk adjustment and cross-subsidization are handled; • Regulation of the insurance schemes; • Policies for opting out of insurance schemes; • Need for government subsidies; • Financing compared to depth of package (focus on "best buy"); • Impact on labor markets; • Other links between political economy and the health sector; • Cost sharing mechanisms; and • The sustainability of financing. 11. Scaling-up the reforms. 12. Key conditions for success. 13. Lessons for other countries. Audience These materials are aimed at governments and international policy-makers, donors, staffs of relevant related stakeholder organizations, and health care analysts working in low and middle-income countries. This report is intended as a practical guide for decision-makers in countries considering health financing and coverage reforms, as well as World Bank operations teams, multilateral development organizations, bilateral aid agencies, private foundations, and other development partners. It is hoped that this very modest attempt to assess successes in major health financing changes will lead to more rigorous evaluations of such changes and important much needed systematic, multi- country contributions to the global evidence base. Good Practice in Health Financing, p. 15 2. HEAL TH FINANCING FUNCTIONS The organization of the case studies in this report focuses on key health financing functions within country health care systems (Appendix 1 provides definitions for key terms). Specifically, the framework includes: (1) Revenue collection; (2) Risk pooling; and (3) Purchasing of health care - including purchasing processes, provider payment systems, and their relationship to health care provision. The report focuses primarily on the expansion of health care coverage, which has at least three separate and inter-related dimensions: (1) the number of people covered by organized (public and private) financing initiatives (the breadth of coverage); (2) the extent (number and type) of services covered (the depth of coverage); and (3) the resulting financial protection against high out-of-pocket costs. Figure 2 provides an overview of how these three functions together result in health care coverage and access at the population level. Arrows in dark blue indicate financing flows. Figure 2: A Framework for Health Financing and Coverage Financing Pooling Provision Ministries of Finance, Health Public / Private Provider Mix Donors Social Insurance Purchasin Organization Private and Mix of Insurance Information Services Systems Quality and Firms Efficiency of HOUSEHOLDS Source: Authors Good Practice in Health Financing, p. 16 In expanding coverage to promote universal access and financial protection, countries need to pursue the following objectives: 1. Raise sufficient and sustainable revenues in an efficient and equitable manner to provide individuals with both a basic package of essential services and financial protection against unpredictable catastrophic financial losses caused by illness and injury; 2. Manage these revenues to equitably and efficiently pool health risks so that individuals are provided with 'insurance' coverage against unpredictable catastrophic medical care costs; and, 3. Assure the purchase of health services in an allocatively and technically efficient manner (Gottret and Schieber 2006). Revenue collection and risk pooling refer to the accumulation and management of sufficient and sustainable revenues to assure that all individuals have access to an essential package of basic services and that these funds are 'pooled' so that the risk of large catastrophic medical expenditures are borne by all the members of the pool and not by each member individually, thus providing all pool members with 'insurance protection' against catastrophic medical care costs. Each pool member's 'contribution' is prepaid through premiums, payroll taxes, and/or general tax payments. The case studies span the range of generic health financing 'models' such as national health services, social health insurance funds, and private health insurance. We use a simplified categorization of health insurance schemes including: (1) General-tax financing, managed by a National Health Service (NHS) or Ministry of Health (MOH); (2) Payroll tax financed social health insurance managed by a quasi public entity; and (3) private-sector based health insurance financed by contributions to private voluntary insurers. Globally, some 100 countries have health financing systems that are predominantly general revenue based, while another 60 have social insurance based systems. Only a few countries have predominantly private health insurance financed systems (e.g., the U.S.). In practice most countries have mixed models. Examples of high-income countries with social insurance systems include Germany and Japan. Most systems based on a national health service have compulsory universal coverage, financed from general government revenues, with provision also predominantly in the public sector; the United Kingdom and the Scandinavian countries are examples. In this study, Sri Lanka is the only pure NHS system, while the others represent SHI models. Chile has elements of both a SHI and private health insurance approach. As discussed below in detail, health care coverage is defined in terms of the breadth and depth of coverage as well as the resulting level of financial protection. Financial protection in health implies that households and individuals: (1) obtain health care when they need it and are not prevented from doing so by excessive costs; (2) do not incur Good Practice in Health Financing. p. 17 costs when they do access health care that prevent them from obtaining other basic household necessities - including food, education, and shelter; and (3) do not fall into poverty due to excessive medical care costs and lost income resulting from illness (Baeza and Packard 2006). There is a plethora of empirical evidence concerning the economic and social impacts of adverse health shocks and the need for policies to provide everyone, but particularly the poor, with financial protection against such large and unpredictable costs (Wagstaff 2005, van Doorslaer, O'Donnell, Rannan-Eliya, Somanathan et al. 2005). Moreover, a substantial body of evidence suggests that for governments to reach the poor and other vulnerable populations, extensive outreach programs and/or demand side subsidies are often required. Subsidized government health services often primarily benefit the better-off, rather than the poor for whom these services are intended. However, governments can adopt targeting measures to increase the proportion of public benefits that flow to the poor; or promote the development of separate, privately-funded healthcare delivery mechanisms that serve the better-off (Gwatkin 2004). Certain demand side approaches have also been found to motivate the poor to seek necessary and often covered care. For example, conditional cash transfers (CCTs) have been found to be an effective demand side measure to encourage the poor to access services. CCTs provide financial support to relatively poor families contingent upon certain behaviors related to household welfare - such as sending children to school or bringing them to health centers on a regular basis. As such, they represent a negative user fee. Evaluation results from a first generation of conditional cash transfer programs show success in improving school enrolment rates, improving preventive health practices, and increasing household consumption (Rawlings 2004). Thus it is important to keep in mind that coverage expansions may need to be accompanied by social marketing, outreach, and other demand side interventions in order to reach the most vulnerable populations. Purchasing is the process by which pooled funds are paid to providers in order to deliver a set of health interventions. Purchasing can be performed passively or strategically. Passive purchasing implies following a predetermined budget or simply paying bills when presented. Strategic purchasing involves a continuous search for the best ways to maximize health system performance by deciding which interventions should be purchased, how, and from whom and at what price. The type of provider payment method constitutes an important part of the purchasing arrangement. The main types of provider payment methods include capitation; fee-for-service; global budgeting; line item budgeting; case-based payment; and Diagnostic-Related Groups (DRGs). Each of these mechanisms has important incentives for the provision of health care, controlling costs, and improving the quality of services provided (Preker and Langenbrunner 2005). Good Practice in Health Financing, p. 18 3. CRITERIA FOR DEFINING 'GOOD PRACTICE' AND CHOOSING COUNTRY CASES Among the most important criteria for determining the country case studies are the following: First Tier Criteria 1. Improvements in health care coverage. The principal criterion for selection was a documented effort to increase the proportion of the population with formal health care coverage and to improve the level of tinancial protection. 2. Applicability and pertinence for low and middle-income countries. 3. Large scale initiatives. The initiatives described and analyzed should be at a national scale, or at least involve significant national policy-making. In most cases, the principal form of expansion of health care coverage is through institutional arrangements - formal-sector employment, tax-based government affiliation, or market-based private insurance. This focus omits countries with significant community health insurance development. However, these experiences have been well documented elsewhere (Preker and Carrin 2003; Preker et a12000, 2002, Gottret and Schieber 2006). 4. Availability of information and data to analyze the case - including the measurability of key outcomes and indicators. The availability of data should not be the sole determinant of whether a country is included - some countries have excellent data available but did not undertake major successful reforms - but it does playa critical role in whether a "story" can be told about the success of a country's reform efforts. The availability of household-level data is particularly important for measuring coverage, access, and financial protection. Second Tier Criteria 1. Health indicators and outcomes. While it is problematic to document a causal association between health system and financing characteristics and population- level health outcomes, the ultimate goals of health coverage reforms are to improve health status and financial protection along with responsiveness (WHO, 2000). Accordingly, key population-level outcomes - including life expectancy, infant and child mortality, and immunization coverage - were used to assess the long-term success of reforms attempted in a specific country. 2. Relationship of expenditures to outcomes. At a national level, the amount of funds devoted to health care reflects a societal commitment to promoting population health as well as financial protection against impoverishment from catastrophic medical care costs. However, spending levels in and of themselves Good Practice in Health Financing, p. 19 provide little information about equity, the efficiency of spending, or the level of financial protection provided to individuals. While many developing countries face severe resource constraints in mobilizing resources, and many poor countries cannot raise sufficient revenues to meet the MDGs, value for money in terms of health outcomes (i.e., allocative efficiency) and production costs (i.e., technical efficiency) is a critical issue in all countries. Thus, it has become increasingly recognized that more money alone is not a sufficient condition for improving health care outcomes. Nevertheless, a direct comparison of health spending per capita across comparable income countries with health outcomes - while confounded by factors well outside the reach of the health system - provides a crude comparative benchmark of how effectively a country is spending its health care resources vis-a-vis its health care outcomes. One critical aspect of this study is the definition used to determine success in terms of coverage expansions and financial protection. Breadth of coverage is readily measured by the number of individuals formally eligible to receive benefits from a particular 'insurance' (i.e., NHS, SHI, or PVHI) mechanism. Measuring the depth of coverage -- the type and number of services covered, which is often defined and measured as the actuarial value of the benefit package per enrollee (and can be approximated by dividing spending less user fees by the number of users), is problematic due to a lack of information on numbers of users. Financial protection can be measured in a number of ways. For the health system as a whole, out of pocket payments (OOP) as a percent of total health spending offers a rough estimate of financial protection, but provides no information regarding the distribution of OOP among population income groups. More revealing measures of financial protection require individual or household level analysis. These may include the number or share of households with OOP exceeding some pre-specified level of total or non-subsistence consumption, the share of households passing this level times the average amount it is exceeded by, or the percentage of households falling below some absolute poverty line due to OOP payments (Wagstaff and van Doorslaer 2003, Xu et al. 2003, Asian Development Bank 2006). These measures attempt to account for both the extent and the severity of OOP payments and to various degrees highlight the underlying notion that using health care often represents sudden shocks that may negatively impact household consumption and welfare. Unfortunately, given the data limitations, as these case studies are all based on ex post evaluations, information on many of these financial protection measures are unavailable for several of the countries. For the purposes of this study: • Breadth of coverage is measured by the number of people with formal coverage. • Depth of coverage is not measured separately. • Financial protection is measured in terms of out-of pocket payments as a percent of total health spending and reflects both the depth and breadth of coverage. Good Practice in Health Financing. p. 20 The other measures of financial protection are not readily available for all the cases and are thus not used systematically, although some of the individual country cases do provide these additional measures. For countries which previously had universal coverage but have been faced with drastic political or economic shocks (e.g., countries of the former Soviet Union), a successful reform is also one in which the country resurrects health sector revenues and expenditures back to a satisfactory pre-economic crisis base level when high levels of coverage and financial protection existed. While success is measured in terms of expansions in coverage and financial protection, it was decided that the initiatives undertaken need not be successful on all of the dimensions outlined above. However, for a given country, the reforms implemented would constitute a success overall based on meeting the first tier criteria of achieving improvements in coverage and financial protection while at the same time exhibiting reasonable (average or lower) overall health spending levels and average or better than average health outcomes (infant mortality, life expectancy, and maternal mortality) for the amounts spent/and or their income and educational levels. Based on these criteria, the Steering Committee recommended nine countries: Chile, Columbia, Costa Rica, Estonia, the Kyrgyz Republic, Sri Lanka, Thailand, Tunisia, and Vietnam. The Kyrgyz Republic and Vietnam are low income countries; Costa Rica, Sri Lanka, Thailand, and Tunisia are middle income and Estonia, Columbia, and Chile are classified as upper middle income. Tables I and 2 below provide information for these nine countries on 13 measures of health system characteristics: per capita GDP, health expenditure to GDP ratio, per capita health spending in US$ and international (purchasing power parity (PPP) adjusted $US), public share of total health spending, health share of the government budget, percent of health spending out of pocket, physician and hospital bed to population ratios, immunization rates, infant mortality rates, and life expectancy. 3 3 Some ofthese statistics may differ from those provided in the individual country case studies because some of the case studies use data from different sources and years. In order to ensure for comparability across countries, Tables 1 and 2 provide data from the same source for the same years for all countries. Good Practice in Health Financing, p. 21 Table 1: Income and Health Spending Levels, 2004 Out-of- Total Total Government Government pocket GDP Total Health Health Health Health Health per Health Spending Spending Spending Spending (% Country Spending capita Spending per per (% Total General (% Total (US$) (%GDP) capita capita Health Government Health (US$) (PPP) Spending) Expenditures) SEendin~l Chile 5,894 6.1 359 720 47.0 4 13.1 24.3 Colombia 2,155 7.8 168 570 86.0 20.9 6.9 Costa Rica 4,349 6.6 290 592 77.0 21.3 20.4 Estonia 8,328 5.3 463 752 76.0 11.5 21.3 Kyrgyz Republic 434 5.6 24 102 40.9 8.4 55.7 Sri Lanka 1,033 4.3 43 163 45.6 8.4 45.7 Thailand 2,539 3.5 88 293 64.7 11.2 26.4 Tunisia 2,832 6.2 175 502 52.1 8.8 39.8 VietNam 550 5.5 30 184 27.1 5.0 64.2 Source: World Bank, World Health Organization 2007. Note: GDP per capita in current $US Table 2: Health Outcomes and Delivery Indicators, 2005 Infant Immunization, Mortality Immunization, Life measles (% of Rate (per DPT(%of Country Expectancy Doctorsll ,000 Bedsll,OOO children ages 1,000 children ages (years) 12-23 live 12-23 months) months) births2 Chile 8 78 1.09 2.5 94 95 Colombia 18 73 1.35 1.2 89 92 Costa Rica 11 79 1.32 1.4 90 88 Estonia 6 73 4.48 5.8 94 96 Kyrgyz Republic 58 68 2.51 5.3 99 99 Sri Lanka 12 75 0.55 2.9 97 96 Thailand 18 71 0.37 2.2 98 96 Tunisia 21 73 1.34 2.1 97 95 VietNam 17 71 0.53 2.3 96 97 Source: World Bank, World Health Organization 2007. 4 However, a large share of health expenditure in Chile counted by WHO as private, based on national health accounts definitions, could potentially be considered public instead since coverage of formal sector workers is mandatory although many are covered by private insurers (lSAPREs). Good Practice in Health Financing, p. 22 Performance of Country Cases Globally We can situate these nine countries by examining some of their health system indicators in a more global context. In general, all case countries have achieved remarkable health outcomes -life expectancy, infant mortality, and access to essential services - despite low or moderate levels of income and total health spending (Figure 3). 5 All but the Kyrgyz Republic have a life expectancy at birth above 70 years, and infant mortality rates below 25 per 1,000 live births. Additionally, all nine countries have immunization coverage rates of at least 88 percent; most are well above 90 percent. Each also performs better than average against comparators with regard to the percentage of births attended by skilled health personnel. Figure 3: Population Health Indicators Relative to Income and Spending ! Q) LIFE EXPECTANCY I r"-- II'EANT MORTALITY I -- C) ~ !!1 : : ~ '" Q) Viet Nam , Q) '..0 ••Sri Lank~ E ' ; -Thailand Kyrgyzsta. .Thailand ~ rEstonia "T~m!$ia Qj .... 'E Q) E c: Q) ro ':= Q) C) Cl ~ ~ « Q) > I g! '0" ~ I "' 3: c. I Q; d; co co Below average Above average LBelowaverage I Above average Attainment relative to health spending per capita Pttainment relative to health spending per capita Source: Wo~d Development Indicators. WHO 2007 5 These graphs plot the residuals for each country from two regressions: the logarithm of the health outcome is regressed first on the logarithm ofGDP per capita alone and then on the logarithm of health spending per capita alone. The scatter plot of the residuals from these regressions is one concise way to assess a country's performance on the outcome compared to the global average for a given income level and total health spending per capita. See Asian Development Bank (2006). Good Practice in Health Financing, p, 23 For five of our nine case countries, health spending as a share of GDP is close to the average for their income levels; however, Estonia, Sri Lanka, and Thailand spend markedly less than their average share of GDP for their income while Colombia spends far more (Figure 4). In per capita terms (US$), almost all countries spend close to the average for their income with Colombia being the only country spending higher than average (Figure 4). 6 Figure 4: Health Spending as Share of GDP and Per Capita vs Income ~ fl' ~ I Colombi - / -Estonia C~~~~ica TuniSia -Colombia ~ Costa~ Kyrgy";$.tanTunii~ile -~ _________ ----!! -Estonia iij'" Viet Nam -Sri Lanka ~ Kyrgyzstan -Thailand ~o I-~ o --,- - - - - -~-- r---' L i ---------r--- - ,- - - - - T ---,-- ,- 100 250 1000 1000(125000 100 250 1000 1000025000 GOP per capita GOP per capita Source: Worid Development Indicators, WHO 2007 Note: GOP per capita in current U5$; Log scale 6 The analysis contains observations for 178 countries for which data were available. This includes 51 low-income countries, 56 lower-middle income countries, 36 upper-middle income countries, and 35 high- income countries. Good Practice in Health Financing, p, 24 The public-private financing mix varies widely across the nine countries (Figure 5). Colombia, Costa Rica, Estonia, and Thailand each have government shares of total health spending well above the average for their income level. 7 As a share of the overall government budget, though, Colombia and Costa Rica spend amounts that are much higher than the average for their income level (Figure 5) with more than 20% of the government budget spent on health. On the one hand, this could signal the importance placed on health in these countries. However, from a financing perspective, this could also raise concerns regarding the sustainability of this situation. Figure 5: Government Share of Health vs Income ~ i '~I "0 ::I CD ! ' III I -Colombia .. C E .c~1 .. *,co '" ::t: , I Costa Rica - -Esto a .. E > 0 0 COIO"'V>f?-0sta Rica iii c!)'" ~ i iii ~gj -Thai nd ~ -", Cl , :.!! 0 ~ Cl , c -Tunisia .6 .!c III _Chile -Sri Lanka . "0 c CI. III 0 -Chil/ Thail~nd ~onia .c~ -Kyrgyzstan '.. Kyrgyz _ -Tunisia *' ::t: " Sri Lanka C E . I ::t: C .. -Viet Nam E. ! -Viet Nam E . gL '~,~," - ,------,-- E ~ C!)o -------,----------- -~- -~---T- 10 100250 1000 1000025000 10 100 250 1000 1000025000 GOP per capita GOP per capita Source: World Development Indicators, WH02007 Note: GOP per capita in current US$; Log scale As shown in Figure 6 below, the majority do well in terms of providing financial protection defined in terms of relatively low shares of out of pocket payments. However, four of the countries - the three poorest - Kyrgyz Republic, Vietnam, Sri Lanka - and Tunisia - have out of pocket payments that are relatively high, at least 40 percent or more as a share of overall health spending, although in Sri Lanka out-pocket expenditure tends to fall on the rich so poses less of a problem from an equity perspective. As discussed above, this also reflects the difficulty poorer countries have in being able to afford both a package of essential services and financial protection. 7In general, as income rises there is a tendency for the public share of health expenditure to rise. This is often attributed to changes in relative prices and to changes in societal preferences as incomes rise. Good Practice in Health Financing. p. 25 Figure 6: Out of Pocket Spending Share vs Income Cl c:: ' "go I ~co1 L i~j: .Viet Nam • Kyrgyzstan ~ ~ .Sri Lanka Co Cl~ ~ ~ .Tunisia .~ i "0 i ~ile. ~ojl U)N - .-.~ Costa Rica --- E m (I) .Colombia J: : a..o~~,------------~----------~-----------'----~I------ o o 10 100 250 1000 10000 25000 GDP per capita Source: World Development Indicators. WHO 2007 Note: GOP per capita in current US$; Log scale Revenue raising capacity increases with income and provides governments with fiscal space that can be used for health spending. Figure 7 shows the very different efforts found in the case study countries. Four of the nine countries - Costa Rica, Chile, Sri Lanka, and Thailand - raise less revenue than the average for their income level. The remaining five countries each raises more than average. However, it is important to note that three of the above average revenue effort countries, Vietnam, Kyrgyz Republic, and Tunisia, also are the countries with the highest OOP spending shares, all of which are well above the global trends. These countries also show lower than predicted government expenditure shares for their level of income. These may reflect the difficulties in terms of political economy of committing governments to cover an essential package of services for their population despite good efforts in revenue collection. Good Practice in Health Financing, p. 26 Figure 7: Revenue to GDP Ratio vs Income Collombla Tunisia .Est Kyrgyzstan Viet Nam Co: • • • ~ Chile ....----~---- Thalla"d _ - - - Sri Lank' I OlL~ __________~__~____________~____~____~_____ 10 100 250 1000 2500 1000025000 GOP per capita Source: World Development Indicators Note: GDP per capita in current US$; Log scale Another notable point is that the former Soviet Union (FSU) countries of Estonia and the Kyrgyz Republic, despite significant reductions in recent years, still have over twice as many hospital beds and physicians compared to the other countries (Table 2) and are far above the global averages as shown in Figures 8. Sri Lanka and Vietnam also have high numbers of hospital beds but these are nowhere near the levels of some of the FSU countries. Interestingly, in eight of these nine success stories, Thailand is the only country with a significantly lower physician to population ratio than the global average. Perhaps a minimal physician to population ratio is a necessary, albeit not sufficient condition, for a successful health insurance coverage expansion. Good Practice in Health Financing. p. 27 Figure 8: Hospital Bed and Physician Capacity vs Income HOSPITAL B8) SUPPLY DOCTOR SUPPLY ~1 co 0 0 I co I Q01 ...... ... ~ ~ 0 0 q ...... I Ul ... Q) • Estorrl a "0 Q.~ I Q) i ... Ul I L~ :J: • Kyrgyzstan • Esto a .9 0 (.) 0 N • Kyrgyzstan / I Sri Lanka . Tunisi Chile Viet Nam • ......- Colom~~. osta Rica J 10 ~---~ I 100250 1000 • . ?-Tunisia an ______.Costa Rica Colombia I GOP per capita I 1000025000 I 0 10 Viet Nam Sri • • -------- 100250 1000 GOP per capita ·Chile nka ·Thailand 1000025000 Source: World Development Indicato",. WHO 2007 Note: GOP per capita in current US$; Log scale Adult literacy is also another important factor that distinguishes most of these countries from others. With the exception of Tunisia, each country has above average rates of both total adult and female adult literacy compared to countries of similar income levels (Figure 9). This is important since the evidence of a strong positive link between higher education and better health outcomes is well documented. In fact, education is more highly correlated with health outcomes than other important factors, like income and health spending. Good Practice in Health Financing, p. 28 Figure 9: Literacy vs Income TOTAL ADUL T LlT~CY ADUL T FEMALE LlT~CY 0, ---II) ~ Q)~ B (IJ "0 / C (IJ LO Costa Rica ~ hile Kyrgyzstan \ tonia -Colomb. Viet Nam _ -- iland - Sri Lan a -Tunisia -Tunisia ~ -~-T- ~---~-- r ---,- - ---~I --~-- -~-- 10 100250 1000 1000025000 100250 1000 1000025000 GOP per capita GOP per capita Source: World Development Indicators, WHO 2007 Note: GOP per capita in current US$; Log scale Good Practice in Health Financing, p. 29 4. SUMMARIES OF COUNTRY CASES 8 The detailed reform experiences and lessons to be learned from each of these 'good practice' cases are discussed in detail for each country in Part 2 of this study. Following is a summary of the main reform elements and lessons from each of the countries. Chapter 5, the concluding chapter of Part 1 of this report, attempts to assess the generalizable lessons from these country cases. Chile Background Chile is an upper-middle income country in the Latin America and Caribbean region with an average-sized population (16.3 million). The country's population is based primarily in urban settings, with only 13% of the population living in rural areas. In 2004, per capita GDP was US$ 5,894. The country devoted some 6.1 percent of GDP or US$359 per capita to health in 2004, which is slightly below average for its income level. Some 53 percent of this amount is paid by private sources, including 24.3 percent paid directly by households as out-of-pocket expenditures. Health spending accounts for some 13.1 percent of the government's budget, which is above average for its income level. In terms of health outcomes, Chile ranks among some of the highest performers in the world. Life expectancy is 78.2 years, and infant mortality is 7.6 per 1,000 live births. In 2005, the number of physicians and hospital beds per 1,000 people were 1.09 and 2.5, respectively, slightly below average for a country of its income level. In Chile, the expansion of a fiscally sound social security system, including health care coverage, has required considerable economic resources and was made possible by an extended period of economic growth and an efficient tax collection system. Chile is an early reformer relative to many of the countries in this study. Its reforms can be traced back to the early 1950s, during which time several reforms culminated in the creation of the national health system. Description of the key reform elements and accomplishments The expansion of health care coverage in Chile can be traced over a period of at least 26 years. The insurance system in place today in Chile was first implemented in 1981 - consisting of a public insurer (FONASA) and private insurers (ISAPREs). The system now covers over 90 percent of the population. Of the 90 percent covered, close to 70 percent are covered through FONASA and some 16 percent are covered through ISAPREs. These programs have undergone several important changes over time. In 8Income levels, health spending and delivery system, and outcomes data in the background section of each summary are derived from the World Bank's World Development Indicators (2007) and WHO World Health Statistics (2007). Good Practice in Health Financing. p. 30 1991 the Superintendent of Health (Superintendencia de Sa/ud), was created to regulate the ISAPRES - and in 2001, the Superintendencia began regulating FONASA as well. Ninety-six percent of the poorest individuals are covered by FONASA, the national Health Fund. Most of the population above 50 years of age (even within the highest income quintile) is affiliated with FONASA. A complete subsidy (without co-payments) is provided for the poor. Financing of these subsidies requires a solid domestic tax base for sustainability. FONASA is also internally equitable - 41 percent of its beneficiaries are indigent and make no co-payments. The share of total health expenditures paid by the public sector increased from 37 percent in 1998 to 49 percent in 2003. The average number of services provided per beneficiary has also increased dramatically over the years both in FONASA and the ISAPREs. To contain public health spending, Chile has set budgetary ceilings for FONASA. Health services are delivered by both public and private providers. Private providers sell their services to insurers and private consumers. Health insurance is compulsory for workers in the formal sector, and voluntary for those in the informal sector. Members of the health insurance system must contribute 7 percent of their earnings up to a ceiling of approximately $2,000 as a premium, and may make additional voluntary contributions, in addition to copayments at the point of service. The required contribution has risen over time, from four percent in 1981. FONASA completely subsidizes health services for the indigent and unemployed. While the health insurance system was launched in 1981, regulation of private insurers did not start until 1991. In 2005, a minimum benefits package for all health plans was mandated. A set of new laws collectively known as Explicit Health Guarantees (GES) requires that both public and private insurers progressively provide coverage for 56 defined conditions - beginning with 25 conditions in July 2005 and culminating with all 56 in 2007. Key Lessons Some of the key lessons of the reforms are: • Expansion of health care coverage requires sustained financial investments into the health sector. In particular, expanding coverage of social insurance systems - either in terms of the population covered or the depth of benefits package - is likely to lead to substantial increases in health care spending, partly due to moral hazard and the need to subsidize the poor and vulnerable. Therefore, it is easier to implement and expand social security coverage when public finances are adequate to face these costs. Implementation of such systems typically occurs during periods of expansion, that is, when there is a high level of economic growth. Good Practice in Health Financing. p. 31 • Increased coverage for low-income groups requires subsidies from government. Subsidies provided to the overall population should also be progressive. • Cost containment measures are important to prevent public sector expenditures from increasing without control due to moral hazard, increased population coverage with subsidies, and the introduction of new technologies. FONASA has co-payments that increase with the level of income of beneficiaries and the public sector has introduced a hard expenditure ceiling on public sector providers. However, an additional reform in Chile is the shift from supply plus demand-side subsidies (budget support to public sector providers plus coverage of the poor population by FONASA) to one in which supply-side subsidies are eliminated. • Insurance expansion leads to increased health spending, partly due to moral hazard, necessitating budget ceilings for public programs. The insured tend to consume more physician visits, medications, and other services than the uninsured. To contain public health spending, the Chilean reform set budgetary ceilings, based on previous years' spending. • Private funding support is essential for a sustainable health system. In Chile, private financing comes mainly from two sources - co-payments, which vary according to services consumed, and monthly contributions, which are set at seven percent of the beneficiary's income. The average monthly salary in Chile is US$490; countries with lower income levels will likely need to charge higher rates to beneficiaries in order to offer a comparable package of benefits. The required contribution has risen over time, from four percent in 1981. • Incomplete regulation and supervision in a competitive insurance system can lead to adverse risk selection, high marketing expenses, and difficulties for beneficiaries in selection of health plans. Moreover, ISAPRES offer over 10,000 health plans - generating asymmetries of information and making it difficult for beneficiaries to make informed choices. • At the same time, the success of the Chilean model is highly dependent on an effective public insurance system. If the Chilean social security system were dissolved in favor of a completely free-market system with only private insurers, it is likely that the process of adverse selection would leave only the lowest-risk, highest-income individuals insured. • Chile has benefited from credible, independent, and efficient institutions. The Superintendencia de las ISAPREs regulates the benefits packages, contribution rates, and services provided by the ISAPREs - providing transparent rules of the game and fostering competition among private insurers. An efficient tax collection mechanism and a low rate of tax evasion have also Good Practice in Health Financing, p. 32 helped to ensure that the necessary resources will be available to finance public spending. • Solid political economy and consistency are important for health reform . The initial health sector reforms in Chile were introduced during a non- democratic period. After 1990, successive democratic governments introduced reforms aimed at increasing investment in the private sector; improving public sector administration; enhancing coverage of the poor; improving regulation and supervision; and promoting consumer protection. However, the current system maintains many of the same financing, organizational and functional characteristics of the model initially established. Moreover, the reforms were introduced only after attaining strong political consensus. Colombia Background Colombia is a lower-middle income country located in the northwestern tip of South America, and is the fourth largest country of the continent, with a size of 1,038,700 square kilometers. It has a population of 45.6 million and varied geography. In 2004, per-capita GDP was US$2,155. In terms of health, total health expenditures were $168 per capita in 2004 - a relatively high 7.8% ofGDP. The health system is largely publicly-financed with 86 percent of total health spending coming from public sources. Only 6.9 percent of total health spending comes from out-of-pocket expenditures by households, a level that is well below the average of countries with similar income levels. In terms of health outcomes, with life expectancy in Colombia at 73 years, and infant mortality at 18 per 1,000 live births, Colombia fares better than other countries of its health spending levels and income category. In terms of its delivery system, Colombia has a relatively high physician supply - 1.33 doctors per 1,000 people. However, with 1.2 hospital beds per 1,000 people, Colombia ranks somewhat below average for its income level. Until the mid 1990's Colombia had one of the most stable economies in Latin America. The country experienced economic growth until 1996, when it entered a period of economic recession. In 1999, unemployment reached 20.1 percent and 58.2 percent of the population was below the poverty line. The economy has been on a recovery trend during the past five years, which - if continued - will favor the sustainability of current health insurance expansion policies. Prior to 1993 Colombia had a three-tiered health care system. One tier consisted of the National Health System, with a network of public facilities for provision of health care - financed from national revenues and providing services to approximately 50 percent of the population. A private healthcare provision market grew in parallel to the public Good Practice in Health Financing, p. 33 network, targeting the population with ability to pay and meeting the demand for services and quality not provided by the public sector. Private providers supplied a large proportion of ambulatory care services in both urban and rural areas - 42 percent and 36 percent of all visits, respectively. A second tier was a mandated social insurance plan for workers in the public sector and formal private sector, financed by employee and employer contributions and covering a total of 20 percent of the population. The third tier consisted of private insurance or health care services paid out-of-pocket by higher income groups. This market was covered by private commercial health insurers, prepaid group health organizations, and worker cooperative organizations. The Colombian health system prior to 1993 resulted in limited access to basic health services for a large proportion of the population, inequities in health expenditures, inefficiency in spending, and shortcomings in the quality of health service provision. About 75 percent of the population was uninsured. In 1992 one out of every six individuals in the poorest income quintile who fell ill did not seek medical care because they could not afford to pay for it. According to national household surveys conducted that year, health care expenditures represented 2.4 percent of total household expenditures - but the lowest income decile spent about 10 percent of their income on health, while the upper income decile spent less than 0.5 percent. Public resources were poorly targeted - as evidenced by the fact that 40 percent of the subsidies to public hospitals benefited the wealthiest five percent of the population, and allocation of health care resources was biased towards more expensive, curative care, despite evidence of low levels of productivity. Allocation of resources followed historical hospital spending trends rather than needs of the population. The poor quality of health services was reflected by low use and acceptance of the network of public providers. Description of Key Reform Elements and Accomplishments In the last 14 years Colombia'S health care reform has expanded insurance coverage to 78 percent of the total population. In 1993, Colombia introduced an ambitious health care coverage reform, which created a National Social Health Insurance (NSHI), organized as a model of "Managed Competition". The NSHI includes two main components. The contributory regime is the mandatory health insurance scheme for formal sector and informal (self-employed) sector workers with the ability to pay, as well as retirees. Health insurance coverage includes immediate family members. As of December 2005, there were a total of 15.5 million people enrolled in the contributory regime (37 percent of total population), including dependents. Although the social health insurance system was initially designed as two different schemes with separate target populations, financing sources, and benefits, subsequent reforms collectively known as Law 100 have established universal health insurance coverage for all Colombian citizens, to be provided by the contributory regime (CR) for Good Practice in Health Financing, p. 34 those with ability to pay, and by the subsidized regime (SR) for the poor. Individuals in the formal sector must contribute 12 percent of their salaries as premiums to the CR - 4 percent is contributed by the employee and 8 percent by the employer. The primary source of financing for the SR is national government transfers to departamientos (departments) and municipalities earmarked for health (56.3% of total resources); a 1 percent solidarity contribution from the CR (34.4% of resources); contributions from revenues of family benefits funds or Cajas de Compensacion Familiar, which are companies that manage benefits in addition to health provided by employers, such as recreational services (0.5% of total resources); and local tax revenues earmarked for health, obtained from "sin" taxes (8.8% of total resources). Solidarity is a key principle of the system - leading to subsidies from the rich to the poor and from the healthy to the sick. The subsidized regime allocates public subsidies to individual insurance premiums for the poor, following a proxy-means testing index known as Beneficiaries Identification System (Sistema de Identificaci6n de Beneficiarios) or SISBEN. SISBEN scores are calculated on the basis of a number of dimensions of poverty - including labor market participation, income, educational attainment, family structure, assets, housing material and crowding, and access to water and sanitation. Health plans are free to establish the level of payments for services they purchase from providers and payment mechanisms. Fees for health services in the NSHI have used the fee schedules that public health plans existing prior to the reform had developed as benchmarks, adjusted for inflation, but little updating for the 'real cost' of services has been carried out. These schedules are used as ceilings for price negotiations between health plans and providers. There are currently strong calls from provider associations for price regulation (setting floors). Two patterns of payment are common among all health plans - preventive and primary care services are mainly contracted through capitation; most specialist care and hospital care is paid on a fee-for-service basis or by service packages. There have been no evaluations of the effect of these payment mechanisms on access and quality. The benefits packages, or "compulsory health plans" (Planes Obligatorios de Salud) are grouped into three levels of ascending complexity that differentiate interventions in terms of the levels of specialization, technology, and the financial resources required for their provision. The first level of complexity includes preventive and emergency care, basic medical, dental and diagnostic services. The second and third levels include specialized and rehabilitation care, hospitalizations and their corresponding diagnostic tests. There are several key challenges remaining in terms of closing the remaining health insurance gaps and ensuring the financial sustainability of the NSHI. A major task is finding mechanisms to enroll the informal sector, which accounted for 58 percent of the working population in 2005. About a third of this population is known as the "sandwich population" which can be sorted out in two groups - one including people who earn less Good Practice in Health Financing, p. 35 than twice the minimum wage, who are not eligible for subsidies but for whom payment of the full contribution to the premium of the CR can be onerous. The second consists of workers with capacity to contribute to the premium but who lack incentives to become affiliated. According to a household survey in 2003, the total sandwich population was about 7.3 million workers and their families - or 17% of the total population. Of these, 2.5 million had some capacity to pay. The administrative costs of enrolling, monitoring, and collecting contributions from this population can be substantial, and the potential for adverse selection is large. The Government is currently piloting a program to provide partial subsidies to individual contributions to the premium. A second challenge for the CR is tax evasion. One of the strategies Colombia has started to implement to reduce evasion is investment in improvements in information systems - establishing links between the pension system data and health insurance contribution data. Increased enrollment of informal workers could complement these measures. Thirdly, in the case of the SR, not only it will be necessary to finalize the transformation of supply subsidies to demand side subsidies by fully eliminating historical budgets to public providers, but it is also likely that it will be necessary to find alternative sources of financing. Under current growth, employment and poverty conditions, and available resources, universal health insurance will not be reached until the year 2010. Key Lessons Some of the key lessons from the reforms are: • The overall level of economic development is clearly a key factor in determining the options for expanding health care coverage. Sustained economic growth for four years prior to and following the implementation of the 1993 reforms was critical for popular acceptance of increases in contributions to the contributory regime, as well as the mobilization of general revenues to finance insurance expansion. • Expansion of health care coverage requires substantial financial investments into the health sector. In particular, expanding coverage of social insurance systems - either in terms of the population covered or the depth of benefits package - is likely to lead to substantial increases in health care spending, partly due to moral hazard but also due to the need to cover premiums for the poor. As Colombia's health care reforms have expanded insurance coverage from 20 percent to 78 percent of the total population total health expenditures have grown from 6.2 percent ofGDP in 1993 to 7.8 percent in 2004, mainly as a result in increases in public expenditures. Good Practice in Health Financing, p. 36 • A clear legal and institutional framework is important in expanding formal sector insurance coverage. The 1991 Constitution sets the legal framework and provides political legitimacy for the National Social Health Insurance scheme with the participation of the private sector. • In Colombia, the previous implementation of a program for the development of community-based health insurers in the rural areas allowed the national social health insurance system to quickly enroll a large number of enrollees throughout the country. The availability of SISBEN has provided the system with an instrument to identify its target population and target subsidies to the neediest. • Cost containment measures are needed to limit the increase in expenditures. These may include full separation of financing and provision through the elimination of supply side subsidies to public providers. • Inclusion of informal sector is a challenge, especially those who are above the poverty line yet insurance premiums constitute an onerous expenditure. Costa Rica Background Costa Rica is a small (population 4.43 million) upper-middle income country in the Latin America and Caribbean region which has a long tradition of political stability. It has a GDP per capita ofUS$4,349, which is about average for an upper-middle income country. In terms of health, Costa Rica devotes some 6.6 percent of its GDP to health, which is above average for its income level and spends about US$290 per capita on health, which is average. Some 77 percent of total health spending is public, which is well above other countries of the similar income levels. Only some 20 percent of total health spending comes from out-of-pocket payments. In terms of health outcomes, with life expectancy at 79 years and an infant mortality rate of 11 per 1,000 live births, Costa Rica fares better than other countries of its income or health spending levels. Its life expectancy places Costa Rica in the ranks of OECD countries. The country has 1.32 doctors per 1,000 people, slightly above average for its income level. However, with 1.4 hospital beds per 1,000 people, Costa Rica is well below average. Both in terms of breadth and depth of the coverage, the health care system provides significant protection to the population. The Costa Rican Social Security Fund (Caja Costarricense de Seguro Social) or CCSS is a public entity that ensures 89 percent of the population and also administers the national pension system. CCSS is also the country's principal curative health care provider. Membership and financing for the public health Good Practice in Health Financing, p. 37 insurance system is based on employment; membership is mandatory for formal sector employees. Dependents and indigent persons are provided free care under special regimes. Access to primary health care services in Costa Rica is universal - 99 percent of the population is able to access primary health services, The good outcomes observed in terms of coverage (both breadth and depth) are the result of a long history of efforts to improve the living conditions of the population that date back to the 19th century. The CCSS was created in 1941. However, before health care reforms in 1994, access to primary care services was restricted to 25 percent of the popUlation. Description of Key Reform Elements and Accomplishments Costa Rica represents an outstanding example of a middle-income country with full health care coverage, considerable financial protection, and an extensive package of services. Both in terms of breadth and depth of the coverage, the health care system provides significant protection to the population. Evidence from national surveys shows that, on average, Costa Rican households allocated 2.6 percent of their income to health spending in 2005. Access to high quality medicines is also universal and exceeds WHO basic drug list standards. To date, the list of drugs includes 460 active principle drugs in 608 presentations, with products for all types of pathologies. In Costa Rica health coverage is financed through a substantial commitment from both government and employers. Following a social insurance model, employers contribute 9.25 percent of the wages paid to CCSS. Workers contribute an additional 5.50 percent of their wages, and the State contributes 0.25 percent of the total national wages. In sum, this amounts to 15 percent of workers , salaries. Self-employed and informal-sector workers are encouraged to join the CCSS' voluntary plans where workers pay between 5.75 percent and 13.75 percent of their salaries depending on income. The health system in Costa Rica is based on solidarity - no matter the size of a person's contribution to the system, he or she has equal access to health care services in the public delivery system. The poor are covered by the "Non-contributory" and the "Insured by State" regimes; it is estimated that approximately 620,000 individuals belong to these two regimes, which are cross-subsidized by formal sector employees. According to CCSS actuarial studies, about 50 percent of the contributions from formal sector employees are used to cover health expenditures of pensioners, independent workers, and poorer households. Contributions to social security are the most important source of financing, representing almost 60 percent of the health sector's total revenues. The second source - sales of goods and services - corresponds to out-of-pocket expenditures by households. When out-of-pocket expenditures and contributions to the social security are considered together, households contribute approximately half of the total revenues. Government contributes 7.3 percent of its total revenues (5 percent in the form of taxes and 2.7 percent in the form of contributions to the CCSS). This figure, however, does not Good Practice in Health Financing, p. 38 consider the contributions of the Government to finance health insurance to poor households. The role of external funds is minimal- less than four percent of total health expenditures. Evidence from national surveys suggests that the system is financially protecting the population, especially poorer groups. The 2006 Survey on Household Revenues and Expenditures estimates that the Costa Rican households allocate 2.6 percent of their income to health spending, or 4.1 percent of current non-capital spending. Most of households' health expenditures go for specialist and hospital services and drugs. The share of health expenditures in the richest quintile is approximately 6.6 percent of income, 3.7 times higher than the share of health expenses in the poorest quintile. In other words, most ofthe out-of-pocket expenditures are from wealthier families. These achievements are due to sustained public health expenditures and political commitment. In the long run, Costa Rica has been capable of applying adequate legislative and institutional arrangements to achieve high rates of health care coverage and to ensure universal provision of health services. When Costa Rica faced difficult obstacles and conditions to achieve its coverage objectives, the country was able to opportunely renew its structure by redefining the network of providers, as well as by integrating private and nongovernmental partners in the delivery of health services. Key Lessons Some key lessons from the reform are: • Political Commitment: both political parties agreed on the need for change. Even though they did not agree on certain details, the common view that a health reform was required prevailed. The relatively rapid approval of the reform loans in the Congress was a signal of such political support. • Transparency and accountability have been critical ingredients for successful expansion of health care coverage. Political parties, the CCSS, healthcare providers and other workers, and MOH staff all had an important voice in shaping the reforms that have resulted in the current health coverage system. Internal institutional reforms have also played role. • Wide participation of main stakeholders: participation in the reform process was not limited to political parties but included CCSS workers (especially medical staff), MoH staff, medical doctors, and communities. Transparency of the information flowing to the different stakeholders was a key aspect. Several groups opposed the reforms, as expected, but continuous negotiation and a consensus-building orientation allowed the implementation of the main reform components. It is important to recognize that, at the end, the original reform agenda has still not been fully implemented, and there are some missing elements. For instance, allowing patients the right to choose their provider is an issue not yet resolved. Good Practice in Health Financing. p. 39 • The most important organizational change introduced in the CCSS as part of the reforms has been the creation of a dedicated purchasing unit, administratively and financially independent from the cess hospitals. Separation of purchasing from provision functions has allowed one single department to fully concentrate on planning, negotiating, monitoring, and evaluating the performance of health providers - and removed conflicts of interest in the purchasing relationship that might had compromised the efficiency of the purchasing process. • Information systems have played an important role in Costa Rica. Investments in computer systems have allowed the financial managers of the CCSS to monitor the flow of revenues on a daily basis, compared to a 30 day delay as was the case previously. These systems have facilitated the implementation of new payment mechanisms designed to enhance provider efficiency and performance, by monitoring management agreements - signed between the CCSS and primary care providers - that explicitly define coverage targets linked to payments. • These reforms require both time and money to produce results, as indicated by the history of health reforms in Costa Rica, including separating purchasing from delivery functions within the CCSS, reorienting a curative-based coverage system towards preventive care, and modernizing medical technology and information. While achieving full coverage with primary care services took the country less than 15 years to achieve, universal health insurance coverage for a full benefits package will take considerably more time. • Important health outcomes were achieved by expanding primary care services as the main channel to provide universal access to the entire population. The share of the population with access to PHC services jumped from 25 percent of the population just before the reform to almost universal coverage by 2005. • Sustainability of the reforms depends on financial support. It is important to have a short-term budget to launch the program, such as the two reform loans approved in Congress, that came from the World Bank (US$ 22 million) and the Inter American Development Bank (US$ 42 million). It is also critical to commit to investing resources over the long-run. Costa Rica is an example that, if sustained over time, health investments pay-off. Good Practice in Health Financing, p. 40 Estonia Background Estonia is a small upper middle income country in the Europe and Central Asia region with a population of 1.3 million and a GDP per capita of US$8,328. It was part of the Soviet Union, gaining independence in 1991. Its GDP per capita declined initially following separation from the Soviet Union but recovered soon thereafter. Since then, Estonia has been one of the fastest growing economies among formerly socialist transition economies in Eastern and Central Europe. Estonia's population, like much of that of Europe, is both declining and ageing: in 2002, over 20% of the population was above 60 and the age dependency ratio is 48%. Almost 70% of the population is urban. Health expenditure is about 5.3% of GDP, lower than that of other countries at similar income levels. The public share of total health expenditure is high (76%), with 87% of funds coming through the health insurance system. About 11.5% of the government's budget (including SHI spending) is devoted to health, about the average for Estonia's income level. 88.3% of private expenditure and 21.3% of overall health expenditure was out-of-pocket in 2004. Life expectancy in Estonia declined following separation from the Soviet Union, reaching a low point of 61 for men and 73 for women in 1994. It has been on the rise ever since: in 2005, life expectancy was 66 for men and 77 for women. Average life expectancy (73 years) is sti11low by EU standards, and remains lower than the average for its income level. Its infant mortality rate is very low: it was only 6 per 1,000 live births in 2005. In keeping with the trend among ex-Soviet republics, the density of physicians and hospital beds per 1,000 is high at 4.48 and 5.3, respectively. Literacy rates are close to 100%. Prior to reforms, Estonia's health system was that of the Soviet Union. Universal health care was centralized and provided nominally free to everybody. However, technology and clinic methods were not as advanced as those in Western countries. Pre-reform, as was the case with other former Soviet regions, the system was characterized by input norms and targets, oversupply of hospital beds, over-specialization, and an out-migration of health personnel. The rationale for reform was to introduce a system that would ensure secure sustainability and predictability of financing for the health sector, especially given the precarious state of the economy. The first Health Insurance Act was approved by the Parliament even before political independence was achieved. Good Practice in Health Financing, p. 41 Description of Key Reform Elements and Accomplishments The reform in 1991 introduced a classic Bismarkian model of coverage with mandatory social health insurance and decentralization. A health insurance tax was introduced with a contribution rate of 13% paid fully by employers on employee salaries. In 1994, the health insurance tax was incorporated into the social tax with an earmarked share going to the health system. Initially, funds were not pooled across the 22 non-competing district based funds. As a result, some of the more deprived areas had lower resources than others. However, this has changed. The revenue collection has been streamlined through the Government Tax Revenue Office and since 2000, the Estonia Health Insurance Fund (EHIF) incorporated under a separate public law, has operated as a single pooled fund with selected administrative and contracting responsibilities delegated to four regional offices. The EHIF is legally obliged to balance yearly revenues and expenditures, and this requirement has been fulfilled in almost all years since the inception of the scheme. Initially, there were no out-of-pocket payments. In 1993, however, Estonia introduced the prescription pharmaceutical reimbursement system which was based on some cost- sharing. Some 23% of out-of-pocket spending goes towards dental care. Flat co- payments are charged for some types of health services such as primary care physician home visits, outpatient visits, and hospital bed-days. The health insurance system is mandatory without an opt-out possibility. Private insurance was allowed to be taken to cover additional expenditure not covered by the mandatory health insurance. Contributions are related to being active in the workforce. Non-contributing individuals (children, pensioners, etc) represent almost half (49%) of the insured population and their expenses are implicitly subsidized by the others. The state officially contributes for only about 4% of the covered population. By the end of2003, 94% ofthe population was covered by Estonia's mandatory social health insurance scheme. The uninsured 6% are those persons who are of working age who are not participating in the formal labor market and are not eligible according to other criteria such as not being registered as unemployed or disabled. The health insurance reform was accompanied by carefully phased changes in the service delivery system. Major delivery system changes included implementation of the family practice model, rationalization of the hospital sector, modernization of the pharmaceutical sector, and the adoption of incentive based provider payment and risk sharing mechanisms. Primary health care is now provided by private family practitioners, and hospital and specialist care predominately by autonomous public hospitals incorporated under private law as stock-companies or not-for profit foundations (trusts). Good Practice in Health Financing, p. 42 Key Lessons Some key lessons from the reforms are: • The reform from a Soviet NHS to a SHI model has been carefully planned with major financing changes including a dedicated 13 percent payroll tax accompanied by carefully phased major changes in the delivery system and regulatory environment. • Health system revenue collection should be in line with overall fiscal policy and should take into account labor market policies, future growth and labor force projections. • The reform has benefited from solid economic growth except for the very beginning. • Implementing the system through the formal employment sector reduced corruption and other distortions. • Streamlining revenue collection through the Government Tax Revenue Office allowed the EHIF to focus more on purchasing health care for its beneficiaries. • Having a single risk pool and clear regulatory frameworks has allowed the EHIF to be an efficient administrator of SHI funds and perform as an effective purchaser of services. • Annual actuarial soundness is assured by legislatively limiting SHI spending to available revenues. • The Government has been very adept at monitoring and undertaking mid- course corrections such as recentralization of risk pools and certain delivery system responsibilities that were initially delegated to local governments, which had in some cases been captured by providers and were unable to deal effectively with certain cross regional service delivery externalities. • Strategically designed out-of-pocket payments - in conjunction with social insurance - can play an important role in ensuring sustainability of health care financing through moderating demand and ensuring that the poor and vulnerable groups receive coverage of essential health care and are still protected from catastrophic expenditures. • The reform has benefited from strong support from the medical community. Good Practice in Health Financing, p. 43 Kyrgyz Republic Background The Kyrgyz Republic is a small (population 5.2 million) mountainous Central Asian country with some two thirds of its population rural. Beyond some gold and agriculture, it is a relatively poorly resource endowed country. In 2004, per capita GDP was $434, about average for a low-income country. In terms of health, the country is devoting some 5.6 percent ofGDP or US$24 per person to health spending. Some 41 percent of total health spending is public, and health accounts for 8.4 percent of the government budget. Out of pocket payments account for 55.7 percent of all health spending, a high level compared to other comparable income countries. In terms of health outcomes with an infant mortality rate of 58 and life expectancy of68, the Kyrgyz Republic compares better than countries of its income or health spending levels. In terms of its delivery system with physician and hospital bed to population ratios of2.51 and 5.3 per 1,000, the country exhibits the typical FSU pattern of high input levels compared to other comparable income countries. Educational levels are also high with adult literacy over 90 percent. Prior to the health sector reform efforts initiated in the mid-nineties, the Kyrgyz Republic had a typical norm driven, centrally planned, general revenue financed health system in which free health care was a right for all citizens. Following the breakup of the former Soviet Union in 1991, the Kyrgyz economy collapsed, and between 1991 and 1996 GDP fell by over 50 percent. The country found itself faced with an over- resourced and unaffordable health system in which 'fixed' salary costs and infrastructure accounted for 75 percent of all health spending, private sector provision was almost non- existent, and private payments were largely under the table 'informal' payments to public providers for preferential treatment in the vast state owned and managed health infrastructure. While in principle the prior Soviet system provided both universal coverage and deep coverage in terms of financial protection, supply side rationing of technologies, diagnostic tests, and devices was rampant and manifested itself in waiting lists and informal payments. With the collapse of the FSU and its drastic economic impact, the Kyrgyz Republic could not longer financially sustain this system and undertook a series of reforms to deal with both funding and delivery system improvements as it was essential that some of the 'financing' be derived from efficiency gains from improvements in the bloated health infrastructure. Description of the Key Reform Elements and Accomplishments Starting in 1996 the Kyrgyz Government undertook a two-phased health sector reform program. The first phase which lasted from 1997-2001 was focused on obtaining additional revenues through the introduction of the Mandatory Health Insurance Fund (MHIF) financed by a small complimentary payroll tax. The second phase of the reform Good Practice in Health Financing, p. 44 was launched in 2001 and was a complete refonn of the funding flows through the system and purchasing mechanisms, explicit specification of the benefit package, and a restructuring of the service delivery system. MHIF initiated in 1997 focused first on the economically active population who paid a 2 percent payroll tax contribution and pensioners and the registered unemployed who were funded out of the pension and unemployment funds. Starting with 30 percent population coverage, coverage reached 83 percent in 2001 with the addition of children funded by the state and social welfare recipients funded by the social welfare funds (Meimanaliev 2003). While the program did not significantly change the depth of coverage from FSU times, the real benefit was to allow the step by step introduction of popUlation and output based purchasing mechanisms (Kutzin 2002). Phase 2, introduced in 2001, marked the introduction of significant provider payment and service delivery refonns including the explicit designation of the benefit package and a subsidized outpatient drug benefit program. The refonn accomplished a number of objectives including a focus on primary care, service delivery rationalization including refonning the bloated hospital system, improved efficiency through provider payment mechanisms, updating of treatment protocols, development of national level risk pooling, diversification of health sector financing, greater consumer choice, encouragement of private sector provision, and the important national political benefit of clarity of entitlements to specific benefits. While the refonn has led to major improvements in efficiency and sufficient revenues to reach pre-breakup spending levels, the evidence on depth of coverage and financial protection as evidenced by the high OOP share which disproportionately affects the poor, shows that the Kyrgyz republic still faces significant financial protection and equity challenges. Key Lessons The key lessons from the reform are: • Successes are in part due to the comprehensive approach rather than on single instruments or major bullets; • Complex reforms require careful sequencing of various reform steps; • Paying attention to institutional aspects was important in order to ensure sustainable benefits; creating the MHIF as a parastatal agency was crucial to adopt strategic purchasing and abandon inefficient input-based budgeting; • Phased implementation and careful sequencing were an effective implementation approach and helped build capacity and stakeholder support as well as learning by doing; • Strong collaboration of the development partners facilitated hannonized support for refonn design and implementation; Good Practice in Health Financing, p. 45 • Poor economic conditions experienced during the implementation limited government's ability to achieve financial protection objectives; • The slow pace of reforms in public financial management created a challenge for achieving health sector reforms; compared to other sectors in the Kyrgyz Republic, the health sector was revolutionary in being the only one to move away from input-based line-item budgets and administrative control mechanisms towards performance based management; • Elimination of copayments without commensurate increases in public funding is leading to the return of 'informal' payments; • Given its income level, the Kyrgyz Republic has a well developed health information system that facilitated policy development, especially prospective provider reimbursement, based on enrollment at primary care facilities, hospital admissions, and outpatient utilization; and, • The Kyrgyz reforms should be replicable in transition economies with excess capacity and reduced fiscal space. Good Practice in Health Financing, p. 46 Sri Lanka Background With a GDP per capita of just over US$l,OOO, Sri Lanka is a lower-middle income country with a population of 20 million. The country is predominantly rural, with only 15% of the population living in urban areas. Total health spending - at US$43 per capita - is about 4.3% of GDP, below average for its income level. The majority of expenditure (54%) is private, of which 48% is out-of-pocket. Inpatient provision is largely public (>95%). The largest part of private sector provision is ambulatory care. Government spending on health is about 8.4% of the overall budget. The country enjoys much better population health indicators compared to other comparable income countries: in 2004, life expectancy was 75 and the infant mortality rate was 12 per 1,000 live births. In 2005, the number of physicians and hospital beds per 1,000 was 0.55 and 2.9, respectively, above the levels found in comparable income countries. Government health spending is financed exclusively from general tax revenues. There is no social health insurance. Private insurance coverage rates are small. Sri Lanka was an early reformer in relation to other countries included in this study. Its reforms can be traced back to 1931 when the country held its first national elections with universal franchise that resulted in the transfer of power to elected Sri Lankan leaders while the country was still a British colony (it achieved full independence later, in 1948). The health system prior to reforms was colonial. The emphasis was on preventive care through public health programs. There were a few modem urban health facilities - funded by a mixture of user fees and general revenue - catering primarily to Dutch and British residents. Most Sri Lankans depended on traditional healers. Motivated primarily by economic and productivity considerations, there was some limited provision of health care provided to plantation workers financed by a fees on exports. Consequently, plantation workers had health outcomes that were, in general, better than their rural non- plantation counterparts. A severe malaria epidemic in 1934-35 had a devastating impact on the rural population. Although by this time Sri Lanka was under elected self-rule, the new administration did little to alleviate the problems created by the epidemic. The general level of resentment among the public that followed the response stimulated policy-makers to reprioritize the role of the state in ensuring the health of the population. Description of Key Reform Elements and Accomplishments The principal scaling-up of health reforms occurred in the 1930s through to the 1950s, following the malaria epidemic of the mid-1930s. Subsequent developments have been relatively minor and incremental to this initial fundamental shift in the structure of the health system. One of the key aspects of the reform in Sri Lanka was the expansion of Good Practice in Health Financing, p. 47 free health care provision to rural areas through the building and staffing of government hospitals and dispensaries, particularly in the 1940s. As a result of the refonns, the emphasis of the health system changed from one that was preventive and sanitation- oriented to one that emphasized universal access to curative health services through hospitals. User fees, another remnant of the colonial health system, were removed in 1951. The health system changed from one in which only the urban rich had access to modem medicine - with the rural population reliant primary on traditional healers - to one in which effectively the entire population has access to modem care. The density of coverage is very high, with most Sri Lankans now living within 2-3kms of a public sector health facility. There are other aspects of the Sri Lankan refonn that are notable. The incidence of health expenditure has been pro-poor and the system of health financing is progressive (Rannan-Eliya, 2001). Catastrophic health spending is very low (van Doorslaer et aI, 2006). Out-of-pocket health expenditure is incurred primarily by the rich who are more likely to seek private care due to its (modestly) higher quality and responsiveness. Efficiency has been an important and critical element in Sri Lanka's success: it enabled it to use a limited budget to reach the poor. Sri Lankan public hospitals are able to deliver inpatient admissions and outpatient visits at a cost per capita and health to GDP ratio that are far lower than those found in other comparable income developing countries. This has been achieved by high bed tum-over rates and short average lengths of stay. Labor productivity is also high, with government doctors and nurses seeing, on average, more patients in both inpatient and outpatient settings than is the nonn for developing countries. As part of the refonns, public sector doctors have also been allowed to practice privately after-hours, thereby increasing their income-generation abilities and providing an incentive for relocation to rural areas (Rannan-Eliya, 2001). Key Lessons Some of the key lessons ofthe refonns are: • The importance of democratic accountability to ensure that the health system is responsive to the needs of the poor. The case argues that a key enabling factor for Sri Lanka is democracy, which has been effective in giving voice to the rural poor. Anti-incumbancy in a functioning democracy made the choice to pursue universal coverage an easy one (politically) in Sri Lanka. 9 9 This is in line with other recent research that suggests that governments are more responsive in setting where the electorate is more informed and politically active [see, for instance, Besley and Burgess (2002)]. Nobel laureate Amarty Sen has often made similar points with regard to the importance of democracy and free press in helping governments prioritize health and education. However, cross-country evidence of the positive impact of democracy on health is weak and inconclusive. A recent article [Safaei, J (2006)] does find a small positive effect of democracies on health outcomes in a cross-country setting, but many other studies have found no impact (Asian Development Bank, 2006). Good Practice in Health Financing. p. 48 • Countries can rely on effective (and free) public provision of health that is funded out of general taxation. Sri Lanka has no social insurance and relatively low levels of labor market formality, again suggesting that these latter two are not necessary for successful health financing reforms. And, therefore, it may not be prudent to push social insurance policies in countries that are not ready for them. • Sri Lanka is not exceptional in terms of its revenue generation and has tended to utilize efficiency gains to effectively finance increases in coverage (along with nudges towards private care for the rich to ease the burden on the system). This suggests the importance of good governance in allowing for effective implementation of such policies. Sri Lanka is not very corrupt and has a history of good public administration -- both are strong pre-conditions, it appears, for the successful implementation of the NHS-type system extant there. • High levels of education appear to have played a facilitative role in Sri Lanka, for instance in terms of influencing demand for health by placing a high priority on this issue and sustaining it through changes in government. Thailand Background With a population of about 64 million and a GDP per capita ofUS$2,441, Thailand is a relatively large lower middle income country in the East Asia and Pacific region. It stands out as being one of the few countries in Asia to never have been colonized. The country was particularly badly hit by the 1997 Asian financial crisis but appears to have recovered in recent years. 68% of its population is rural. At 3.5%, the proportion of spending on health relative to GDP in Thailand is very low for its income level. In 2004, the government's share of health expenditure was 64.7%. The remainder was largely out-of-pocket, with 74.8% of private spending being out-of-pocket. Even though overall health expenditures are low, a relatively large proportion (11.2%) of the government's budget is devoted to health. Health provision is mixed, with both public and private providers. Thailand has high population health outcomes. Life expectancy in 2005 was 71 years and the infant mortality rate was 18 per 1,000 live births. There are 2.2 hospital beds and 0.37 doctors per 1,000, the latter being somewhat low given Thailand's income level. Literacy rates, both for adults in general and for adult females, are higher than 90%. Prior to initiation of recent reforms, Thailand pursued attainment of universal health coverage through four different schemes. The first of these was the Medical Welfare Scheme (MWS), which was introduced in 1975 with the aim of providing free medical care to low-income groups. The scheme was funded by general taxation and covered the poor, the elderly, the disabled, and children aged below 12. An estimated 11 million Good Practice in Health Financing, p. 49 people were eligible for this scheme. They were provided a comprehensive package of services without user charges at public facilities. There were some problems related to funding and targeting, and the quality of care provided was questionable (Pannarunothai 2002). The second scheme was the Civil Servant Medical Benefit Scheme (CSMBS) which was non-contributory and was designed to provide health benefit coverage to civil servants and their dependents (a total of7 million people). The scheme was also funded through general taxation. Beneficiaries were free to choose private or public providers, although only 50% of costs incurred not exceeding 3,000 baht were reimbursed at private health facilities. Compared to the other schemes, CSMBS had the highest expenditures as a result of its fee-for-service reimbursement model (Sriratanaban 2002). The Social Security Scheme (SSS), launched in 1990, covered about 8 million formal- sector employees (but not their dependents). It followed a capitation model, but there remain concerns with regard to quality of care provided (Tangcharoensathien et al 2002). The Voluntary Health Card Scheme (VHCS) was introduced in 1993 to cover those who were not eligible for any of the other schemes. The scheme collected premiums from households, the Ministry of Public Health, and from an Asian Development Bank loan. However, there were concerns that the scheme was not reaching the intended target population and suffered from a classic adverse selection problem. The VHCS was not successful in expanding coverage due to financial non-viability (Donaldson et aI1999). As of 1998, about 80% of Thailand's population was insured. In 2001, Thailand introduced a Universal Coverage Scheme (UCS) (Tangcharoensathien et al 2002). The rationale for introduction was to deal with some of the problems of the earlier schemes and to increase coverage among the uninsured. Description of Key Reform Elements and Accomplishments UCS began in 2001 as a mandatory scheme funded by general taxation and a 30 baht co- payment. It merged MWS and VHCS and covered the remaining uninsured. Unlike many other countries of similar income levels seeking to expand coverage, Thailand's reform was rapidly implemented and scaled up nationwide within one year. All Thai citizens are eligible for the scheme which provides a standard benefits package, and all contract provider networks - both public and private - are required to provide these services to registered beneficiaries. The number of uninsured has decreased from 20% in 1998 to 5% in 2003 and UCS alone now covers 74.7% of the population (or 47.7 million people) (Vasavid et aI2004). More recently, the poor, elderly, children, and disabled have been granted exemption from copayments. The results of the reform have been remarkably pro-poor. Following the implementation of the UCS scheme, there has been a 25% increase in outpatient care and a 9% increase in hospitalizations, both of which are concentrated among the poor. Additionally, government health subsidies are now effectively reaching the poor at health centers, district hospitals, and provincial hospitals for both outpatient and inpatient care and Good Practice in Health Financing. p. 50 reducing inequality in living standards among the population (Limwattananon et al 2005). This has also translated into improved financial protection; the rates of catastrophic spending have decreased, and fewer people are impoverished due to OOP spending than before the reform (Limwattananon et al 2005). Regarding equity in the financing of health care, the better off pay more in general taxes as a share of their income than the worse off (O'Donnell et al 2005). Accordingly, DeS has achieved a strong pro-poor focus through improvements in the equity of utilization, public subsidies, and the health system's progressive financing structure. Although the DeS scheme is widely regarded as having been successful, especially in increasing access and providing financial protection, there remain concerns that it has led to an unsustainable increase in demand, resulting in a rapid increase in the workload of health personnel. In addition, there are no earmarked funds for the DeS and it is vulnerable to budgetary competition on a year-to-year basis. Key Lessons Some of the key lessons of the reforms are: • While sustained growth may help lay the foundation for expansions in coverage, major health reforms can still occur shortly after a recession if fiscal space exists. Even after GDP growth rates had become positive again in 1999 and 2000 following double digit declines, total health spending still shrank slightly over this period. There was also a significant reduction in defense spending that created budgetary room for the health sector's expansion in 2001. • Strong political and social support for universal coverage provided the necessary conditions for a rapid scale up. The motto "30 baht treats all diseases" proved to be simple and popular and the National Health Security Bill was the first bill sent to parliament because its proposal had been signed by more than 50,000 citizens. • The sustained development of primary care since 1980 and establishing its importance under ues has supported efficiency. The gatekeeping role is especially salient since there are concerns that the system is under-funded and capitation rates are too low. • Setting appropriate provider reimbursement through a combination of prospective payment methods has also improved efficiency. Thailand's experience with a variety of payment schemes over the last three decades helped inform the reimbursement design in DeS; however, the Civil Servant Medical Benefit Scheme pays providers based on fee for service and has the highest per capita spending of any of the insurance schemes; the MWS had been conducting systematic monitoring and evaluation on DRGs since 1998; capitation has been successful at containing costs for SSS. As a result of these experiences, DeS was designed to use a mix of capitation, global budgets, and DRGs to put financial risks Good Practice in Health Financing. p. 51 on providers to contain expenditures. Reform of the payment mechanism of the Civil Service Medical Benefits scheme is an ongoing political challenge. • General revenue financing has been pragmatic and beneficial to the equity in health care financing. VHCS suffered from adverse selection, did not expand coverage, and was not financially viable. By contrast, the UC system is funded by progressive general taxation and does not rely on household contributions, which were difficult to collect in the informal sector. • The rapid implementation of ues and its achievement of national coverage within one year were possible partly because reforms were built on previous investments in the delivery system. Thailand has had a sustained investment in the health system infrastructure for decades, including gradual development of facilities and human resources in rural areas. This network, coupled with some utilization of the private sector, greatly enabled the increase in access immediately after the big bang reform. • Good governance arrangements have built stakeholder participation. Partly due to the strong social movement, civil society has an influential position with the top decision making authority for UCS. The committee, comprising all key stakeholders and chaired by the deputy Health Minister, meets weekly to deal with policy implementation. • An extremely high level of technical capacity and a strong evidence base, especially given Thailand's income level, was important to inform policy making and facilitate implementation. Tunisia Background Tunisia is a small (popUlation 9.91 million) lower middle income country in the Middle East North Africa (MENA) region with a per-capita GDP of$2,832. Total per-capita health expenditures were US$175 in 2004 - 6.2% ofGDP. Approximately half of health spending comes from private sources and the vast majority (83 percent) of private spending comes from out-of-pocket expenditures by households. In terms of health status, Tunisia is one of the best performers in the MENA region, and is also above average in comparison to other countries' with similar income levels. Life expectancy is 73.5 years, and infant mortality is 21.0 per 1,000 live births. With respect to the delivery system, Tunisia has 1.34 physicians per 1,000 people, and 2.1 hospital beds per 1,000 people, which place the country slightly above and below average, respectively, in comparison to other countries of its income category. Disparities in health status between urban and rural areas as well as among different socio-economic groups persist: the population's health status is significantly better in urban areas than in Good Practice in Health Financing, p. 52 rural areas. However, the severity ofthese disparities is less in Tunisia than in other countries of the MENA region, especially in tenns of infant mortality and life expectancy. Description of Key Reform Elements and Accomplishments Over the course of the past three decades, Tunisia has developed a health system that covers nearly its entire population and which compares favorably with that of other middle-income countries. During this time period, Tunisia's population has undergone dramatic demographic, socioeconomic, and health status changes. The health system has had to confront multiple challenges arising from this triple transition. Health care coverage has been expanded through substantial increases in health workers; health facilities; modem medical equipment; and health care coverage. Consequently, health care expenditures have surpassed five percent of GDP since the second half of the 1990s. Until the 1980s, the health care system evolved based on the colonial tradition of a hospital-centered health infrastructure, concentrated in large urban areas. Over the course of the past three decades, Tunisia has progressively put into place a health system that covers nearly the entire population. These efforts have included the implementation of mandatory health insurance for substantial segments of the population - including government and other fonnal sector workers. At the same time, the health system has placed a strong emphasis on preventive programs, financed completely by the government. This has led to the eradication of some infectious diseases and significant reductions in the incidence of others. The Tunisian strategy of coverage expansion has emphasized the expansion of geographic coverage of the population - through primary health facilities that are both geographically and financially accessible to a large majority of the population. The strategy also focuses strongly on qualified human resources in the health sector. Tunisia has followed a consistent strategy in training and deploying medical personal since the 1960s - first concentrating on the capital, Tunis, and now decentralized, especially for the training of nurses. Additionally, more than 80 percent of Tunisians now have health insurance coverage - through either a health insurance scheme or a medical assistance program. There are two main types of social security systems, covering approximately seven million persons (71 percent of the population). The CNSS (Caisse Nationale de securite Sociale) provides health care coverage to 1.2 million private sector enrollees and their families - including employees, independent workers, and other categories such as students, the disabled, non-salaried agricultural workers, and Tunisians working abroad. The payroll contribution rate for this system is 4.75%. The CNRPS (Caisse Nationale de Retraite et de Prevoyance Sociale) covers approximately 0.6 million public sector employees and retirees and their families. The contribution rate for this scheme is 1 percent each from employers and employees. These two schemes are mandatory. Relatively poor households are covered by two subsidized medical assistance programs provided to the enrollees and their families, together known as Assistance Medicale Gratuite (AMG). Good Practice in Health Financing. p. 53 The poorest are exempt from all fees, while others who are eligible for a reduced fee scale contribute 10 TND annually. The expansion of insurance coverage in Tunisia has featured the early introduction of a mandatory health insurance for civil servants and formal sector employees, gradually extended to additional groups. At the same time, additional reforms have placed a strong emphasis on preventive programs and the extension of coverage through primary health facilities, geographically and financially accessible for the entire population. Private sector providers have been increasingly incorporated in the health care delivery system. Private hospital bed capacity has doubled in the past ten years, and the number of medical examinations provided in the private sector approaches that of public providers. However, the public sector remains the main health care service provider- supplying about 85 percent of total hospital beds and more than 55 percent of medical personnel. Since 1990, the government has pursued a health sector strategy emphasizing: (1) the continuation and consolidation of gains in primary health care services; (2) improvements in hospital care - particularly in university hospitals - through structural and institutional reforms; and (3) legislative reforms to permit greater private investment in health service provision. In terms of health financing, the extension of the health care coverage has been accompanied during the last sixteen years, by a growth in the relative household contribution, a decreased in the government contribution, and a slight increase in the social health insurance contribution. The health status of the Tunisian population continues to improve, with the further development of preventive and curative health care services, and a continued decrease in the birth rate. Tunisia demonstrates better health outcomes than other average middle- income countries, particularly in terms oflife expectancy, infant mortality, nutritional status, and reproductive health. Key Lessons Some key lessons from the reforms are: • A strong initial commitment to primary health care, associated with the gradual development of more intensive and more costly hospital care, influenced by the regulated introduction of new medical technologies contributed to the success of the Tunisian case. • The extensive growth of the private sector, associated with the quantitative and qualitative change in health care demand, has contributed to a significant growth in household health expenditures - posing a potential burden on poorer households. Since the end of the 1980s, there has been a rapid development of the private health care provision sector, in terms of number of providers and of health Good Practice in Health Financing, p. 54 care categories. Private hospitalization, especially in surgery, has developed strongly. • The dimensions of the large-scale mandatory health insurance reform should not be underestimated when initially implemented. Coverage expansions are complex and entail extensive financial commitments. • Regulatory and administrative capacity - backed up by a commitment of access to quality health care services - is critical to the successful expansion of health care coverage. The development of public hospitals' management capacity and the promotion of participatory management will lead to improvements in hospital performance. • Likewise, the development of information systems is essential. Despite substantial investments in information technology, Tunisia currently faces extensive demands in this area. VietNam Background With a population of about 84 million and a GDP per capita ofUS$550, Viet Nam is one of the larger low-income countries in the East Asia and Pacific region. It is notable as now being one of the most dynamic economies in the world, having posted average economic growth rates on the order to 6-8% per year over the past decade or so after initiation of economic reforms in the mid-1980s. The country is predominantly rural: only 25% of the population currently lives in urban areas, but the ratio of the rural to urban population is rapidly falling. In 2004, it spent about 5.5% of its GDP on health- about US$30 per capita - which was about the average for its income level. The bulk of health spending is private (72.9% of total health spending), with a high proportion (74.2%) of that being out-of-pocket. The government spends a relatively low 5.0% of its budget on health. Viet Nam has excellent health outcomes, especially in relation to its income level. In 2005, life expectancy was 71 years and infant mortality was only 17 per 1,000. Hospital beds and physicians per 1,000 people are 2.1 and 1.34, respectively, somewhat higher than the average for other low-income countries. Both adult literacy and female literacy rates are high. Prior to the reforms, Viet Nam's health system was geared towards provision of "health for all" via an extensive network of community health services and inter-communal polyclinics for primary care and government hospitals for higher levels of care (MOH, 2001). Coverage was fairly widespread. However, the quality of care was low primarily due to a lack of resources which resulted in shortages, especially of modem medical Good Practice in Health Financing, p. 55 drugs and equipment with estimates suggesting that only 8% of perceived need for medical equipment was met. Health financing, already at low levels, suffered even more so at the end of the 1980s as a result of the collapse of the Soviet Union, a country on which Viet Nam was quite dependent for foreign aid. There became an urgent need to mobilize alternate sources of finance for the health sector to function effectively. Description of Key Reform Elements and Accomplishments Health system reforms were undertaken in the backdrop of broader economic and social reforms (Doi Moi) which began in 1986 with a gradual transition from a centrally- planned Soviet-style economy to a socialist-oriented market-based one. In 1987, private health and pharmaceutical sectors were officially sanctioned. This resulted in a burgeoning small-scale private health sector in the urban areas, often staffed by moonlighting government health workers. Recent estimates suggest that almost 30% of all provision is now private, mostly concentrated in the urban areas. Partial service fees and charges for drugs and diagnostics were introduced in 1989 in all public facilities, with the revenues being utilized for improved service provision. This "commercialization" of the health sector had an inimical impact on utilization rates among the poor, triggering the introduction of health insurance reforms in 1992: both as a means to reduce the burden on the government budget and to improve access among the poor. Insurance coverage rates increased among those with compulsory, employment-related coverage and voluntary student insurance. However, extensions of coverage to the general population have not met expectations. As a result, the government has made several attempts - such as provision of subsidized health insurance cards for the poor and mandating free health care for children - in an attempt to improve coverage among vulnerable groups. The government's goal is provide universal insurance coverage by 2010 through a mix of social health insurance, targeted health insurance for the poor, and voluntary insurance schemes. In 2004, about 60% of the population was covered by health insurance. Compulsory insurance contribution rates are set at a modest level of 3 percent of contractual salary and basic allowances, pension, social insurance payments, scholarship or minimum wage depending on which entitlement group the insured individual belongs to. For the employed, workers pay 1 percent and their employers pay 2 percent of salary. For retirees and people receiving social insurance benefits, the contributions are paid by VSS. For the other groups the government budget pays the contributions. For the poor and the elderly aged 90 and older, the contribution is a fixed amount of 3.1 USD per person per year with the contribution paid from the state budget. The reforms in Viet Nam have improved access to care and health outcomes. Inequalities remain an issue, and the government is making a special effort to target the poor in order to mitigate those. Good Practice in Health Financing. p. 56 Key Lessons The key lessons from the reform are: • Social insurance may not be a panacea, especially in a low-income setting. Viet Nam's experience indicates the difficulty of reaching high levels of coverage using social insurance alone. There is a large level of informality and, in order to cover informal workers and other vulnerable groups, the reforms had to be adjusted to incorporate special targeting programs to reach those not covered by social msurance. • Viet Nam's experience stands quite in contrast with that of Sri Lanka, suggesting that there may be a multiplicity of modalities for successful reform. Whereas the latter was a democracy, the former is a one-party state. Whereas the latter favored universal coverage using tax-revenue funded public provision, the latter has favored an approach that is more reliant on use of cost-sharing, user fees, and social insurance with adjustment to reach those not covered. Initially, the poor suffered in Viet Nam as a result of the reforms (high out of pocket payments and catastrophic expenses) and the focus on equity came about later in the reforms, unlike Sri Lanka which started with an initial focus on equity. • High literacy levels seem to have played an important role in Viet Nam, as in Sri Lanka. The population's awareness of the importance of health is high and that appears to have helped facilitate the choice and implementation of reforms. • Good governance has been important factor. This has been true both in terms of the implementation of social insurance as well as in the success of targeting programs such as the special funds for the poor. Decentralization in the implementation of the programs appears to have helped as well, especially in targeting. • Historical inheritance of a network of primary health care facilities helped play an important enabling role. This has been true for some of the other cases as well: initial conditions seem to matter. • The case also mentions the importance of strong economic growth in helping sustain the reforms. Sri Lanka, by way of contrast, has had more moderate levels of growth but still saw reforms through. At least in these two cases, some amount of economic growth seems to have helped the reforms along. • The gradual implementation of reforms - as opposed to a "big-bang" approach -- appears to have helped in Viet Nam's case in instilling a learning-by-doing influence into policy making. The deficiencies of initial reforms triggered the initiation of several adjustments aimed at solving the problem. Good Practice in Health Financing. p. 57 5. CONCLUSION - ENABLING FACTORS FOR EXPANDING COVERAGE Introduction While the nine countries described in this report have all made significant gains in expanding health care coverage, as discussed in Chapter 4, the countries themselves are quite different and each has a different story to tell. This final chapter of Part 1 attempts to draw generalizable lessons across the nine countries. Perhaps most importantly, the one factor that the countries all have in common is that they show that successful expansions of health care coverage require a considerable amount of time. During this time, such expansions also require extended political stability; economic growth; financial investments in human resources, physical infrastructure and information systems; sound planning; and institutional strengthening. Generalizable Conditions for Success Trying to generalize enabling conditions for successful health financing reforms from these nine case studies is difficult, particularly as these nine cases are hardly representative of all global experience and do not constitute a random sample. Nevertheless, these cases all represent countries that have undertaken serious reforms and based on our definitions of successes have done relatively well. These nine cases include two low-income countries and seven-middle income countries which span the entire middle-income country spectrum. Sri Lanka is the only pure NHS model, while the other countries represent SHI models, and Chile represents a hybrid SHI - private insurance model. However, even this categorization is not straightforward as Costa Rica's SHI system very much resembles a NHS with one fund, which owns its own facilities and is heavily subsidized from general revenues. In fact most of the SHI systems with the notable exception of Estonia are heavily subsidized by general revenues. Estonia is one of the few countries in the world that funds virtually its entire SHI system exclusively from payroll taxes. We have argued elsewhere that the models are not important, but what is important is the basic objectives attached to the three health financing functions as discussed in Chapter 2. While much of the detailed information needed to assess the equity, efficiency, and sustainability of revenue collection, the efficiency and equity of risk pooling, and the technical and allocative efficiency of purchasing is not available, we have in the analysis presented some indications of the sustainability of spending by looking at health shares of the national government budgets, assessed the level of risk pooling by the level of non-pooled OOP payments, and looked at allocative and technical efficiency crudely by comparing delivery system capacity, health outcomes and overall spending levels to global trends. We have looked at success in terms of the breadth of coverage - defined as the percent of the population with insurance coverage and financial protection - by looking at out of pocket spending as a share of total spending and in some cases as a share of household income. More importantly, an attempt is made to see what common underlying factors appear to have enabled these results. Good Practice in Health Financing. p. 58 Motivations for undertaking these reforms were very different for Estonia and the Kyrgyz Republic, both of which faced major financial crises following the breakup of the FSU - resulting both in revamping their Soviet style systems with SHI systems but even more fundamentally making drastic changes to their inefficient delivery systems. Vietnam, previously ideologically a communist state, was affected financially in much the same way as it was heavily dependent on Russia for subsidies prior to the breakup. Other countries were motivated by various humanitarian/solidarity factors and in most cases important rural representation in the political process, which highlighted equity considerations vis a vis both physical and financial access. It is clear that the paths taken by these nine countries are very different and heavily contingent on the political economy and institutional arrangements in the individual countries. The question is: as health systems and health reforms are very country- specific, is it possible to find a common set of enabling conditions in these nine successes or is the conclusion simply everything is country-specific and one cannot generalize? Based on the lessons summarized in Chapter 4, the cross-country analysis in Chapter 3, and the detailed studies in Part 2 of this report, there would appear to be a number of commonalities among almost all these countries. In addition, for several different smaller groupings of these countries, a number of other factors are highlighted. While one might argue that these results could have been posited ex ante, the findings are based on a standardized definition of success in terms of coverage expansions and financial protection in the context of average or below average spending for countries which have achieved good health outcomes. In addition, development assistance for health is minor or absent as a health financing source in most of these countries. General 'Enabling' Conditions In virtually all of the country cases several common characteristics are apparent: • Robust economic growth during the bulk of the reform period; • Long-term government commitment; • Strong institutional and policy environment; • High population literacy levels; • Rural focus; • Coverage expansions accompanied by carefully sequenced and significant delivery systems (including human resources for health) and provider payment changes; • Flexibility and mid-course corrections; • Strong and explicit primary care focus; and • Strong fiscal commitment to cover the poor. Good Practice in Health Financing, p. 59 From subsets of these nine cases other factors emerged including: • The importance of good information systems and evidence based decision- making; • Strong stakeholder support; • Efficiency gains and copayments used as financing mechanisms; • Consolidation of risk pools; and • Limits to decentralization This chapter discusses these 'enabling' conditions as necessary and perhaps sufficient conditions for successful health financing reforms. In this context, the unmet goals of these reforming countries are also highlighted. The chapter concludes with suggestions about next steps for developing the global health financing evidence base. Broadly Shared Enabling Conditions Robust Economic Growth during the Bulk of the Reform Period Given the length of the reform periods, the nine case countries faced varying economic circumstances. However, for the most part, as shown below in Figure 10, virtually all the countries had extended periods of favorable economic growth during much of their reform efforts. Figure 10 - Real GDP per capita trends, 1960-2005 Sri Lanka Viet Nam Thailand "'0 i§ "'0 i§ i8 ~fil ~O ~O ~8 2tfil 2tfil 2tl() ~'" (90 ~'" (90 ~'" (90 ~ 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 Year Year Year Estonia Tunisia Kyrgyz Republic "'0 "'0 ~8 %§ J %§ ~fil ~~ ~~ ~§ ~'" (90 1960 1970 1 ~e1990 2000 2010 ~'" (90 --------- 1960 1970 1~ 1990 2000 2010 ear ~'" (90 1960 1970 1~ ----- 1990 2000 2010 ear Colombia Chile Costa Rica "'0 "'0 i§ i8 ~ ~§ ------- ~fil ~o 2tfil ~'" ~~ ~'" ~§ ~'" ~ (90 (90 (90 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 201 0 1960 1970 1980 1990 2000 2010 Year Year Year Source: World Development Indicators Good Practice in Health Financing, p. 60 Even the Kyrgyz Republic, perhaps the least favored among the nine countries, achieved improved economic circumstances after the breakup of the FSU. Estonia's reform, in which virtually the entire SHI system which covers the whole population is funded from a 13 percent payroll tax, has been enabled by strong economic growth, Economic factors have also been particularly important in Colombia, where sustained economic growth for four years prior to and following the implementation of the 1993 reforms was critical for popular acceptance of increases in contributions to the National Social Health Insurance as well as the mobilization of general revenues to finance insurance expansions for the poor. The enrollment of a large fraction of the population through coverage of the families of contributing individuals allowed the public sector to free additional resources for health care provision for the poor. Similarly, the case study of Vietnam shows that sustained economic growth over the past 15 to 20 years has enabled the government to undertake ambitious health financing reforms, currently reaching over 60 percent of the population with a goal of universal health insurance by 2010. In real terms, per-capita spending on health has increased by two-thirds during this time period. Moreover, increases in public revenues have enabled the government to significantly increase its share of health spending, creating the possibility of new programs of free health care for the poor and for children. In Chile, the expansion of a fiscally sound social security system has required considerable economic resources and was made possible by an extended period of economic growth and an efficient tax collection system. But the experiences of Chile also demonstrate clearly that the extent of formality in employment is equally important. Registration in social security systems and other formal labor market mechanisms can be important facilitators for extending health care coverage in SHI systems but - as the cases of Sri Lanka and Viet Nam demonstrate - are not always necessary precursors. There is a strong correlation between the extent of formality ofa country's labor force and overall economic conditions. Chile has been able to combine growing public sector revenues with private investment to expand its social security system, including health coverage. Continuous and steady real growth of 5.5 percent over the past 30 years has enabled Tunisia to increase coverage and at the same time develop the necessary delivery capacity to accommodate these expansions. It has also allowed Tunisia to reduce poverty and have a continuous improvement of the living conditions of its population. Economic liberalization has helped spur economic growth in Sri Lanka, as in Vietnam. Such growth has been an important enabling factor for the government to expand coverage and enact its pro-poor reforms, particularly as sustained growth can provide the basis for enhanced government revenues and facilitates formalization of the labor market, which can further improve the collection of revenues in SHI-based financing systems. Good Practice in Health Financing, p. 61 Although Thailand suffered a recession between 1996 and 1998, its economy experienced very strong growth in the early 1990s. Thailand's economy also rebounded quickly from the shock and appears to have since recovered. During the first two years of positive GDP growth after the recession, total health spending growth remained negative. However, the positive economic growth coupled with decreased health spending during these years may have helped enable the dramatic increase in health spending that accompanied the introduction of universal coverage in 2001. As these countries have experienced strong economic growth, government per capita health expenditures have also been rising in real terms over the last decade. Rates of increase have been especially high in recent years in Chile, Colombia, Estonia, Kyrgyz Republic, Sri Lanka, and Thailand. The sustainability of government spending will be a key issue as countries move forward with their reforms. For example, it is not clear if the very high shares of the government budget devoted to health in Costa Rica and Columbia can be sustained. Some countries have demonstrated fiscal discipline, such as eliminating supply-side subsidies (such as historical budget support to public providers) and establishing budget ceilings as in Chile in order to contain costs. In some others, the sustainability of government spending will need to be addressed in the future through supply side controls or demand side cost sharing. Long-term Government Commitment In all cases government commitment to the reform lasted for many years or even decades. In no case presented here has the reform been ongoing for less than 10 years. Chile has been pursuing its reform for over 25 years. Given the complexity of many of these reforms in terms of scope, sequencing and concomitant implementation of major delivery system changes, and frequently development and adoption of modern health information and provider payment systems, it is not surprising that any success would require long-term government commitment. The adage that health reform is a perpetual process certainly comes to mind. Hsiao and Shaw (2007) make this same point with regard to SHI implementation experiences. Thailand, for example, introduced its health insurance scheme for the poor (MWS) in 1975. In particular, expansion of health care coverage requires substantial financial investments into the health sector. Expanding coverage of social insurance systems - either in terms of the population covered or the depth of benefits package - can lead to substantial increases in health care spending, partly due to moral hazard as well as unmet need. To contain public health spending, Chile has set budgetary ceilings for FONASA, the National Health Fund. Similarly as Colombia's health care reforms have expanded insurance coverage from 20 percent to 78 percent of the total population, total health expenditures have grown from some six percent of GDP in 1993 to some eight percent in 2003, mainly as a result in increases in public expenditures. However, the cases documented in this report show that significant reforms have been feasible even in countries that spend relatively small amounts on the health sector (Table 1). On a per-capita basis without adjusting for Purchasing Power Parity, the Good Practice in Health Financing, p. 62 countries documented in this report spend from $24 per capita (Kyrgyz Republic) to $463 per capita (Estonia) on health care - combining public and private sources. Thailand, Sri Lanka, and Vietnam in addition to the Kyrgyz Republic all spend less than $100 per person per year on health care. As a percentage of GDP, health care spending ranges from 3.5 percent in Thailand to 7.8 percent in Colombia. Sri Lanka, which spends just $43 per capita annually for health care, has been able to overcome some of the difficulties imposed by limited health financing by realizing improvements in efficiency - achieving efficiency gains in the public health system of one to four percent per year. Sri Lanka's life expectancy of 75 suggests that health spending alone is not the most important factor in determining population health outcomes. Sustaining health care coverage expansions requires sustained financing over the extended period oftime needed to put into the place the necessary institutional and administrative arrangements. The countries that have made the most progress in expanding coverage have also steadily increased financing in the health sector, through a combination of public and private sources. The Chilean model - which combines aspects of a private insurance system with social insurance - suggests that households' contributions, linked to health care coverage, are essential for a sustainable system. Strong Institutional and Policy Environment Institutional strength and political stability are also important enabling factors for a successful reform. Efficient tax collection and the ability to enforce mandatory membership in social insurance programs are essential functions for expanding coverage; both depend critically on institutional strength and stability. Likewise, the existence of competent, independent regulatory agencies is a prerequisite for the provision of both insurance coverage and health services. All these countries had when they started their reforms - and currently have - better than average institutional and policy environments as measured by the World Bank's Country Policy Institutional Assessment Index (CPIA). The index is composed of some 16 components measuring a wide range of institutional and policy management issues including macroeconomic and trade policy, financial sectors, public sector management, social safety nets, poverty focus, environmental management, and other factors. Strong institutional environments were found for both democratic and non-democratic countries as shown by the polity index of democracy in Figure 11. Higher values of the polity index signify greater degrees of democracy and changes over time reflect changes in the nature of the polity (Marshall and Jaggers 2005). Good Practice in Health Financing, p. 63 Figure 11 Political freedom trends in case countries, 1900-2004 Sri Lanka Viet Nam Thailand 0 0 0 a: (])~ ~ (])~ I!!~ ~ g'" 0'" (/)0 » ~o ~'" ~, """, ~, """, ~, a.~ 0. 0 , ~ 1900 1920 1940 1960 1980 2000 1900 19201940 1960 19802000 1900 1920 1940 1960 1980 2000 Year Year Year Estonia Tunisia Kyrgyz Republic 0 0 0 (])~ (])~ I!!~ 5", 5", 0'" ~o 0 &l ;0 »0 :;LO ~'" ~'" a.' ~ a.' ~ ~ a.' ~ 1900 1920 1940 1960 1980 2000 1900 1920 1940 1960 1980 2000 1900 1920 1940 1960 1980 2000 Year Year Year Colombia Chile Costa Rica 0 0 0 a: (])~ (])~ -utI I!!~ g'" (/)0 ~'" ~, a. g JlJ » """, ~, a.~ §'" »0 ""'", ~, a. g , 1900 1920 1940 1960 1980 2000 1900 1920 1940 1960 1980 2000 1900 1920 1940 1960 19802000 Year Year Year Source: POLITY IV data set Note: A polity score of +10 represents a fully-functional democracy and -10 represents an autocracy. For the nine country cases, there are highly democratic countries like Costa Rica, Estonia, and Sri Lanka, and less democratic countries like Tunisia, the Kyrgyz Republic and Vietnam, The others have all gone through periods of democracy and non- democracy, although the recent trend appears to be toward democratization, Of course, the institutional features of the different country cases are very country- specific. An important component of the Vietnamese experience has been the high level of both economic and political stability. The one-party structure of the Vietnamese governance system has enabled the country to develop policies that have been allowed to be implemented over an extended period of time - a potential advantage compared to the complexity of reaching agreements on health policy reform measures that span the political divide in multi-party democracies. Such political stability helps obviate the risk that health reform measures may be precluded from running their course over a protracted time period as a result of frequent changes in government. Competent regulatory bodies are an important institutional element. This is true for both public and private health insurance. Chile in particular has benefited from credible, independent, and efficient institutions. The Superintendencia de las ISAPREs regulates the benefits packages, contribution rates, and services provided by the ISAPREs, which Good Practice in Health Financing, p. 64 are private health insurers - providing transparent rules of the game and fostering competition among these insurers. At the same time, the success of the Chilean model is highly dependent on an effective public insurance system. If the Chilean social security system were dissolved in favor of a completely free-market system with only private insurers, it is likely that the process of adverse selection would leave only the lowest- risk, highest-income individuals insured. To avoid this problem, FONASA, the public insurer in Chile, is not permitted to employ discriminatory practices, making it possible to achieve broad coverage of the population. Similarly, in Colombia a clear legal and institutional framework has proven to be important in expanding formal sector insurance coverage. The 1991 Constitution sets the legal framework and provides political legitimacy for the National Social Health Insurance scheme with the participation of the private sector. Formal sector workers contribute through payroll taxes, and in return have access to a minimum benefits package defined as the Contributive Mandatory Health Plan (Plan Obligato rio de Salud). Workers outside the formal sector, and individuals eligible for public subsidy, have access to the Subsidized Mandatory Health Plan (Plan Obligato rio Subsidiado de Salud). Contributors can purchase additional coverage from for-profit Health Insurance Institutions (Empresas Promotoras de Salud). The system promotes solidarity among payroll-tax payers through a fund which collects all payroll-tax contributions and then distributes resources among the EPS on a per-capita basis using a risk adjustment mechanism based on age and sex. The system is regulated by an agency under the Ministry of Health. In Costa Rica, transparency and accountability have likewise been critical ingredients for successful expansion of health care coverage. Political parties, the Costa Rican Social Security (CCSS), health care providers and other workers, and MOH staff all had an important voice in shaping the reforms that have resulted in the current health coverage system. Internal institutional reforms have also played a role. The most important organizational change introduced in the cess as part of the reforms has been the creation of a dedicated purchasing unit, administratively and financially independent from the CCSS hospitals. Separation of purchasing from provision functions has allowed one single department to fully concentrate on planning, negotiating, monitoring, and evaluating the performance of health providers - and removed conflicts of interest in the purchasing relationship that might had compromised the efficiency of the purchasing process. Although difficult to measure, social solidarity is also an important facilitating factor. In general, the greater the tradition of solidarity within a country, that is, the greater the number of systems and institutions working towards solidarity in a society, the better the chances of success for pooling arrangements, including social security and health msurance. The case studies of Sri Lanka and Thailand show that democratic accountability can be a key facilitating factor, encouraging health systems to be responsive to the needs of the poor. Democratic accountability, achieved through free and fair elections, should not be Good Practice in Health Financing, p. 65 confused with political decentralization, community participation, cost-sharing or political decentralization. Sri Lanka has had little success with community participation in local government and in the running of individual facilities - but the most basic decisions about the health system are not taken at the local level, but nationally, where the poor do have a voice. In Thailand, the proposal for universal coverage (UCS) was the first bill sent to the national parliament because it had received more than 50,000 signatures from citizens. The 1997 Constitution gave civil society this power. High Literacy Rates All these success stories had high adult literacy rates in absolute terms as well as relative to other comparable income countries. While one cannot attribute causality, one can speculate that high literacy rates made it easier to sell the reform to the population, resulted in better demand side response to coverage expansions with respect to health seeking behaviors, and enabled individuals to be more informed consumers of both public and private care. Rural Focus With the exceptions of Tunisia and Chile (which was heavily urban at the start of the reform), rural representation and/or basic socialist equality objectives (Kyrgyz Republic, Estonia, and Vietnam) resulted in governments assuring that coverage and delivery system expansions encompassed rural areas and the rural poor. This undoubtedly had important impacts on health outcomes and certainly in heavily rural countries - such as Thailand, Vietnam, the Kyrgyz Republic, and Sri Lanka -led to high levels of population support for health reforms. The Sri Lanka, Thailand, Costa Rica, Vietnam, and Columbia reforms specifically focused on rural areas or on the poor, who disproportionately reside in rural areas. Given the high correlation between rural residency and poverty, such a focus undoubtedly contributed to the favorable health outcomes in the study countries. Accompanying Delivery System Reforms Perhaps it is obvious that significant coverage changes would need to be accompanied by serious delivery system reforms, and this is certainly the case with all nine of the cases. Interestingly the reforms comprise two types: (1) enhancements in delivery capacity to accommodate the increases in coverage; and (2) significant downsizing, rationalization, and upgrading of medical practice standards in the FSU countries. Sequencing of delivery system changes with financing reforms has been an important element in all the reforms. Dealing with physical infrastructure and more importantly human resources for health (HRH) have gone hand in hand. Several of the countries, perhaps most notably Thailand, have implemented advanced HRH policies to staff rural areas. Implementation of modern contracting techniques, provider payment systems, and risk sharing mechanisms has been an important element of all the reforms. Pharmaceuticals, given their high share of total and out pocket health spending, in terms Good Practice in Health Financing, p. 66 of essential drug lists, practice patterns (particularly in the FSU), and pricing have also been an important foci of the reforms. Flexibility and Mid -Course Corrections Given the long implementation periods of all reforms in all of these countries, and the careful sequencing of delivery system changes with the coverage expansions, it is not surprising that both evidenced based policies and mid-course corrections have impacted implementation. Thailand prides itself on the use of evidence-based policies in all aspects of its reform efforts. It was one of the earliest developing countries to develop and implement comprehensive national health accounts (NHA), and it has both used and significantly contributed to the global HRH evidence base. Thailand has for several of its insurance programs implemented modem provider payment and risk sharing approaches. It has undertaken detailed studies on equity and impoverishment due to high medical expenses. Estonia has exhibited marked flexibility in its reform and has for example reversed polices of decentralization of risk pools and delivery system responsibilities. Several of the countries (Colombia, Thailand, Sri Lanka, Vietnam, Kyrgyz Republic) made significant and explicit choices in the way that the care for the poor is subsidized through the SHI schemes. Primary Care Focus Every country case, with the exception of Sri Lanka, cites primary care and/or prevention as an important delivery system change. In some cases like Estonia and the Kyrgyz Republic, this entailed a complete re-orientation of the medical system toward a family physician model. While the Sri Lanka case argues that one of the keys to their success was a hospital-based system geographically distributed in all areas of the country, without a specific focus on cost-effectiveness, hospital outpatient units certainly likely proved to be effective purveyors of primary care. In other words the real issue may be that appropriate care can just as readily be efficiently distributed through hospital-based settings as the more usual rural clinics and community health worker approaches. Such a focus on primary care helps explain the good health outcomes found in these countries. Strong Fiscal Commitment to Cover the Poor The governments of nearly all of the nine countries have demonstrated strong fiscal commitment to finance coverage expansions for the poor through a combination of dedicated revenues sources and cross-subsidization through risk pooling. Many have subsidized both insurance contributions and copayment rates of the poor. For example, in Chile, the public insurer (FONASA) covers 96 percent of the poor and fully subsidizes health services for both this group and the unemployed. Colombia uses proxy-means testing to allocate subsidies to the poor in its social health insurance system. Thailand set up a new completely revamped general revenue funded program with no copayments for the poor, elderly, children, and disabled. The poorest are also fully exempt from copayments in the Kyrgyz Republic and Tunisia. Even in Vietnam Good Practice in Health Financing. p. 67 where targeting and implementation of pro-poor policies is weaker, there remains strong government support to subsidize the poor's contributions. Additionally, creating larger risk pools to facilitate cross subsidization from the rich to the poor has also demonstrated the pro-poor fiscal commitment. The transition in the Kyrgyz Republic from a system characterized by fragmented revenue collection and pooling to one with larger risk pools -first at the regional level and in 2006 at the national level - has provided the foundation for pro-poor fiscal redistribution. This move offers the prospect to improve both vertical and horizontal equity in health financing, and perhaps especially so in a country with a small population. Benefit incidence analysis has indicated that the poorest quintile receive more than 20 percent of government health expenditures in the majority of these countries (O'Donnell et a12007, Davoodi et a12003, Trejos 2002). Furthermore, in some countries, such as Costa Rica and Colombia, the poorest 20 percent receive close to 30 percent of total government health spending. Enabling Conditions Shared by Several Countries A number of other enabling conditions were highlighted for smaller groups of countries in the country case studies. It is quite likely that these conditions were also important determinants of success in many of the other cases. However, as they were not consistently highlighted as such, they are discussed separately. Importance of Good Information Systems and Evidence-Based Decision-Making As previously discussed, evidence-based decision-making was highlighted as one of the key factors underlying the Thailand reform. It was also important in Estonia and the Kyrgyz Republic as they adopted modem treatment protocols and, along with several other countries, modem provider payment reforms. Intrinsically related to evidence-based policy, information systems were highlighted by a number of countries. Tunisia's reforms were embodied in its investments in administrative capacity - particularly related to management and information technology in the public hospital sector. Information systems also have played an important role in Costa Rica. Investments in computer systems have allowed the financial managers of the CCSS to monitor the flow of revenues on a daily basis, compared to a 30 day delay as was the case previously. These systems have facilitated the implementation of new payment mechanisms designed to enhance provider efficiency and performance, by monitoring management agreements - signed between the CCSS and primary care providers - that explicitly define coverage targets linked to payments. In Colombia, the previous implementation of a program for the development of community-based health insurers in the rural areas allowed the national social health insurance system to quickly enroll a large number of enrollees throughout the country. The availability of SIS BEN, an information system providing updated information on Good Practice in Health Financing, p. 68 beneficiaries' economic status, has provided the system with an instrument to identify its target population and target subsidies to the neediest. Estonia in particular has taken a lead in terms of investments in information technology. The case study of Estonia shows the impact of these investments. In the mid 1990s, bills from providers were transferred to the health insurance fund on paper. All health care providers are now equipped with personal computers; electronic communication links the Estonian Health Insurance Fund (EHIF) with its patients, employees, and providers. In addition to facilitating efficiency in provision, electronic data transmitting has enabled the collection of high quality data for detailed analysis including benchmarking of providers and monitoring service utilization by specific vulnerable population groups. The Kyrgyz Republic also has a well-developed health information system that has facilitated policy development, especially the new provider mechanisms. Strong Stakeholder Support Strong stakeholder support from the bureaucracy, consumers, medical societies, and/or other stakeholders were identified as important elements of the reform processes in a number of the countries. The Estonian reform was strongly supported by the medical association, which had much to gain from the movement away from the old Soviet system. The support and voice of the rural poor were highlighted as important factors in the success and sustainability of the Sri Lanka reform. Strong support from the people through their political leaders has been a hallmark of the reforms in Thailand. The Costa Rican reforms have had strong support from the political parties, the CCSS itself, medical care providers, workers, and MOH staff as they all had important voices in shaping the reform. Efficiency Gains and Copayments Used as Financing Mechanisms Efficiency gains are an important means of both financing a health system and providing fiscal space for coverage expansions. Efficiency gains were a critical element in the expansions in the FSU countries and Sri Lanka. Thailand has implemented modern provider payment systems to assure efficiency in its new UCS program. Costa Rica has also focused on efficient purchasing arrangements to reduce outlays. Copayments have been used as both a financing mechanism and to preclude moral hazard in Chile, Vietnam, and Estonia, but less so in Sri Lanka, the Kyrgyz Republic and Thailand. Consolidation of Risk Pools The trend of consolidation of risk pools which is increasingly apparent in both high- income and middle-income countries is also found in several of the case studies here. Estonia and the Kyrgyz Republic have consolidated geographically defined pools into single national level pools. Thailand has reduced its number of major risk pooling schemes from four to three. Sri Lanka has a single national risk pool. There are no Good Practice in Health Financing, p. 69 cases of expansions in the number of risk pools among the nine cases, only consolidations. Limits to Decentralization Estonia and the Kyrgyz Republic both established decentralized risk pools at regional levels then recentralized them at the national level to improve both efficiency and equity. Estonia likewise decentralized certain delivery system functions then recentralized some of them as some local governments had been captured by medical care providers and from a provision perspective certain economies of scale required larger geographic entities to be fully exploited. Conclusion This chapter has attempted to identify the enabling factors for successful expansions of health insurance coverage. The factors identified make sense on conceptual grounds and would all be likely hypotheses for any rigorous evaluation of health financing change impacts. While these findings are useful, the more important contribution for the current evidence base is not these findings per se, but rather the need for the international community to do a better job in evaluating and disseminating the global evidence base. To do this, at least five important actions need to be taken. First, we need to rigorously define what we mean by successes in health financing. Second, we need to collect standardized and appropriate qualitative and quantitative information. Third, health systems characteristics must be described in enough detail so that all the critical components and their interactions can be clearly identified and assessed. Fourth, rigorous evaluations need to be undertaken. Fifth, these evaluations need to be disseminated to all stakeholders in a policy-relevant and user-friendly manner. Up to now, we in the global community have done a rather poor job in this area. However, with the growing importance attached to health financing and coverage by the G-8 as well as the countries themselves, and with a single magic bullet approach being pushed by certain organizations, it is incumbent for the global community to take charge and move this agenda forward in an analytically sound and policy relevant manner. It is hoped that global stakeholders through their various aid effectiveness fora as well as the G-8 and Paris Declaration Process will rise to the occasion and put this desperately needed effort on track. Good Practice in Health Financing, p. 70 APPENDIX 1- DEFINITIONS Adverse selection - a process that occurs when individuals with different expected losses are charged the same premium, whereby those with low expected losses drop out of the insurance pool, leaving only individuals with high expected losses, Adverse selection can make it difficult to sustain private insurance markets. In health care, relatively healthy individuals will self-insure or use cheaper plans, leaving higher-risk individuals in the risk pool. Allocative efficiency - requires that an economy provide its members with the amounts and types of goods and services that they most prefer. In standard economic theory in a perfectly competitive economy, it occurs when resources are allocated in such a way that any change to the amounts or types of outputs currently being produced (which might make someone better off) would make someone worse off. This is sometimes also called "Pareto efficiency". Other definitions of allocative efficiency include: • Optimization of health service mix to maximize health status. • Distribution of resources between health interventions to provide the most cost- effective mix. • Allocation of resources to maximize community welfare. Coinsurance - the share of costs for covered services that the consumer is required to pay out-of-pocket at the point of service. Coinsurance places additional administrative requirements on insurers and providers. Cost-sharing - a provision of health insurance or third-party payment that requires the individual who is covered to pay part of the cost of medical care received. This is distinct from the payment of a health insurance premium, contribution, or tax, which is paid whether medical care is received or not. Cost-sharing may be in the form of deductibles, co-insurance or co-payments (user fees). Community rating - a situation in which all members of an insurance pool are charged the same premium, regardless of their risk status and is sometimes geographically based. A pure community rating implies rates based on the cost of care for all those in the pool. A modified community rating includes demographic factors but not individual differences in health status. Copayments - a specific fee paid for each covered health service. Cream-skimming - a process whereby an insurer tries to select the individuals with lowest expected losses (i.e., health risks) in order to increase profits. Cream-skimming can make it difficult or impossible for individuals with high expected losses to purchase private insurance. Good Practice in Health Financing, p. 71 Deductibles - require the insured to pay all charges for covered services out-of-pocket until the total cost reaches the deductible amount. Financial protection in health - the concept that households and individuals should (1) be able to access health care when they need it and not be prevented from doing so by excessive cost; and (2) when they do access health care, not incur costs that prevent them from obtaining other basic household necessities - including food, education, and shelter And drive them into poverty. In low and middle-income countries, most studies that have approached the concept of financial protection in health have compared households' out-of-pocket spending to total household expenditures. Horizontal equity - the principle that says that those who are in identical or similar circumstances should pay similar amounts in taxes and should receive similar amounts in benefits. Load factor - administrative and other costs associated with underwriting insurance. The loading factor is the amount by which the actual premium charged exceeds an actuarially fair premium based only on service use. Moral hazard - for a given health status, individuals use more health care services when insured than when not insured, and more health care than they would if they were paying for it themselves. This raises the possibility of over-consumption or inappropriate consumption. Pooling is a key function of a health system. Pooling allows for sharing of risk, cross- subsidization from relatively wealthy to relatively poor and from well to sick. To be feasible, contributions to the pool must cover losses plus administrative costs. In general, prepayment and pooling of financial resources in the health sector offers several very significant advantages over individual-level (out-of-pocket) purchasing of health care services, including: (1) spreading financial risk; (2) promoting efficiency through pooled purchasing: and (3) enabling cross-subsidization from the relatively rich to the relatively poor, from well to sick, and over the life cycle. Purchasing - the process by which pooled funds are paid to providers in order to deliver a specified set of health interventions or assume the risk for such for an individual for some period of time .. Purchasing can be performed passively or strategically. Passive purchasing implies following a predetermined budget or simply paying bills when presented. Strategic purchasing involves a continuous search for the best ways to maximize health system performance by deciding which interventions should be purchased, how, and from whom. Reinsurance - allows insurance companies to spread the risks they assume by themselves taking out insurance for high cost cases, etc.. Reinsurance is rare in countries with weak capital markets, limiting the coverage and benefits packages of private health insurers. Good Practice in Health Financing, p. 72 Social health insurance -systems in which membership is publicly mandated for a designated population. There is a direct link between the payment of contributions which are generally based on a dedicated payroll tax to finance the system and the receipt of medical care benefits, even though there are exceptions to that rule (the contributions of people with insufficient income may be financed by the government in order to include them in the system). Finally, the management of social health insurance involves some degree of autonomy from the government, often through quasi-independent organizations in charge of the system. Technical efficiency - technical efficiency requires that for any given amount of output the amount of inputs used to produce it is minimized. Underwriting - evaluating individual health status and either rejecting potential insurees or placing them together with others at same level of assessed risk. Private insurers use underwriting to avoid adverse selection into their insurance pools. Vertical equity - the principle that says that those who are in different circumstances with respect to a characteristic of concern for equity should, correspondingly, be treated differently - those with greater economic capacity to pay should pay more; those with greater need should receive more. Good Practice in Health Financing, p. 73 REFERENCES Abel-Smith, Brian (1992). "Health Insurance in Developing Countries: Lessons from Experience." Health Policy and Planning, Vol. 7, No.3, pp. 215-226. Adams, Susan (2005). Vietnam's Health Care System: A macroeconomic Perspective. Prepared for the International Symposium on Health Care Systems in Asia, January 2005. International Monetary Fund. Antioch KM, Walsh MK (2004). The risk-adjusted vision beyond casemix (DRG) funding in Australia - International lessons in high complexity and capitation. European Journal of Health Economics. 2004 Jun;5(2):95-109. Aris B (2004). Slovenia's health system out-performs EU neighbours. Peace and prosperity have helped Slovenia ministers push through effective reforms. Lancet. 2004 Jun 26;363(9427):2143-6. Asfaw A, Braun Jv J (2004). Can community health insurance schemes shield the poor against the downside health effects of economic reforms? The case of rural Ethiopia. Health Policy. 20040ct;70(1):97-108. Asian Development Bank (2006), Measuring Policy Effectiveness in Health and Education, Manila: Asian Development Bank. Baeza, C et al (2002) "Towards Decent Work: Social protection in health for all workers and their families". Strategies and Tools against Social Exclusion and Poverty (STEP). Baeza, Cristian and Fernando Montenegro Torres (2004). Selecting Health Care Providers. Health, Nutrition and Population (HNP) Discussion Paper. The World Bank, Washington, DC. C. Baeza and T. Packard, (2006) Beyond Survival: Protecting Households from Health Shocks in Latin America, Washington, DC: The World Bank and Stanford University Press. Balabanova D, McKee M (2004). Reforming health care financing in Bulgaria: the popUlation perspective. Soc Sci Med. 2004 Feb;58(4):753-65. Good Practice in Health Financing. p. 74 Barnum, Howard, Joseph Kutzin, and Helen Saxenian (1995). Incentives and Provider Payment Methods. World Bank Human Resources Development and Operations Policy Working Paper No. 51, Washington DC. http://www.worldbank.orglhtmllextdrlhnp/hddflashlhcwp/hrwp043 .html Bassett, Mark (2005). Background Paper for Conference on Private Health Insurance in Developing Countries. Wharton Business School, 15-16 March 2005 - Frameworks for analyzing health systems, health financing and the regulation of health insurance. The World Bank. Besley, T and R Burgess (2002). "The Political Economy of Govt Responsiveness: Theory and Evidence from India." Quarterly Journal 0/Economics. 117(4): 1415-1451. Bhushan,I., S. Keller, and B. Schwartz (2002). "Achieving the Twin Objectives of Efficiency and Equity: Contracting Health Services in Cambodia." ERD Policy Brie/No. 6. Manila: Asian Development Bank. Blau, David M., (1985), "Self Employment and Self-Selection in Developing Country Labor Markets", Southern Economic Journal 51, no. 2, pp. 351-63. Boncz I, Nagy J, Sebestyen A, Korosi L (2004). Financing of health care services in Hungary. European Journal of Health Economics 2004 Oct;5(3):252-8. Bosch X (2004). French government approves unpopular health reforms. Lancet. 2004 Jun 26;363(9427):2148. Carrin and Hanvoravongchai (2003). Provider payments and patient charges as policy tools for cost-containment: How successful are they in high-income countries? Human Resources for Health; 1(6). Carrin, Guy and Chris James (2004). Reaching universal coverage via social health insurance: key design features in the transition period. World Health Organization, Geneva EIP/FERlDP.04.2. Castaneda, Tarsicio (2005). Targeting Social Spending To The Poor With Proxy-Means Testing: Colombia's SISBEN System. Social Protection Discussion Paper Series No. 0529. Social Protection Unit, Human Development Network, The World Bank. Charoenparij, Sriracha et al (1999). Health Financing in Thailand: Final Integrated Report. Thailand: Health Management and Financing Study Project. Management Sciences for Health, ADB No. 2997-THA. Cheng (2003). Taiwan's new National Health Insurance Program: genesis and experience so far. Health Affairs; 22:61-76. Good Practice in Health Financing, p. 75 Chollet, Deborah and Maureen Lewis (1997). "Private Insurance: Principles and Practice". In Schieber, George (ed.) Innovations in Health Care Financing, World Bank Discussion Paper No. 365, pp.77-103. Clark, Mary (2002). Health Sector Reform in Costa Rica: Reinforcing a Public System. Prepared for Woodrow Wilson Center Workshops on the Politics of Education and Health Reforms, Washington D.C., April 2002. Colombo, Francesca and Nicole Tapay (2004). Private Health Insurance in OECD Countries: The Benefits and Costs for Individuals and Health Systems. Organisation for Economic Cooperation and Development DELSAIELSAlWDIHEA(2004)6. Paris, France. Commission on Macroeconomics and Health (2002). "Macroeconomics and Health: Investing in Health for Economic Development." Report of the Commission on Macroeconomics and Health. Commission on Macroeconomics and Health - World Health Organization. Geneva. Cornell J, Goudge J, McIntyre D, Mbatsha S. 2001. South African National Health Accounts: The Private Sector. National Health Accounts Research Team and National Department of Health, Pretoria. Csonka A (2004). Hungary makes slow progress in health-system reform. Private funds, which could rescue Hungary's ailing health-care system, have been slow to materialise. Lancet. 2004 Jun 12;363(9425):1957-60. Davoodi, H et al. (2003). "How useful are benefit incidence analyses of public education and health spending?" IMF Working Paper. Washington, DC. Docteur E and Oxley H (2003). Health-Care Systems: Lessons from the reform experience. Organisation for Economic Co-operation and Development. Directorate for employment, labour and social affairs. DELSAIELSAIWDIHEA(2003)9. Doherty, Jane, Stephen Thomas and Debbie Muirhead (2002), Health Financing and Expenditure in Post-Apartheid South Africa, 1996/97 - 1998/99. The National Health Accounts Project. Centre for Health Policy, University of the Witwatersrand and Health Economics Unit, University of Cape Town. Donaldson, D., Pannarunothai, S., &Tangcharoensathien, V. (1999). Health Financing in Thailand Summary Review and Proposed Reforms. Nonthaburi: Health Systems Research Institute. Ensor, Tim (1999). "Developing Health Insurance in Transitional Asia". Social Science and Medicine, Vol. 48, No.7, pp. 871-879. Good Practice in Health Financing. p. 76 Ensor, Tim and Robin Thompson (1998). "Health Insurance as a Catalyst to Change in Former Communist Countries". Social Science and Medicine, Vol. 43, No.3, pp. 203- 218. Ensor, Tim and Stephanie Cooper (2004). Overcoming Barriers to Health Service Access and Influencing the Demand Side through Purchasing. Health, Nutrition and Population (HNP) Discussion Paper. The World Bank, Washington, DC. European Observatory on Health Care Systems (2000). Health Care Systems in Transition: Estonia. WHO Regional Office for Europe, Government of Norway, Government of Spain, European Investment Bank, the World Bank, London School of Economics and Political Science, London School of Hygiene & Tropical Medicine, in association with Open Society Institute. European Observatory on Health Systems and Policies (2005). Health Care Systems in Transition - HiT summary: Kyrgyzstan. WHO Regional Office for Europe, the Governments of Belgium, Finland, Greece, Norway, Spain and Sweden, the Veneto Region ofItaly, European Investment Bank, the World Bank, London School of Economics and Political Science, London School of Hygiene & Tropical Medicine, in association with Open Society Institute. Figueras, Josep, Ray Robinson and Elke Jakubowski (eds). Purchasing to improve health systems performance (2005). European Observatory on Health Systems and Policies Series. Frenk, Julio, Jaime Sepulveda, Octavio G6mez-Dantes, and Felicia Knaul (2003). Evidence-based health policy: three generations of reform in Mexico. The Lancet, Vol. 362:1667-71. Gertler, P. and Van der Gaag, 1. (1988) "Willingness to pay for social services in developing countries." LSMS Working Papers. The World Bank. Gertler, Paul and Jonathan Gruber (2002). "Insuring Consumption against Illness." American Economic Review No. 92 (1): 51-76. Gertler, P. (2004). "Do Conditional Cash Transfers Improve Child Health? Evidence from PROGRESA's Controlled Randomized Experiment." Health, Health Care, and Economic Development. 94(2): 336-341. Gill,1. and Ilahi, N. (2000). "Economic Insecurity, Individual Behavior, and Social Policy." Background Paper for the Regional Study on Economic Insecurity. Office of the Chief Economist, Latin America and the Caribbean. The World Bank. Washington, DC. Gill,1. Packard, T and Yermo, 1. (2004) Keeping the Promise of Social Security in Latin America, Latin American Development Forum, World Bank, Stanford University Press. Good Practice in Health Financing, p. 77 Goglio, Alessandro (2005). In Search of Efficiency: Improving Health Care in Hungary. OECD, Paris, France. Gomez-Dantes 0, Gomez-Jauregui J, IncIan C (2004). [Equity and fairness in the Mexican health system reform]. Salud Publica Mex. 2004 Sep-Oct;46(5):399-416. Gottret, Pablo, and George Schieber (2006). Health Financing Revisited. HDNHE. The World Bank, Washington D.C. Government of Taiwan (2003). Statistical Yearbook. http://www.gio.gov.tw/Taiwan- website/5 -gp/yearbooklchpt02.htm# 1. Gwatkin, Davidson (2004). Are Free Government Health Services The Best Way to Reach to Poor? Health, Nutrition and Population (HNP) Discussion Paper. The World Bank, Washington, DC. Hlavacka S, Wagner R, Riesberg A (2004). Health care systems in transition: Slovakia. Copenhagen, WHO Regional Office for Europe on behalf of the European Observatory on Health Systems and Policies. Hsaio, William (2000). What Should Macroeconomists Know About Health Care Policy? A Primer. IMF Working Paper WP/00/136. International Monetary Fund: Washington, DC. http://www.imf.org/external/pubs/cat/longres.cfm?sk&sk=370 1.0 Hsiao, W. and P. Shaw. (2007). Social Health Insurance for Developing Nations. Washington, DC: The World Bank. Hu TW (2004). Financing and organization of China's health care. Bull World Health Organ. 2004 Jul;82(7):480. Ikegami N, Campbell JC (2004). Japan's health care system: containing costs and attempting reform. Health Affairs. 2004 May-Jun;23(3):26-36. Institute of Policy Studies (2001). Equity in Financing and Delivery of Health Services in Bangladesh, Nepal and Sri Lanka: Results of the Tri-country Study. Nepal Health Economics Association. Data International Ltd. Institute of Policy Studies of Sri Lanka (2001). Sri Lanka National Health Accounts - Sri Lanka National Health Expenditures 1990-1999. Ministry of Health. Health Policy Programme supported by the IDAlWB Health Services Project. International Labor Organization (2002). Extending Social Protection through Community Based Health Organizations - Evidence and Challenges. Geneva: International Labor Organization. Good Practice in Health Financing, p. 78 Jegers M, K Kesteloot, D De Graeve, and W Gilles (2002). A Typology for Provider Payment Systems in Health Care. Health Policy 60(3): 255-73. Jesse M, Habicht J, Aaviksoo A, Koppel A, Irs A, Thomson S. (2004). Health care systems in transition: Estonia. Copenhagen, WHO Regional Office for Europe on behalf of the European Observatory on Health Systems and Policies. John T, Charles S, Nathan R, Wilczynska K, Mgalula L, Mbuya C, Mswia R, Manzi F, de Savigny D, Schellenberg D, Victora C (2004). Effectiveness and cost of facility- based Integrated Management of Childhood Illness (lMCI) in Tanzania. Lancet. 2004 Oct 30-Nov 5;364(9445):1583-94. Joint Learning Initiative. (2004). Human Resources/or Health: Overcoming the Crisis. Cambridge: Harvard University Press. Karashkevica J (2004). Latvia's health care system on the move. Cah Sociol Demogr Med. 2004 Apr-Jun;44(2):221-42. Katharina Hauck, Peter C. Smith and Maria Goddard (2004). The Economics of Priority Setting for Health Care: A Literature Review. Health, Nutrition and Population (HNP) Discussion Paper. The World Bank, Washington, DC. Kuszewski, Krzysztof and Christian Gericke (2005). Health Systems in Transition - Poland. Edited by Christian Gericke and Reinhard Busse. European Observatory on Health Systems, Berlin. Kutzin, Joseph (2001). "A descriptive framework for country-level analysis of health care financing arrangements." Health Policy, Vol. 56, pp. 171-204. Kutzin, Joseph. 2002. Health Care in Central Asia. Policy Brief. Copenhagen: European Observatory on Health Care Systems. La Forgia, Gerard (ed.) (2004). Health System Innovations in Central America- Lessons and Impact of New Approaches. World Bank Working Paper No. 57. The World Bank. Washington D.C. Langenbrunner John C, Xingzhu Liu (2004). How to Pay? Understanding and Using Incentives. Health, Nutrition and Population (HNP) Discussion Paper. The World Bank. Washington D.C. Levine, R. (2007). Case Studies in Global Health: Millions Saved. Washington, DC: Center for Global Development. Levine, R. and M. Kinder. (2004). Millions Saved: Proven Successes in Global Health. Washington, DC: Center for Global Development. Good Practice in Health Financing. p. 79 Lim MK (2004). Shifting the burden of health care finance: a case study of public- private partnership in Singapore. Health Policy. 2004 Jul;69(1 ):83-92. Limwattananon, Set al. (2005). "Equity in financing of health care in Thailand, 2000-2004." Eguitap Project Working Paper #16. Erasmus University, Rotterdam nd IPS, Colombo. Liu, Xingzhu and Sheila O'Dougherty (2004). Purchasing Priority Public Health Services. Health, Nutrition and Population (HNP) Discussion Paper. The World Bank. Washington D.C. Londono, J (1996). "Estructurando pluralismo en los servicios de salud: la experiencia colombiana." Revista de analisis econ6mico 11(2): 37-60. Londono, J. and Frenk, J. (1997) "Structured pluralism: Towards an innovative model for health system reform in Latin America". Health Policy, 41(1): 1-36. Maceira, Daniel (1998). Provider Payment Mechanisms in Health Care: Incentives, Outcomes, and Organizational Impact in Developing Countries. Major Applied Research 2, Working Paper 2. Bethesda, MD: Partnerships for Health Reform Project, Abt Associates Inc. http://www.phrproject.com/publicat/arp/download/mar2/m2wp2fin.pdf Marek, Tonia, Rena Eichler, and Philip Schnab (2004). Resource Allocation and Purchasing in Africa: What is effective in improving the health of the poor? Africa Region Human Development Working Paper Series. The World Bank. Washington D.C. Markota, Mladen et al (1999). Slovenian Experience on Health Care Reform. Institute of Public Health of the Republic of Slovenia, Ljubljana. Croatian Medical Journal. 1999 Mar;40(2). Marshall, M. G. and Jaggers, K. (2005), Polity IV Project: Political Regime Characteristics and Transactions, 1800-2004, University of Maryland and George Mason University. Meimanaliev, Tilek. 2003. Kyrgyz Model of Health Care System. Bishkek: Uchkun. Meng Q, Rehnberg C, Zhuang N, Bian Y, Tomson G, Tang S (2004). The impact of urban health insurance reform on hospital charges: a case study from two cities in China. Health Policy. 2004 May;68(2):197-209. MOH (2001). 55 Years of Developing Viet Nam 's Revolutionary Health System. Hanoi: Medical Publishing House. Good Practice in Health Financing, p. 80 Ministry of Public Health, Thailand (2002). Thailand Health Profile 1999-2000. Bureau of Policy and Strategy. Printing Press: Express Transportation Organization. Montenegro, F. and Nazareli, R. 2004. "Health and Household Income and Consumption: A review of the literature." Background Paper. Washington, DC. Morlock Laura, Hugh Waters, Alan Lyles, Haluk Ozsari, and G5ksenin Aktulay (2004). Health Care Reform in Turkey - Charting the Way Forward. Final report submitted to: The Turkish Industrialists' and Businessmen's Association Turk Sanayici ve Isadamlari Dernegi (TUSIAD). Morlock Laura, Alan Sorkin, and Hugh Waters (eds) (forthcoming). National Health Insurance in Taiwan: What Lessons May Be Learned? New York: Elsevier Press. Mossialos, Elias, Anna Dixon, Josep Figueras and Joe Kutzin (eds) (2002). Funding Healthcare: Options for Europe. European Observatory on Health Care Systems Series. Open University Press. Musgrove, P. (1996) "Un fundamento conceptual para el rol publico y privado en la salud". Amilisis Economico. Noviembre, 11 (2). Musgrove, Philip, Riadh Zeramdini, Guy Carrin (2002). Basic Patterns of Health Expenditure. WHO Bulletin, Vol. 80, pp. 134-142. O'Donnell, O. et al. (2007). "The Incidence of Public Spending on Health Care: Comparative Evidence from Asia." World Bank Economic Review. 21: 93-123. O'Donnell, O. et al. (2005). "Explaining the Incidence of Catastrophic Expenditures on Health Care: Comparative Evidence from Asia." EQUITAP Project: Working Paper #5. Erasmus University, Rotterdam and IPS. Colombo. Organisation for Economic Cooperation and Development (2002) "The Reform of Health Care: A Comparative Analysis of Seven OECD Countries". Health Policy Studies. Paris. Organisation for Economic Cooperation and Development (2003). OECD Reviews of Health Systems - Korea. ISBN 92-64-00892-6. Paris, France. Organisation for Economic Cooperation and Development (2004). Proposal for a Taxonomy of Health Insurance. OECD Study on Private Health Insurance. Paris, France. Organisation for Economic Cooperation and Development (2005). OECD Reviews of Health Systems - Mexico. ISBN 92-64-00892-6. Paris, France. Good Practice in Health Financing, p. 81 Packard, T. and Barr, A. (2002) "Revealed Preference and Self-Insurance: Can We Learn from the Self-employed in Chile?" World Bank Policy Research Working Paper No. 2754, Washington, DC. Pan American Health Organization (1999). Cuba Profile of the Health Services System. Program on Organization and Management of Health Systems and Services, Division of Health Systems and Services Development. Pan American Health Organization (2002). Profile of the Health Services System of Costa Rica. Program on Organization and Management of Health Systems and Services, Division of Health Systems and Services Development. 2nd edition. Pannarunothai, S. (2002). Medical Welfare Scheme: Financing and targeting the poor. In : Pramualratana, P.& Wibulpolprasert, S. eds. Health Insurance Systems in Thailand. Nonthaburi: Health Systems Research Institute, 62-78. Pannarunothai S, Patmasiriwat D, Sritharnrongsawat S (2004). Universal health coverage in Thailand: ideas for reform and policy struggling. Health Policy, Apr;68(1):17-30. Pauly, Mark, (1986), "Taxation, Health Insurance, and Market Failure in the Medical Economy", Journal of Economic Literature, Vol. 24, No.2, pp. 629 - 675. Phelps, C. E. (1978) Illness Prevention and Medical Insurance, Journal of Human Resources, Vol 13, Supplement: National Bureau of Economic Research Conference on the Economics of Physician and Patient Behavior, pp, 183 - 207. Preker AS, Carrin G, Dror D, Jakab M, Hsiao W, Arhin-Tenkorang D. (2002). Effectiveness of community health financing in meeting the cost of illness. Bull World Health Organ. 80(2): 143-50. Preker, A. and Langenbrunner, J. (2005)."Spending Wisely: Buying Health Services for the Poor". The World Bank. Washington, DC. Preker, A. S. and G. Carrin (2003). Health Financing for Poor People: Resource Mobilization and Risk Sharing. Washington, DC, World Bank. Preker, Baeza, Jakab, and Langenbrunner (2000). The World BanklILOIWHO Resource Allocation and Purchasing (RAP) Project - Arrangements that Benefit the Poor and Excluded Groups. Rannan-Eliya, RP (2001), Strategies for Improving the Health of the Poor - The Sri Lankan Experience, Colombo: Institute of Policy Studies Health Policy Programme. Good Practice in Health Financing, p. 82 Rawlings Laura B (2004). A New Approach to Social Assistance: Latin America's Experience with Conditional' Cash Transfer Programs. Social Protection Discussion Paper Series No. 0416. The World Bank, Washington, DC. Rodwin VG, Le Pen C (2004). Health care reform in France--the birth of state-led managed care. N Engl J Med. 2004 Nov 25;351(22):2259-62. Rodwin VG, Le Pen C (2004). Health care reform in France--the birth of state-led managed care. N Engl J Med. 2004 Nov 25;351(22):2259-62. Rosa RM, Alberto IC (2004). Universal health care for Colombians 10 years after Law 100: challenges and opportunities. Health Policy. 2004 May;68(2): 129-42. Safaei, 1. (2006). "Is Democracy Good for Health?" International Journal of Health Services. 36(4): 767-786. Schieber G, Baeza C, Kress D, and Maier M. (2006). Financing Health Systems in the 21st Century. In Dean Jamison, George Alleyne, Joel Breman, Mariam Claeson, David Evans, Prabhat Jha, Anthony Measham, Anne Mills, Philip Musgrove (Eds.), Disease Control Priorities in Developing Countries, second edition. Oxford and New York: Oxford University Press for the World Bank. Schieber, George (ed.) (1997). Innovations in Health Care Financing. The World Bank. Schieber, George and Akiko Maeda (1999). "Health Care Financing and Delivery in Developing Countries." Health Affairs, Vol. 18, No.3, pp. 193-205. Schieber, George and Nicole Klingen (1999). Islamic Republic of Iran - Health Financing Reform in Iran: Principles and Possible Next Steps. Prepared for Social Security Research Institute, Health Economic Congress, October 1999. The World Bank. Schieber, G., L. Fleisher, and P. Gottret. "Getting Real on Health Financing." Finance and Development. 43. No.4 (2006): 46-50. Sriratanaban, J. (2002). Civil Servant Medical Benefit Scheme: Unregulated fee-for- service and cost escalation. In : Pramualratana, P.& Wibulpolprasert, S. eds. Health Insurance Systems in Thailand. Nonthaburi: Health Systems Research Institute, 43-50. Somanathan, Apamaa (2003). Beneficiaries from Public Health Care in Five Asian Countries. Equity in Asia-Pacific Health Systems Network. Presentation at World Bank conference Reaching the Poor. Suraratdecha, Chutima (2003). Thailand's Universal Insurance Coverage Initiative. Thailand Health Systems Research Institute. Presentation at World Bank conference Reaching the Poor. Good Practice in Health Financing, p. 83 Szende A, Mogyorosy Z (2004). Health care provider payment mechanisms in the new EU members of Central Europe and the Baltic states. Current reforms, incentives, and challenges. Eur J Health Econ. 2004 Oct;5(3):259-62. Tabor, Steven R (2005). Community-Based Health Insurance and Social Protection. Social Protection Discussion Paper No. 0503. The World Bank, Washington, DC. Tangcharoensathien V, Wibulpholprasert S, Nitayaramphong S (2004). Knowledge- based changes to health systems: the Thai experience in policy development. Bull World Health Organ. 20040ct;82(l0):750-6. Tangcharoensathien, V., Srithamrongsawat & Pitayarangsarit, S. (2002). Overview of Health Insurance Systems. In: Pramualratana, P.& Wibulpolprasert, S. eds. Health Insurance Systems in Thailand. Nonthaburi: Health Systems Research Institute, 28-38. Taylor Rob, and Simon Blair (2003). Financing Health Care - Singapore's Innovative Approach. Public Policy for the Private Sector. The World Bank. Washington DC. Thomas S, Gilson L (2004). Actor management in the development of health financing reform: health insurance in South Africa, 1994-1999. Health Policy Plan. 2004 Sep; 19(5):279-91. Towse A, Mills A, Tangcharoensathien V (2004). Learning from Thailand's health reforms. BMJ. 2004 Jan 10;328(7431):103-5. Trejos, Juan Diego. (2002). "La Equidad de la Inversion Social en el 2000". Paper presented at the Octavo Informe sobre el Estado de la Nacion en Desarrollo Humano Sostenible. San Jose, Costa Rica: UNDP. Trujillo AJ, McCalla DC (2004). Are Colombian sickness funds cream skimming enrollees? An analysis with suggestions for policy improvement. J Policy Anal Manage. 2004 Aug;23(4):873-88. Tunisia Health Sector Strategy Report. The World Bank, Washington, DC. van Doorslaer, E, 0 O'Donnell, RP Rannan-Eliya, A Somanathan, SR Adhikari, CC Garg, D Harbianto, AN Herrin, MN Huq, S Ibragimova, A Karan, BR Pande, CW Ng, R Racelis, S Tao, K Tin, K Tisayaticom, L Trisnantoro, C Visasvid, and Y Zhao (2006). "Effect of Payments for Health Care on Poverty Estimates in 11 Countries in Asia: Analysis of Household Survey Data." Lancet (9544): 1357-1364. Good Practice in Health Financing, p. 84 van Doorslaer, E., E. O'Donnell, R. Rannan-Eliya, A. Somanathan et al. (2005) "Paying out-of-pocket for health care in Asia: Catastrophic and poverty impact." EQUITAP Project: Working Paper #2. Erasmus University, Rotterdam and IPS. Colombo. Vasavid, c., Tisayaticom, K., Patcharanarumol, W., & Tangcharoensathien, V. (2004). Impact of universal health care coverage on the Thai households. In: Tangcharoensathien ,V.& Jongudomsak, P. eds. From Policy to Implementation : Historical Events During 2001-2004 of Universal Coverage in Thailand. Nonthaburi: National Health Security Office, 129-149. Wagstaff, (2005). The economic consequences of health shocks. World Bank Development Research Group, Policy Research Working Paper no. 3644. Washington DC. Wagstaff, Adam and Menno Pradhan (2005). Health Insurance Impacts on Health and Nonmedical Consumption in a Developing Country Development Research Group and East Asia Human Development Unit. The World Bank, Washington DC. Wagstaff, Adam and Eddy van Doorslaer. (2003). "Catastrophe and impoverishment in paying for health care: with applications to Vietnam." Health Economics. 12: 921-934. Waters, Hugh (1999). Measuring the Impact of Health Insurance with a Correction for Selection Bias - A Case Study of Ecuador. Health Economics, Vol. 8, No.5, pp. 473- 483. Waters, Hugh and Peter Hussey (2004). Pricing Health Services for Purchasers - A Review of Methods and Experiences. Health Policy, Vol. 70, No.2 pp. 175-184. Waters, Hugh, Gerard Anderson, and Jim Mays (2004). Measuring Financial Protection in Health in the United States. Health Policy, Vol. 69, No.3, pp. 339-349. Waters, Hugh, Laura Morlock, and Laurel Hatt (2004). How Healthcare Purchasers Can Influence Quality - A Conceptual Framework and Comparative Analysis of Contextual Factors. International Journal of Health Planning and Management, Vol. 19 No.4, pp. 365-381. Waters, Hugh, Axel Rahola, and Jennifer Hulme (2005). Health Insurance Expansion in Morocco. The World Bank, Washington, DC. Watson N (2004). Poland's broken promises. Lancet. 2004 JuI17-23;364(9430):235-8. Webster P (2004). Reforms mean 25 million Russians lose free health care. CMAJ. 2004 Nov 9;171(10):1157. Good Practice in Health Financing, p. 85 Williamson, O. (1985) "The Economic Institutions of Capitalism." The Free Press. New York, NY. World Bank (2002). HNP Strategy Paper. Washington, D.C.: World Bank. World Bank (2003). Project Appraisal Document - Republic Of Ghana Health Sector Program Support Project. Report No: 24842-GH. Human Development II. Africa Regional Office. World Bank (2004a). 2004 World Development Report. Washington, DC: The World Bank. World Bank (2004b). Croatia - Health Finance Study. Report No. 27151-HR. Human Development Sector Unit, Europe and Central Asia Region. World Bank (2005). Beyond Survival- Protecting Households from the Impoverishing Effects of Health Shocks. Latin America and the Caribbean Region. The World Bank, Washington, DC. World Bank. (2007). World Development Indicators 2007. World Bank: Washington, DC. World Bank. (2007a) "Healthy Development: The World Bank Strategy for Health, Nutrition, and Population Results." Washington, DC: The World Bank. World Bank (2007b). World Development Report 2007: Development and the Next Generation. The World Bank, Washington, DC. World Health Organization (2000). World Health Report 2000 - Health Systems: Improving Performance. Geneva: WHO; 2000. World Health Organization (2007). WHO Statistical Information System (WHOSIS). Online database. http://www.who.int/whosis/database/core/core select.cfm Wouters, Annemarie (1999). "Alternative Provider Payment Methods: Incentives for Improving Health Care Delivery." Partnerships for Health Reform and Abt Associates Inc., Primer for Policymakers Series (no. 1). http://www.phrproject.com Yepes, F (200 I). Health Reform in Colombia. Health Care Reform and Poverty in Latin America. P. Lloyd-Sherlock. Washington, London & Brookings Institute Press: 232. Xu, Ke et al. (2003), "Household catastrophic health expenditure: a multicountry analysis." The Lancet, 362: 111-117. Good Practice in Health Financing, p. 86 Zum, Pascal and Orvill Adams (2004). A Framework for Purchasing Health Care Labor. Health, Nutrition and Population (HNP) Discussion Paper. The World Bank, Washington, DC. Good Practice in Health Financing, p. 87 PART 2 COUNTRY CASE STUDIES Good Practice in Health Financing. p. 88 Chile Ricardo D. Bitran Gonzalo C. Urcullo Good Practice in Health Financing, p. 89 Abstract This document describes the evolution of the Chilean health system and analyzes the implications of its current design, extracting lessons learned to be replicated in other countries. In 1981, Chile implemented the social security system in place today. The system consists of a public insurer (FONASA) and various private insurers (ISAPRES) that compete amongst themselves for members. Health services are furnished by public and private providers, who sell their services to the insurers and private consumers. The Chilean health system is noteworthy for its broad coverage of the population, including the following features: • Insurance for workers in the formal sector is compulsory • Insurance for workers in the informal sector is available on a voluntary basis • Members must contribute 7% of their earnings as a premium, and may make additional voluntary contributions • Members pay for a percentage of the cost of the services they consume (co- payments) • FONASA completely subsidizes health servIces for the indigent and unemployed • In 1991 an overseer, or Superintendent of Health (Superintendencia de Saiud), was created to regulate the ISAPRES. In 2001, the Superintendent of Health began regulating FONASA as well. • In 2005, a basic package of illness to be covered by all health plans was created. The basic package applies to both FONASA and the ISAPRES. Some of the lessons learned include: • Introducing public health insurance could result in major health spending increases; Chile set a budgetary ceiling to help control spending. • The predominantly formal labor market facilitates wide coverage by the social security system. • It is possible to mix public and private providers within the same system, but competition is only effective among the population with the highest income. • Private providers only cover 30% of the population. • Private contributions in the form of co-payments or premiums are essential for the system's sustainability and to promote efficient use. • If the State subsidizes part of the population, public finances must be stable, and there must be an efficient system to collect payments. • Although some questions remain regarding the equity and efficiency of the system, proposals for reform are aimed at improving the current system rather than replacing it. Good Practice in Health Financing, p. 90 Acronyms AFP Administradora de Fondo de Pensiones, or pension fund administrator CASEN Socioeconomic Census CEPAL Economic Comission for Latin America and the Caribbean FONASA National Health Fund EHG Explicit Health Guarantees CPI Consumer Price Index ISAPRES Instituciones de Salud Previsional, or private health insurance companies WHO World Health Organization GGP Gross Global Product GDP Gross Domestic Product SEREMIS Secretarias Regionales Ministeriales de Salud, or Regional Secretaries of the Ministry of Health SERMENA Servicio Medico Nacional de Empleados, or Employees' National Medical Services SNS Servicio Nacional de Salud, or National Health Services SNSS Sistema Nacional de Servicios de Salud, or National System of Health Services SUS Servicio Unico de Salud, or Single Health Care Service HIV/AIDS Human Immunodeficiency Virus I Acquired Immunodeficiency Syndrome Good Practice in Health Financing. p. 91 1. Background Economic context Figure 1. Economic growth 1810 - 2005 (%) Economic growth. On the whole, Chile has achieved positive economic growth rates throughout its history, with periods of recession and expansion along the way. Figure 1 shows Chile's economic growth, measured as Gross ~~~mm~~~~~~~~~mm~~~~~ ~~~~~~~~~mmmmmm~mmmoo Domestic Product (GDP), from 1810 ~~~~~~~~~~~~~~~~~~~~~ oooooooooooooOOOOO~~M through 2005. Average growth over oo~~~~~~~~g~~~~~~~~~~g ~~~~~~~~~~~~~~~~~~~~N each ten-year period between 1810 and 1979 was always positive. Growth was Source: Own elaboration based on data from the Central Bank of Chile relatively stable between 1840 and 1910. There was one decade of more rapid growth (1920-1929) and two decades of Figure 2. Growth of real GOP 1997-2005 (%) relative stagnation (1810-1819 and 1910-1919). The country experienced a 8% period of unusually rapid growth from 1985 to 1997, after a period of slow growth from 1980-1984. The economy decelerated from 1998 to 2002, and then recovered again in 2003 onward. ~ '" 2% 0: 6% 4% 0% . .···1:1. . . - - - ~I - - - . - - -- - - - - - - - - - - - - - - - Figure 2 provides a more detailed look 1997 1998 9 2000 2001 2002 2003 2004 2005 -2% at real GDP growth rates from 1997 through 2005. As shown, only one year Source: Own elaboration based on data from the Central Bank of Chile showed an economic downturn, and growth was rapid in 2004 and 2005. Figure 3. GOP per capita 2004 (international dollars) As a result of its sustained economic Peru growth, Chile's GDP in international Venezuela ~m:!!l!~B!!:!Il!!eriJ dollars has become one of the largest Colonnia ~lZm!!ill!lEl~~!l!:!' ~m::ii[3'i2il!.. in Latin America, as shown in Figure 3. Olile • _ _ _. -_ _ _ _ _ _ _. . Public finances. Chile has reported significant budget surpluses in 2.000 4.000 6.000 8.000 10.000 12.000 14.000 recent years and continues to project surpluses for the future. These hll $ surpluses are due mainly to the price Source: World Health Organization (2006). of copper, Chile's main export product, climbing to record levels. As the global economy has grown, demand for minerals and metals, especially in India and China, has increased. Furthermore, Chile's internal economy has performed well. In 2005, for example, internal demand rose 10% and investments grew by 26%, representing a record 26% of the GDP. Good Practice in Health Financing, p. 92 In 2005, projected central government revenues were US$ 28,634 million. The size of this figure is attributable in part to growing income from copper exports, but it is also due to increased tax revenues. Tax revenues, not including those related to copper exports, represent about 70% of the total central government revenues, while revenues related to copper exports represent about 10% of the total. The tax revenue structure is shown in Table 1. Table 1. Net tax revenue structure 1996-2004 (%) Type of tax 1996 1997 1998 1999 2000 2001 2002 2003 2004 Income tax 23,59 22,95 24,05 22,60 24,97 26,70 27,74 27,07 25,72 Value·added tax 47,90 48,07 47,81 48,44 48,46 47,12 48,37 50,48 51,50 Taxes on specific products 11,03 11,79 12,43 14,09 13,72 13,74 13,67 13,41 12,17 Taxes on legal documents 3,81 4,12 3,85 4,22 3,64 4,10 4,39 4,46 4,36 Import taxes 11,86 10,86 10,30 9,22 8,28 7,02 5,67 3,91 2,85 Other 1,81 2,20 1,56 1,43 0,92 1,32 0,17 0,68 3,40 Source: Own elaboration based on data form the Central Bank of Chile External debt. The fiscal surpluses noted above have gradually reduced the Figure 4. Stock of external debt 1996 - 2005 (millions of relative weight of the public debt and dollars) increased the resources available to the 50.000 Collective Capitalization Fund of the social security system. Figure 4 shows .,. 40.000 ----------------- - - - ~ 30.000 - - - - - - - - - - - the evolution of the stock of the gross <= ~ 20.000 - external debt in millions of dollars over E - - - - - - - recent years. Debt has increased from 10.000 - .. - - - - - - - year to year, although the rate of 0 increase has decelerated. According to 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 available reports, "the central Source: Central Bank of Chile government's debt has decreased from 11 % of the GDP to around 8.4% of the GDP, a record 10w."JO Most of Chile's external debt is owed to private creditors, as shown in Figure 5. 10 In 2005 the Collective Capitalization Fund amounted to US$ 1,055 million. Good Practice in Health Financing, p. 93 Macroeconomic performance. Chile enjoys striking macroeconomic stability, in terms of economic growth, external debt sustainability, internal capacity for Figure 5. Composition of the Chilean external debt 2004 (millions of dollars) savings, attractiveness to investors, and Other IBRD; high import and export levels due to a $ MJltilateral; 1% $573; Bilateral; growing commercial and financial 1% $ 321; 1% integration with the rest of the world. Table 2 shows the evolution of the major macroeconomic variables of the Chilean economy from 2000 to 2005. Private; The years 2004 and 2005 were 35.016; especially outstanding, particularly in 80% I ------~ ---- ------------- terms of gross capital as a percentage of Source: World Bank 2004 GDP and import and export levels. The increased trade is due in part to increases in the price of copper. Furthermore, the budget moved from deficit to surplus. Table 2. Chilean macroeconomic performance 2000·2005 Macroeconomic variable 2000 2001 2002 2003 2004 2005 GOP (% real annual variation) 4.5 3.4 2.2 3.9 6.2 6.3 Internal Demand (% real annual variation) 6.0 2.4 2.4 4.9 8.1 11.4 National income (% real annual variation) 4.4 2.4 2.8 3.9 8.6 9.1 Gross Capital Formation (% GOP) 21.9 22.1 21.7 22.0 21.4 23.0 National savings rate (% GOP) 20.6 20.6 20.7 20.7 23.0 23.6 Inflation (December of each year) 4.5 2.6 2.8 1.1 2.4 3.7 Unemployment rate (%) 9.2 9.1 8.9 8.5 8.8 8.0 Real salaries (% annual variation) 1.4 1.6 2.1 1.0 1.8 1.9 Exports (US$ millions) 19,210 18,272 18,180 21,664 32,215 40,574 Imports (US$ millions) 17,091 16,428 15,794 17,979 23,020 30,394 Budget deficit or surplus (US$ millions) -897 -1100 -580 -964 1586 702 Price of copper (US$ Cents /pound) 82.3 71.6 70.7 80.7 130.0 166.9 Source: Central Bank of Chile. INE Cochilco Demographic, epidemiological, and social context Demographic context. Chile is a tricontinental country, with land in America, Antarctica, and Oceania. The main body of the country stretches along the Pacific border of South America, with territories including part of Antarctica as well as Easter Island in Polynesia. Other territories include the Juan Fernandez Islands, the San Felix Islands, the San Ambrosio Islands, and the Sala y Gomez Island. Chile's surface area covers 2,006,096 square kilometers, including South American, Antarctic, and island territories, but not including its seazone (Exclusive Economic Zone and corresponding continental shelf). The country is divided into 13 regions, 51 provinces, and 342 communities. According to national census data, in 1835 the population of Chile was 1,010,336, growing to 2,695,625 in 1895; 5,023,539 in 1940; 8,884,768 in 1970; and finally 15,116,435 in 2002 (the last year for which census data is available). Until 1930, the population lived mainly in rural areas. By 1940, the majority of the population had Good Practice in Health Financing, p, 94 become concentrated in urban areas, and in 2002 the 2,026,322 people living in rural areas represented only 13.40% of the country's total population. Two features define the Chilean demographic situation, which is categorized as being in the advanced transition stage. First is the projected decrease in gross reproductive rate from 18.2 per 1000 population in 2000-2005 to 13.4 per 1000 population in 2045-2050. Second is the growth of the elderly population. In 2000,36.9% of Chileans were under the age of 19; in 2050, only 22.6% of the population will be under 19. During the twentieth century Chile experienced a 45% decrease in birth rate and 85% decrease in mortality rate. In the first half of the century, the population growth increased at a moderate pace, accelerating in the 1950s. After 1960, the birth rate quickly began to decline, finally stabilizing in the 1980s. This shift was due to changing marriage trends as well as increased individual control over family planning. This latter phenomenon occurred more rapidly than in developed countries. Between 1971 and 1992, the number of children per woman dropped from 3.80 to 2.39. The decrease was most pronounced in the rural sector, even though historically this population has tended to have higher fertility rates. Figure 6 shows the Chilean population pyramids in 1990, 2005 and a forecast for 2020. The comparison of these three periods illustrates the aging of the population structure. In general: i) the percentage of middle age people (30 to 44 years old) its relatively constant, ii) the percentage of young people (0 to 29 years old) has decreased, specially in the younger groups; and iii) the percentage of old people (more than 46 years old) has increased, specially in the older groups. Good Practice in Health Financing, p. 95 In the middle of the twentieth century, Chile was a young country, with 40% Figure 6. Structure of the Chilean population 1990, 2005 and 2020 of its population under the age of 15 _ _ _ •• _ _ _ _ 0_ _ _ _ _ -------- - - - - - and only 3% over the age of 65. By the Chile, 1990 end of the century, only 29% of the 90-94 population was younger than 15, and 75-79 7% were 65 or older, marking it as an 60.04 aging country. In half a century, the " 45-49 ~ average age of the population increased 30-34 by nearly 5 years, from 26 in 1952 to 15-19 31 in 2002.11 0-4 15 10 0 5 10 15 Other markers of the changing structure Population (%) of the Chilean population are the - - - - ---_.---- ---- indices of dependency and of aging. Chile, 2005 The dependency index is expressed in terms of unproductive population (those 90-94 under 15 and over 65 years of age) 75-79 divided by the productive population . '" 60-04 « 45-49 (those between the ages of 15 and 65). 30-34 The index of aging represents the 15-19 number people over 65 divided by the 0-4 i number of children under 15. In the 15 10 5 10 15 ! Population (%) last 50 years, the dependency index has decreased by nearly a fourth, reaching 1- --- --- ----- --- -------- Chile, 2020 50%. Meanwhile, the index of aging has increased by over half, reaching 90-94 nearly 20%. Furthermore, the latest 75-79 census data indicate that there are more . III 60.04 women than men in Chile and that the ~ 45-49 30-34 sex ration (number of males per 100 15-19 females) is greater in rural than in 0-4 urban areas. 15 10 5 0 5 10 15 Epidemiological context. The latest Population (%) available study on Chile's disease Source: National Statistical Institute of Chile burden was published in 1996 using 1993 data. At the risk of ignoring epidemiological changes that might have occurred after 1993, the authors selected this study as the most reliable reflection of Chile's epidemiological context. 12 The results of this study showed that Chile lost 1,769,557 years of healthy life (DALYs), representing a loss of 128.5 DALYs per 1000 population. The disease burden varies by age and sex. The disease burden in Chile was 1.25 times higher for men than for women. In men, most of the disease burden was attributable to premature death, while for women most was attributable to disability. In both sexes, disease burden was highest for the very young and very old, and men lost more DALY s than women at all ages. Disease burden was II The data used to describe Chilean demographics in the middle twentieth century were taken from the 1952 census, and the data to describe the end of the twentieth century were taken from the 1992 census. 12 Chilean Ministry of Health (1996). Good Practice in Health Financing, p. 96 mainly attributable to premature death among the very young and very old, and mainly to disability among those of intermediate ages. Table 3 lists the 15 diseases the produced the greatest death and disease burden in Chile in 1993, by DALYs lost, the percentage of lost DALY s attributable to that disease versus the total DALY s lost in Chile that year, and the DALY s lost attributable to that disease per 1000 population. 13 Table 3. Disease Burden in Chile 1993 (by DALYs lost) Illness DALYs % Ratio/1000 Congenital anomalies 103.654 5.86 7.53 Acute lower respiratory infections 73.234 4.14 5.32 Ischemic heart disease 67.534 3.82 4.90 Hypertensive disease 60.172 3.40 4.37 Cerebrovascular disease 57.700 3.26 4.19 Asthma 55.118 3.11 4.0 Traffic accident 53.692 3.03 3.90 Alcohol dependence 53.498 3.02 3.88 Bile duct disorders 53.361 3.02 3.87 Major depressive disorder 53.279 3.01 3.87 Arthritis and similar diseases 48.452 2.74 3.52 Alzheimer's and other dementias 42.889 2.42 3.11 Perinatal infections 41.710 2.36 3.03 Psychosis 32.474 1.84 2.36 Cirrhosis 32.172 1.82 2.34 Total 828.941 46.84 60.19 Source: Ministry of Health These rankings vary by sex and age. In men, alcohol dependence places second, followed by traffic accidents and ischemic heart disease. In women, depression places second, followed by bile duct disorders and acute lower respiratory infections. Neoplasms do not make the top fifteen due to low life expectancy. Cervical uterine cancer produces the most lost DALY s. However, among those over the age of 45, various cancers begin to rank higher on the list, including gallbladder, breast, and stomach cancer in women, and stomach, lung, and prostate cancer in men. Among the very young and the old - those under the age of five or over the age of 60 - the disease profile is similar among males and females. The largest sex differences in disease profile occur between the ages of 15 and 44, mainly due to the higher rates of injuries and alcohol dependence and abuse among men. 13 A DALY, or Disability-Adjusted Life Year, represents one lost year of healthy life, due to premature death or disability. Good Practice in Health Financing, p. 97 Health indicators. Infant mortality has gradually decreased. The most dramatic advances occurred between 1940 and 1950, when the infant mortality rate decreased by 58%. As shown in Figure 7, the rate continued to decrease from 1960 to Figure 7. Infant mortality per 1000 live births 1960·2002 2002, although at a decelerated pace. 01~~--------------------------. The Chilean infant mortality rate g 120 decreased from 234 to 8 per 1000 live births from 1930 to 2003, a 96.6% reduction. Decreased infant mortality rate tends to correlate with a number of other factors: improved living ONvW~ON~W~ONvW~ONvW~ON conditions, especially in terms of wwwww~~~~~~~~~~~~mmmoo mmmmmmmmmmmmmmmmmmmmoo ~~~~~~~~~~~~~~~~~~~~NN improved basic sanitation, availability of potable water, and sanitary waste Source: National Statistical Institute disposal systems; better delivery of public services; better education, greater availability of food; and better access to medical care. In 2000-2005, life expectancy was 74.8 years for males and 80.8 years for females. This represents a 21.9 year increase in life expectancy for males and 24.03 year increase for females since the middle of the twentieth century. The difference in life expectancy by sex held constant throughout all periods Figure 8 Life expectancy, by historical period and sex studied, as shown in Figure 8. 1950·2025 Chile enjoys one of the lowest infant mortality rates and highest life expectancies in Latin America. Figure 9 shows 2004 data for infant mortality rate per 1000 live births and life expectancy for several Latin American 60 - - - 55 50 , ,lif" ."~m >- 65 - - - . -- - , Ht*1 10 .. ----------- o 0 10 - -- - - - - . - - - - -. .. - - - - - U') o ~ ~ , - - - ~ 0 ID - - - - - . ~ ~ , - - ----~-- ~ ~ ~ ~ :g en r-... en CD ~ co g; en0) CJ) countries, listed in order of GDP per o Lb 0 J, 0 J, 2> Lh 0 co co ~ r;; ~ ~ 10 10 CJ) capita in international dollars. The two 0> 0> CJ) ~ ~ trend lines show that as economic _____ Both genders -+-- Men -...- Women development increases, infant mortality Source: National Statistical Institute of Chile decreases and life expectancy increases. Chile is the second-most-developed Figure 9. Infant mortality and life expectancy 2004 country, has the longest life expectancy, and has the second-lowest infant S M ------------------------~ ~ + 80 mortality rate among the Latin ; ~~ 1 75 American countries shown here. 'Ii 50 ~ 40 + 70 I!! 65 ~ j Social context. Chile's solid economic ~ ~~ :E 10 ~ 60 performance throughout its history and 0 ~~',!1~~~~"T~'~ ~. ~ 'g' 50 especially in recent years has allowed ~ ~ ~ ~ ~ ~8~ t <5 ~:5g ~ ~ ~ 8' 1] ~ ~ ~ <3 .~m~ ~~~~~ ~~ ~ I the country to make significant gains in ":er ca:i: GDP'" Z alleviating poverty. According to 2003 =+=Clilidiliortalityra~i¢Jt!C~__ CASEN census data, 80% of the Source: National Statistical Institute population is classified as "not poor." The number of people in Chile living on less than one dollar per day was 843,000, or about 5.6% of the total popUlation. Between 1987 and 2003, the population living below the poverty line in Chile decreased from 45.1 % to Good Practice in Health Financing, p. 98 18.8%. Poverty in Chile is much less severe than in most Latin American countries. Figure 10 shows that in 1999, only 20% of the Chilean population was poor, while in most Latin American countries over 40% of the population was poor. The average percentage of the population that was poor in this region was 43% in 1999. Figure 10. Poverty in Latin America 1999 (% population Although Chile has made important that is poor) strides against poverty, the country's Honduras Nicaragua wealth remains inequitably distributed. 6::uador Chile is considered one of the most Paraguay Bolvia inequitable countries in Latin America. Guat8l'lll1a According to 2003 CASEN census 8 Co""",, Salvador data, in the average income of the Mexico ! Latin Arrerica richest quintile was 16 times higher Brazil than that of the poorest quintile. The Argentina , Olile richest 10% of the population earned 41.2% of the country's total income, 10 20 30 40 '" 60 80 70 90 Source: Universidad Emilio Maspero. while the poorest 10% earned only 1.2%, or 34.3 times less. Political Context Brief history. During the nineteenth and twentieth centuries, Chile slowly consolidated a highly stable democratic political system. Key milestones include: • 1810: first elections held • 1822: first Constitution drafted • 1828: second Constitution drafted • 1833: third Constitution drafted • 1871: presidents prohibited from running for a second consecutive term • 1874: voted rights expanded • 1890: secret vote established • 1925: fourth Constitution drafted • 1935: women given the right to vote in municipal elections • 1949: women given the right to vote in presidential and parliamentary elections • 1972: illiterates given the right to vote From 1935 to 1888, candidates put forth by the State Party regularly won presidential and parliamentary elections. After 1888, there began to be more transparency in the electoral process. However, while government intervention in elections decreased, irregularities in the electoral process remained common, and political power continued to be concentrated in one party. After 1920, social movements began to erupt; more of the population began to participate in elections; and leftist political parties developed, including a Communist (1922) and Socialist (1933) party. The political mobilization of Good Practice in Health Financing, p. 99 the population in the 1960s and early 1970s, in a highly polarized political climate, came to an abrupt end after the 1973 coup d'Etat. The autocratic government ended in 1990, followed by a series of democratically elected administrations. The latest election was in 2006. 14 Government administration. Chile's current system of government is organized in the following way: the Executive Branch is directed by the President of the Republic, who is also the Chief of State; the Legislative Branch consists of a two-chamber congress (representatives and senators); and the Judicial Branch is led by the Supreme Court. Functions of each branch are clearly delineated among the three branches, each of which operates independently. Each region of the country is led by a Superintendent, who represents the President of the Republic in that region. The Regional Governments also include a Regional Council, which serves to approve regional development plans and allocate resources assigned for regional investment, among other duties. Within the regions, each province is led by a Governor, who is subordinate to the Superintendent, but functions in a decentralized manner. The Governor presides over an Economic and Social Provincial Council. Finally, on the community level, a Mayor leads presides over the Economic and Social Municipal Council that serves to allocate resources and make community-level decisions. Local leaders are chosen by popular election, with 4 year terms 15. 2. Overview of health care coverage and financing in Chile The defining feature of Chile's health care system is the mandatory social security system, which covers about 90% of the population. This section provides an overview of the Chilean social security system as well as a more in-depth discussion of health care financing, social security coverage, basic package coverage, and the equity and efficiency of the Chilean social security system. Chile's Mandatory health insurance (MHI) system Chile's MHI system calls for the existence of a single non-profit public insurer (known as FONASA) and of multiple for-profit or non-profit private insurers (known as ISAPREs), all operating in competition. By law all formal sector workers who are not self-employed, retired workers with a pension, and self-employed workers with a retirement fund must enroll with MHI by making a monthly contribution e~ual to 7 percent of their income or pension, up to a monthly ceiling ofUSD 2,000. 1 Other individuals may enroll as well. They include independent workers, who can voluntarily enroll with FONASA or an ISAPRE conditional on their 7 percent contribution; and legally certified indigent citizens and legally unemployed workers, who are entitled to free coverage by FONASA. ISAPRE beneficiaries may voluntarily make an extra contribution to their insurer to purchase additional coverage. Up until 2005 there was no such thing as a basic benefits package required of FONASA or ISAPREs, although the 14 This historical timeline is based on: www.memoriachilena.cl 15 See website: www.gobiemodechile.cl 16 The maximum legal monthly contribution to FONASA or an ISAPRE is therefore USD 140. Good Practice in Health Financing, p. 100 latter could not provide less financial coverage than the fonner. Starting in 2005, a set of new laws collectively known as Explicit Health Guarantees (GES) mandated the public and private MHI insurers incrementally to provide coverage for 56 legally defined health problems, starting with 25 of them in July 2005 and ending with all 56 problems in 2007. FONASA must by law purchase the majority of the health services it covers from public hospitals and health centers, although it also provides a modest subsidy to its beneficiaries willing to purchase private health care. Public health care providers must by law sell the majority of their services to FONASA, while being subject to tight limits on the kinds and volumes of services they may sell to private patients and to ISAPRE beneficiaries. Figure 11. Chile's Mandatory Health Insurance System 2006 , , , , , , , ~: ~: EI EI , , ~: ~: , , ,.51 ..at EI .§: : ____________ ~!!'!.d~'2~!,!o ,_~a_n~~t~~Z~ ______ ! ~: ~I , : .----------------h .-~----~------~ Financing Health care service") -- ------- -------- - --~ Source: Authors. The National Health Superintendence is responsible for regulating FONASA and the ISAPREs. Main regulatory functions include beneficiary protection, financial solvency ofISAPREs, and compliance by both FONASA and ISAPREs with the provision of benefits contained in the GES law. Chile's MHI system is also characterized by a split in financing and provision. In the public sector, FONASA acts solely as a financing agent, and the provision is left in the hands of the Ministry of Health; in the private sector, by law ISAPREs are not allowed to provide health services directly (i.e., they cannot be vertically integrated). This split has given beneficiaries a wider choice of provider alternatives. FONASA beneficiaries may choose to seek care from any provider, public or private, as long as they are registered with FONASA. If the provider is public (called Good Practice in Health Financing, p. 101 Modalidad de Atencion Institucional, or MAl) co-payments are small or none. If the provider is private (called Modalidad de Libre Eleccion, or MLE) co-payments are larger. ISAPRE beneficiaries also have similar choices, but almost always seek care in the private sector. Key indicators Health care spending in Chile amounts to 6% of the GDP and is financed by both public and private sector entities. 17 As shown in Table 4, public sector health spending has increased in recent years, although until 2004 public spending remained lower than private spending. Spending decreased between 1999 and 2003 and then increased again in 2004 to over the 1998 amount. Spending as a percentage of GDP dropped about 1% from 1998 to 2004. Table 4. Key health spending indicators 1998-2004 INDICATOR 1998 1999 2000 2001 2002 2003 2004 Health spending (in US$ millions) 5635.5 5182.7 4587.0 4251.3 4170.5 4475.6 5741.7 Health spending as a % of GDP 7.1 7.1 6.1 6.2 6.2 6.1 6.1 Public health spending (US$ millions) 2068.2 2021.2 2128.4 2044.9 2001.8 2184.1 2664.1 Private health spending (US$ millions) 3567.3 3161.4 2458.6 2206.4 2168.6 2291.5 3077.5 Percentage of health spending by public sector 36.7 39.0 46.4 48.1 48.0 48.8 46.4 Percentage of health spending by private sector 63.3 61.0 53.6 51.9 52.0 51.2 53.6 Out-of-pocket spending (US$ millions) 2190.3 1915.8 1158.0 1059.1 1025.8 1058.7 1412.6 Out-of-pocket spending as a % of total health spending 38.9 37.0 25.2 24.9 24.6 23.7 24.6 Spending on private health plans 1377.0 1245.6 1300.6 1147.3 1142.9 1232.8 1664.9 Spending on private health plans as a % of total health spending 24.4 24.0 28.4 27.0 27.4 27.5 29.0 Total per capita health spending (en US$) 375.0 342.0 299.0 272.0 265.0 282.0 355.0 Source: Own elaboration based on data from the World Bank 17 Some experts feel that offical data on health spending as a percentage ofGDP underestimate out-of- pocket spending on medications, and that the true figure is closer to 8% of the GDP. Good Practice in Health Financing, p. 102 Private health spending consists of out-of-pockets payments on health care and payments for health insurance premiums. In 1998, spending on the former was 59% higher than on the latter. By 2000, the trend had reversed, and spending on health insurance premiums had eclipsed out-of-pocket spending on health care. The decrease in private health spending as relative to public spending, therefore, can be explained as a decrease in out-of-pocket spending on health care. Figure 12 shows the composition Figure 12. Structure of health spending 1998·2004 (%) of spending on health care in Chile, expressed in terms percentage of 80% spending by source. The figure demonstrates how public spending and 60% private spending on health insurance 40% premiums have increased relative to 20% out-of-pocket spending on care. Private spending on health as a 0% 1998 1999 2000 2001 2002 2003 2004 percentage of total health spending is greater in Chile than in most other Latin • PUblic Expendnure • Out of Fbcket PayrTent El A"ivate A"epayrTent American countries at similar levels of Source: Own elaboration based on data from the World Bank development. Compared to Brazil, Costa Rica, and Mexico, Chile has the lowest level of public spending on health as a percentage of GDP and the highest level of private spending, as shown in Table 5. Table 5. Health spending as a % of GOP, 2004 Brazil Costa Rica Mexico Chile Public 3.6 4.9 2.8 2.6 Private 3.6 3.4 3.2 3.8 Total 7.2 8.3 6.0 6.4 Source: Pan American Health Organization In 2002, public spending on health as a percentage of total public spending as 10.2%. From 2000 to 2005, public spending on health increased relative to total public spending on social welfare programs by about 2%, reaching almost 15%. This amounted to an average annual increase of 100 million dollars. Public spending on education also increased over this period, while spending on social protection and other social welfare programs decreased, as shown in Table 6. Table 6. Composition of spending on social welfare programs 2000·2005 (%) Function 2000 2001 2002 2003 2004 2005 Health 12.8 13.3 13.4 14.0 14.5 14.9 Education 16.7 17.4 17.8 17.8 18.5 18.8 Social protection 35.5 35.5 34.7 34.5 34.1 33.6 Other social welfare programs (1) 2.2 1.9 1.9 1.8 1.5 1.6 Total spending on social welfare programs 67.2 68.2 67.8 68.1 68.6 68.9 Other 32.8 31.8 32.2 319 31.4 31.1 Source: Ministry of the Interior. Budget office. www.dipres.cl (1) Includes: Environmental protections, housing and community services, recreational activities, culture, and religion Good Practice in Health Financing, p. 103 The basic benefits package In 2005, Chile developed a benefits package with four basic guarantees: guaranteed coverage for eligible public and private sector beneficiaries; guaranteed quality of care, achieved through accreditation of participating health providers; guaranteed access to care, achieved by establishing maximum wait times and processes for enforcement; and guaranteed financing for care, achieving through a government fund to cover specified serious health conditions. 18 These promises are called the Explicit Health Guarantees (EHG), and contain the following specific provisions 19: • Primary care. The EHG follows the Integrated Family Care Model, continuing the coverage provided as part of the previous Family Health Plan. Coverage includes certain preventative and curative care provided by a family medicine team led by a physician. The EHG covers acute illnesses, preventative health care, health screenings, and special programs for mental health, cardiovascular health, specialist referrals, and home visits. • Emergency care. The EHG guarantees an emergency care network covering the entire country that provides pre-hospital care, transportation, diagnostic care, stabilization, and treatment for life-threatening or serious emergencies. • Targeted diseases. The EHG guarantees effective treatment, according to current standards of practice, for certain targeted diseases in Chile. These diseases are selected according to the seriousness of the illness, the disease burden produced, the cost of care, considerations regarding equity, and other factors. The EHG began by guaranteeing covering for 25 health problems in 2005. Coverage was expanded to 40 health problems on July 1, 2006, and as of July 1, 2007, coverage is set to expand to 56. Table 7 lists the health problems covered by the EHG since 2005. 18 Before the EHG were established, there was no National Benefits Package. FONASA did not guarantee specific rights to its beneficiaries. 19 Before the EHG were established, there was a Family Health Plan that covered certain preventative and curative care in municipal facilities. This plan became part of the GHS. Good Practice in Health Financing, p. 104 Table 7. Health problems covered under the EHG 2005·2007 2005 2006 2007 No. Health problem No. Health problem No. Health problem 1 End-stage renal disease 26 Preventative 41 Hearing loss in individuals over 65 years cholecystectomy for gallbladder cancer 2 Operable congenital cardiopathies 27 Stomach cancer 42 Leukemia in adults in children under 15 years 3 Cervical uterine cancer 28 Prostate cancer 43 Eye trauma 4 Pain relief and palliative care for 29 Refractive disorders in 44 Cystic fibrosis advanced cancer individuals over 65 years 5 Acute myocardial infarction 30 Strabismus in children under 45 Severe burns 9 years 6 Type I Diabetes Mellitus 31 Diabetic retinopathy 46 Drug and alcohol dependence in adolescents from 10 to 19 years 7 Type II Diabetes Mellitus 32 Detached retina 47 Complete prenatal and delivery care 8 Breast cancer in individuals over 15 33 Hemophilia 48 Rheumatoid arthritis years 9 Spinal defects 34 Depression in individuals 49 Mild and moderate osteoarthritis of the hip in over 15 years individuals over 60 years; mild and moderate osteoarthritis of the knee in individuals over 65 years 10 Surgical treatment for scoliosis in 35 Benign prostatic hyperplasia 50 Ruptured aneurysms; ruptured ateriovenous individuals under 25 years malformations 11 Surgical treatment for cataracts 36 Acute cerebrovascular 51 Central nervous system tumors and cysts accident 12 Total hip replacement for advanced 37 Chronic obstructive 52 Herniated disks osteoarthritis in individuals over 65 pulmonary disease years 13 Cleft palate 38 Bronchial asthma 53 Dental emergencies 14 Cancer in children under 15 years 39 Infant respiratory distress 54 Dental care for adults over 65 years syndrome 15 Schizophrenia 40 Orthotics and technical 55 Multitrauma support for individuals over 65 years 16 Testicular cancer in individuals over 56 Traumatic brain injury 15 years 17 Lymphoma in individuals over 15 years 18 Acquired immunodeficiency Syndrome (AIDS) I HIV 19 Outpatient treatment for acute respiratory infection in children under 5 years 20 Walking pneumonia in individuals over 65 years 21 Primary (essential) arterial hypertension in individuals over 15 years 22 Nonrefractory epilepsy in children 1 to 15 years 23 Complete oral health care for children under 6 years: prevention and education 24 Prematurity - Retinopathy of Prematurity - Hypoacusia of Prematurity 25 Major conduction disorders requiring a pacemaker in individuals over 15 years Source: Bitran y Asociados (2005) Good Practice in Health Financing, p. 105 Social security coverage in Chile The social security system covers a large proportion of the population, reaching 90% coverage in recent years. Over two-thirds of the population is covered by public health insurance (FONASA), while a fifth to a sixth is covered by ISAPRES. A tiny percentage is covered by private insurance through the Armed Forces or universities. Table 8 shows the Chilean social security system's health coverage in 2005, by insurer2o. Table 8. Health coverage provided by the Chilean social security system, 2005 Insurer Persons covered Percentage of population covered FONASA 11,329,481 69.65% Open ISAPRES 2,521,444 15.50% Closed ISAPRES 138,894 0.85% Uninsured 1,701,648 10.46% Others 575,771 3.54% Total population 16,267,278 100.00% Source: Superintendent of Health The public system is the main provider of health insurance in Chile. All the ISAPRES combined have never covered more than 26% of the total insured population. Average ISAPRE membership from 1984 to 2005 was 17.8% of the population. Figure 13 shows the evolution of coverage under the Chilean social security system. As Figure 13. Coverage of social security system in Chile 1984-2005 (%) seen, the public system has many more beneficiaries than the private. For a 90% .. 80% - -- - - - --~----.--- - time, the number of ISAPRE 70% beneficiaries was increasing and the 60% 50% number of FONASA beneficiaries was 40% decreasing. This trend peaked in 1995- 30% 20% 1997 then reversed itself. The number 0% ~rr 11 of beneficiaries of other systems peaked 10% jf If m ~~ ~~ ~II II II II m ~ ~ ~ ~ If if ~ 0 ~ II 1111 111111 N M • ~ ~ ~ ~ 0 ~ N M V ~ in the late 1980s, then dropped by about ~ ~ ~ ~ ~ ~- - m m m m m m m m m moo ~ ~ m m m m m m m m m m m m m m m moo ~ ~ ~ ~ ~ ~ ~ _ - _ _ _ N N 0 0 N 0 0 N 0 0 N 0 0 N half in succeeding years. • FON'\SA .ISAPRES D Others When the system was established in Source: Superintendent of Health 1981, there were only 5 participating ISAPRES. Many more ISAPRES quickly entered the market, peaking at 35 in the 1990s. Later, the number of participating ISAPRES began to drop back down. In 2006 there are 15 ISAPRES, 7 of which are closed and 8 of which are open. Figure 14 shows the number of beneficiaries (subscribers plus dependents) in the ISAPRE system in April 2006. Over two thirds of the total beneficiaries belong to three ISAPRES: Banmedica, Consalud, and ING Salud. 20 There are two types of ISAPRES, open and closed. Open ISAPRES are available to anyone working in a given sector, while closed ISAPRES are available only to workers in certain companies, including both public and private companies. Good Practice in Health Financing. p. 106 Type of care provided. The most common type of care provided to both Figure 14: Coverage of open ISAPRES 2006 FONASA and ISAPRE beneficiaries (beneficiaries and %) were diagnostic exams. Table 9 shows Nonnedica 49,594 - Sfera the services provided to FONASA Vida Tres 135,440 2% 1% beneficiaries. As seen, although Mas Vida 192,967---- 5% Banmedica 610,250 25% diagnostic exams and clinical support 8% services are the most common types of Colmena 373,453 care provided, physician visits and days 15% Consalud 563,509 hospitalized account for a larger 22% INGSalud proportion of the cost. Table 10 shows 547,653 22% the evolution of services provided by Source: Superintendent of Health (2006) ISAPRES. Most striking is the nearly 50% increase in average annual services per beneficiary from 1996 to 2004, especially the number of laboratory tests. Table 9. FONASA: Services provided and average annual spending per beneficiary 2004 TYPE OF SERVICE PROVIDED Number of services provided (%) Spending (%) ($MM) Physician visits 5.14 33.7 20.15 24.0 Diagnostic testing and clinical support services 6.65 43.6 16.79 20.0 Diagnostic and therapeutic procedures 1.38 9.0 7.97 9.5 Hemodialysis and peritoneal dialysis 0.16 1.0 5.74 6.8 Surgical interventions 0.10 0.7 10.16 12.1 Days hospitalized 0.14 0.9 1.88 2.2 Deliveries and obstetric surgeries 0.65 4.3 19.21 22.9 Other services 1.04 6.8 1.97 2.3 Total services 15.26 100 83.87 100 Source: FONASA Table 10.ISAPRES: Average number of health services provided per beneficiary 1996-2004 Type of service provided 1996 1997 1998 1999 2000 2001 2002 2003 2004 2004 (%) Physician visits 3.34 3.42 3.62 3.96 4.13 3.82 3.97 4.00 4.53 26.8 Diagnostic testing 3.74 3.84 4.01 4.68 5.35 5.25 5.39 5.54 6.48 40.9 Clinical support services 1.87 1.86 1.97 2.44 2.80 2.98 2.70 2.72 3.22 20.3 Surgical interventions 0.09 0.09 0.09 0.10 0.11 0.11 0.11 0.11 0.12 0.8 Other services 1.26 1.36 1.47 1.41 1.17 0.91 0.88 0.85 0.85 5.4 Not classified 0.96 0.69 0.57 0.66 4.2 Total services 10.30 10.56 11.16 12.58 13.56 14.03 13.74 13.78 15.86 100 Source: Superintendent of Health Table 11 shows the number of hospital admissions by type of insurance. In 2003, FONASA patients accounted for 78% of hospital admissions. Given that in 2003 FONASA covered 68% of the population, it seems that FONASA beneficiaries are more likely to be hospitalized than the rest of the population. This may be due to the fact that FONASA has a greater proportion of elderly and poor beneficiaries than do the ISAPRES. For example, according to 2003 CASEN census data, 18.1 % ofISAPRE beneficiaries were females between the ages of20 and 39, while only 5.9% were females Good Practice in Health Financing, p. 107 over the age of 65. Furthennore, ISAPRE only covers 3.2% of the poorest quintile of the population. Table 11 also shows that the number of hospital admissions for individuals with unknown insurance status dropped significantly from 2001 to 2003, likely due to improved registration systems. Finally, the table shows that the number of individuals belonging to other insurance companies dropped by nearly half between 2001 and 2003. Table 11. Number of hospital admissions 2001 2002 2003 FONASA 1,117,826 1,139,727 1,248,869 ISAPRE 245,553 261,579 252,879 Private 62,884 62,875 28,201 Other 121,343 128,926 68,599 Unknown 18,581 5,968 732 TOTAL 1,566,187 1,599,075 1,599,280 Source: Authors, based on information from the Ministry of Health Sources of funding In Chile, the entire population is guaranteed access to the public health system, whether or not they have the resources to pay premiums. By law, dependent workers must enroll in FONASA or an ISAPRE by making a contribution equal to 7% of their income to the health system, up to a ceiling of 60 UF (an inflation-indexed unit) per month. 21 These monthly contributions account for about a third of FONASA's funding. About half of FONASA's funding come from Sate funds. The ISAPRES, in contrast, are funded mainly by their beneficiaries' monthly contributions. Funding for social security health benefits in Chile, not including co-payments, comes from contributions to ISAPRES (35%), contributions to FONASA (24%), and government subsidies (41 %). Funding for health care in public facilities comes from four main sources: government subsidies, monthly contributions, operational income, and co-payments. The percentage of funding for health care in public facilities from government subsidies increased by 10% from 1990 to 1993. In the following years, this figure remained constant at about 55%, before dropping down to about 50% by 2002 and 2003. Operative income and co- payments represented a low proportion of funding throughout the period studied, and the proportion represented by co-payment decreased slightly (see Figure 15). For example, in 2003, the public sector spent US$ 2300 million on health care, of which 51 % came from government subsidies, 35% from monthly contributions, 6.4% from operative income, and 7.3% from co-payments. 21 UF, or Unidades de Fomento, are units of account linked to the Chilean Consumer Price Index. UF are used to fix a price that will not devalue over time as a result of inflation. Currently, one UF is about US$ 35. Good Practice in Health Financing, p. J08 Equity FONASA and ISAPRE membership vary by level of income. In general, Figure 15. Structure of financing for public health spending individuals with lower incomes tend to 100% subscribe to FONASA, while those with higher incomes tend to subscribe 90% 80% 70% '" '" to an ISAPRE. Within FONASA, the 60% _. services provided to its beneficiaries 50% ~. 40% also vary according to level of income. 30% -, 20% FONASA Group A contains the poorest 10% beneficiaries, while those with 0% progressively higher incomes belong to ,, 140 E 0 12:) ~ 0 100 0 . ~ CI> c. 00 00 ., i" 0:: 40 2:) 0 15-19 20-24 Source: ORC Macro, 2006. MEASURE DHS STATcompiler. http://www.measuredhs.com Good Practice in Health Financing, p. 137 Figure 1. Population pyramid, Colombia 2005 asy eo.a. 75079 70-7. 65-69 I Females eo.&t I. Males ss-se ..,... 50-5/0 100'% , • Ir d3per dErt • Farrel ! Source: 1993 (Data from Ministerio de Salud 1994; 1997-2005 authors calculations with data provided from FOSYGA The growth in enrollment is notable in the SR, from 0 to 40% of the total population between 1993 and 2005. SR insurance coverage has reached the population most vulnerable to economic shocks. For example, a comparison of the proportion of insured individuals by Good Practice in Health Financing, p. 168 income quintile in 1992 and 2003 shows an increase of37 percentage points or a variation of 444% for the first quintile (9% in 1992 to 48% in 2003), whereas in the fifth quintile the increase was 21 percentage points ( 60% in 1992 to 81 % in 2003) as seen in Graph 10. Graph 10. Insurance coverage by income quintiles, 1992-2003 90 - i!i~ ! -'~ !:" =--- ...- . -- o .---- 1992 1997 2000 2003 ______u .:;g i-+-1 2 -n--3 -~4 ~5 quintiles Source: Based on MPS calculations EH92, ECV2003 Graph 11 illustrates how rural and urban disparities in insurance coverage have been progressively bridged. In the top panel, it is seen how in 1993 there was a gap of 26% points in total insurance coverage between urban and rural areas and how this gap was decreased to 13% in 2003. For this same period, the bottom panel shows the percent population insured by type of regime in urban and rural areas. Health insurance coverage in rural areas is mainly at the expense of the SR. Good Practice in Health Financing, p. 169 Graph 11. Growth of total insurance coverage and of insurance coverage by type of regime regime, by urban and rural areas 1993-2003 (percent population enrolled) 1- + - Rural - - - UrtJan I 2s:.... .:.... -:: .i ~ 0 ~I 60 ~ 1993 .. 1997 2000 2003 50 40 i• 30 20 10 0 1993 1997 2000 2003 I_ U1Jan ~ - R.raI ~- U1Jan SR 0 RJraI SRI Sources: 1993: Ministerio de Salud (I994); 1997-2003 based on calculations by Panagiota Panopolou with ECV 97and 2003, and EH 2000 The observed coverage trends for the SR have been a response to availability of resources, explained further in the following section. 4.4.2 Changes in financing 4.4.2.1 New sources of funding and reallocation of existing resources To increase health insurance coverage the 1993 legislation introduced new sources of funding for each regime. Funds for the CR were obtained by mandating wage contributions for all those with capacity to pay, and also, by increasing the amount of the contribution as percent of salary from 11 to 12%. With the initial favorable macroeconomic and labor market assumptions these resources would provide for CR expenditures and for SR solidarity funds. To finance the SR existing national and local resources allocated to health were to be increased and directed towards demand subsidies. For example, between 1994 and 2000 the percent of total national revenues for health transferred to local governments would be incremented from 8.7 to 10.5, and supply subsidies to public providers were to be gradually Good Practice in Health Financing, p, 170 transformed into demand subsidies. New sources of funding for the SR were bundled into FOSYGA's solidarity sub-fund, which included: • The solidarity point from the CR and contribution from Cajas, which were to be matched one to one by resources from the national budget • Financial returns ofFOSYGA surpluses • Expected revenues from taxes on oil exploitation Altogether, national and local resources represented 40% of funding for the SR and solidarity resources 60%. 4.4.2.2 Results Health expenditures Between 1993 and 2003 total health expenditures as % of GDP grew from 6.2% in 1993 to 8.1 % in 2003, mainly as a result in increases in public expenditures (1.4% in 1993 to 3.1 % in 2003). Social security expenditures grew from 1.6 in 1993 to 3.7% until 1999 when they began to decrease to 3.0% in 2001. Total private expenditures remained around 3% from 1993 to 1997 and then decreased to 1.2% in 2003. Direct out of pocket expenditures (OOP) also decreased from 2.7% to 0.6% over the period, most likely as a result of health insurance. Graphll summarizes the distribution and trend of health expenditures as percentages ofGDP for 1993-2003. Good Practice in Health Financing. p. 171 Graph 12. Distribution and trend of national health expenditures as % of GDP, Colombia 1993-2003 12r-------------------------------------------------~ 10 8~------------~-- __~~__--~----------------_=--~ DID 6+a---11- 2~ __1-- o -- ------.-----------.--.:----1 l O!c.oTotal h..ealth ex. penditure./GOP l1li. II % Social security expenditure/GOP . - ----- ----"--- - - - . II %Total public expenditures/GOP ---- II % Private expenditures/GOP - - - - - "_ . - - - - - ._---, , ! Note: According to NHA methodology, total expenditure includes direct public expenditures and public expenditures on insurance (subsidized regime) , social insurance expenditures which are essentially contributory regime expenditures, and private expenditures (including both voluntary private insurance and families' out of pocket expenditures) Source: Baron ( 2006) Financial balance of the CR The financial equilibrium of the CR is reflected in the balance of the FOSYGA account, which is obtained basically by subtracting all yearly payments to EPS56 from total revenues from wage contributions. Graph 13 shows how between 1996 and 1998 there was a positive balance which then began to decrease to negative values between 2000 and 2001 and afterwards recovered (Restrepo 2006). 56 Recall payments include premium payments, maternity and sickness leave and payments to cover expenses of medications and procedures not covered by the POS. Good Practice in Health Financing. p. 172 Graph 13. Balance of FOSYGA compensation fund, 1996-2005 (Millions of Col. Pesos for the year 2005) ... 1_ .... .. S..... ... l"~~------------'-~~~--- 1+---- ·llflMl ,~ 1996 I'" 1M 191J no lJll 21m 1113 l»I .xo5 Source: Grupo de Economia de la Salud 2006. Resultados financieros del segura publico de salud en Colombia 1996-2005. Observatorio de la seguridad social. No. 13. Medellin, Universidad de Antioquia, Facultad de Ciencias Economicas The observed trend can be linked to the macroeconomic conditions of the country during this period, which had a direct impact on the behavior of key variables influencing FOSYGA revenues and expenditures. Graph 14 shows unemployment rates, GDP real growth and population in poverty between 1991 and 2005. The period began with favorable real GDP growth rates around 5% and unemployment rates less than 10%, which explain the optimistic projections of health reformers. However, between 1996 and 2001 average real GDP growth was only 1.1 %, hitting negative values in 1999, and the unemployment rate soared to 20%. By 2002 the economy began to recover, and although unemployment rates have decreased they have not reached pre-recession levels. The percent population in poverty increased from values below 50% to a peak of 57% between recession years and has also been recovering. Table 12 presents the values of key determinants of the FOSYGA balance for the period 1998-2002. It can be seen that the enrollment in the CR decreased, along with the ratio of contributors to beneficiaries, family density increased, and average salaries decreased. The behavior of these variables are related not only to the low quantity of employment during this period, but also to increased informality. All this resulted in an increase in payments to health plans that surpassed revenues from wage contributions, and in negative balances of the FOSYGA account that required use of the surplus revenue that had been collected in previous years. In subsequent years the balance recovered, but its dependency on the behavior of factors not under control of the health sector shows the balance is fragile. Good Practice in Health Financing. p. 173 Graph 14. Unemployment rates, GDP real growtha and % population in povertyb, Colombia 1991-2005 15 10 5 o ~r---------------------------~ -10"---------------------' '" '" Mr------------~--------4r----~ 52 Sources: a. Lasso 2004 (data for years 1991-2000) and Encuesta Nacional de Hogares (2000-205); b. Calculations by DNP-SISD Table 12 Selected determinants of FOSYGA balance 1998-2002 KEY DETERMINANT 1998 1999 2000 2001 2002 % total population enrolled in CR 39 33 32 30 29 Ratio of beneficiaries/contributors 1,27 1,46 1,47 1,49 1,51 Family density 2,27 2,46 2,47 2,49 2,51 Average salary (# minimum wages) 2,18 2,09 2,07 2,10 2,00 Total revenues 3.314.565 3.841.614 4.096.909 4.539.282 4.741.443 Total expenditures 2.980.795 3.845.181 4.159.935 4.497.739 4.594.324 Balance 333.770 -3.567 -63.026 41.543 147.119 Notes: Total revenues=contributions +financial returns, millions of pesos; Total expenditures=(premiums, maternity and sickness leaves, 1% solidarity and prevention fund) Source: Acosta, Ramirez, Canon 2004 It is also important to mention that another factor that has had a negative impact on FOSYGA revenues is evasion of premium contributions, in particular by independent Good Practice in Health Financing, p. 174 workers. Health plans directly collect premium contributions from this population and have no incentives to promote accurate income reporting nor mechanisms to verify income. Bitnin (2002) estimates that in 2002 only 65% of potential contributors were actually paying their obligations, and furthermore, those who did contribute paid much less than they should. The combined effect of evasion and elusion could be decreasing revenues of the CR by 35%. In the absence of this problem in 2000 resources to enroll an additional 1.5 million people to the SR would have been available. Availability of resources for the SR The total amount of funding for the SR has did not meet the projected level needed to insure all eligibles in 2000. Graph 15 compares the projected level of resources and and the actual level of expenditures in 1997 and 2000, and shows how the latter barely reached a third of total expected resources. Graph 15. SR projected resources vs actual expenditures, 1994-2000 (in millions of Col. pesos for year 2000) Projected resou-ces Actual expenditU'eS 3!DJOOO - , - - - - - - - 3!DJOOO -,------------, 300000O + - - - - - - 300000O + - - - - - - - - - - - l 2!iJOOOO + - - - - - - 2!iJOOOO + - - - - - - - - - - - l 200000O + - - - - - - 200000O + - - - - - - - - - - - l 19XJOOO +---- 19XJOOO + - - - - - - - - - - - l 1000000 + - - - - 1000000 + - - - - 9XJOOO + - - - - j 500000~~-- o+-~~~~.-~~ O+------.---'---'----.---'----'-----l 1994 1997 2000 1994 1997 2000 Solidarity D National and municipal resources • Source: Based on data in Martinez, Robayo and Valencia, 2002. The budget constraints to reach universal health insurance have several explanations. First, the unfavorable macroeconomic conditions of the country after the reform also affected the availability of key sources of funding for the SR, such as the solidarity contributions and national tax revenues. This is illustrated in Graph 14, where the solidarity point is shown to have reached only 70% of its expected level in 2000, not only because of the decrease in contributions to the CR explained above, but also because for political reasons and fiscal pressure other projected sources of funding such as matching government funds to the solidarity point and oil revenues were not allocated to the SR .. Second, the transformation of supply subsidies into demand subsidies has taken place at a slower pace than expected. Not Good Practice in Health Financing. p. 175 only has there been political opposition from public hospitals, but also, the persistence of uninsured population and the design of the POSS set the public network in the role of satisfying the demand for a large portion of second and third level care, which has been used as a justification to continue financing public hospitals with supply subsidies (Saenz 200 I; Giedion, Lopez, Marulanda 2000; Giedion, Morales, Acosta 2000). Graph 16. SR projected resources vs actual expenditures of solidarity funding sources, 2000 (in millions of Col. pesos) ~.-----------------------------------------~ ~+-----------------------. ~+----------------------- 300000 250000 200000 150000 100000 50000 o+--- 1%soIidarity Cajas Matching RrBlCiai oontribUion fu1ds retuns I[] Projected • Actual expel dhns I Source: Based on data in Martinez, Robayo and Valencia, 2002 Despite these constraints the growth of resources for the SR has been substantual. Between 1995 and 2005, resources to finance SR premiums have increased to from 0.75 to 1.1 % GDP. Growth of resources for the SR came to a halt during the economic crisis and then have followed a positive trend. The sources of financing for the SR have been changing, largely in response to the behavior of the solidarity account of FOSYGA. Also, in 2001 Law 715 was enacted, which sought to correct problems of previous decentralization policies and to give greater stability to national resources available to local governments. This law resets the amount of national fiscal resources for health and the parameters used for their distribution, basing the latter on territorial indicators of equity and efficiency. It also redefines responsibilities more in accordance to territorial capacity, although it allows decentralized municipalities to maintain competences and authority over resources and service provision as long as they meet performance indicators designed for this purpose. Table 13 shows how at the beginning of the period the main source of financing was the Good Practice in Health Financing, p. 176 solidarity contribution, and how it has gradually been replaced by resources from the national budget, which reached 56% in 2005, about 783 million USD. Table 13. Sources of funding as % of total resources for the SR, 1995-2005 Source 1995 1996 1997 1998 1999 2000 2001 2002 2003 2005 85,7 63,1 46,7 51,3 38,0 34,6 36,0 29,4 31,70 34,40 Fosy:ga % % % % % % % % % % National 14,3 36,9 44,0 37,2 50,0 51,9 51,0 67,9 transfers % % % % % % % % 62.0% 56.3% Local resources 3,8% 4,5% 3,9% 5,1% 3,0% 2,0% 4.2% 8.8% Sin taxes 1,8% 2,2% 3,4% 2,1% 9,0% Cajas 3,7% 4,8% 4,8% 6,2% 2,0% 0,6% 1.8% 0.5% Source: GES 2006 In 2005 the MPS began to explore providing central government matching funds as incentive for departmental and municipal governments to finance insurance premiums for the poor from their own resources. During this year 6 out of 32 departments, which include 250 municipalities, and a total SISBEN 1-2 population of about 2 million signed three year commitments to provide resources to reach universal health insurance for the popUlation eligible to join the SR. Most of the local governments that have opted to participate in these initiatives are the wealthiest or those for which new sources of revenues from exploitation of natural resources such as oil and coal have appeared in the last years. Administrative costs Due to the complexity of the system, the administrative costs of health plans can be large and highly variable, ranging from 4 to 60% of the value of the premium (Cendex 2000). More than 50% of the administrative costs are spent on support activities for daily operation (financial, personnel and information management) and enrollment processes; very little is spent on risk management and quality assurance (Cendex 2000). 4.4.3 Changes in the organization and administration of health care delivery 4.4.3.1 Development of insurance markets Before 1993, health insurance to employees was provided almost exclusively by public insurers, the largest being Instituto de Seguros Sociales (ISS). Private insurance offered supplemental coverage or coverage for family members of employees. The reform brought the public health insurance monopoly to an end by promoting the creation and entry of health plans (EPS and ARS) as organizations that would articulate the financing and delivery of services for the populations of the CR and SR. The CR health insurance market- developed rapidly, thanks to the fact that there was a supply of commercial health insurers, prepaid group health organizations and benefits funds, Good Practice in Health Financing, p. 177 with infrastructure, pre-established networks of providers, and specific experience in managing health care services that were ready to enter the CR as EPS. Between 1995 and 1998 the market share of private health plans increased from 11 % to 30 % of the population in the CR, a net increase of 261.1 % (Cespedes, Ramirez, and Reyes 1998). Competition by private health plans has reduced the public sector's CR market shares in terms of the percent of the CR population covered dramatically, from 100% in 1993 to 18% in 2005. Development of the health insurance market of ARS was favored by the implementation of a government program to support the creation of community based insurance organizations, or ESS, which began before Law 100 was passed. NGOs were contracted to provide technical assistance to local governments to establish ESS and central government and local municipal funds were allocated for this purpose. Between 1993 and 1994 172 ESS contracts were established, to cover 1.48 million people (Harvard School of Public Health 1996). ARSs were authorized to operate in 1995, and as of December 1999 there were 239 ARS in the market, with a total of9 million enrollees, or 22 % of the total population (Cardona 1999). Participation was not restricted to ESS and other types of health plans also entered the market, in that year the share of total enrollment in the SR of was distributed among public (10%), private for profit (30%), ESS (40%) and private not for profit (20%) firms. ARS size ranged between a couple hundred to a million enrollees. The large number of ARS did not permit adequate risk pooling, generated large inefficiencies due to large transaction costs, and made oversight very difficult. Furthermore, there were areas of the country were ARS did not operate. This motivated the government to start implementing a set of measures to consolidate the ARS market in 2001, first by requiring a minimum size of enrollees, and then by requiring regional operation of ARS as of November 2005 (MPS 2005b; 2005c). Through a public competitive process ARS were assigned to five different regions of the country to operate during a period of 4 years. Regions were defined in such a way that there could be a total number of 15 ARS, each serving a population of 200,000. according to their by potential number of enrollees, to operate during a period of 4 years, at the end of which the results of these measures in terms of efficiency, financial stability and improved access remain to be evaluated. 4.4.3.2 Integration of public and private providers and preparation of public providers to compete At the time of the reform Colombia had a well developed private supply of health services and a wide network of public providers set up during the 1970' s as part of a national strategy to increase coverage of health services. The NSHI model required the participation of both public and private providers. The reforms included provisions that required public institutions to transform their management and budgeting structure in order to improve the efficiency and quality of their services. For this purpose since 1999 the MPS since has been implementing a program to guide and support public institutions' efforts to improve productivity and sustainability57. 57 This is known as "Programa de modernizaci6n de hospitales" Good Practice in Health Financing, p. 178 So far 172 public institutions of all levels of complexity have participated in the restructuring program on a voluntary basis. Among the observed results of the program are reductions in total expenditures and personnel expenditures, increases in the number of services provided and in bed-occupancy rates. Some of these institutions were on the verge of closure and were able to recover and maintain their operation. However, many of these providers still struggle with liquidity due to problems in cash flow and have not been able to payoff previously acquired pensions debts with their employees (MPS 2005). 4.4.4 Administrative and regulatory changes In the managed competition model, the role of the government and its affiliated regulatory bodies is to provide information and formulate, monitor and enforce regulations directed to minimize market failures and to devise mechanisms to guarantee equitable access to healthcare services. For this purpose the 1993 reforms defined and assigned functions to the following governing institutions of the health sector: • Formulation and monitoring of national policies, regulations~ Ministry of Health, • Inspection and oversight of health plans and providers~Superintendencia Nacional de Salud (SNS). • Pooling and management of the system's revenues~National Health Insurance Fund denominated the Fondo de Seguridad y Garantia (FOSYGA) • Creation of the National Council of Social Security in Health (Consejo Nacional de Seguridad Social en Salud -CNSSS), a policy making body with regulatory and policy making authority over benefit packages, premium payments, copayments and measures to avoid adverse selection, tariffs, operation rules of the SR and management of the FOSYGA. These institutions have had to face the typical problems of public, bureaucratic institutions, including lack of qualified personnel and frequent rotation of upper level management, which affected the speed with which they were able to focus on priority tasks and develop the skills to carry out their assigned functions. For instance, many of the foundations ofa quality assurance system were outlined in the reform, and the regulatory basis was formulated in 2002, just until 2004 provider certification processes began to be implemented and are just beginning to be evaluated. Perhaps one of the major handicaps faced by all these institutions to be able to carry out their responsibilities has been the lack of an accurate information system. For example, it took FOSYGA almost ten years to refine the procedures to be able to effectively detect duplicate enrollees and multiple affiliations in its databases. The amount of resources that may have been wasted due to this problem could be substantial. Or as to date, there is no reliable information to calculate the costs of providing the benefits packages in order for the CNSSS to make sound decisions regarding the affordability of premium adjustments and updates to the benefits packages. Although not part of the 1993 reform, it is important to mention that in 2003, with the transformation of the former Ministry of Health into the Ministry of Social Protection, NSHI Good Practice in Health Financing, p. 179 has acquired new significance as a policy to achieve goals that are beyond the scope of the health sector, such as the alleviation of poverty. 5. Challenges and lessons for other countries The achievements of the 1993 Colombian health care reform in terms of health insurance coverage for the population have been large, placing the country as one with the most extensive social health insurance coverage in Latin America. This case study concludes by discussing major future challenges for NSHI in Colombia, derives lessons for other countries from the circumstances that can be highlighted as having contributed to the achievements of the reform and discusses replicability in other low and middle income countries. 5.1 Future challenges Closure of the remaining health insurance gap and financial sustainability ofNSHI face multiple challenges. First, with respect to the CR a major task is finding mechanisms to enroll the informal sector, which in 2005 was 58% of the working population (DNP 2006). About a third of this population is known as the "sandwich population" which can be sorted out in two groups: one includes people who earn less than 2 minimum wages, who are not eligible for subsidies, but for whom payment of the full contribution to the premium of the CR could be onerous, the other is people with capacity to contribute to the premium but who lack incentives to become affiliated. According to a household survey in 2003 total sandwich population was about 7.3 million (workers and their families) or 17% of the population (calculations by Panopolou 2004). Of these 2.5 million had some capacity to pay (Bitran, Giedion and Munoz 2004). The administrative costs to enroll, monitor and collect contributions from this population can be substantial, and the potential for adverse selection is large. Among the alternatives being explored by the MPS are providing partial subsidies to individual contributions to the premium. A pilot program will be launched in 2007 to evaluate the feasibility of this strategy. Another alternative is devising temporary income protection mechanisms, within the framework of an integrated social protection system that the MPS seeks to develop. A second challenge for the CR is reduction of tax evasion. One of the strategies Colombia has started to implement to reduce evasion is investment in improvements in the information system by establishing cross links between the pension system data and health insurance contribution data. Increased enrollment of informal workers could complement these measures. Third, in the case of the SR, not only it will be necessary to finalize the transformation of supply subsidies, but it is also likely that it will be necessary to find alternative sources of financing. For example, under current growth, employment and poverty conditions and available resources, universal health insurance would not be reached by the year 2010, as shown by Graph 17. Good Practice in Health Financing. p. 180 Graph 17. Projections for universal health insurance, 2005-2010 oo.~.~ .---------------------------------------------------, 45.~.~ 40.~.~ 35.~.~ 3O.~.~ 25.~.~ 20.~.~ 15.~.~ 10.~.~ 5.~.~ 9.843.944 10.195.313 8.180.622 6.823.269 5.015.667 --"--_. .- ---- --~ 2.136.436 2.763.723 3.209.770 3.835.849 - - ---- 16.462.532 18.183.393 .048 20.681.666 21.124.593 ...- - - - ~--- 17 15.906.388 16.288.141 16.679.057 . 17.079.354 Source: MPS calculations, 2005 As already mentioned, reaching universal health insurance by 20 lOis in the current government's agenda, but at the time of writing of this case study, stable sources of funding for this purpose have not been clearly delineated. For example, although there has been some degree of success in engaging local governments to provide own resources to co- finance demand subsidies, issues have been raised regarding the risk of local governments not meeting their obligations and the subsequent need for government to intervene with own funds. Fourth, there is much progress to be made in terms of the design and implementation of mechanisms to obtain accurate information on enrollment, population eligibility, and on the costs and effectiveness of covered services. Quality of services evaluation is also a priority. The affordability and sustainability ofNSHI could be improved by reducing the inefficiencies generated by these information gaps. Finally, there is a need to evaluate the implications in terms of equity and efficiency of maintaining two separate insurance schemes and to begin to consider feasible alternatives towards convergence. For example, due to the gaps in coverage of services in the SR this population is more likely to experience discontinuities in care that have the potential of reducing effectiveness of treatments. 58 Also, since this population has access to health services both from health plans (covered services) as well as from local governments and 58 For example, a woman with a positive PAP smear for cervical cancer, which is covered by the POSS, has to payout of pocket to have the diagnosis confirmed by a colposcopy, which is not covered Good Practice in Health Financing, p. 181 public hospitals, the monitoring of access, quality and efficiency has become especially cumbersome. 5.2 Lessons from the Colombian Reform- enabling factors 5.2.1 Prior legal framework The decentralization laws prior to the reform, plus the formulation ofthe1991 Constitution provided a legal framework for many of the strategies of the 1993 reform. Decentralization laws did the job of mobilizing national resources for health, which were then to be reallocated by Law 100 for the purposes of the reform. Decentralization conferred increased autonomy for local governments to manage resources and set the stage for public hospitals to respond to market mechanisms, conditions that were necessary for insurance contracts. The 1991 Constitution gave political legitimacy for the provision of health services under a National Social Health Insurance scheme with the participation of the private sector. 5.2.2 Institutional arrangements The existence of a reasonable number of private and public institutions that met the organizational requirements for the provision of services under the managed competition model or that were able to adapt to the fulfill these conditions, plus previous implementation of a program for the development of community based health insurers in the rural areas, allowed NSHI to quickly enroll a large number of enrollees throughout the country. The fact that many of these organizations had pre-established provider networks, summed to the existence of a public-private mix of health care providers facilitated the response of NSHI to the demand for health care services. The availability of SISBEN provided the reform with an instrument to identify its target population and provide subsidies to the neediest. 5.2.3 Financial aspects Positive economic growth conditions prior to and during the first four years of the reform favored acceptance of increases in contributions to the NSHI and mobilization of large sums of general revenues to finance the insurance expansions. The enrollment of a large fraction of the population through coverage of families of contributing individuals allowed the public sector to free additional resources for health care provision for the poor. 5.2.4 Political support Anecdotal evidence suggests that it is likely that the potential to use the SR to gain electoral support has played a role in the uptake of the SR by local governments. Expansion ofNSHI has been a predominant topic in the presidential agenda in the last two elections. There is also evidence from household surveys of increased valuation by the population of being enrolled in the SR, very likely as a result of the expected learning curve about eligibility and benefits after 13 years of implementation (Fresneda 2002). Good Practice in Health Financing, p. 182 5.3 Replicability in other countries The replicability of Colombia's coverage expansion experience in poor countries is low. The main obstacle would be that a large portion ofthe population would have to be subsidized, and it would not be feasible to mobilize large sums of national resources as these countries have small tax bases. This group of countries may have supply constraints that would need to be addressed and breadth of the SR benefits package would have to be reduced. The administrative costs of running the system may be better invested in improvements in quality of public services. Middle income countries considering NSHI should evaluate whether they meet key conditions to replicate the Colombian experience successfully, including some to avoid some of the problems that Colombia faces, which include: • Favorable economic growth rates and a reasonable size of general revenues and potential enrollee contributions to finance the desired level of benefits • Existence of quality assurance regulations that guarantee minimum safety and quality standards • The existence, or the possibility of setting up in a parallel fashion with the reform accurate and reliable information systems with population, financial, accounting and clinical and service use data. • Well developed insurance and provider markets • Institutional capacity for oversight and enforcement Good Practice in Health Financing, p. 183 REFERENCES Acosta, O. L., M. Ramirez, and C. 1. Canon. 2004. "Principales estudios sobre sostenibilidad financiera del SGSSS." Working paper, Corona Foundation [Fundacion Corona]. Bogota. Asociacion Colombiana de Facultades de Medicina-Ascofame. 2000. Recurso Humano en Medicina: formacion, distribucion y bases para una propuesta politica. Ascofame, Bogota. Baron, G. 2006. "Cuentas de salud de Colombia: Aspectos conceptuales y analisis de resultados 1993-2002." Working paper, Health System Reform Support Program of the Ministry of Social Protection and Social Development Division of the National Department of Planning [Programa de Apoyo a la Reforma de Salud del Ministerio de la Proteccion Social y Direccion de Desarrollo Social del Departamento Nacional de Planeacion,]. Bogota. Bitran and Associates, Econometria and Superior School of Public Administration [Bitran y Asociados-Econometria and Escuela Superior de Administracion Publica] 2002. Propuestas de reestructuraci6n de los procesos, estrategias y organismos publicos y privados encargados de la afiliaci6n, pago y recaudo de aportes al SGSSS. Health System Reform Support Program of the Ministry of Social Protection [Programa de Apoyo ala Reforma de Salud del Ministerio de la Proteccion Socia1.] Project 03-99. Bogota. Bitran, R., U. Giedion, and R. Munoz. 2004. Risk pooling, ahorro y prevenci6n: Estudio regional de Politicas para la protecci6n de los mas pobres de los efectos de los shocks de salud. World Bank, Washington, DC. BDO Y CCRP (2000) "Evaluacion del Sisben. Eficiencia, Eficacia Institucional de los proceso de clasificaci6n y selecci6n de beneficiarios. " Informe final. Ministerio de Salud, Bogota. Cardenas MS. (2006). Introducci6n ala Economia Colombiana. Alfaomega - Fedesarrollo. Bogota, Colombia. (in press) Cardona, A. 1999. "Participacion de las entidades promotoras de salud (EPS) en el mercado de aseguramiento de salud." Revista Facultad Nacional de Salud Publica 17 (1): 52- 62. Cendex-Ministerio de Salud. 2001. Los Recursos Humanos de Salud en Colombia: balance, competencias y prospectiva. Editorial Javegraf, Bogota. Good Practice in Health Financing, p. 184 Cespedes, J.E., M. Ramirez, M., and A. Reyes. 1998. AnaIisis de la encuesta de calidad de vida del DANE. Santafe de Bogota: Econometria S.A. Departamento Nacional de Estadistica (DANE). Encuesta de Hogares 1992, Encuesta Nacional de Calidad de Vida, 1997 and 2003. Encuesta de Hogares 2000, modulo de salud. Departamento Nacional de Estadistica (DANE). Vital statistics registry for 2002, Census results 2005. Labor statistics 1990-2005. Available at http://www.dane.gov.co, last accessed November 2006. Departamento Naciona1 de Planeacion. 2006. Bases del Plan Nacional de Desarrollo 2006-2010. Chapters 1,3 and 4. available at http://www.dnp.gov.co/paginas detalle.aspx?idp=890. Last accessed November 2006. Departamento Nacional de Planeacion y Ministerio de Salud. 2003. "Qui en se beneficia del SISBEN. Evaluacion Integral". Photocopy. Escobar ML. 2005. Health Sector Reform in Colombia. Development Outreach, May. The World Bank Institute. Available at www1.worldbank.org/devoutreachimay05/artic1e.asp?id=295 - 28k Fresneda, Oscar. 2003. "Focusing on the subsidized health regime in Colombia". Rev. salud publica. 5:(3):209-245. Fresneda, Oscar. 2002. "La vision de la poblacion sobre el sistema de sa1ud". Ha mejorado e1 acceso en salud?" In Evaluacion de los procesos del regimen subsidiado. Arevalo D, Martinez F Eds. Dniversidad Nacional de Colombia, Bogota. Profamilia. 2005. "Sa1ud Sexual y Reproductiva en Colombia. Informe de resultaados de la Encuesta Nacional de Demografia y Salud". Profamilia, Bogota. Grupo de Economia de 1a Sa1ud-GES. 2006. Resultados financieros del seguro publico de sa1ud en Colombia 1996-2005. Observatorio de 1a seguridad social. No. 13. Medellin, Dniversidad de Antioquia, Facultad de Ciencias Economicas. Giedion, D., H. R. Lopez, and 1. A. Marulanda. 2000. Desarrollo institucional del sector salud en Bogota. Coyuntura Social Vol.23, pp. 57-81. Giedion, D., L. Morales, and O. L. Acosta. 2000. Efectos de la reforma en salud sobre las conductas irregulares en los hospitales publicos. Coyuntura Social No. Vol. 23. p.98-126. Good Practice in Health Financing. p. 185 Harvard School of Public Health. 1996. Colombia Health Sector Reform Project. Final Report. Photocopied. Lasso, F. 2004. "Incidencia del gasto publico social sobre la distribucion del ingreso y la reduccion de la pobreza. " Working paper, Mission for the design of a strategy for poverty and inequality reduction, National Department ofPlannning [Mision para el disefio de una estrategia para la reduccion de la pobreza y la desigualdad, Departamento Nacional de Planeacion], Bogota. International Monetary Fund. Article IV Consultation, Poverty Reduction & Growth Facility Reports. Washington DC. Martinez F, Robayo G and Valencia 0.2002. Porque no se logra la cobertura universal de la seguridad social en salud? Bogota, FEDESALUD. Ministerio de la Proteccion Social. 2006. Informe de Actividades al CNSSS 2005-2006. Ministerio de la Proteccion Social, Bogota. Ministerio de la Proteccion Social. 2005. Politic a Nacional de Prestacion de Servicios de Salud. Ministerio de la Protecci6n Social, Bogota. Mision para el Diseiio de una Estrategia para la Reduccion de la Pobreza y la Desigualdad MERPD. 2006. "Metodologia de Medicion y Magnitud de la Pobreza en Colombia ". Departamento Nacional de Planeacion, Bogota. ORC Macro. 2006. MEASURE DHS STATcompiler. http://www.measuredhs.com. Last accessed November 2006. Pinto ME, Vergara A and Lahuerta Y. 2005. "Cuanto ha perdido Colombia por el Conflicto?" Estudios Economicos, Documento 277. Departamento Nacional de Planeacion, Bogota. Restrepo JH. 2002. "El seguro de salud en Colombia. Cobertura Universal?" Revista Gerencia y Politicas de Salud. Volumen 2: 25-40. Restrepo, J. H., M. Arango, and L. Casas. 2001. Estructura y conducta de la oferta del seguro de salud en Colombia. Observatorio de la Seguridad Social Vol 1: pp. 1-10. Good Practice in Health Financing, p. 186 Saenz, L. 2001. Modernizacion de la Gestion Hospitalaria Colombiana: Lecciones aprendidas de la transformacion de los hospitales en Empresas Sociales del Estado. Regional Initiative of Health Sector Reform in Latin America and the Caribbean, Document No 46. electronic document available at http://www.1achsr.org/documents/187- modemizaciondelagestionhospitalariacolombianaleccionesaprendidasdelatransforma c-ES.pdf, last accessed October 23, 2006. World Bank. 2006. World Development Indicators. Washington, DC: World Bank. WHO Global Burden of Disease (GBD) Estimates, 2002 Good Practice in Health Financing. p. 187 Costa Rica James Cercone Good Practice in Health Financing. p. 188 Introduction Costa Rica represents an outstanding example of a middle income country with full health care coverage, considerable financial protection, and an extensive package of services. Both in terms of breadth and depth of the coverage, the health care system provides significant protection to the population. Evidence from national surveys shows that an average Costa Rican household allocated 2.6 percent of its income to health spending in 2005 (INEC 2006). Between 1988 and 2004, the financial burden over poorer families decreased from 2.1 percent of the total family's income to 1.8 percent in 2004. Access to high quality medicines is also universal and exceeds WHO basic drug list standards. To date, the list of drugs includes 460 active principles in 608 presentations, with products for all types of pathologies. The good outcomes observed in terms of coverage (both breadth and depth) are the result of a long history of efforts to improve the living conditions of the population that date back to the 19th century. In 1941, with the creation of the Caja Costarricense del Seguro Social (CCSS), a health insurance system began operating in the country, one of the first systems in Latin America. To date, 88 percent of the population is affiliated with the health insurance scheme, and around 99 percent of the people have adequate access to primary care services. Before the 1994 health care reform, access to primary care services was restricted to 25 percent of the popUlation. The health system has achieved some other notable outcomes. For instance, between 1970 and 2005, infant mortality rates declined from 68 infant deaths per 1,000 live births in 1970 to 9.6 deaths in 2005. Life expectancy increased steadily over time, being 78 years in 2005 (82 years among women), and immunization rates exceeded 90 percent of the target population. All this was achieved due to sustained public health expenditures (5.5% of GDP). In summary, the paper presents the case of a country that, in the long run, has been capable of applying adequate legislation and institutional arrangements to achieve high rates of health care coverage and to ensure universal provision of health services. When Costa Rica faced difficult obstacles and conditions to achieve its coverage objectives, the country was able to opportunely renew its structure by redefining the network, as well as by integrating private and nongovernmental partners in the delivery of the health services. The paper is structured as follows. Section one provides a brief profile of the country in terms of economic, demographic, epidemiological, and political characteristics. Section two is devoted to the analysis of the financing and payment mechanisms of the Costa Rican health sector. Then, section three describes the delivery system including a characterization of the network of providers, an overview of the inputs of the system, and a brief discussion about regulatory issues. Section four then goes inside the health sector reform emphasizing policies aimed at improving coverage depth and breadth. Finally, in section five, key lessons for policymakers are discussed. Good Practice in Health Financing, p. 189 Section 1. Background 1.1. Introduction This section provides a brief overview of Costa Rica in three main areas: economics, demographics and epidemiology, and the political environment. The section describes the main trends observed in the last 15 years as a way to provide adequate background information for issues addressed later in the document. 1.2. Economic Profile Three key aspects characterized the Costa Rican economy between 1990 and 2005: 1) a stable macroeconomic context with persistent problems that weakened over time; 2) a stagnated poverty rate after the mid-1990s with increasing inequality after 1998; and 3) decreasing dependence on external funds and declining external debt. In general, Costa Rica has been classified as an upper middle income economy, with a GDP per capita ofUS$ 4,327 in 2004 (US$ 9,481 in PPP terms). Between 1990 and 2004, the GDP per capita grew at a 4.7 percent rate per year, but between 2000 and 2004 the growth rate decelerated to an average 3.3 annual rate. Between 1991 and 2005, the per capita real disposable income grew at 2 percent per year, but in the last six years the rate fell to just 1 percent per year. Although the unemployment rate has been increasing during the last decade, it remains at relatively low levels when compared to the regional levels that are close to 10 percent. Over the last 15 years, the share of poor people declined from 32 percent in 1990 to 21.7 percent in 2004. Despite substantial progress in this issue, the last 10 years show that poverty has practically stagnated around 20 percent of the population. Fiscal problems are a constant in the Costa Rican economy. As shown in the table below, the gap between revenues and expenditures has been permanent, with fiscal deficit averaging 3 percent of GDP. Tax revenues to GDP have remained between 12 and 13.5 percent, while expenditures to GDP have stayed in a 14 to16 range. Recurrent deficits are frequently blamed for causing inflation rates above international averages. For almost two decades, inflation rates in Costa Rica have been higher than 10 percent, currently being the second highest rate in Latin America. Moreover, external debt has shown steady signs of decline, and by 2005 the ratio of external debt to GDP was less than half the ratio in 1990, not representing a significant threat to the macroeconomic situation of the country. Similarly, Official Development Assistance decreased in the overall context, representing in 2004, just 0.07 percent of GDP. Good Practice in Health Financing. p. 190 Table 1. Costa Rica: Economic Indicators 1990-2004 Year GOP per GPO Headcount Revenue to Tax Expenditures capita PPP Growth Ratio GOP Ratio Revenue to GOP Ratio (current Rate % (excluding to GOP international grants) $) 1990 5,219.6 3.6 31.9 23.1 13.1 14.2 1995 6,630.2 3.9 20.4 20.3 12.1 16.1 2000 8,621.3 1.8 20.6 20.9 12.1 15 2004 9,481.4 4.2 21.7 22.4 13.4 15.9 Year External Oebt OOAas a Unemployment (% of total Inflation, to GOP Ratio Percent of labor force) consumer prices GOP (annual %) 1990 NA 4.20 4.5 19.0 1995 36.6 0.30 5.1 23.2 2000 19.9 0.08 5.2 11.0 2004 21.1 0.07 6.5 12.3 Source: World Development Indicators 2006. Revenue to GDP ratio comes from MIDEPLAN www.midepian.go.cr 1.3. Demographic and Health Profiles According to population pyramids for 1990,2005, and 2020 (see figure below), the Costa Rican population is getting older, and it is projected for the next 15 years that this trend will continue. The number of children between 0 to 4 years old is expected to decrease from 13.4 percent of the population in 1990 to 7.2 percent in 2020. Similarly, the 0 to 14 years old group is expected to fall from 36 percent to 22 percent of the total population. On the contrary, older groups will represent an ever increasing share ofthe population. For instance, while in 1990, women 60 years of age and older made up 7.2 percent of the female population, in 2020, women in this age bracket will make up 14.2 percent of the female population. That is, in general terms, older groups will practically double their participation between 1990 and 2020. By residence, the urban population accounts for 61 percent of the total population. Good Practice in Health Financing. p. 191 Figure 1. Costa Rica: Population Pyramids 1990, 2005, and 2020 Costa Ri:a, 2005 Costa Ri:a. 1990 90·94 90·94 75·79 75·79 60·64 60·64 0 Ol> = "" 45·49 "" 45·49 30·34 30·34 15·19 15·19 0-4 0·4 15 10 10 15 15 10 10 15 Popubtion(%) Popuhtim (%) Costa Rica, 2020 90-94 75-79 60-64 " OJ) -< 45-49 30-34 15-19 0-4 15 10 5 0 5 10 15 Population (%) Source: United Nations Population Division (2004). Parallel to changes in the demographic structure, the population growth rate (see following table) is also decreasing. During the 1990s, the overall population growth rate was 2.4 percent, but near the turn of the century, the growth rate began declining; by 2020, it is expected to be close to 1 percent per year. Good Practice in Health Financing. p. 192 Table 2. Costa Rica: Population Growth Rates, 1990-2020 Period Average Population Growth Rate % 1990-1995 2.4 1995-2000 2.4 2000-2005 2.0 2005-2010 1.5 2010-2015 1.3 2015-2020 1.1 2020-2025 1.01 Source: United Nations Population Division (2004). Furthennore, as can be viewed in the table below, it is expected that the child dependency ratio will decline from 61 to 31 percent betweeri 1990 and 2025. On the other hand, the old age dependency ratio will increase from 8 to 16 percent. Table 3. Basic Demographic Indicators Year Child Dependency Old Age Dependency Ratio Ratio 1990 61 8 1995 57 8 2000 50 9 2005 43 9 2010 38 9 2015 35 11 2020 32 13 2025 31 16 Source: United Nations Population Division (2004). Overall, life expectancy and infant mortality rates have improved over time, as is depicted in the next table. Life expectancy, for instance, increased almost two years between 1990 and 2004 to almost 79 years. Infant mortality rates have decreased over the last 15 years from 15 deaths per 1,000 live births to 9.2 deaths in 2004. While, infant mortalities have decreased, maternal mortalities have seen a significant increase in the past 15 years. In fact, the maternal mortality rate doubled during the assessed period, although improvements in maternal death recording and registration partially explained such an increase. In addition, almost 98 percent of all births were attended by skilled staff. Immunization rates showed a V-shaped pattern falling from 95 percent coverage of children aged 12 to 23 months to 85 percent in 1995, and then rising to 90 percent in 2004. Good Practice in Health Financing. p. 193 Table 4. Basic Health Information Year Life Expectancy Infant Mortality (per Maternal Mortality Immunization, at Birth 1,000 live births) (per 10,000 live OPT (% of births) children ages 12- 23 months) 1990 76.8 16 1.5 95 1995 76.7 14.3 2 85 2000 77.7 12.5 3.6 88 2004 78.7 11.3 3 90 Source: World Development Indicators 2006. Maternal Mortality was provided by MIDEPLAN. The mortality and morbidity profile of Costa Rica shows the characteristic pattern of a developed economy (see table below). In general, half of the deaths in 2002 were due to circulatory diseases and cancers; non-communicable diseases represented 76 percent of all deaths. In terms of DALYs, the top five causes accounted for 60 percent of total DALY s, led by neuropsychiatric conditions and unintentional injuries (38 percent of total DALYs). By disease category, around 70 percent of the Burden of Disease was caused by non- communicable diseases, while communicable and matemo-infant diseases represented 14.5 percent of total episodes; one out of each seven cases of illness were due to injuries. Good Practice in Health Financing, p. 194 Table 5. Leading Causes of Morbidity and Mortality in Costa Rica 2002 (% of total) Leading 5 Burden of Disease Causes Leading 5 Causes of Mortality (DALYs) Cause Percent Cause Percent Neuropsychiatric conditions 28% Cardiovascular diseases 31% Unintentional injuries 10% Malignant neoplasms 21% Infectious and parasitic 9% diseases Unintentional injuries 8% Infectious and parasitic Cardiovascular diseases 7% diseases 7% Malignant neoplasms 7% Digestive diseases 7% Source: World Health Organization. 1.4. Political Profile Costa Rica is a small Republic in Central America occupying 51,100 sq Ian of area. The country is organized around seven Provinces, 81 Cantons, and 459 Districts. The Government of Costa Rica is democratic and structured in three powers: Executive, Legislative, and Judicial Power. Moreover, the Presidential term lasts for four years, and re- election has been permitted since 2003. The country has a long tradition of political stability, and since 1889 elections have taken place normally except for two instances: the 1917 to 1919 tyranny and the 1948 to 1949 transitional period when a new Constitution emerged and the army was abolished. The last two elections (2002 and 2006) broke the hegemony of the two main political forces, and new actors emerged with a solid support from the voters. The political, administrative, and fiscal systems are centralized, and the allocation of resources is centrally done as well, usually following historical patterns. The excessive number of regulations creates a very rigid budget, leaving the Ministry of Finance (MoF) with little degree of freedom to re-allocate resources according to needs or to control the fiscal deficit. High centralization is clearly observed in the low participation of the 81 municipalities in the total budget; just 2 percent of the funds are managed at the municipal level. The National Plan of Development (NPD) elaborated by each government is the main political, programmatic, and budgetary tool. The Ministry of Planning is in charge of monitoring the adequate implementation of the National Plan through the National Evaluation System. Social policies are also part of the NPD. Besides the three powers, the institutional framework is complemented by three key entities: the General Controller Agency, the Constitutional Court (Sala IV), and the Ombudsman Office. In particular, the General Controller is the public auditing institution mainly in charge of evaluating administrative contracting processes (like purchasing). As for the Constitutional Court, it is entitled to revoke any unconstitutional decision of the Congress. Lastly, the Ombudsman supervises the quality of public services and protects citizens against any abuse of power. Good Practice in Health Financing. p. 195 Section 2. Health Financing and Coverage 2.1. Introduction This section analyzes the main financing and coverage characteristics of the Costa Rican health system. As such, the section describes the evolution of health expenditures in the last decade and the distributional features of these expenses. In addition, a brief analysis of the package of benefits and the payment mechanisms is included, along with an evaluation of the equity issues. 2.2. Health Expenditures Total health spending in Costa Rica, which is detailed in the following table, was US$ 1,276 million during 2003, 48 percent higher than in 1998. In the same year, the country allocated around 7.3 percent ofGDP to finance health services, a proportion that has been growing over the years. Private health outlays increased at a faster rate (8 percent on average) than public health spending, allowing private expenditures to jump from 25 percent to 29 percent of total health expenditures (Picado and Saenz 2000). Despite this upward trend in private health expenditures, Costa Rica has one of the lowest private participations and one of the lowest out-of-pocket expenditure coefficients (25 percent of total health spending) in Latin America. Per capita expenses also went up 33 percent between 1998 and 2003, from US$ 230 to US$ 305 (US$ 616 in current international US$); per capita public health expenditures exceeded US$ 200 in 2003. Public health expenditures accounted for more than one-fifth of total government expenditures, out of which 52 percent was allocated to finance hospital activities in 2003. Good Practice in Health Financing, p. 196 Table 6. Health Expenditures Indicators 1998, 2000 and 2003 SECTOR 1998 2000 2003 Total Health Expenditures (million US$) 859.85 1,004.62 1,276.44 Health Expenditures as % GOP % 6.1 6.3 7.3 Private Health Expenditures/GOP % 1.3 1.3 1.5 Public Health Expenditures/GOP % 4.8 5.0 5.8 Per capita Health Expenditures (current US$) 230 258 305 Public Health Expenditures as % of Total Health 77.9 79 78.8 Expenditure Public Health Expenditure as % of Total Government 21 21.7 22.8 Expenditure Private Health Expenditures as % of Total Health 25 28 29 Expenditures Participation of Hospitals in Public Health 51.1 49.6 51.6 Expenditures Source: Based on WDI, MoH-PAHO (2003); cess (2005). 2.3. Benefit Incidence Analysis In 2001, the distribution of public health spending by income category reflected a positive redistribution of resources in favor of poorer families, as is seen in the next figure. For instance, regarding the cess spending, the poorest 20 percent of the population (that received 4.7 percent of national income), received almost 30 percent of the cess health related expenditures. The wealthiest 20 percent of the families that represented 48 percent of the national income received 11.1 % of the Social Security resources. Good Practice in Health Financing, p. 197 Figure 2. National Distribution of Public Spending on Health by Income Category (2001) 60 50 § 40 '! 1 30 cf. 20 10 o Quintile I Quintile II Quintile III Quintile IV Quintile V Quintile -=- % Public Heahh Expenditures - % National Income Source: Prepared based on data from the cess and !NEe. Trejos (2002) estimated the proportion of public health expenditures associated with the different income groups, by program. On average, the poorest 20 percent of the families received 29 percent of public health in contrast to the 11 percent received by the richest 20 percent. Individually, poor families (i.e., Quintile I) tended to benefit from nutritional programs (55 percent of the correspondent budget), primary care (37 percent), outpatient services (26.5 percent), and hospitalization (26.3 percent). In no case did the poorest quintile receive less than 22 percent of the budget. 2.4. Benefit Package The package of benefits is defined and provided by the CCSS and includes: a) General medical assistance, specialized services, and surgery; b) Hospital assistance; c) Pharmacy services; d) Dentistry services; e) Cash subsidy for direct affiliates; f) Funeral cash stipend; and g) Social provisions. The package of benefits is highly inclusive and just a few interventions (like some plastic surgeries) are excluded from the list. The Law of Constitution of the CCSS establishes that the Board of Directors (BOD) is fully entitled to define the components of the package. However, recent resolutions of the Constitutional Court suggest that other entities (like the Court itself) are able to modify the package. The absence of a specific reference regarding the right to health in the Constitution has created a vacuum in the Costa Rican legislation. In those cases where conflicts arise, the Constitutional Court, at the last minute, will decide about the specific situations. The resolutions of the Constitutional Court must abide by the CCSS and create jurisdiction for future cases, which ultimately has an enormous impact on the definition of the package Good Practice in Health Financing, p. 198 of services. Resolutions have been as broad as the one establishing that the package of benefits cannot incorporate explicit restrictions to the access of health services, or as narrow as the recently issued resolution forcing the CCSS to provide a specific drug to a patient with breast cancer. In practice, the BOD has only partial control on the definition of the package of services and drugs; the Board is unable to exclude health services or drugs from the package regardless ofthe contributory condition of the person or the type of disease that affects the patient. 2.5. Financing and Payment 2.5.1. Financing Sources Revenues to finance health care and health related activities come from three sources: taxes, as in the case of the Ministry of Health; contributions [especially payroll contributions] to finance health insurance (CCSS); and fees, as in the case of the water and sewage entity (the AyA). The CCSS is the most important health institution representing 76 percent of public health expenditures. The contributory base of the CCSS is the result of three parties: workers, employers, and the State (see Table 7). The legislation also regulates revenues collected from independent workers. The BOD is entitled to determine premiums according to actuarial analyses, and under no circumstances the premiums paid by the workers shall be higher than premiums paid by employers. The mandatory health insurance financing relies on a 15 percent payroll tax for formal employees, a 10.25 percent rate for independent workers (over reported income), and a 14 percent rate for pensioners (over the pension of the person). As expected, the most important source of financing for the CCSS is contributions that represent 88 percent of total revenues (employers, workers, and State contributions together). Over time, contributions from workers and employers increased from 83 percent (1990) to 87 percent (2004), and collected revenues from workers increased faster than employers' revenues. Table 7. Payroll Fees by Insurance Scheme percent contribution by type of contributor Type ofHealth Employee Employer State Pension Total Insurance Re~ime Salaried 5.5 9.25 0.25 15.00 Independent 4.75 5.5 10.25 Voluntary 4.65 5.5 10.15 Contributory Pensioner 5.00 0.25 8.75 14.00 Non-contributory 0.25 13.75 14.00 Pensioner Insured by State 14.00 14.00 Source: cess. Good Practice in Health Financing, p. 199 Furthermore, poor households are covered by the State. The Ministry of Finance collects revenues for this purpose from taxes to luxury goods, liquors, beer, colas, and other similar imported goods. The budget allocated to cover poor household expenditures follows a complex procedure. That is, based on the Household Survey, the MoF estimates the number of expected poor people, then the average "contribution" per poor person is figured by applying a 14 percent rate to the average premium paid by the rest of contributors at the national level. The expected number of poor people times the average contribution is the budget the MoF plans to transfer to the CCSS. In practice, however, planned figures differ significantly from actual numbers. The CCSS does not have an explicit program to identify and register poor households as beneficiaries of the health services, meaning that affiliation of poor families mostly follows an as-you-go basis. Although no person can be denied health care in Costa Rica in public facilities, current regulation establishes that non-affiliated persons must pay for health care at subsidized rates. To avoid paying for care services, poor persons must register at the CCSS or at the facility when services are provided. Exemption is effective when the CCSS validates the socioeconomic conditions of the person; however, due to lack of adequate staff and resources to track all the requests, the CCSS has problems detecting fraudulent cases of non-poor people registering as state-beneficiaries. 2.5.2. Payment Mechanisms In the mid-1990s, the BOD of the CCSS introduced the Management Agreement (Compromisos de Gesti6n), a legal and managerial tool for planning and improving resource allocation. Basically, the Management Agreement is a means to regulate the relationship between the CCSS and health service providers, involving both primary care facilities and hospitals. These contract-type of agreements include key indicators for assessing quality (intra-hospital infection, mortality, and re-admission rates, for instance) and utilization (waiting lists and the average length of stay). The agreements also aim at improving coverage and quality of health services by linking resource allocation with the consecution of pre-negotiated health outcomes, and there is an implicit objective for cost containment. The Management Agreement is also a regulatory instrument for the CCSS. It defines, for instance, the health services to be supplied by providers throughout the country. Services included in the Management Agreement are Integral Care for Children under 10 years old, Integral Care for Teenagers 10 to 19 years old, Integral Care for Women 20 to 64 years old, Integral Care for Adults 20 to 64 years old, and Integral Care for the Elderly Population (over 65 years old). For each area, the Management Agreement also incorporates pre- negotiated health goals in terms of accreditation, quality, efficiency, and coverage. Moreover, hospital performance contracts were implemented in 1997 and by 2005, 29 hospitals and 104 health regions had signed one with the CCSS. Additionally, the CCSS has signed Agreements with non-profit associations, universities, and cooperatives that provide services to more than 400,000 persons, being one of the mechanisms, starting in the 1990s, utilized to increase coverage. Good Practice in Health Financing, p. 200 Finally, payment mechanisms differ for primary and tertiary care facilities. Primary care services are paid on a historical capitation basis. For hospitals, a payment unit was defined known as the Hospital Production Unit (UPH in Spanish) for the purposes of establishing rates for hospital activities, being patient days, consultations, or emergencies. The original payment scheme considered a variable component, generally 10 percent of total budget, to reward the consecution of objectives (incentives) and to cover unexpected events. The entire hospital payment system was originally expected to evolve into a results-based payment model similar to the related diagnostic groups; despite some significant progress in this issue, to date there is no full link between resource allocation and performance. In fact, the incentives fund never came into action. By 2005, a total of29 hospitals and 104 health areas had signed a Management Agreement with the cess. 2.6. Equity: Health Indicators, Outcomes, and Their Distribution 2.6.1. Health Outcomes The following table presents a list of key health indicators related to health processes and health outcomes; the information is presented in an averaged biannual basis from 1990 to 2004. In general terms, health outcomes improved in mortality related indicators and life expectancy; the balance in morbidity diseases and immunization is not fully satisfactory. Life expectancy, for instance, increased almost two years, and it is one of the highest in Latin America, while infant mortality rates fell 32 percent during the assessed period. Universal access to potable water was almost achieved and measles disappeared. On the contrary, maternal mortality increased, vaccination rates remained the same or slightly decreased, and the incidence of some diseases like dengue and HIV went up significantly. It is worth noting that the increasing rates of maternal mortality are mostly attributable to the improvement of recording systems. Good Practice in Health Financing, p. 201 Table 8. Costa Rica: Health Indicators 1990-2004 Indicator 1990-91 1995-96 2003-04 Situation 1990-2004 Gross birth rate 26.4 22.7 17.3 Decrease Infant mortality rate (per 1,000 live births) 14.3 12.5 9.7 Decrease Life expectancy at birth 76.7 76.5 78.6 Increase Men 74.7 74.3 76.4 Increase Women 78.9 78.8 80.9 Increase Maternal Mortality Rate (per 100,000 live 19.8 22 25.3 Increase births) % Children with low birth weight 6.3% 7% 7% Increase Dengue per 100,000 inhabitants 0.0 109.7 347.0 Increase Measles per 100,000 inhabitants 103.2 1.4 0.0 Decrease AIDS per 100,000 inhabitants 2.9 4.7 3.7 Increase Vaccination SRP-measles (% 1 year) 91.0 88.0 89.0 Decrease Vaccination VOP3-poliomielitis (% 1 year) 92.0 86.0 89.0 Decrease Total population served by water system ND 92% 99% Increase Source: Based on Ministry of Planning (www.mideplan.go.cr); Proyecto Estado de La Nacion (2005); OPS (www.paho.org). 2.6.2. Equity Problems: Gaps As it was mentioned before, Costa Rica achieved mixed outcomes during 1990 to 2004. Some of the major gains were in line with the objectives of the health reform. The new delivery model emphasized primary care and prevention at the community level. The significant decline in infant mortality or even the improvements observed in recording and registering maternal mortality can be explained, at least partially, by the new policies accompanying the reform. In other areas, the outcomes are not satisfactory. For example, morbidity related conditions and vaccination coverage do not yield the best outcomes, despite the fact that vaccination is part of the Management Agreements signed between primary care providers and the CCSS. In balance, however, there is the perception that the country has improved overall living conditions. One of the main achievements of the reform is that health outcomes were disseminated to practically every population group, income group, and region, though not uniformly. In some cases, poorer regions got most of the benefits but in others, the gap between the best performer and the rest of the regions or groups tended to increase. This section evaluates equity gaps in four areas: A) health insurance coverage; B) health service utilization; C) infant mortality rates and D) per capita budget allocation. Health Insurance Coverage By 2004, 88 percent of the Costa Rican population was covered by social health insurance. Differences by income quintiles are significant, as shown in the next table. For instance, while 85 percent of the Q-V population was covered by health insurance, 70 percent of the Q-I population was in the same condition. Even more important, only 10 percent of the poorest 20 percent group was a direct contributor, a number that increased to 49 percent among Q-V. Gaps by region existed, but differences tended to be small in the range of 2 to Good Practice in Health Financing, p. 202 4 percentage points. This can be considered a major strength of the health system: disparities by region have not been as wide as in other countries. Additionally, the former results clearly evidence the existence of a cross-subsidy that goes from richer groups to poorer families. As it will be described in the next paragraph, this condition is clearer given the different patterns of utilization by income group. Good Practice in Health Financing, p. 203 Table 9. Health Insurance Coverage Rates by Income Group 2005 Indicator -v National % of the population that is insured 70.7 77.6 82.5 86.3 84.7 % of the population that is direct contributors 10.6 21.3 31.1 39.5 49.2 % of the population insured as family dependents 34.4 46.1 45.2 44.0 33.5 % of the population insured by the State 19.3 6.2 3.7 l.9 0.7 % of the population insured by other plans 6.35 3.94 2.53 0.95 1.36 Urban % of the population that is insured 71.96 77.93 83.78 86.54 87.82 % of the population that is direct contributors 13.23 26.09 34.46 42.14 53.99 % of the population insured as family dependents 35 42.63 45.62 41.46 35.62 % of the population insured by the State 16.18 4.27 1.83 1.93 0.31 % of the population insured by other plans 7.55 4.94 1.87 0.9 Rural % ofthe population that is insured 68.18 75.27 81.06 85.48 82.43 % of the population that is direct contributors 7.96 16.64 23.76 33.83 39 % of the population insured as family dependents 30.2 47.37 49.22 46.12 40.59 % of the population insured by the State 25.04 8.35 4.85 4.5 1.52 4.96 2.91 3.23 1.03 1.32 Source: Authors' estimations based on the INEC 2005 Household Survey. Health Service Utilization In general, poorer groups tended to use public facilities more than richer groups (Figure 3). There was a clear tendency among higher income groups to use private services instead of public ones, despite the previously mentioned fact that richer groups presented higher rates of affiliation. Only hospital services showed similar rates of utilization between poorer and richer groups. Good Practice in Health Financing, p. 204 Figure 3. Costa Rica: Outpatient Consultations per Habitant by Income Decile: 1998 and 2001 59 1998 2001 -- -- - ------------- ---- ---- Total ~~~~ 2. j § Public j 2. :E ublic '" ..= ... 8.1. ~ (l) 0- ~ "'§ ~ 1. 0 '" I': 0 U u O. O. O. ----------- dl d2d3 d4 d5 d6 d7 Decil d8 d9 dlO O. dl d2 d3 d4 Decil d5 d6 d'7 d8 d9 diD Source: Author estimations based on the Costa Rican Household Surveys for 1998 and 2001. Infant Mortality Rates Health outcomes, as noted earlier, practically benefited the entire population regardless of the person's residence. Regionally, with the exception of the Chorotega Region, child mortality rates and the under-5 mortality rates have shown a decreasing trend in the different provinces, and the gaps between the highest and the lowest rates fell during this period (see table below). At the Canton level, differences in infant mortality rates were as wide as 5.5 times between Talamanca -17.8 deaths per 1,000 live births-, with the highest rate, and Turrubares -3.2 deaths- with the lowest; however, dispersion is falling. In total, 95 percent of the Cantons possessed infant mortality rates below 20 deaths per 1,000 live births. Table 10. Child and Infant Mortality Rates by Health Region 2001 2004 Less than Less than Less than Less than 5 years Region one lear 5,Years one ,Year Chorotega 9.5 8.2 12.8 10.7 Central 11.5 9.7 9.4 7.8 Central Pacific 12.0 9.8 8.8 8.2 Brunca 16.9 12.2 12.0 10.0 Atlantic Huetar 16.7 12.7 13.1 10.0 North Huetar 14.6 9.7 10.7 9.1 Source: Author estimations based Centro Centroamericano de Poblaci6n database. 59 Income deciles are defined using the 1998 and 2001 Household Surveys. Good Practice in Health Financing, p. 205 Per Capita Budget Allocation The last equity issue deals with the distributional problems that the current payment schemes generate. Although the reform has modified the way health services are paid and goes a step further toward a performance based system, the allocation of funds is creating enormous gaps between regions. The MIS-PARO (2004) reported that some regions receive up to 60 times more funds (in per capita terms) than others without any particular reason (i.e., no significantly different health profiles). This situation is especially true for the capitation formula used in primary care. The existing capitation scheme still allocates funds based on historical trends, not on needs. Thus, since the formula does not fully take into account differences between regions (e.g., demographic profiles, distance, higher administrative costs, etc.), then the mechanism is preserving the bad incentives generated by input based models. 2.7. Efficiency 2.7.1. Spending by Source and Use As can be viewed in the next table, the CCSS, the biggest health institution of Costa Rica, contributes two out of three colones spent in the health sector60 • The second most important source of financing is households, with almost one-fourth of the total spending. Most of the families' expenditures are allocated to drug purchasing and outpatient visits; when needed, families make use of public hospital services given the burdensome costs of private hospitalizations. Nongovernmental organizations and private insurance represent a very low share of total health expenditures (less than one percent each). Table 11. Health Spending in Costa Rica: Sources National cess Insurance Private Source MoH Households NGOs Institute Insurance (INS) Participation 5.2 65.0 4.2 24.5 0.4 0.7 Source: MoH-PAHO (2004). The composition of public health expenditures, by level of care, has two main features. The first characteristic points to the fact that the participation of hospital spending remained practically unaltered between 1997 and 2004 (see following table). Approximately half of the public health outlays have financed hospital services. During the post reform period, hospital expenditures have been growing at 5 percent per year in real terms, being the most dynamic component among the three levels of care. Arce (2002) estimates that, if this trend 60 The colon is the Costa Rican currency. Good Practice in Health Financing. p. 206 continues, in 20 years the system will be forced to allocate three times more resources to effectively cover all the expenses generated by the sector. For Saenz and Arce (2002), this behavior reflects the failure of the current hospital payment mechanism to control costs because the model incentives extended lengths of stay and the reform has not been able to strengthen the link between hospitals and primary care centers. The second relevant feature is the increasing share of primary care services in total public expenditures, a good signal that the system is moving in the right direction. However, if hospital reforms are not completed, it will be difficult to observe the full benefits of incorporating more primary care services in the health sector. Table 12. Health Spending in Costa Rica by Level of Care Level of Care 1997 2000 2004 Primary Level 18.8 22.4 21.8 Outpatient Level 30.1 28.0 26.6 Tertiary Level 51.1 49.6 51.6 Total 100.0 100.0 100.0 Source: cess (2005). 2.7.2. Sources of Financing Contributions to the social security are the most important source of financing, representing almost 60 percent of the health sector's total revenues, which is depicted in the following table. The second source, sales of goods and services, corresponds to out-of-pocket expenditures by households. If out-of-pocket expenditures and contributions to the social security are considered together, then households contribute approximately half of the total revenues. The Government contributes with 7.3 percent of total revenues (5 percent in the form of taxes and 2.7 percent in the form of contributions to the CeSS). This figure, however, does not consider the contributions of the Government to finance health insurance to poor households. Moreover, the role of external funds is minimal and do not exceed 4 percent of total health revenues. External grants and donations mainly finance medical equipment purchasing and salaries of staff working in the reform program. Good Practice in Health Financing. p. 207 Table 13. Sources of Health Financing Financing Sources Participation Taxes 5 Contributions to Social Security 59.7 Contributions to INS 4.8 Grants and Donations 3.4 Sales of Goods and Services 27.1 Source: MoH-PAHO (2004). 2.7.3. Health Spending Equity and Sustainability In Costa Rica, public health spending is the most progressive social spending component due to the fact the poorest families receive resources in a proportion greater than their weight in the population. Since health expenditures tend to favor poorer families more than non-poor families, health spending is one of the key factors that improves living conditions of the poorest and reduces income inequality (Trejos 2003). The following Lorenz Curve clearly depicts this conclusion61 . In the figure below, curves above the 45° line are said to be progressive. In these circumstances, only two sectors, health and housing, appear above the income line. Since health is above housing, it is evident that the health sector is, of all the expenses evaluated, the most progressive within the social area. 61The Lorenz Curve is a graphical representation generally used to visually present the degree of inequality in the distribution of a determined variable such as income, social spending, public subsidies, assets, etc. In the x-axis, the graph shows the cumulative share of households/families, while the y-axis presents the cumulative proportion of the assessed variable. Perfect equality is represented by a 45° line. If the final curve lies above the 45° line, then the variable presents a progressive distribution. If, on the contrary, the curve is located below the 45° line, then the variable is regressive. Good Practice in Health Financing, p. 208 Figure 4. Income and Social Expense Distribution by Function, 2000 'OOr-------------------------------------------------~~~~ 90 ao 70 eo 50 4() 30 2() Socicl Secu.-it.y 10 o~~~==~~--~--~---- o 10 __ --~--~---- __ eo --~--~ ACClUnulated % of fam:i1ies Source: Trejos (2002). The former pattern is also observed among regions. For all the regions except the Central one, the poorest 20 percent of the families receive, at least, 29 percent of the public resources invested in health (see next table). In some cases, like the Brunca and the Chorotega regions (the two poorest regions of the country), the families of Quintile I receive between 38 percent and 42 percent of the health expenditures. Figure 5. The Distribution of Public Spending on Health by Income Group, Nationally and Regionally (2001) l 1 1 1 1 1 1 1 1 1 1 National -1 I I Central I ~ I 1 - Chorotega 1 I 1 - Central Facific 1 I I 8runca Atlantic Huetar North HLetar -1 1 0% 10% 20% ::0% 40% 50% 60% I I 70% 30% 1 1 I I 90% -100% - Quintile 1_ Quintile 20 Quintle 3D Quintile 4- Quintile 5 Source: Prepared based on data from the cess and INEe. The following Lorenz Curve shows the distribution of health insurance contributions and the associated level of spending by level of income. The principles of solidarity and equity in access are clearly presented. First, the expenditure curve is very similar to the 45° line representing a high level of equity in access by all the income groups. Solidarity is observed by the differences in terms of expenditure participation and contribution participation. Good Practice in Health Financing. p. 209 Decile I, for instance, absorbs 12 percent of the health insurance expenditures, but its contributions represent just 0.6 percent of total contributions. On the other hand, Decile X families (richest 10 percent of the families) contribute with 35 percent of the contributions but account for only 8 percent of the expenditures. Figure 6. Health Expenditures and Contributions to CCSS by Income Decile 100 § ] 80 a c 70 8 '" § 60 ~ :a 50 40 ! .... 0 30 ! 20 10 0 0 10 20 30 40 50 60 70 80 90 100 Share of Families --"-Expenditures ---Contnbutions - Source: Duran (2005). A final element that deserves special attention is the financial sustainability of the health sector. At the present time, there is an important list of factors that place the financial sustainability of the sector into risk, being the most important: 1. Progressive aging of the population: people over 60 years old represent a significant financial burden for the population. In 2002, a person over 60 years old was 3.4 times more expensive than a person aged 15 to 59 years old and 4.4 times more expensive than a person aged 0 to 14 years old. 2. Low rates of affiliation among Economically Active Population (EAP): only 52 percent of total EAP (61 percent of the salaried EAP and 38 percent of the non- salaried EAP) are affiliated with the system. 3. Evasion and default revenues are still high: despite recent measures aimed at controlling these two problems, the levels of evasion and default continue to remain high. According to the cess, evasion represents 30 percent of regular revenues, while default accounts for 19 percent of regular revenues. It is estimated that 40 percent of salaried workers and 60 percent of non-salaried workers underreport their wages. 4. Government continuing default conditions: the Government usually pays a fraction of their mandatory contributions (both the ones of public employees and the Good Practice in Health Financing, p. 210 contributions to protect poor families). An increasing accumulated debt between the State and the CCSS is part of the financial limitations faced by the institution. 5. Incentives for evasion and inconsistencies between contributions and costs: differential rates in favor of independent workers are an incentive for evasion. In addition, lower contributory rates are dissociated from the real costs that these contributors represent, creating a permanent "deficit" in special groups like pensioners, independent workers, and others. 6. Poor mechanisms to verify the socioeconomic conditions ofstate-subsidized families: the CCSS does not have adequate material conditions to monitor if the statements declared by the State-subsidized beneficiaries are true. Instead, the CCSS usually approves the conditions without applying the correct administrative process. 7. Increasing hospital costs: as it was shown before, hospital costs (in real terms) are growing at a faster rate than the rest of the spending components. Duran (2005) presented the expected financial balance of the CCSS by 2050 on the basis that certain revenue and expenditure conditions occur. The main lesson to be learned, in spite of the assumptions about future expenditure behavior, is that the financial sustainability of the CCSS depends on how well the institute will be able to increase revenues. Even under an optimistic expenditure scenario (see table below), if collected revenues do not grow appropriately, then the expected deficit would be 8.1 percent of total expenditures. Table 14. Estimated Health Insurance Financial Balance in 2050 according to Different Scenarios (% of total expenditures) Revenues Pessimistic Base Optimistic Pessimistic -31.8 -6.1 2.3 Expenditures Base -21.4 8.0 17.8 Optimistic -8.1 26.6 37.8 Source: Duran (2005). 2.8. Other Issues 2.8.1. Risk-sharing and Cross-subsidization The health system in Costa Rica is a solidarity-based system. No matter the size of a person's contribution to the system, he or she will have equal access to health care services. This condition where risks, benefits, and contributions are not aligned one-to-one with the level of contributions permits the existence of cross-subsidies in favor of those persons with a lower income or higher health risks. Poor people are covered by the "Non-contributory" Good Practice in Health Financing. p. 211 and the "Insured by State" regimes; it is estimated that approximately 620,000 individuals belong to these two regimes. Cross-subsidization goes from contributors (i.e., formal employees) to the rest of the groups. According to actuarial studies of the CCSS, about 50 percent of the contributions from formal employees are used to cover health expenditures of pensioners, independent workers, and poorer households. 2.8.2. Insurance Regulation The system is not overseen by a formal health insurance regulator. In effect, regulatory tasks are undertaken by the Ministry of Health (health issues) and the General Controller Agency (budget and financial management). More specifically, the Controller has extensive power to force changes or refuse decisions, and it frequently makes use of this power, being the most powerful supervisory entity of Costa Rica. The Controller is entitled to approve or reject the institutional budget or to recommend modifications to contracts that the CCSS signs with private firms, for instance. Additionally, the Controller is the final step for approving any investment project (especially if private firms are involved), drug purchasing, and acquisition of medical equipment. Last but not least, the Controller is entitled to recommend changes in the internal processes of the CCSS, like contracting, purchasing, and planning. Although most of the resolutions of this Agency are "suggestions" to the CCSS, at the end there is an implicit incentive to obey these resolutions because final approvals are generally linked to the inclusion of such recommendations. 2.8.3. Centralization and Decentralization In 1998, the Ministry of Health launched a process of self-decentralization to improve its stewardship functions at the local level. The structure consisted of one Minister Office, one General Directorate, six Central Directorates, nine health regions, and 81 individual health stewardship areas. The CCSS also drove a process of decentralization, due to the fact excessive centralization was one of the key problems facing Costa Rica before the health care reform. Several measures were undertaken in the last decade aimed at reducing the level of centralization of the cess. For example, in 1998, the Congress approved the Act of Decentralization No. 7852 (1998) that provided enhanced autonomy to hospital directors in terms of budgetary procurement and administrative and human resource management. The Act also set up local Health Boards for enhancing community participation in local decisions. Additionally, some administrative changes were also adopted. For instance, the revenue collection process is now done at regional centers so contributors can go and pay in local CCSS headquarters. Currently, the CCSS is organized around six central managements, seven sanitary regions, and 103 health areas. Good Practice in Health Financing, p. 212 2.8.4. Cost-sharing Mechanisms Costa Rica does not possess a system of co-payments or any other cost sharing mechanism. It is well known, however, that informal co-payments are common in the system, especially in those areas where waiting lists are long or urgent interventions are required. 2.8.5. Unions in the Health Sector Medical labor unions are strong. In total, there are six labor unions that have a direct relationship with the CCSS, being the most representative the Sindicato de Profesionales en Ciencias Medicas (SIPROCIMECA) and the Union Medica Nacional. Influence of medical unions can be observed in a wide range of areas. For instance, doctors have a seat on the BOD and have a structure of monetary incentives that considerably differs from the payment scheme of the rest of public employees. 2.8.6. Freedom of Choice In Costa Rica, users have no freedom of choice of health care providers within the MIH system. The CCSS defines providers according to geographic location (residence) of the person. However, as established by Arce and Mufioz (2001), there is a "de facto, non- regulated freedom of choice", in which persons with some social capital or wealth may get health services in a public facility where they are not registered. This system resembles the characteristics of a black market. In 1992, the CCSS introduced a special program that offers its affiliates the chance to select a private provider (under the responsibility of the own user), while the institution partially covers the selected services. Services may range from surgeries to childbirths, but the final approval of the service and the economic subsidy is in the hands of the Medical Division. Interaction with Other Programs The Costa Rican health sector should be viewed as a health system in the sense that it is composed of many other institutions with health-related functions. One of these institutions is the National Insurance Institute that administers the state monopoly on private insurance. The institute participates in the health sector by protecting workers against the risks of labor accidents and traffic incidentals, for instance. The Costa Rican Institute of Water and Sewer Systems provides potable water supply services and sewage and industrial waste liquid collection and treatment services. This particular institution establishes the norms on how the rain water systems operate in the urban areas, and to finance its activities, it charges user fees. Finally, the University of Costa Rica contributes to the improvement of the health sector in three different ways: (i) by educating and training professionals and technicians in the health area, (ii) by hosting and participating in research and social action projects in the health setting; and (iii) by providing health services to the CCSS as part of Management Agreements signed between the two institutions (see Section 2.5.2). Its health professional teaching and training actions are financed through the general budget, which in tum is financed through direct transfers that it receives annually from the national budget. In addition, the CCSS is in charge of administering one of the existing Good Practice in Health Financing, p. 213 pension regimes in Costa Rica (Invalidez, Vejez y Muerte). For many years, the institution managed the health insurance system and the pension system in a single administrative body, and revenues collected from both systems went to a single collector. There was no transparency in the use of these resources, and it was not clear the level of resources that each system individually collected. The health reform separated the two systems and created the figure of a Pension Division, exclusively dealing with this issue. Presently, revenue collection is still in the hands of one entity, the Financial Division, which collects funds through the SICERE system. Section 3. Overview of Health Delivery System 3.1. Presentation This section presents a brief characterization of the network of providers involved in the Costa Rican health system. As such, this part includes a description of the availability of health related inputs, a presentation of the regulatory framework governing providers and the insurance market, the role of other components of the delivery system (like medical education and information systems), and an assessment of the most important weaknesses and strengths of the system. 3.2. Supply and Organization of Hospitals, Physicians, and Public Health Programs 3.2.1. Organization of the Health Sector and the Network of Providers Institutionally speaking, the health system of Costa Rica is organized around the Ministry of Health (the health policy governing, regulating, formulating, oversight, and regulatory body of the sector) and the Costa Rican Social Security Institute (CCSS in Spanish), in charge of administering health insurance. Health insurance, as a basis for attending to people's health, is complemented by the risk labor insurance and the traffic accident insurance, administered by the National Insurance Institute (INS, in Spanish), which is the second institution that finances, contracts, and provides health services in the country. The private sector,62 which is relatively small, has been growing in the latter years. Its services are primarily concentrated on ambulatory care and on marketing pharmaceutical products. Furthermore, system oversight and regulation have been assigned to the Ministry of Health, which is currently undergoing a reorganization to take on these tasks. According to Garcia G. (2004) the health sector in Costa Rica" ... is made up of a group of institutions and organizations that are part of the public and private sectors whose direct or indirect purpose is to contribute to improving the health of people, families, and communities, whether the institutions are from the health or other sectors." The proposed distinction is found in two executive decrees: (i) the promulgation of Executive Decree No. 14313 SPPS-PLAN entitled "health sector constitution", which formally created the health sector in Costa Rica and which regulates its structure and organization; and (ii) Executive Decree No. 19276-S of November 9, 1989, which created the National Health System and establishes the General 62 Diverse estimates exist, each with a specific methodology offering different results (e.g., Klevsen estimates that the relative weight of the private sector is 20%, while Saenz and Leon calculate it at 23%, and Duran and Herrero at 31 %). Good Practice in Health Financing, p. 214 System By-Laws, where the Ministry of Health became the governing institution. The National Health System groups together the health sector institutions and four additional types of organizations, such as municipalities, communities, institutions in charge of training health staff, and the public or private institutions that provide health services. The network of providers of Costa Rica is mainly publicly owned and organized around three levels of care. With the health sector reform that started in the mid-1990s, the MoH became the oversight and regulatory institution of the sector, while the CCSS gained control of administering the three levels of care. The total consolidation of the network in one institution was consolidated after many years of effort. Before 1977, hospitals and secondary level facilities were in hands of the MoH or charity entities, but after that year the CCSS began managing them. Then, the reform completed the process by integrating primary care centers under the umbrella of the CCSS. Currently, the network of public providers is organized in the following way: The first level of care is the gateway of the system, and it is currently organized around 103 health areas that provide primary care services in five programs of integral care to children, teenagers, women, adults, and elderly people. Primary care services are provided in health centers and clinics where 895 Basic Teams of Health Integral Care (Equipos Basicos de Atencion Integral en Salud -EBAIS-) work, delivering health services to an average 4,705 persons per EBAIS63. According to the CCSS, 99 percent of the population was covered by primary care services in 2005, making the establishment of the EBAIS one of the most successful coverage oriented policies of the health reform process. The second level of care supplies outpatient services, medical treatments, surgeries, and hospitalization in a broad range of areas including internal medicine, pediatrics, gynecology and obstetrics, and surgery in a network of 10 major clinics, 14 peripheral hospitals, and six regional hospitals. Finally, the third level is in charge of delivering complex health services and surgeries and highly specialized treatments in a network of three general hospitals and six specialized hospitals (children, women, geriatrics, psychiatrics, and rehabilitation). 3.2.2. Supply of Resources In general, the availability of health related inputs has improved in the last decade. The total number of physicians and pharmacists (per 1,000 habitants) increased more than 50 percent between 1995 and 2002, while the number of dentists went up by one third. Only nurses seemed to decline over time, but this situation is explained by three factors. First, the figure represents only nurses hired by the CCSS, an entity that has been representing a lower share of the total health workforce (though still the biggest). Second, in the same line, many nurses are working as independent nurses or for the private sector. Third, there has been a brain drain of nurses migrating to other markets (especially to the United States) where better working opportunities appear. Despite this, the MoH-PAHO (2004) estimated there 63 Each EBAIS has a basic structure of one physician, one nurse, and one primary care technician. Good Practice in Health Financing. p. 215 were 1.1 nurses per 1,000 habitants in Costa Rica, a figure that is higher than similar countries like Chile (1.0 nurses per 1,000 habitants), Argentina (0.52 nurses), and Uruguay (0.7 nurses). Table 15. Inputs ofthe Costa Rican Health System 1995-2002 (per 1,000 habitants) 1,995 2,000 2,002 Hospital Beds a/ 1.78 1.50 1.40 Physicians 0.85 1.32 1.75 Dentists 0.39 0.43 0.52 Phannacists 0.35 0.45 0.55 Microbiologists 0.28 0.29 0.31 Nurses bl 0.56 0.32 NA Total workforce cl 8.21 8.18 8.77 a/ Last value corresponds to 2003 bl Figures referred to the cess only cl Last figure corresponds to 2004 Source: Infonnation on hospital beds and physicians was taken from WDI 2006; MIDEPLAN (web page); cess (2005). Total staff working for the CCSS (still the most important health related employer of Costa Rica and the second biggest employer of the country) increased 27 percent (1995 to 2004) due to the significant growth rate observed among health professionals (3.2 percent per year). Policies oriented to increase primary care coverage are a critical factor for understanding these accelerated dynamics. As part of the consolidation of the EBAIS, the CCSS hired a significant number of general practitioners and primary care technicians during the first years of implementation of the project. Between 1995 and 1997, the CCSS created 1,180 positions to launch the first EBAIS, and in 2005 the number of employees working in EBAIS exceeded 3,000 persons. In contrast, hospital beds have been declining in the last years. Between 1995 and 2005, the number of beds (in CCSS hospitals) fell from 6,035 to 5,823 beds. This situation goes in line with the objectives of the reform of improving efficiency by limiting the capacity of general and specialized hospitals, ultimately giving primary care more predominance. 3.3. Regulatory Framework The General Act on Health entitled the MoH to regulate the health sector through the Central Directorate of Health Services, the Directorate of Protection to the Human Environment, and the Directorate of Records and Controls. The Ministry has the right to control benefits, to set up tariffs, and to define which providers can participate in the market. The MoH has organized a system of health regulation that consists of three subsystems: regulation of health services, regulation of health inputs, and regulation of the environment. Good Practice in Health Financing, p. 216 The first subsystem regulates providers by establishing a minimum set of standards to which providers must adhere to before being licensed and accredited. This regulation applies to both public and private providers and is mainly oriented to define standards in terms of infrastructure and quality of health services. The MoR is also in charge of regulating pharmaceuticals (regulation of health inputs), and for this reason the entity approved a regulation regarding the appropriate recording of medicines. Jointly with the laboratories of the University of Costa Rica and the CCSS, the Ministry controls the quality of medications. In terms of environmental regulation, serious problems were observed in the last decade regarding the pollution of water (explained by inappropriate waste management in coffee factories), soil erosion, and progressive deterioration of the forests. The MoR with the Ministry of Environment implemented some specific measures to reduce environmental problems during the last decade, and some improvements have been observed. For instance, the coffee related pollution, that used to represent 21 percent of industrial pollution, fell to 5 percent in the last years. Despite significant progress in the field, regulation is still incomplete. First, the regulation of private providers is weak, unclear, and does not protect the patient adequately. Minimum standards in terms of equipment, infrastructure, staff, quality of health services, protocols, monopolistic competition, and market control are not sufficient to effectively regulate private providers. Similarly, the Management Agreement, the key tool in the hands of the CCSS to regulate public and non-profit providers working for the institution, is a good instrument but incomplete due to the absence of a sanction in case the provider fails to comply with the negotiated terms. 3.4. Other Components 3.4.1. Medical Education Medical education has a long tradition in the country, and the School of Medicine of the University of Costa Rica is one of the oldest faculties in the nation. Traditionally, a medical degree has enjoyed a high reputation, being very attractive to high school students and to private universities 64 as well. After 1985, the number of medical schools started growing, and by 2005 the country jumped from having two faculties to having 11 faculties involved in teaching medicine or any other health related major (like physiotherapy). The number of graduates also increased at a faster pace. Between 1990 and 1995, the number of graduated doctors multiplied by 2.2 times, while the total population just increased 10 percent. The system has been unable to absorb the new doctors, especially because the CCSS (that used to hire more than 80 percent of the national doctors) is now hiring around 50 percent of the physicians. Proliferation of new schools reflects the absence of a needs based human resource national plan and the existence of a poor regulation on the issue. Accreditation of new schools (in the hands of two National Councils where the MoR has no participation) is done according to administrative criteria (e.g., fulfillment of faculty conditions, laboratories standards, etc.) 64 In Costa Rica, there are four public universities and more than 40 private universities. Good Practice in Health Financing. p. 217 and not according to real needs. This situation also creates problems in the market. The CENDEISS (the center in charge of planning human resource needs and training requirements within the CCSS) is unable to provide a space to all new doctors. The group of graduates without a post usually starts working on their own, pushing down the cost of the consultancies. 3.4.2. Pharmaceutical Supply and Distribution Drug spending in Costa Rica exceeds US$ 200 million. The CCSS being the most important single purchaser (38 percent of total drug expenditures). Both supply and distribution are highly concentrated in a few companies (oligopolistic markets), and the top seven suppliers of the CCSS account for 50 percent of the institutional drug purchases. In the private market, competition is higher, and the top five laboratories concentrate 28 percent of private drug purchases. Such high concentration has a negative impact on the cost of the medicines. In order to improve access to some key medicines, the MoH has exerted a partial control over some drugs. During 2000, the Ministry made available an 800-phone service to orient users about the price of certain drugs and to provide therapeutic information. Distribution is even more concentrated in a few companies. Although the market has more than 40 drug distributors, the top two firms account for 42 percent of drug sales, and the top four companies represent 61 percent of total sales. The College of Pharmacists reports approximately 850 pharmacies in the entire country. The CCSS has defined a list of official drugs (399 active principles classified in 54 therapeutic groups) that must be part of the inventory in all the pharmacies. In addition, every pharmacy must have a graduated pharmacist that should provide free services to the clients of the establishment. 3.4.3. Medical Technology Research and innovation practice is regulated by the MoR. The general opinion of public health experts is that the system devotes few resources to improve research and innovation. The country has a wide range of institutions in charge of different tasks in the health sector. For example, public health research is developed by the Institute of Health Research and by the Clodomiro Picado Institute (University of Costa Rica). Additionally, there are three more public entities with very different profiles: a research institute on nutrition, a national drug center, and an institute on alcoholism and pharmacodependency. 3.4.4. Information Systems Despite the existence of many Heath Information Systems (HIC)(( in the CCSS, they playa minor role in policymaking processes. The absence of strategic planning, one of the core problems faced by the CCSS previous to the reform, has been just weakly incorporated into the daily tasks of the institution. The information collected by the HIS is mostly used for following up the performance of contracts but rarely for estimating population needs. The high rates of evasion and default force the CCSS to improve collection using modem tools. In 2001, the CCSS installed the Sistema Centralizado de Recaudacion (Centralized System Good Practice in Health Financing. p. 218 of Collection, SIC ERE) aimed at easing management procedures regarding billing and contribution collection. 3.5. Strengths and Weaknesses of the System The table below summarizes the major strengths and weaknesses of the Costa Rican health system. The list mostly includes those aspects assessed in the previous sections plus some other issues that are derived from the analysis. Table 16. Major Strengths and Weaknesses of the Health Sector Strengths Weaknesses Solid institutions with more than 50 years of operations; Falling rates of contributory coverage among EAP Strong National Health System Strong political support from all sectors Weak regulation in key areas like medical education Solidarity principle works Increasing hospital costs Universalized access to primary care centers and high Problems with quality of services, especially in level of access to public hospitals hospitals Positive incidence of benefits among poorer Income Persistent equity and health outcome gaps by gender groups, whose benefits exceed their financial and between regions contributions to the system High level of financial protection, especially to poorer Lack of adequate long-run strategic planning households; low levels of out-of-pocket expenditures and no cost sharing mechanisms (formal) Extensive Package of Benefits Slow transition to a full performance based system High levels of equity between rural and urban areas Weak internal organization of the CCSS Stable levels of public financing over the years Evasion and defaulting are high Strong external control from other public institutions to No freedom of choice avoid corruptive practices Adequate number of health professionals Lack of a consolidated medical technology program Low dependence on external funds Excessive number of Medical Schools Cumulative experience that is transformed into concrete Not a patient centered system outcomes Source: Author elaboration based on previous sections Good Practice in Health Financing, p. 219 Section 4. Health Coverage Reform 4.1. Introduction This section aims at analyzing the most important policies implemented in Costa Rica during the last 30 years to increase coverage. The section argues that, between 1943 and 1990, policies in this regard were mainly oriented to increase the breadth of coverage (i.e., the affiliation of the population with health insurance). Then, the health reform (1994) shifted to target depth of coverage; that is, the increase of the share of the population with access to primary care services. 4.2. Policies Oriented to Increase Health Coverage Since the 1940's, with the creation of the CCSS, efforts were made to make social security coverage in Costa Rica universal and to maintain the health achievements already achieved. Although the health insurance scheme was initially conceived as a program for covered public employees and manufacturing workers only, during the next decades the institution expanded the breadth of coverage to include farmers, independent workers, and unemployed/poor people. In 1961, the Congress approved Act 2,738 and established the universal health insurance for workers and their families. This reform represents the breaking point in terms of health insurance coverage by including family dependents as beneficiaries of the insurance; this reform became the most important step toward universalization of the health insurance. In 1973, Act 5,349 passed, transferring all the public hospitals under control of the Ministry of Health to the CCSS. The Act defined the first bases of a social assistance program in which the Government engaged in paying for health services delivered to poor people. Although this effort was not a formal health insurance program for the poor, it at least guaranteed higher access to health services for poorer families. Later, in 1975, the CCSS expanded the health insurance to cover farmers and independent workers. Just one year later, the BOD established the obligation of all inactive pensioners to be insured under the sickness and maternity scheme. In 1978, the CCSS created and regulated the voluntary health insurance scheme for independent workers. Then in 1984, the Government of President Monge created the State-subsidized insurance plan for covering poorer families and unemployed persons. Contrary to the program established in 1973, this was a permanent insurance program with all the characteristics of a health insurance plan. The last two milestones in the long history of coverage related policies occurred in 2000 and 2006. In 2000, the Act of Protection of the Worker declared the obligation of all independent workers to affiliate with the health insurance. More recently, the CCSS reformed the health insurance existing regulation for the first time in the last 10 years. The reform has allowed affiliated workers to cover handicapped brothers or sisters not capable of working. Also, the affiliated can insure any brother or sister who is taking care of their parents under the condition that the parents are handicapped and over 65 years old. Finally, the reform expands the health insurance protection to students below 25 years old registered Good Practice in Health Financing, p. 220 in non-university tertiary level education centers (like university colleges and para- university institutes). Formerly, only students below 25 years old studying at universities were protected by the legislation. The economic crisis at the end of the 1970's and the beginning of the 1980's caused a major deterioration to the country's public finances, so that investments allocated to the health sector witnessed an abrupt stagnation that translated into the inability to broaden the sector service coverage. In addition, within the framework of reforming the Costa Rican state at the beginning of the 1990's, a health sector assessment identified the most relevant problems in which resolutions implied achieving efficiency and effectiveness objectives in the public function. The diagnosis of the health sector yielded various characteristics about the system, including: (i) The existence of a fragmented health system in terms of structure, organization, and functions (ii) Deficiencies in the regulatory framework (iii) Inefficiency and inequity in allocating the sector's resources (iv) Deficiencies in scheduling services, arising from the process being too centralized (v) Non-existence of demand self-regulation mechanisms, favored by the privilege given to tertiary care over integral individual care (vi) Lack of continuity in user care, caused by the non-existence of a reference and counter-reference framework, and by the over-specialization of medicine (vii) Dissatisfaction by the service users and providers (viii) Little training and management capacity for the sector's human resources, among other things A critical problem experienced by the sector before 1990 was the low rates of coverage with primary care services. Before 1990, around 82 percent of the population was covered by health insurance. Despite this, the traditional PHC model was already eroded and the share of total population with physical access to a PHC facility (within 30 minutes from the user residence) fell from 46 percent of the population (1990) to only 25 percent of the population (1995). In other words, the main challenge was not to increase the breadth of coverage but to increase the depth of coverage. 4.3. The Health Sector Reform: Policies and Evaluation To deal with the problems described in the previous page, the health sector reform was structured based on three large action areas: the financing model, institutional reorganization, and modernizing management of the service provider network. Good Practice in Health Financing, p. 221 Financing Model For almost 40 years, social security financing experienced several legal constraints to enforce the application of mandatory affiliation. The reform sought to reduce this problem by promoting two critical elements: (i) increase contribution coverage by designing means to ensure that compulsory contribution to the system is enforced, establishing systems for better evasion tracking, and extending contribution coverage into the independent worker sector; and (ii) redesign of the financing model, including redefining contribution rates for independent workers and pensioners, and modifying the calculation system for state contributions to the CCSS to cover the poor and the uninsured. The main results from the reform period may be summarized as follows: The contributions and real expense per insured showed a trend toward deterioration in the system's balance. As a percentage of the GDP, health insurance contributions showed a stable trend between 1990 and 2004 at around 5 percent. The differential between real income per direct insured and the real expense per direct insured has been falling. Thus, while between 1990 and 1999 the "surplus" per direct insured was 21 colones, after 2000 it declined to 10 colones per insured person, indicating that the subsidy to the non-contributors has fallen to less than half. In 2003, the real contribution was insufficient to provide a response to the real expense that was implied by the demand for health services by the insured parties. One of the reasons that can explain this spending behavior is the phenomenon of the demographic transition that Costa Rica is undergoing. According to Arias (2004), there are other factors that affect the trend toward the previously mentioned financial imbalance. Among the factors are evasion, contributions in arrears, insurance administration deficiencies, and inappropriate use of medical insurance by the insured's family members. According to calculations by Arias, for 2003, the estimated evasion in the non-contributing, independent, and retired regimens came to 143,179 million colones (US$ 359 million 65 ), which represented 51 percent of the collections from the salaried worker regimen. The changes introduced into the contributions have not influenced the lack offinancing that still affects the system. Modifications to the fees always represent a significant change, since approximately more than 90 percent of the health insurance income is made up of contributions by members. Although contributions increased beginning in 2000, other unsolved problems related to financing appeared, such as: • State contributions for financing the regimens only cover 14 percent of the effective costs • Contribution evasion is estimated at around 30 percent of the regulatory health insurance income • The high level of defaulting payments from employers and even the State (as an employer) also affects the financial structure of the CCSS. According to data for 2005, the state and employers owe a total of 50.2 billion colones (US$ 106 million) 65 Average reference exchange rate for 2003 by the BCCR: one dollar was equivalent to 398.6 colones. Good Practice in Health Financing, p. 222 due to failure to make their health insurance payments - of that amount 86 percent . to state debt. 66 pertams The cess contribution coverage has showed a decreasing trend. Both salaried and non- salaried worker coverage has decreased in the last decade, as shown in the following figure. For non-salaried workers, a change in methodology explains the sharp drop between 2000 and 2001, when the non-salaried coverage went from 75.3 percent in 2000 to 43.12 percent in 2001 67 . However, a decreasing trend can be seen in the first years of the decade that gave rise to a slight recovery in the last two years. For salaried workers, although not as pronounced, the trend is also to drop with relative stability in the last biennium. In the last two years, the CCSS was able to increase the number of workers contributing to health insurance by 114,000 more people. Thus, for example, in September 2004, 1,076,028 workers were contributing to insurance, while in 2003 the number of members was 961,266 people, implying a 12 percent increase in that period. Figure 7. Costa Rica: EAP Health Insurance Coverage: 1990-2004 90 80 ~82.51 -...75.34 70 i' 60 ~ 50 "" 40 ~14 30 .-. g; ~ ~ <"l ~ ~ ~ ~ ~ ~ !;; ~ ~ ~ ~ ;Q ~ ~ - E: .-. g .o'.£"~ ~Salaried --6-- N on salaried Source: State of the Nation. Results on Financial Protection Evidence from national surveys suggests that the system is financially protecting the population, especially the poorer groups. The recently published Survey on Household Revenues and Expenditures -SHRE- (2006) estimated that the Costa Rican households allocate 2.6 percent of their income to health spending, or 4.1 percent of current, non-capital spending. Most of the households' health expenditures go to specialist/hospital services and drugs, and such expenses affect urban households (5.1 percent of their income) more than rural families (3.6 percent). The share of health expenditures in the fifth Quintile is 66 According to data published by La Nacion newspaper (June 27, 2005), current debt could cover almost all the health needs for one year, taking into account, among other items, purchasing medicine for all members, building new facilities, and social benefits that are given to specific groups such as senior citizens. 67 The significant reduction observed between 2000 (75 percent coverage) and 2001 (43 percent coverage) is explained by a depuration process of the cess databases. The Law of Protection of the Worker forced the cess to extend, in a five-year term, compulsory health insurance to independent workers, so only a fraction of the original group remained as voluntary affiliates. Good Practice in Health Financing. p. 223 approximately 6.6 percent of their income, 3.7 times higher than the participation of health expenses in poorer households (Quintile I). In other words, most of the out-of-pocket expenditures are in the hands of wealthier families. Poorer households are not only less affected by health spending, but their burden decreased between 1998 and 2004. In 1988, the SHRE reported that poor households allocated 2.1 percent of their income to cover health expenditures. By 2004, that share declined 15 percent and represented 1.8 percent of poor household expenditures. Institutional Reorganization The primary characteristics of institutional reorganization are: The cess and MoH function specialization was strengthened. Beginning in 1996, with the separation of functions between the CCSS and the MoH, a health sector function separation process occurred in the system. The Ministry assumed the sector's tasks relating to oversight, coordinating, and regulating the system, and the CCSS took over the insurer and health service provider functions for people in the three traditional levels, including health promotion and prevention services that were previously carried out by the MoH. This process went into operation beginning in 1998. One of the more important actions was the establishment of the Sector Council as a coordination body under the supervision of the MoH to be used to manage and lead sector policies. This is despite the fact that the CCSS' autonomy provides a fair amount of discretion in relation to its own government and formulating its own polices and plans. The MoH regulatory function development has been clearly visible. The MoH institutional reorganization went into operation by establishing four strategic functions: political direction and leadership, health oversight, technological research and development, and health development regulation. 68 Some major progress in the Ministry's oversight and regulatory function included the formulation of the National Health Policy 2002 to 2006, the publication of the health sector's joint agenda for the 2002 to 2006 period, implementation of the Institutional Evaluation and Development System (SEDI in Spanish), design and implementation of the national health account system, and evaluation of integral first level care. The cess institutional reform made effective advances in relation to the separation of functions. The administrative separation of managing and financing health and pension benefits and the creation of a pension fund management area were some of the more outstanding features of the CCSS institutional reorganization. Within the new health insurance structure, three basic functions were defined: (i) health service financing, (ii) health service purchasing, and (ii) health service provision. Carrying out these new functions required certain changes, such as: (i) the creation in 1999 of the health service purchasing management area, responsible for service planning and purchasing activities; and (ii) approval in 1998 of the Law of Decentralization of the CCSS Hospitals and Clinics. As a result of this process, the presence of subsidies overlapping between the two types of 68 Ministry of Health, 2002. Good Practice in Health Financing. p. 224 insurance (health and pensions) was eliminated, and the two regimens' administrative costs were efficiently allocated. The process, however, is currently undergoing some drawbacks and various challenges still lay ahead, including: • The cess reorganization has faced resistance by interest groups within itself. • The health service purchasing management area was assigned to the cess administrative division with no clear, direct link to the financial division, so the planning and purchasing processes lack an appropriate interrelationship to make it possible to efficiently carry out financial and budget scheduling for the entity and its relationship with resource allocation and service provision. • The institutional reorganization needs to be redefined based on processes and not on structures, as was done in the last several years. A central setting is needed for both the MoH and the cess based on awareness and analysis of the national reality (Arce and Saenz 2002). • Although the MoH's oversight role has made major advances, the type of oversight, the conceptualization, and putting this function into practice still needs to be clarified. It still needs to be established whether the oversight process will be based on norms or on taking into account incentives that regulate and encourage competitiveness and coordination among the providers, as well as strengthen its actions in the environmental and technological area. • The regionalization used needs to be brought back to the table because the cess and the Ministry currently use two different regionalization systems that decrease efficiency when coordinating sector policies and unnecessarily duplicates staff, financial resources, and materials. The Law on Decentralization has been a key instrument in encouraging decentralization. Law 7852 of 1998 contributed to the regulation process in three essential aspects of the health reform process. In first place, the Law promoted administrative decentralization in the cess. In second place, by creating the Health Boards as community groups that encourage participation of citizens in local health decisions. Finally, several clauses of that Law were addressed to create specific management commitment as an instrument to encourage efficiency and effectiveness in providing health services. Service Provider Network Management Modernization In this action area, the reform intended to replace the traditional financing plan based on the historic budget with one that was more prospective, where service production and compliance with a series of quality standards were indicators that defined budget allocation. A first attempt occurred at the end of the 1980' s with the introduction of cooperatives and predefined single payments that predefined a coverage region. Later, in 1997, Management Agreements were introduced, with their characteristics already commented on in previous sections. The inclusion of Management Agreements has generated major changes, for example: Good Practice in Health Financing. p. 225 The first care level has increased its share in the total expense. Between 1997 and 2004, the proportion of resources allocated to the first care level varied from approximately 19 percent to 22 percent. Redistribution of these resources among the respective care levels, however, had a larger impact on the second care level since hospital expenses practically stayed the same during the period studied. The impact that resource allocation has had on the first care level is even clearer when the real expenses for the period are observed. For ambulatory services, the real expense increased approximately 78 percent, versus a 55 percent increase in hospital expenses. The reform began in the poorer regions. The fact that the health reform was to concentrate in the relatively less developed geographic areas has translated into better health services for these less privileged populations. In these outlying areas in the country, health posts, community doctor offices, and small clinics predominate (Rosero 2000). From the health service supply perspective, there is a larger number of health establishments (for example EBAIS) that were created during the reform in the less favored areas of the country.69 Results on Primary Care Coverage Readjusting the health care model favored an integral and social health promotion approach. The readjustment pivotal point was the creation of the health areas (103 areas in 2005) based on geographic-population criteria and criteria pertaining to accessibility, population distribution, and the country's political-administrative division. Each of the health areas was subdivided into segments that have been assigned at least one EBAIS. The number of primary care establishments that have an EBAIS team has grown constantly from 1995 to date. Thus, for example, while 232 establishments were founded in 1995, by 2004, there were 855 EBAIS, (i.e., 3.7 times more EBAIS that at the beginning of the program). Priorities were determined for each region in the program based on the population's epidemiological profile, which was a major step in defining plans based on particular needs. The new model that accompanied the health reform in 1994 strengthened the role of primary care through the EBAIS. In terms of population covered by the EBAIS, the rate of coverage jumped from 36 percent in 1995 to 69 percent in 2000, and to 86 percent in 2003, as can be viewed in the following figure. According to the Presidency of the Republic, in 2005, practically all the Costa Rican population was covered by the PRC program. 69 The proportion of the population with deficient access to health services went from 30 percent to 22 percent in the first three geographic areas (Buenos Aires, Perez Zeled6n, Golfito, Turrubares, etc.) that began the reform in 1995 to 1996 (Rosero, 2000). Good Practice in Health Financing, p. 226 Figure 8. Population Covered by the Primary Health Care Program 1990-2003 100% -: i 90% J 80% -1 .,. i !S:: 70%"'; B I 60% ~ ! 50% ..J § 400/0 1 J 30% - Year Source: cess. In addition to the reorganization of the network by health areas and the creation of the EBAIS, another strategy developed by the CCSS to increase primary care services delivery entailed the introduction of special partnership models between the cess and non-profit institutions. The first attempts to revert the problems observed in primary care coverage started in 1988 when the CCSS broke the traditional public based delivery model and began purchasing health services from health cooperatives. Since the first experiences were very successful in terms of quality and cost, the CCSS decided to expand the program during the reform period. By 2004, the CCSS had established Management Agreements with four cooperatives and one foundation (belonging to the University of Costa Rica), covering more than 10 percent of the Costa Rican population. Good Practice in Health Financing. p. 227 Box 1. Cooperatives as Health Care Providers: the Case of Costa Rica The introduction of market like mechanisms began with efforts in 1988 that introduced the first health care cooperatives. Each health care cooperative was founded by the employees of a primary health care clinic. They formed autonomous, legal entities that assumed responsibility for management of the facility. The facilities were leased from the CCSS to the cooperative for a yearly fee of roughly US$ 1, and all equipment and infrastructure was transferred to the cooperative. From this point onward, the cooperative assumed full responsibility over the maintenance of existing equipment and the purchase of new equipment (unless cooperatives could prove that damage or deterioration was due to defects in the construction or quality of materials used in the installations). The cooperatives purchased inputs, such as drugs and medical supplies, from CCSS centers, at cost plus 15 percent for administrative and shipping expenses, or, if necessary, directly from the market. The services provided were the same as those in public clinics, and the catchment population continued to be the population living within the geographic area served by the cooperative. The cooperatives received a yearly capitation fee based on the estimated number of members in the geographical area. Furthermore, no additional services were provided or fees charged by the cooperatives, as the aim of the new model was to maintain an identical package of services with a change in management responsibility. The cooperatives were, however, allowed to charge non-insured individuals seeking care on equal terms with the CCSS clinics. Several groups had strongly opposed the introduction of cooperatives because they suspected that it was effectively a privatization, which was seen as a renunciation by the state of its responsibility for the provision of free or low cost health services. In 1988 a self managing cooperative, COOPESALUD R.L., was awarded the administration of services and placed in charge of the management of the Pavas Clinic of the CCSS. Cooperative management was selected by an administrative council, which was itself elected every two years by the general assembly of all cooperative members. Legally, all workers in the cooperatives became shareholders, and periodically shareholders received earnings generated by the cooperative. The cooperatives operated under private law, without public encumbrances related to contracting, firing, and management of personnel and resources. The cooperatives enjoyed income tax exemptions and considerable political support. The cooperatives had full autonomy to manage the capitated payments they received from the CCSS, but were obligated to present annual financial statements to the CCSS for external audit purposes. All profits generated by the business were either reinvested in new equipment or infrastructure, or distributed to cooperative members. In return, the cooperatives were required to follow the guidelines and policy objectives set by the CCSS and the Ministry of Health. The contracts signed with the CCSS obliged cooperatives to provide the following services: general and specialized medicine, emergency care, minor surgery, dental care, pharmaceuticals, laboratory and radiology services, biopsies, laboratory smear tests, social work services, services related to rights and benefits verification, transportation for patients, and support services. The results of a model developed by Gauri, Cercone, and Briceno (2004) showed that cooperatives conducted an average of 9.7 to 33.8 percent more general visits, 27.9 to 56.6 percent more dental visits, and 28.9 to 100 percent fewer specialist visits. Numbers of non-medical, emergency, and first-time visits per capita were not different from the traditional public clinics. These results suggest that the cooperatives substituted generalist for specialist services and offered additional dental services, but did not tum away new patients, refuse emergency cases, or substitute nurses for doctors as care providers. Source: Gauri, Cercone, and Briceno (2004) Separating financing from provision: evidence from 10 years of partnership with health cooperatives in Costa Rica. Health Policy and Planning 19 (5): 292- 301. Good Practice in Health Financing. p. 228 4.4. Key Conditions for Success In both breadth and depth of coverage, Costa Rica is a successful country; however, this does not mean the system is free of problems. In terms of breadth of coverage, the country is close to universalization, with almost 90 percent of the popUlation affiliated to the CCSS. The health insurance has important internal problems regarding financing and administration, but over the years the CCSS has been transformed into one of the most solid institutions in Costa Rica. In terms of depth of coverage, the reform has also been quite successful. In less than 12 years, the system moved from a lowest 25 percent coverage rate (primary care services) to 99 percent in 2005. Again, the EBAIS model improved access and the role of primary care in the system, although some problems persist in terms of quality of services and immunization coverage, for instance. At the end, results of the reform suggest a net positive balance. Main factors explaining the success of the reform are: 1. Political Commitment: both political parties agreed on the need for change. Even though they did not agree on certain details, the common view that a health reform was required prevailed. The relatively rapid approval of the reform loans in the Congress was a signal of such political support. 2. Origin of the Reform: contrary to what happen in many countries, the idea ofa health reform emerged from inside the health sector, not from outside. This special feature has allowed the process to flow smoothly (not without difficulties). In addition, this situation also has facilitated the process to avoid confrontations between key institutions like the CCSS and the MoH. 3. Availability of Funds: political support was transformed into economic support. The reform program was financed with external loans and local contributions. The first loan amounted to US$ 22 million from the W orId Bank, and it was part of a US$ 32 million project aimed at financing the core component of the reform agenda. A second loan was approved by the Inter-American Development Bank for US$ 42 million. The total cost of the second project amount to US$ 60 million. Most of the funds were allocated to the modernization of the primary care network. In total, US$ 17 million from Project I and US$ 19 million from Project 2 were assigned to the primary care transformation. 4. Wide participation of main stakeholders: participation in the reform process was not limited to political parties but included CCSS workers (especially medical staff), MoH staff, medical doctors, and communities. Transparency of the information flowing to the different stakeholders was a key aspect. Several groups opposed the reforms, as expected, but continuous negotiation and a consensus-building orientation allowed the implementation of the main reform components. It is important to recognize that, at the end, the original reform agenda has still not been fully implemented, and there are some missing elements. For instance, allowing patients the right to choose their provider is an issue not considered yet. Good Practice in Health Financing. p. 229 5. Rural and poorer areas, first: one of the distinctive characteristics of the health reform regarding health coverage is that the first EBAIS were launched in rural areas, particularly in poor regions. Practically all the first 229 EBAIS were established in poor, rural areas, following a social equity principle. This has allowed the system to gain quick benefits and to create a solid basis for further changes. 6. Strong technical team: both the group of professionals that design and implemented the reform and the "political team" (the BOD and the Executive President of the CCSS) were all highly qualified persons with a solid background in similar tasks, in the understanding of the national problems and in the negotiation of key changes. 7. Clarity of the objectives of the reform: the CCSS was clear in what objectives it wanted to pursue. Although there were many problems at the time of the assessment, technical and political efforts needed to be concentrated in just some of them. Putting so many objectives and "components" in the reform agenda may dilute the strength of the reform. 8. Transformation of the CCSS for a sustained reform process: changes in the model needed some internal reorganization in the way the institution functions. Some measures were implemented, like the creation of a Purchasing Unit and the establishment of the CENDEISS, the department in charge of training CCSS staff. 9. Continuous monitoring and evaluation: the reform was constantly monitored to track if the different measures were adopted and to observe the outcomes of the decisions. 10. Focus on access to primary care to achieve rapid gains: the health sector reform greatly benefited from expanding the PRC services rather than concentrating on hospital care. The new strategy moved to a PRC centered delivery model that increased coverage in rural areas and improved physical access to basic health services throughout the country, yielding concrete benefits during the first two years of implementation. In almost a decade, the new PRC model practically achieved full coverage. As expected, the outcomes of the new strategy were the result of multiple concurrent factors: institutional transformations, community participation in decision making, increasing budgets, enhanced participation of profit and non-profit sectors, adequate planning, and incorporation of quality standards. Good Practice in Health Financing. p. 230 Sections 5. Lessons for Policymakers 5.1. Introduction The last section analyzes the enabling factors (financial, institutional, others) that create adequate conditions for success in the field of health coverage. The section evaluates some elements of political economy that were presented in the Costa Rican situation and a set of lessons for policymakers in low income and middle income countries. 5.2. Enabling Technical and Political Factors From the Costa Rican experience previously analyzed, the key enabling factors for the success of a coverage oriented reform are: • National Commitment: in order to have a successful health reform it is essential that the different political actors involved compromise in implementing, ideally, a consensuated reform. In Costa Rica, the reform was design by one government but it was implemented by the opposition in the next presidential term. Achieving commitment implies a constant dialogue, negotiation, and consensus building orientation with the main stakeholders. • Technically sound strategies: the design of a health reform is a complex exercise that requires an adequate diagnosis of the sector, an economic and political strategy of implementation, and the incorporation of those conditions that are country-specific. When designing a reform, it is important to concentrate on some key aspects where most of the efforts will be conducted, rather than trying to cover all the existing problems of the country. Also, it is very relevant to discuss the appropriate phases of the reform. • Resources: sustainability of the reform depends on the financial support to implement the required measures. This implies two things. First, it is important to have a "short-term" budget to launch the program, such as the two loans mentioned before. Second, it is critical to commit to investing resources over the "long-run". Costa Rica is an example that, if sustained over time, health investments pay-off. Since 1950, the country has engaged in a compromise of allocating an ever increasing amount of money into the health sector. As a result, life expectancy increased from 45 years old in the mid-40s to almost 80 years in the mid-2000s. • Empower communities: it is also important to consider, as part of the reform design, the active participation of local communities in decision making. In that way, the final user will comprehend the scope of the changes, and opposition (that is particularly high in health related issues) will slow down. • Quick benefits: when implementing the reform, it is important to consider which measures will yield positive benefits in the short term. It is hard for the different involved agents to accept any change in the status quo if some "short-term" benefits Good Practice in Health Financing, p. 231 are not observed. In the case of Costa Rica, the first steps were oriented to increasing coverage in rural and poor communities. Role of non-public sector: despite the opposition in many countries to accept other service provision than the public one, it is important to have private sector participation. In countries where the network of providers is scarce and there are urgent needs, the contribution of the private sector may be critical in order to avoid bureaucratic regulations that usually delay infrastructure projects, for instance. 5.3. Institutional-Financing Arrangements Parallel to changes in the health sector, some internal administrative and organizational transformation should be implemented in the institution: • Internal reorganization: the most important organizational change introduced in the CCSS during the reform period was the creation of the Purchasing Unit. Separation of purchasing from provision functions has allowed one single department to fully concentrate on planning, negotiating, monitoring, and evaluating the performance of health providers. At the same time, this separation has permitted the Purchasing Unit to focus on coverage issues. • Training staff: the CENDEISS aimed at training both clinical and non-clinical staff of the CCSS on the new orientations of the model. It could be a good idea not only to implement a training program but to give it a permanent status, with its own budget and with clear long-run objectives. • Information systems: adequate information goes hand-in-hand with the transformation of the sector. Having good information systems will permit the institution to improve its capability to track contributions, evasion, consultancies, costs, and reporting, and to improve the planning cycle. The SICERE system, for example, allows the Financial Management of the CCSS to know the flow of revenues on a daily basis, not until 30 days as before. • New payment mechanisms (oriented to enhance performance): restructuring payment mechanisms could be a powerful tool to increase coverage. In the particular case of Costa Rica, the Management Agreements signed between the CCSS and primary care providers explicitly define coverage (immunization, smear, etc.) as a target to which payments will be linked. • Set clear incentives: incentives should be set up aimed at promoting efficiency among administrative staff. Good Practice in Health Financing, p. 232 5.4. Lessons for Low Income and Middle Income Countries 1. Reforms can be applied even if the context is far from ideal. In effect, the creation of the mandatory health insurance in Costa Rica, at the beginning of the 1940s, faced the opposition of all the involved sectors, including workers. Despite this, the President insisted on the need for having a health insurance system to protect workers. After intensive negotiations, the project was approved in the Congress with the support of the Roman Catholic Church, the governing party and the Communist Party. Completely opposite forces, like the Church and the Communist Party, strongly supported the initiative. 2. Health reforms take timefor implementation andfor yielding outcomes. Changing a model of health service provision takes time, not only because the model itself demands new staff and equipment, but because, inherent to the model, there are several pervasive practices among doctors and the population. In Costa Rica, one of the major barriers faced by the CCSS was the transformation of the model from a curative based model to a preventive oriented scheme. The experiences vary from country to country and from topic to topic. While achieving full coverage with primary care services took the country less than 15 years, fully universal health insurance coverage has still not been achieved after 63 years. 3. Expand primary care to reach universal coverage. The main focus of the health sector reform in Costa Rica was the expansion of primary care services as the main channel to provide universal access to the entire population. The new strategic orientation yielded important health outcomes. The share of the population with access to PHC services jumped from 25 percent of the population just before the reform to almost universal coverage by 2005. Additionally, immunization rates have recovered, 97 percent of pregnant women now receive pre-natal care (52 percent in the pre-reform period), and 70 percent of the women that are 49 years old or older have received a pap smear. It is quite important to mention at this point that the relative success of the PRe oriented strategy is the result of multiple factors that converged at the same time, including: i. Internal institutional transformations in the CCSS. The top authorities were aware of the need of the new model 11. Increasing budgets focused on PHC activities lll. Intelligent incorporation of cooperatives and the private sector in health service delivery, regulated by Management Agreements that incorporate specific coverage and quality targets IV. Adequate planning of the new model that favored the division ofthe country in small health areas, each one with at least one EBAIS v. Small country with relative easy access to rural areas VI. High level of investments in infrastructure during the first stages of the reform VB. Low PHC coverage before the reform, so the need of a new model was practically supported by all relevant stakeholders Good Practice in Health Financing. p. 233 4. No success or sustainability of the outcomes without enough money. Money is relevant in two senses. First, to trigger the reforms and to consolidate a solid basis for further transformations. Second, to strengthen primary care, one of the pillars of the Costa Rican pro-coverage strategy. Costa Rica, for instance, increased 22 percent the participation ofPHC in the total health budget within 10 years. Additional money allowed the CCSS to establish more than 900, generally well-equipped, local posts (EBAIS) by 2006. Such a big number of EBAIS covered the entire country, taking care of 1,000 households per EBAIS (4,000 persons), an adequate proportion of families per post. 5. Both the public and the private sector have critical roles to play in improving health conditions; together, they can achieve more than iffunctioning alone. Costa Rica broke with the traditional public centered delivery model and incorporated new models of provision in which the private sector had an important role to play. Both the international experience and the Costa Rican case show that the private sector can be an extraordinary means to increase coverage, but in particular service coverage. Participation of non-public entities should be limited to private firms; in Costa Rica, the spectrum of potential providers range from cooperatives to universities to non-profit associations and NGOs. The incorporation of cooperatives and private foundations was a key step to complement the efforts of the CCSS to extend primary care throughout the country. By incorporating the Management Agreements and creating an adequate regulatory framework that normalize the relationships between the CCSS and private providers, the model was able to launch EBAIS in poorer districts. In short, it is important to mention that the participation of the private sector should be oriented to service provision, not to financing. 6. Think in poorer sectors first. As it was explained in the previous section, the first EBAIS were launched in rural and poorer regions, a decision that yielded multiple benefits to the reform process. First, it increased equity and access in traditionally underserved communities. Second, starting with poorer regions most likely will generate quick benefits, an enabling factor for success. Finally, quick benefits create a solid base for continuing with further reforms because it will possible to gain the recognition of different social sectors. 7. Be transparent. Transparency and accountability mechanisms should be implemented at the beginning of the project in order to enhance communication with all interested parties. 8. Pay attention to the financial health of the system. In short, Costa Rica is not a good example of cost-containment, but rather a good case of financial sustainability. In effect, costs per affiliated have been increasing, as well as contributions, keeping a positive margin between both financial figures. In a report prepared by the World Bank (2003), the study revealed that the real expenditure per direct contributor increased 5 percent during the 1990s, while the real contributions per direct contributor increased 11 percent. The difference has been enough to finance the deficit generated by other groups like Good Practice in Health Financing, p. 234 pensioners, voluntary affiliates, and state beneficiaries. Several issues explain the dynamic behavior of contributions vis-it-vis the decelerated growth in expenditures: 1. The CCSS implemented specific plans to reduce evasion and defaulting 11. Waiting lists were used as cost-containment mechanisms iii. Increasing funds to PHC activities resulted in high cost-effective interventions that allowed the system to achieve interesting results while keeping expenditures partially controlled IV. The existence of a General Controller Office (outside the organizational framework of the CCSS) is important for avoiding irrational spending by health authorities Good Practice in Health Financing, p. 235 References Amemibar, Ana Victoria C. "Industria Farmaceutica en Costa Rica: EI negocio de los medicamentos." Revista Actualidad Economica N° 308-309. Ano XVIII December 2004 - January 2005. Arce, Claudio and Munoz, Carlos. 2001. "La Reforma Pendiente: Introduccion de Libre Eleccion en el Seguro de Salud'. San Jose, Costa Rica: CCSS. Arce, Claudio and Saenz, Luis. 2002. "Hallazgos preliminares, logros y desafios de la reforma en el sector salud de Costa Rica durante los afios noventa". Report prepared for Proyecto Estado de la Nacion. San Jose, Costa Rica. Arce, Claudio. 2001. "El desempefio hospitalario: entre luces y sombras". San Jose, Costa Rica: CCSS (Unidad Coordinadora Proyecto de Reforma, Gerencia de Modernizacion y Desarrollo). Banco Mundial, Unidad Sectorial de Desarrollo Rumano Departamento de America Central Region de America Latina y el Caribe. 2003. Costa Rica: El Gasto Social y la Pobreza. Washington D.C.: The World Bank Castro, Carlos and Saenz, Luis. 1998. "La Reforma del Sistema Nacional de Salud'. San Jose, Costa Rica: Coleccion Tiempos de Cambio, MIDEPLAN. CCP (Centro Centroamericano de Poblacion). Informacion Censal de los Cantones de Costa Rica (INFOCENSOS). San Jose, Costa Rica: UCR. http://infocensos.ccp.ucr.ac.cr/ Duran, Fabio. 2005. "(,Racia donde va la Seguridad Social de Costa Rica?" In Ensayos en Honor a Victor Hugo Cespedes Solano, ed Grettel Lopez and Reinaldo Rerrera. San Jose, Costa Rica: Academia de Centroamerica. Gauri, Varum, Cercone, James and Briceno, Rodrigo (2004). "Separating financing from provision: evidence from 10 years of partnership with health cooperatives in Costa Rica". Health Policy and Planning 19 (5): 292-301. Instituto Nacional de Estadistica y Censos (INEC). 2005. Encuesta de Hogares de Propositos Multiples. San Jose, Costa Rica. Instituto Nacional de Estadistica y Censos (INEC). 2006. Survey on Revenues and Expenditures. San Jose, Costa Rica. Ministry of Planning of Costa Rica. Sistema de Indicadores sobre Desarrollo Sostenido, Variables e Indicadores Sociales. San Jose: MIDEPLAN. http://www.mideplan.go.cr/sides/sociallindex.html MoR-PARO. 2003. Gasto y Financiamiento de la Salud en Costa Rica: Situacion Actual, Tendencias y Retos. San Jose: Organizacion Panamericana de la Salud. Good Practice in Health Financing. p. 236 MoH-PAHO. 2003. Las desigualdades de salud en Costa Rica: una aproximacion geografico - poblacional. San Jose: Organizacion Panamericana de la Salud. MoH-PAHO-CCSS. 2004. Perfil del Sistema de Servicios de Salud en Costa Rica. www.cor.ops-oms.org Picado, Gustavo. y Saenz, Katia. 2000. Estimacion del Gasto Privado en Servicios de Salud en Costa Rica. San Jose, Costa Rica: Documento de Trabajo de la Direccion Actuarial y de Planificacion Economica N° 25, CCSS. Proyecto Estado de la Nacion. 2005. 11 Informe del Estado de la Nacion, Anexo Estadistico. San Jose, Costa Rica: UNDP. http://www.estadonacion.or.cr/Compendio/ind compendio.html Saenz, Alberto. 2005. "The role of equity and solidarity within social health insurance: Chances and risks of the Costa Rica way towards universal coverage." Paper presented in the International Conference on Social Health Insurance in Developing Countries. http:Uwww.shi-conference.de/contribut.html Trejos, Juan Diego. 2002. "La Equidad de la Inversion Social en eI2000". Paper presented at the Octavo Informe sobre el Estado de la Nacion en Desarrollo Humano Sostenible. San Jose, Costa Rica: UNDP. http:Uwww.estadonacion.or.cr/info2002/nacion8/ponencias.html United Nations Population Division. 2004. World Population Prospects: The 2004 Revision Population Database. New York: United Nations. http://esa.un.org/unpp/index.asp WHO (World Health Organization) Department of Measurement and Health Information. Global Burden ofDisease Estimates. Geneva: WHO. www.who.intlhealthinfolbodestimates/enlindex.html World Bank. 2006. World Development Indicators 2006. Washington, D.C.: The World Bank Estonia Triin Habicht J amo Habicht Good Practice in Health Financing, p. 238 1. Background 1.1. Economic Situation and Main Social Trends Radical economic reforms in Estonia started in late 1980s with the aim to restore a market economy after the planned economy model under Soviet Union. This was supported by the re- establishment of independence of Estonia in August 1991 and the following monetary reform, as a first step to modernize economy. This meant the introduction of a own currency (Estonian Kroon) with a fixed exchange rate and a currency board arrangement, liberal external policies, and, most importantly, an immediate abolishment of most restrictions on capital account transactions. The main characteristic features of Estonian reforms have been radical and rapid introduction of market oriented institutions. By now the privatization of the economy has been completed, with only a few strategic companies (e.g. Estonian Energy Company) left under the governance of the State. Estonia's GDP decreased rapidly in the first years of independence in 1991-1994, but the recession period ended by 1995. In the following years, the economy recovered quickly and GDP has been growing since then in each year. In 2004, the Estonian economy grew by 7.8%, being by 5.6 percentage points higher than the European Union average economic growth. Rapid growth impacts positively to the convergence process - as a result, the country is witnessing a remarkable convergence in real terms to EU levels, with purchasing power parity per capita GDP increasing to 51% of the EU level in 2004 from 35% in 1995. The GDP per capita in 2004 were risen to the level as 14555 international dollars compared to 6289 in 1995. The estimated growth rate of economy in following years is expected to be around 6%, which is about three times faster than EU average. However, the vulnerability to external shocks is high as Estonia is small open economy and, with a persistent and large current account deficit (1.4 billon US dollars in 2004) and a rapidly expanding stock of gross external debt (about 10 billions US dollars in 2004). The purchasing power of population has increased by 30% between 1992 and 2002, but this growth has not evenly distributed among different socioeconomic groups. Income inequality, as measured by the Gini coefficient, widened substantially in the 1990s (reached to the level 0.38), but began to decline steadily in the recent years in 2004 being at the level of 0.36. Still high level of differences in average monthly income per household member by different income groups persist where income of the richest 20% of the population is 6.1 times larger than that of the poorest 20%. However, the proportion of people living below the national poverty line has decreased tremendously from 36.1 % percent 1997 to 17% in 2003. The groups at highest risk of poverty are unemployed (in particular long-term unemployed), large families and single-parent families. The poverty headcount ratio at $1 a day and $2 (in international dollars) were respectively 2 and 8 percent of the population in 2003. Estonian fiscal policy can be accounted as conservative one and yearly balanced budget policy is followed to guarantees a favorable and stable environment for economic development. Government has declared that the long-term sustainability of fiscal policy in conditions of ageing population as one priority of Estonia's macroeconomic policy. This conservative approach is also pursued in health insurance budget, which has been so far yearly balance without remarkable deficit. 238 Good Practice in Health Financing, p. 239 The Estonian taxation system is counted as simple and transparent, with as few exceptions and differentiations as possible. The Estonian personal income tax system with its flat rate is counted one of the most liberal tax regimes in the world. The goal of taxation policy is to motivate entrepreneurship and initiative, and to rather tax consumption instead of earnings. This has guaranteed the economic growth and there is no plan for fundamental changes in the near future. But compared to neighboring countries, Estonian labor is still highly taxed 7o • To reduce the taxation oflabor, the personal income tax rate is reduced starting from 2004 where it was 26% to 20% by 2009. This reform supports the Government low taxation of earning policy. However, the high tax burden on labor is mainly related to social insurance tax paid by employers on behalf of employees as 33% of the salary of each employee, 13% whereof is marked for health insurance and 20% is for the pensions of today' s pensioners. 1.2. Demographic and Epidemiological Situation The size of Estonia is 45 227 square km. It has more than 3 500 km of coastline and borders with Russia in the East and Latvia in the South. The population size was 1,351 million in 2005.71 Compared to the similar size of European countries the density is rather low as 30 inhabitants per square km. By ethnic groups the popUlation comprises 69% Estonians, 26% Russians and 5% other nationalities (as Ukrainians, Belarusian, and Finns). Similarly to countries across Europe, the Estonian population is ageing. The proportion of the population over 60 years old was around 20% in 2002 and it is expected to be more than one quarter by 2025 (see population pyramid in figure 1). Among other trends usual for Europe as ageing population the urbanization has taken place over last decades where in 200569.3% of the population lived in the cities. The population of Estonia is declining since 1991 steadily, mainly due to negative natural growth and emigration, mainly to Russia. Estonia has a crude birth rate in 2005 of 10.7, a crude death rate of 12.9, and a fertility rate of 1.37, which is lower than required to maintain current population level. Hence, there is negative population growth with a rate of natural increase at -3.8, leading to a declining population. Total popUlation decreases to 2025 more than 10% compared to 2000, which is mainly result of low fertility rate (see figure 1). The dependency ratio (the proportion of population aged 0-14 or over 65 years) is currently at 48%, which is lower than average EU rate. However, this is likely to increase in the future mainly due to the population aging. The demographic trends in Estonia such as declining and aging population have serious consequences for the health sector due to the reduced tax base, as health expenditures are mainly financed through labor related taxes and children and elderly are insured where no contributions are made on behalf of these groups. 70 The tax wedge for low income workers was 37.2% in Estonia in 2004, at the same time the European Union average was 36.4% (Source: Eurostat 2006, europa.eu.intlcommleurostatl) 71 Source: Statistics Estonia 2006, www.stat.ee 239 Good Practice in Health Financing, p. 240 Figure 1. Population pyramids 2004 and 2025 in Estonia Year 2004 Year 2025 80+ i 80+ 75- 79 ! 75- 79 S~- 74 70- 74 - 69 65- 69 60- 64 60- 64 55- 59 I 55- 59 50- 54 50- 54 45- 49 45- 49 40- 44 40- 44 35- 39 35· 39 30- 34 30- 34 25- 29 25- 29 20- 24 : 20- 24 15- 19 15- 19 10-14 10-14 5-9 5·9 0- 4 0- 4 4 2 0 2 4 6 4 2 0 2 4 Population shares, Ofo Population shares, 0/0 Source: Eurostat 2006, europa.eu.int/comrnleurostat/ The overall trend is that the average age of mothers giving birth, including the average age at first birth, has continuously increased since 1995 from 23 years to 24.8 years in 2003 72 . At the same time, birth activity among women under 25 has continuously decreased. Bearing witness to positive changes in family planning habits is the fact that in recent years, the absolute number of induced abortions and the number of abortions per 1 000 women between the ages of 15-49 have steadily decreased from 50.0 in 1995 to 31.0 in 2003 73 • In 2000, the number of induced abortions fell below the number of live births for the first time and this tendency of decrease has continued ever since. The number of deaths has been quite stable in recent years, only slightly crossing the 18 thousand margins. Over half of deaths are caused by diseases of the circulatory system, followed by cancer and accidents, poisonings and acute injuries. The causes of death differ by sex - in 200363% of women's deaths were caused by diseases of the circulatory system (for men 47%), 17% by neoplasm (for men 20%) and less than 5% by accidents, poisonings and acute injuries (for men 16%). The mortality and morbidity patterns are similar to Eastern European countries. According to the latest Burden of Disease study in Estonia (Ministry of Social Affairs and University of Tartu 2004) main causes of the burden fall under cardiovascular diseases, neoplasm and external causes following the mortality patterns. Still, due to the burden from disability or decrease on health status other disease groups as joint and muscle diseases, pulmonary diseases and mental health disorders rank high (see figure below). In addition, more than half of burden of disease is concentrated to the working age population, where among men it reaches 58%; less burden is among over 65 years old 39% and young people aged 0-19 as 8%. There is clear difference among men and women as only two causes in top ten are same and among men external causes are much more prevalent formulating 11 % of the total disease burden. 72 Source: Ministry of Social Affairs 2006, www.sm.ee 73 Source: Ministry of Social Affairs 2006, www.sm.ee Good Practice in Health Financing, p. 241 Figure 2. Burden of Disease in Estonia 35% ~ o Years lived with disability • Years of life lost (mortality) co 'c 30%- ~ w .= :I I 25% " . :B '6 '0 20% r- i "E I . " .Q 1i 15% L B E g c 10%- o 1:: &. e 0. 5% ~ I 8,0% , 6,9% ~ '" "' "' E "' " Ul :; "' (.)" "' gJ Ci '" l& '" '" ~ .~ g '5 ~-c Z ~ '" (J .~ e ::J " Z Source: Ministry of Social Affairs 2004 74 The demographic trends in Estonia such as declining and aging population have serious consequences for the health sector due to the reduced direct tax base, as health expenditures are mainly (65%) financed through labor-force related taxes where children and elderly are covered with health insurance without any contributions made on behalf of these groups. In addition, there is a pressure to health expenditures due to epidemiological transition as an increasing burden of chronic illnesses which are more costly to treat. The average life expectancy decreased in the first half of 1990-s, reaching the low turning point in 1994, when it was 61 for men and 73 for women. This was the result of radical changes in a country due to regained independence, which was also expressed in increase of mortality due to external causes. Since 1995 the life expectancy has slowly increased and was respectively 66 for men and 77 for women in 2003, which is still among the lowest among European Union (EU) countries (see figure 3). 74 Raport "Years lost due to burden of disease: links with risk factors and cost effectiveness to reduce risks." (available in Estonian). Good Practice in Health Financing, p. 242 Figure 3. Average life expectancy at birth in Estonia and European Union in 2003. 85 -: 82.0 80 T -- -- 78~4 ---- ! 77.2 75-'- - - ~ I "' > GI , 70 -7- -- 65 ..;- I 60 Estonia EU members before EU members since May 2004 May 2004 • female 0 male Source: Health for All Database 2006 Over the years the infant mortality rate has fallen from 14.8 in 1995 to 5.7 in 2002. Maternal mortality rate 23 per 100000 live birth (in 2002 7.7) (number is very low, as in absolute terms 1 to 3 death per year and therefore the annual rate is varying). The five main risk factors creating the disease burden are smoking (counting for 8.3% of disease burden), low physical activity (7.4%), high level of alcohol consumption (6.6%), overweight (5.1 %), and low fruit and vegetable intake (3.8%). The risk factors as smoking, alcohol consumption are mainly concentrated among men causing both more 12% of disease burden; smoking alone is causing 40% of the cardiovascular diseases and 40% of cancers in Estonia 1.3. Political Environment Estonia is a parliamentary republic. The Parliament (Riigikogu) consists of one chamber with 101 members. It is elected for a period of 4 years. Since 1992, when the first elections in independent Estonia were held, all governments have been coalition governments of two or three political parties. Although none of the coalitions has governed for a full term, they have been stable enough to launch and implement economic and social reforms. Estonian political parties tend to be at the centre or to the extreme right of the political spectrum. To date, governments have been on the right, although social democratic values and ideology have become more visible in recent years. It is also shown that in early nineties the health and social sector policies development was driven by social values. The current coalition has agreed to lower the proportional rate of personal income tax and to make a Good Practice in Health Financing. p. 243 one-shot increase in state pensions. In terms of health policy the coalition agreement enables wide interpretation and is not very articulately formulated. However, the priorities are the HIV 75 and drug dependency prevention, remuneration and access to care. The second political layer in Estonia consists of 227 municipalities, which have budgetary autonomy and local tax-raising powers. However, administratively Estonia is divided into 15 counties, each run by a governor and having an administrative structure known as the county government. Both the governor and the county government staff members are civil servants of the central administration. However, many state agencies, including those engaged in health care administration and finance, operate not on a county basis but through regional departments that cover two to four counties. The role of the counties and municipalities in the health sector has varied. During the decentralization era in early nineties a number of responsibilities were distributed to counties similarly to the fragmented health insurance system at this time. In addition the planning and oversight on health care services were the role for counties where special staff was recruited. During the centralization in late nineties of the planning and development, the role of planning primary health care and liaising with municipalities on behalf of government remained to counties. The municipalities have also some role in health sector as running local public health activities and supporting health care sector (as well owing some of the hospitals) but the practice varies over Estonia. 2. Overview of Health Financing and Coverage The following chapter describes the health financing system in Estonia in terms of collection of funds, pooling, benefit package, cost sharing and purchasing of the health care services. Finally, the performance of the current system is described as well its impact to efficiency and equity. 2.1. Collection of Funds In 2004, Estonian total health expenditures were 5.5% of gross domestic product (GDP). Among the European Union counties the same figure varies from similar to over 10 percent being around 6.4% on average. Over the time the health expenditures level in Estonia has been quite stable, with small variations due to changes in economic environment. During the last years the health expenditures share of GDP has increased slightly, which has been due to more favoring economic environment ensuring stable collection of taxes and increasing out-of-pocket payments. The annual increase of the total health expenditures in absolute terms has been around 4% in late nineties and during the last years about 14%76. In 2004, total health expenditures constituted 618 million US dollars. The public expenditures share of total health expenditures is relatively high forming 76% of total health expenditures and 4.2% of GDP in 2004. From total Government budget public health expenditures have been over last years around 11 %. Majority of public funds comes through health insurance system - 87% of health expenditure in the public sector and 66% 75 According to latest data available (10.2006) there were 5550 HIV positive in total (Source: Health Protection Inspectorate 2006, www.tervisekaitse.ee). 76 Inflation rate was 3.0% in 2004 and 4.1 % in 2005 (Source: Ministry of Finance 2006, www.fin.ee). Good Practice in Health Financing, p. 244 of total health expenditure in 2004 (see table 1). When health insurance system was introduced in 1992, it was expected that municipal and central government financing will grow in line with health insurance funding. In spite of that, the share of central government and municipalities funding has been slightly decreasing and the role of financing trough health insurance is increasing (see figure 4). Table 1. Total health expenditures by sources, in 2004 (US Dollars) Expenditures Source Share, % (USD, million) Central government 52 8.5% Municipalities 8 1.3% Social insurance 406 65.7% Out-of-pocket 132 21.3% Other private 17 2.7% External sources 3 0.4% Total 618 100.0% Source: Ministry of Social Affairs 2006, www.srn.ee Figure 4. Distribution of funding sources, in 1999 and 2004. 100% 80% 60% 66. 40% 20% 0%-1----""""- 1999 2004 mI Central government o Municipalities EJ Health Insurance Fund ED Private household out-of- pocket tl Other private sector • External sources Source: Ministry of Social Affairs 2006, www.srn.ee The main source of health insurance revenues (about 98-99%) is a social health insurance contribution paid by salaried workers and self-employed people. This is also known as social tax, which covers health and pension contributions respectively 13% and 20% of employee wages and of self-employed individuals' earnings. Good Practice in Health Financing, p. 245 As the contributions are related to active workforce, Estonia has about half of insured equalized and having the health insurance coverage. The non-contributing individuals represent almost half (49%) of the insured population and their expenses are implicitly subsidized by the other categories that shows high level of solidarity in the system. These groups, including children, pensioners, those receiving a disability pension and students (during their nominal study time) are eligible to the same benefit package as everybody in the insurance pool without any contribution from either themselves or the state. The state officially contributes for only a small proportion of the covered population (4%), including individuals on parental leave with under 3-year old children, those who are registered as unemployed (eligible for cover for up to nine months) and those caring for disabled people. The state's contribution for these groups is defined annually at the same time as the state budget is approved and is dependant of number of persons eligible. However, the contribution rate is very low and compared to average contribution made by workers and self-employed it is 15 times lower. Currently, there is a plan to increase the contribution rate per person to the level, where it would be equal to the health insurance tax paid on minimum salary. This would decrease the difference compared to the average contribution to the level 2.5 times. The system applied in Estonia has relatively high solidarity level, as half of the insured population contributes and at the same time all insured are entitled to same benefits (meaning thus cross-substitution inside the pool). In 2004, Central Government funding accounted for 8.5% and local municipalities 1.3% of total health expenditures. Central Government mostly finances ambulance services, emergency health care services 77 for uninsured people and public health programs (e.g. HIV/AIDS; tuberculosis prevention; cardiovascular diseases prevention). Local municipalities have no clear responsibility to cover health care expenditures and therefore the financing practices vary a lot 78. Mainly municipalities spend on covering health expenditures of uninsured (some extra care in addition to Central Government covered emergency care), travel support to the health care facilities, support to the selected public health programs and high out-of-pocket payments of people facing high health expenditures. At the same time other municipalities are covering some costs of family physician (FP) services in their region, but also in-kind contributions for FP-s are quite common (e.g. rooms to establish practice). In addition, municipalities owning hospitals are making contributions to cover some capital costs of these hospitals. Out-of-pocket payments comprise statutory cost-sharing for Estonian Health Insurance Fund (EHIF) benefits, direct payments to providers for services outside the EHIF benefits package or from non-EHIF-contracted providers and informal payments. Out-of-pocket payments constituted about 21 % of total health expenditures in 2004. Most out-of-pocket payments go to pharmaceuticals (53% of out-of-pocket payments) and dental care (23% of out-of-pocket payments). According to the National Health Accounts, households paid 45% of total pharmaceutical expenditure and 61 % of dental care expenditure out of their pockets in 2004. Both shares have been increasing due to introduction of new rules of drug reimbursement 77 According to Health Care Services Organization Act (200 l) in Estonia the emergency care means health services which are provided by health care professionals in situations where postponement of care or failure to provide care may cause the death or permanent damage to the health of the person requiring care. 78 In 2005 the share of health expenditures from total budget varied from 8% to 0% by municipalities. Good Practice in Health Financing. p. 246 (introduction of reference prices) and exclusion of adult dental care from EHIF's benefit package. The share of out-of-pocket payments of total health expenditures has been increasing over time (see figure 4) and it is expected to increase further in the near future. Still, the share of out-of-pocket payments is relatively low (21.3%) compared to similar countries in Eastern Europe region (e.g. 45.9% in Latvia and 24.2% in Lithuania) and some extent higher compared to Western European countries (see figure 5). Figure 5. Out of pocket payments in 2004 £ 25.0 ~- 22.8 1 .c 21.3 E 20.0 - - -- o .... III 15.5 ~ til I o § 15.0 - .... .- III .... C'tl GI C E ~ 10.0 >)( ~GI 8 5.0 ~- 0.0 ------,- - Estonia EU members EU members before May since May 2004 2004 Source: Health for All Database 2006 Other private expenditure includes employer-paid health care travel insurance, employer- paid health check-ups and pharmaceuticals (mainly bought by foreign visitors but also by corporations). However, the share of these expenditures is rather low comprising less that 3% of total health expenditures. According to the recent studies 79 informal payments are not common in Estonia and do not constitute important source of out-of-pocket payments. This on one hand may be due to the introduction of formal co-payments in 2002, but on other hand to the low level of corruption and informal sector in general. Private insurance currently plays a very small part in funding care in Estonia. Until recently, most of the private insurance market comprised employer-paid health care travel insurance. Since 2002, there is one private insurer providing health insurance, but it is not very popular (currently few hundred enrollees) as their benefit package is very limited. In addition, the EHIF scheme enables people who otherwise would remain uninsured to enroll in the public health insurance scheme. Eligibility for voluntary coverage is restricted to residents of 79 losin M. 2004. Korruptsioonija varimajanduse levik Eestis [Corruption and the black market in Estonia]. Tallinn, Estonian Institute of Market Research.; CIET International. 2002. Curbing system leakages: the health sector and licensing in Estonia. Tallinn, Community Information, Empowerment and Transparency (CIET) International. Good Practice in Health Financing, p. 247 Estonia who receive a pension from another country or people who are not currently eligible for membership but who have been members for at least 12 months during the last two years prior to applying for voluntary membership. The voluntary members pay a contribution based on the previous year's average national salary and benefit from the same coverage as other insured people. The funds are pooled together with those who cover the rest of the insured people. 2.2. Allocation of Funds and Purchasing Services The schematic overview of the organization of health financing arrangements in Estonia is given in figure 6. As mentioned above, the core purchaser of health care services is EHIF, who purchases most of care for insured persons (94% of total population), except ambulance care. Ambulance care is financed directly through state budget and this is administered by Health Care Board, which is an agency of Ministry of Social Affairs. Emergency care cost for the uninsured persons are covered through the state budget, but the administrative tasks are delegated to the EHIF to make the administration more effective, as same kind of payment methods and tariffs are used for uninsured and insured persons. The EHIF funds are collected centrally to balance the regional disparities in income levels. Then pharmaceutical and temporary sick leave benefits (these are open-ended obligations for EHIF) are administered centrally, but most of health care services funds are allocated to EHIF's regional departments (currently four) according to crude capitation without any need-adjustments. The exemption is primary health care for which more sophisticated formula is used 8o • Overall, 98% of the EHIF health care services funds are allocated to the regional branches. The rest remain centrally managed for a small range of expensive or infrequent procedures for which regional allocation would not be feasible. These include bone marrow transplants, peritoneal dialysis, some areas of oncology and hematological treatment. 80 Age-adjusted capitation plus additional allocations according to fee-for-service and lump sum payments. Good Practice in Health Financing, p. 248 Figure 6. Overview of the health financing system in Estonia (shares of different revenue sources are given for 2004) Ministry of Social Affairs _ _ _ _ _ _ _ _ _ _ _ _ _ __ and its agencies 8.5% ~ ________ ~ Municipalities ~ 1.3% Estonian Health Insurance Fund (EHIF) 65.7% EHIF branches Hospitals a -----t - ---:-- --I I I I I I I Ambulatory specialist ----.., I I I _____ :___ .J care b I I I I ____ oJ I I I I I I I Public health I programmes ----------..1 - bold line represents main source of revenue Ambulance services d 1-01.1 . . . .- - - - - - - Notes: a fee-for-service + daily rate + some per case payments; 50% of each case is reimbursed using DRG prices; contracts close-ended case-volume contracts; b fee-for-service; close-ended case-volume contracts; C weighted capitation + fee-for-service + additional fixed payments; d general budgets and fixed payment per provider unit Source: Jesse et al (2004)81, with author modifications The allocation of funds is further refined during the contracting process. The EHIF is contracting with providers (not individual doctors) for all types of care such as primary care, specialist outpatient and inpatient care (except ambulance care and emergency care for uninsured), and all other services in the benefit package. The EHIF's contracting process is shown in figure 7. 81 Jesse M, Habicht J, Aaviksoo A, Koppel A, Irs A, Thomson S. 2004. Health care systems in transition: Estonia. Copenhagen, WHO Regional Office for Europe on behalf of the European Observatory on Health Systems and Policies Good Practice in Health Financing, p. 249 Figure 7. Contracting process ofEHIF Pooling capitation- Needs selection of Annual capped cost Contract in the based assessment partners and and volume performance EHIF allocations in each negotiations contracts and for monitoring f-t to EHIF f-t region I-t about contract f-t strategic hospitals a .... and utilization regions volumes five-year framework review (regional basis) contract /' l' I Standard contract conditions negotiated and agreed Adjustments of between EHIF. Estonian contract volume Hospital Union. Estonian according to Society of Family Doctors utilization Source: Couffinhal and Habicht 2005 82 Since 200 I, EHIF has gradually introduced the need assessment as input for purchasing decisions. Currently it bases mainly on historical data analysis about health care services utilization and existing queues or waiting times, but there are further plans to use more sophisticated methods. Since 2005, EHIF started to negotiate with medical specialist associations 83 about their assessment on population needs in terms of their specialty to have detailed input from specialties and involve them into planning process. The results of these negotiations are taken into account in EHIF's budgeting and contracting processes and when benefit package is updated. The aim is to increase the role of medical profession in long term direction setting for purchasing decisions. Next step in contacting process is selection of providers which are to be contracted. Contracts are only made with providers who are licensed to work in Estonia by Health Care Board. EHIF is practicing selective contracting (choosing not to contract with a given providers) to some extent. Main objectives of the selective contracting are to motivate service quality improvement and to introduce mild market competition into health care provision, but also to buy services in areas that are perceived less attractive by providers 84 • Nevertheless, EHIF is required to contract with all Hospital Master Plan hospitals (less than 20 strategic acute care hospitals) and they have guaranteed contract volume according to historical volumes. The main exception is dental care, which do not systematically have a contract with the EHIF and provide services privately. Also, 20% of outpatient care is purchased using selective contracting, where EHIF announces public procurement and all providers can submit their offerings. 82 Couffinhal A, Habicht T. 2005. Health System Financing in Estonia: situation and challenges in 2005. HSF Working Document, Health Systems Financing Programme. Copenhagen: WHO Regional Office for Europe 83 In 2005 pulmonology and oncology were priority areas. 84 The criteria for providers' selection are: the proximity of service provision to patient; the share of provision of services in day-care; the experience of last contracting period (rejected claims for reimbursement, complaints by patients); and the price of services - lower price offered to EHIF giving preference. Good Practice in Health Financing, p. 250 EHIF's contract standard conditions are negotiated on yearly basis with providers' representing associations as Society of Family Physicians and Hospital Association. This assures that once contact conditions are agreed between EHIF and providers associations, these are universal and have to be accepted by all providers. Once providers which are to be contracted are selected and contract standard conditions are agreed with providers' associations, further negotiations with selected providers in specialist care continue to determine the volume of services by number of cases as well as average case prices by specialty. These negotiations do not determine the actual payment method but rather constitute a planning element aimed at, among other things, containing costs for each case. In terms of coverage more important is the agreement of the number of cases. This supports the implementation of the EHIF objective to keep the access to care at least on the level of previous year at macro level at provider level. As a result of these negotiations the contract volumes with each provider are agreed. If provider exceeds agreed volume during the contract period, EHIF will not reimburse additional expenditures. This principle is very strictly followed to ensure the balance between EHIF's available funds and expenditures. If there are additional revenues or need to change the number of cases, the contracts can be renegotiated and agreed during the year. To monitor the real access, EHIF performs monthly monitoring process via Management Info System across all providers on the contract implementation. Described process holds only for specialist care, as for primary care the contract volume is not an object of negotiations being determined fully in accordance of payment methods. As mentioned above, the actual payment methods, service prices and benefit package (in Estonia these are usually seen all together as there is "health service list", which determines all three) are not determined during the contract negation process. All payment methods that EHIF uses are regulated by health care service lists which is approved by Government. Process of revision of payment methods is following the same process as revision of prices. Inpatient and outpatient providers of specialized care are paid using a range of payment methods that depend on the type of services provided: fee-for-services, visit fees, per diem, DRG-based and case-based complex prices. During the late 1990s, a decision was made to move away from detailed fee-for-service payments and their perverse incentives and to increase case based payments. This was motivated by the fact that the average length of stay was stagnant and remained well above the proclaimed objective of 4.5 days for acute-care cases. In addition, the bed occupancy rate in acute-care hospitals decreased from 80% to 65% during the 1990s, which indicated that hospital capacity was being inefficiently used. The DRG system started to be implemented in April 2004, and it is planned to be gradual. The proportion of DRG payment for each case was initially set at a low 10% to minimize possible risks of a new system. In 2005, the share was raised to 50%. However, there is a mix of several payment methods with aim to balance different incentives (see figure 8). Good Practice in Health Financing, p. 251 Figure 8. Distribution of different payment methods in monetary terms, in-patient and out-patient specialist care in 2005. 90% 80% 70% III _GI 60% J9 ~ .s~ 50% ... 'tI o C GI 40% :::eCL o >< GI 30% 20% 10% 0% Fee-for- Case based Per diem DRG service (old) • Out-patient 0 In-patient Source: EHIF (personal communication, July 2006) FP -s are paid through a combination of age-adjusted capitation and other types of payment (fee-for-service, lump sum payment to cover fix costs) that make up the practice budget (see figure 9). In practice for the capitation payments the FP practices receive monthly pre- payments, which are recalculated twice a year to reflect changes in the patient list. Additional development from 2005 (to be implemented 2006-2007) is the movement to develop additional payment instrument - a quality bonus system - where targets are set at practice level to improve the overall health care quality85 in primary care and promote good performers, especially those dealing with prevention services and managing chronic conditions. 85The expected outcome is the improved quality and effectiveness of preventive services, as well as better monitoring of chronic diseases. Specific outcome indicators have not yet been defined, but the aim is to reduce morbidity and hospitalization rates. Good Practice in Health Financing, p. 252 Figure 9. Distribution of different payment methods in 2005 in monetary terms for FP- in Estonia. 24/7 GP line, 0.4% Lump sum, Other, 2.0% 10.6% Fee-for- service, 14.0% Capitation (age- adjusted), 73.0% Source: EHIF 2006, www.haigekassa.ee Health care service prices are identical for all providers, and there are no adjustments for region or hospital characteristics such as teaching status. In principle, health service prices cover all costs related to providing services except those related to scientific and teaching activities, which are funded separately. All prices approved are maximum prices and providers and the EHIF can agree on lower prices for the contracts. Revision of service prices and payment methods can be initiated by provider or specialist associations or by EHIF. Each service should be evaluated in terms of four criteria: medical efficacy (evaluated by relevant medical specialist association), cost-effectiveness (evaluated by health economist), appropriateness and compliance with national health policy (evaluated by Ministry of Social Affairs), and the availability of financial resources (evaluated by EHIF). The process to update the service list is usually applied annually. 2.3. Benefit Package and Patient Co-payments In Estonia health insurance coverage is mandatory without opt-out possibility and aimed at the whole population. Entitlement to health insurance coverage is based on residence in Estonia and membership of main categories defined by law. In 2004, the health mandatory insurance covered 94% of the population in four main categories: (a) those who pay their own contributions as employees or self-employed, (b) those eligible for coverage without contributing, in particular children and pensioners, (c) those covered by contributions from the state, and (d) those covered by voluntary agreements. There have been some minor alterations in entitlement rules, but these have not caused changes in overall coverage level. Good Practice in Health Financing, p. 253 The EHIF provides for insured two kinds of benefits: cash benefits (about 20% of total EHIF expenditures) and benefits in kind (rest of the budget - health care services paid by EHIF). The cash benefits (reimbursement for services) include the costs of dental care for adults and some reimbursement in case of high pharmaceutical expenditure: insured people who spend more than a given amount out of pocket on listed drugs during a given year can claim partial reimbursement 86 . In addition EHIF pays for temporary health related incapacity for work in different forms as sickness benefits, maternity benefits, adoption allowance, and care allowance 87. Since the end of 2002, voluntary EHIF coverage has been extended to those who might otherwise remain uninsured. Voluntary members pay a contribution of 13% on the national average salary of the previous year, as published by the Statistical Office and they are entitled to the same benefits as compulsory members. However, the number of voluntary members is still very low. Uninsured are (6% of total population) persons in working-age who are not participating in formal labor market and are not eligible according to other criteria as being registered unemployed or disabled person. EHIF is not reimbursing any kind of health care cost of uninsured. Uninsured have the right for emergency care, which is reimbursed trough the state budget. However, administratively this is organized by EHIF having lowest transaction costs, but all necessary costs are covered from the state budget. Earlier, the emergency care costs of uninsured were covered by local municipalities, but this led to the situation where the access to treatment was different by municipalities. Currently there are ongoing discussions, whether local municipalities should start to cover the primary care and essential pharmaceutical costs of uninsured. Another option on table is that municipalities can start to buy the regular EHIF coverage to selected vulnerable groups in their constituency. The EHIF covers a broad range of health care benefits (so called in-kind benefits for patients) as family physician services, in- and out-patient specialist care, long term care, rehabilitation, dental care for children and prescription drugs. This feature is in part inherited from the old system in which the state funded and provided universal, comprehensive health care. Some health care services are excluded from the benefit package, e.g. cosmetic surgery, alternative therapies and opticians' services. During recent years, explicit rules for adding new services to the benefit package and establishing the appropriate level of user charges have been tried to introduce (see above). In addition to the health care services, EHIF finances prevention and promotion programs. This includes special health promotion fund under EHIF, from which public health institutions, NGOs and local activists can apply for funds for one year projects on pre-set priority areas. In terms of disease prevention several programs are funded as school health, reproductive health, screenings (breast cancer, phenylketonuria and hypothyreosis, cervical cancer, hearing on newborns), specific immunizations and other activities. This has 86 Since 2003, additional financial protection has been provided to those who face high phannaceutical expenditure: the EHIF reimburses 50% of a yearly cost between US$460 and US$767, and 75% beyond, up to a limit ofUS$1534. Any additional cost is not covered. 87 See details in Health Insurance Act (http://www.legaltext.ee/failid/findfile.asp?filename=X60043) Good Practice in Health Financing, p, 254 broadened the scope of services supported by EHIF from pure curative side to preventive ones. Dental care benefits included in the benefit package differ for children and adults. Since 2002, the EHIF has guaranteed dental care free of user charges to children and adolescents up to 19 years of age, including preventive and curative services. The cost of dental care for adults must be paid for out-of-pocket but is subject to partial reimbursement by the EHIF (in general, most people can expect to be reimbursed US$I1.5 per year). The reimbursement rate is higher for some population groups with greater needs, such as pregnant women, mothers in the first year after childbirth and people suffering from certain diseases that affect their need for dental care. The pharmaceuticals covered by the EHIF are defined by a positive list 88 where all medicines have to be prescribed by physician. During recent years, clearer guidelines were developed for adding new pharmaceuticals to the positive list and reference prices were introduced. In addition, efforts were made to introduce more generics on the market and to promote their utilization. This has been done with the objective of cost containment for the insurance, but also the aim has been to promote cheaper but same efficiency medicines for patient, still with mixed outcome (see the share of out-of-pocket payments under performance section). Medical devices for certain diseases are also included as in-kind benefits and are subject to a co-insurance rate of90% up to a yearly ceiling ofUS$1534. Prescription drugs are generally subject to a US$3.8 deductible, and a further co-insurance percentage usually applies (50% up to US$15.4). Beyond this ceiling, the user covers all costs. Additional measures aim at limiting the burden for some categories of patients: a positive list of drugs for chronic conditions 89 with 75% and 100% co-insurance rates and a lower deductible ofUS$I.6. Additional exemptions apply for children and retired people. Flat co-payments are charged for some type of health care services - primary care physician home visits, outpatient care visits, and hospital bed-days. The general principle has been to move to free primary care without any patient side cost sharing. Since 2002, the only co- payment that FP-s are allowed to charge is a US$3.8 home visit fee. Earlier small visit fee, US$0.6 was applied with some exemptions as retired, disabled persons and children. For outpatient specialist care, the consultation fee is US$3.8 and this is applied to office visits as well. These fees are defined as maximum, but the provider can decide the actual amount between zero and US$3.8. Providers are not allowed to charge visit fee from under 2 years old children and pregnant women. Inpatient care providers can charge patients a per diem rate for up to 10 days with a limit ofUS$2.2 per day. Children, pregnancy or delivery- related conditions and emergency care are exempt. These rules apply to care that is reimbursed by EHIF with the contracted providers. If costs of care are covered fully by patient (as those without insurance or paying for full service out-of-pocket), then providers are allowed to establish different user charges to cover the cost of treatment. The only requirement for these established charges is that they have to be "reasonable". Before 2002, 88 Developed by Ministry of Social Affairs and updated according to need. 89 List of conditions is approved by Government (available In Estonian at https:1Iwww.riigiteataja.ee/ert/act.jsp ?id= 1008084) Good Practice in Health Financing, p. 255 there were no such an explicit rules and different providers applied different user charges even in cases where EHIF was reimbursing the costs. The health care service list approved in Government also sets a maximum ceiling for co- insurance per type of service, which according to Health Insurance Act can be up to 50%. Most services have no co-insurance except for in vitro fertilization (30% co-insurance), abortion without medical indication (30%) and rehabilitation per diem for some illnesses (20%). 2.4. Performance of the Health Financing System in Terms of Efficiency and Equity In 2004, Estonia spent 5.5% ofGDP on health, which is relatively moderate. At the same time, the public share of total health expenditures remains high -76% of total health expenditures. This enables to ensure good financial protections for most of population. However, out-of pocket payments have been increasing from 13.7% in 1998 to 21.3% in 2004 and the financial protection is increasing concern. The share of households facing high health expenditures have increased, where in 1995 and 2002 respectively 0.3% and 1.6% of households faced health expenditures, which were higher than 40% of their budget after food needs were covered 9o • The main objective of health financing system is to improve the population's health status. As present health financing organization has been in place for 15 years, its impact to the population's health status is rather ambitious. Therefore, intermediate objectives as access to care and quality are more informative indicators of health financing performance. Access to care in terms of waiting times and number of treatment cases has been the most explicit and important objective of EHIF. The EHIF is legally obliged to balance yearly revenues and expenditures and until today this requirement has been fulfilled, except in 1999 when due to economic crisis and reduction on health insurance revenues running year expenditures exceeded revenues 91 . On one hand this is related to the EHIF's liberty to adjust contract volumes of providers according to diminished revenues (decreasing the volume), but as yet more important reason has rather been the rapidly increasing health insurance tax revenues (leading to increase in volume of the contract). In addition, according to the Health Insurance Fund Act, EHIF has to constitute adequate reserves to minimize the potential financial risks. EHIF has to have two types oflong term reserves -legal reserve and risk reserve. Risk reserve has to 2% and legal reserve 6% ofEHIF's budget. The constitution of reserves started in 2002 and already in 2003 the reserve requirements were fulfilled In terms of income inequalities Estonia is characterized as country where income is very unequally distributed: in 2004, the average income in the lowest decile was one tenth that in 90 Habicht J, Xu K, Couffinhal A, Kutzin J. 2006. "Detecting Changes in Financial Protection: Creating Evidence for Policy in Estonia." Health Policy and Planning. Forthcoming 91 The deficit was then covered by fund's own reserves and no extra allocations from state budget were made. Good Practice in Health Financing. p. 256 the highest decile. Health expenditures follow same pattern and probably the low level of out-of-pocket expenditure for the poorest individuals reflects financial barriers in access to care. For instance, the usual co-payment for a specialist visit (50 Estonian kroons, which is about US$4), which represents more than 6% of the monthly income in the lowest decile92. This assumption is su~ported by empirical research correlating health care utilization with socioeconomic status 3. Controlling for health status, lower-income individuals utilize fewer FP and specialist services as well as dental care. It is expected that pharmaceuticals follow the same pattern. Estonian health insurance coverage is mandatory and targeted to the whole population. Still, about 6% of population remains uninsured. As Estonian health insurance is closely related to labor market, the share of insured is dependent on employment level. Since 2000, the share of uninsured has decreased to some extent from 6.6% to 5.5% in 2004, which is mainly related to increased labor market participation where the unemployment rate has decreased over the same period from 13.8% to 9.9%. The uninsured are among working age population as children and pensioners are all covered. According to Household Budget Survey 2004 94 , the uninsured are mainly men (share of uninsured among men and women is 6% and 3% respectively) in age group 35 to 54. By socio-economic background uninsured are mainly with lower education and they live in rural areas. The uninsured tend to live also in the poorer municipalities, where the average income level of inhabitants is below country's average. 3. Overview of Health Care Delivery System The health care delivery system can be divided into primary care and specialist care. The natural parts of the primary care are well developed family medicine system, ambulance services and prescription medicines. On specialist care the out-patient and in-patient care mostly provided by hospitals is described below. In addition nursing homes, rehabilitation services and long tenn care development is currently in the early stage in Estonia, but has not been discussed below. 3.1. Primary Care - Family Medicine The introduction of the primary care started in 1991 to redesign the Soviet type of polyclinic model, where primary care was provided mainly in polyclinics. There were working medical school graduates without additional specialist training, pediatricians, and other specialists. The first step in primary care reform was to introduce the family medicine as medical 92 According to Income and Living Conditions Survey 2004, 7% of adult population did not get needed FP care because of economical reasons (mostly uninsured, who face out of pocket payments), long waiting times and long distance. In case of specialist and dental care these shares were accordingly 8% and 15%. The access barriers are bigger for lower socio-economic groups and those living in rural areas. 93 Habicht J, Kunst A. 2005. "Social inequalities in health care services utilization after eight years of health care reforms: a cross-sectional study of Estonia, 1999." Social Science & Medicine, 60:777-787 . 94 Statistical Office of Estonia. 2004. Household living niveau 2003. Tallinn, Statistical Office of Estonia. Special data query provided by Statistical Office. Good Practice in Health Financing. p. 257 specialty and to start post-graduate training on that in early nineties. The number of family physicians has increased covering the whole population in 2003 (see figure 10). Since 1997, all citizens were required to register with family physician patient lists and family physicians were entitled to become independent service providers. In the first year the enrollment was done with both empowering the doctors and citizens to do so (leading to written consent), and for doctors a motivation was the capitation fee defined by the size of the registered list. Those people not enrolled over time the family doctor was assigned (coordinated by country doctors) with the possibility of change the provider in the future. This was possible as critical amount of family physicians were available to scale up the system over Estonia. When 1996 only 8 percent of the population used family doctor 95 then in 2003, the country reached already full coverage. Current regulative framework became in force in 2002, which sets that primary care is the first level of contact 96 with the health system for most of the conditions. Direct access is maintained for the list of out-patient specialists, e.g. gynecologists. However, as all family physicians are required to work with at least one family nurse to be contracted by EHIF there is a shortage of family nurses (as the number in 2004 was below 700), who have better employment prospects abroad or in other industries 97 . This has set some limitation on further development of family medicine functions in Estonia. Family medicine includes a range of services as diagnosis, laboratory tests, investigations, treatment, follow-up, and also prevention and promotion activities. Further development in 2004 has been introduction of the family medicine call centre funded by EHIF to provide 2417 access to family doctor advice. This line is free of extra charge for whole population and only the regular call fee is applied. There were already more than 36 thousand calls to this line during the first quarter of year 2006. Figure 10. Number of family physicians, in 1993-2004. 1000 ~ - - - - 996 900 ~ - - 907 800+----- 861 700 +- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 600 ~ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ~a _6?~ _____ _ 500 T I - - - - - - - - - - - - - - - - - - 462 - - - - - - - - - - - - 400 ;.. - - - - - 404 300 ~- 290 o .-< N M .,. o o o o 0 0 g o o N N N N N Source: Ministry of Social Affairs 2006, www.sm.ee 95 Survey "Arstiabi kasutamine elanike poolt 1996" (Utilization of health care services by population. in Estonian) EMOR, Tallinn. 96 If specialist is accessed directly, patient has to bear the full cost of provided services (except some narrow specialties as selected infectious diseases, fertility and gynecology related cases, traumas etc). 97 Atun RA. 2004. Advisory Support to primary health care evaluation model: Estonia PHC evaluation project. Final Report. Copenhagen: World Health Organization Regional Office for Europe. Good Practice in Health Financing, p. 258 Family physicians can operate as sole proprietors or companies (general partnerships or limited partnerships) with a list of registered patients. Even there is an option to work purely on private basis and without EHIF contract, majority of family physicians have a contract with EHIF98. As mentioned above, each family physician has a list of registered patients, which cannot contain fewer than 1200 or more than 2000 patients (exemptions are allowed if county governor and EHIF agree otherwise in some specific regions). On average the practice size is around 1600 persons in 2005. The practices are organized around one family doctor, but there has been trend over past years on the creation of the group practices for efficiency gains and better use of available resources (equipment and human resources). The patient is free to choose and change a family physician. In 2005, only 13% percent of the population had changed their doctors and main reason was the change in the living place. The family physician can refuse to register a person only if the practice list size has reached the maximum number (2000 persons) and if person does not permanently reside in the practice service area. Family physicians are required to have at least 20 visiting hours a week, and practices should be open for at least 8 hours a day99. Patients should be able to see their family physician within one day for acute problems and within three days with chronic conditions. In reality over last years around half of the persons requested appointment can access to the family doctor in the same day (regardless of the conditions) and majority will have the visit during the first three day. The access and services available at the family medicine level has been resulted also on the increasing satisfaction with the family medicine system over the years - in 1999 about 80% and in 2005 about 90% of population was satisfied with family medicine system 100, The family physicians reimbursement system is designed to provide incentives to orient more to preventive care, as well to take more responsibility for diagnostic services and treatment. Therefore, mix of payment methods are used (capitation, fee-for-service, lump- sum, etc) to balance the financial risk associated with different incentives (see chapter 2 for detailed description). 3.2. Primary Care - Ambulance Services The ambulance services organization in Estonia were inherited from the previous system with strong emphasis to pre-hospital care as in many Eastern European countries. This is the mixed model, where ambulance has both roles: to diagnose and to provide on site treatment to the patient and to provide transportation to the hospital. Part of ambulance teams includes also doctors and currently 40% teams are leaded by doctor and other 60% by specialized 98 IfFP has no contract with EHIF, patient bears the full cost of service. 99 Monitored by trustee doctors - a special system developed by EHIF to perform quality control (visiting practices, reviewing clinical work, aUditing, access checks etc) 100 Annual health care access, utilization and population satisfaction surveys 1999-2005 (by EMOR or Faktum), commissioned by Estonian Health Insurance Fund and/or Ministry of Social Affairs Good Practice in Health Financing, p. 259 nurse. Thus, in many cases the service is provided at the place and further transportation to the hospital is not needed. As provider organizations 80% of ambulance service providers belong to other institutions (e.g. hospitals) and 20% are specialized to emergency care. The ambulance services have went through several stages in terms of financing schemes and services. In the early nineties, until 1996, ambulance services were financed by regional sickness funds with some additional transfers from the state budget. This defined the entitlement to the services by the insurance coverage. The basis of the payment was the bills sent by the providers for services provided. In 1997, the financing was centralized to the Central Sickness Fund, but coverage was still related to the insurance status. However, centralization removed regional differences in service financing practices. In addition, the incentive to provide services were changed as the remuneration of each visit and became related to the preparedness of the emergency care teams consisting both doctors and nurses. The general budget is used and depends on number of teams and is adjusted to the compositions of the teams. Since 1998, the ambulance services have been financed from the State budget (managed by the Ministry of Social Affairs). This ensured that all people in Estonia (both citizens and those temporary) are entitled for the services according to need, as before access to ambulances were related to insurance status. Even if the number of teams (90 teams in 2005) and cars (126 in 2005) has been slightly decreasing over the years, the service availability has not decreased considerably (as per annual visits). In late nineties 230 thousand visits were performed and in 2005 in was around 240 thousand. Since 2002, additional change in purchasing ambulance services took place due to the establishment of Health Care Board (agency under Ministry), when before the resource allocation was performed directly by Ministry. One of the responsibilities of Health Care Board is to purchase and monitor ambulance services in Estonia. It is evident that ambulance services and family medicine balance and substitute each other to some extent to provide comprehensive access to the servIces. 3.3. Out-Patient and Hospital Specialist Care Since beginning of nineties, Estonian specialist health care has undergone broad reforms including centralization of highly specialized services and decentralization of outpatient specialist services. Currently, outpatient specialist care is provided by health centers, hospital outpatient departments and specialists practicing independently. Both public and private specialists can be contracted by EHIF. Estonia inherited a significant overcapacity of hospital beds from 50 years of Soviet era. In early nineties when health insurance system was established, the clear purchaser-provider split was introduced. The hospital sector was restructured and lower level or quality facilities were closed (see figure 12) in the first half of nineties. This was implemented setting quality standards and introducing licensing system. Still, many aspects remained unsolved in the changing regulatory framework and therefore hospital sector reform was re- initiated in late 1990s, when the Ministry of Social Affairs commissioned the development Good Practice in Health Financing, p. 260 of the Hospital Master Plan 2015 101 to make projections about future hospital capacity. The plan predicted the need to reduce the number of acute inpatient beds by two thirds and to concentrate acute inpatient care in 15 larger hospitals, decreasing the total number of hospitals, through mergers and other types of restructuring, by three quarters (from 68 to 15) by 2015. The plan was to decrease number of acute-care beds considerably and to develop system of rehabilitation and long-term care. Criteria used for planning hospital capacity included sufficient population pools to support minimum service volume for quality and efficiency, development of medical technology, demographic and epidemiological projections and a requirement that a hospital should not be further away than 60 minutes travel time by car (70 km). The recent study using data from late nineties has shown that the barriers to access to care was lower for around 3.4% of the population living currently farer than 30 minutes from acute care hospitals 102. Even variations exists the hospital care is well accessible for different population groups in rural and urban areas (see figure 11). Further reforms would influence one quarter of the population when centralizing high technology and services. At the same time access to services will decrease for 10% of popUlation, but can be compensated with the development of out-patient care and strengthening the primary care already available. 101. The HMP 2015 was done by Swedish consultants in 1999 and published in April 2000 (HiT 2004). Criteria used for planning hospital capacity included sufficient population pools to support minimum service volume for quality and efficiency, development of medical technology, demographic and epidemiological projections and a requirement that a hospital should not be further away than 60 min travel time by car (70 km). The plan has been used as basic planning exercise developed further in 2002 and as second wave currently in 2006. 102 Roovali L, Kiivet R.A. 2006. "Geographical variations in hospital use in Estonia." Health and Place, 12(2), 195-202 Good Practice in Health Financing, p. 261 Figure 11. The proportions of respondents (age 25-74) in 1999, who have had telephone consultation with a doctor, visit to a doctor, visit to a specialist, visit to a dentist (all during last 6 months) or have been hospitalized during last 12 months, according to residence. ~ 50 ! 45 o:r 40 " In I 35 N 30 -g 01 25 ... : C 20 15 GI 'tI 10 C o 5 a. 2! o ... o OJ c: c: .2 tJ 0 cot) 0= ..... co co ..... .8~ c: o p 0 ..... 0 .... 'u ..... c: co ..c19-o 'iii Ql 'iii Ql ~ g. ~ co .- 0. > !II 5"0 .l9 Qjc:.c: '0. f-8j !II o J: i • Tallinn. Other urban area 0 Rural area 103 Source: Kunst, A E. et al. 2002 This hospital master plan was also important input to the development of new Health Service Organization Act 104 (2001), which elucidated the legal environment of service provision (hospitals but also all service providers in the health care sphere). According to that, all hospitals (incl. public) have to operate under private law as foundations or joint- stock companies. This means, that even when hospital's ownership remains to the public sector (to central or local government), they are ran as private companies full managerial rights over assets, residual claimant status and access to financial markets. Hospitals are governed by Supervisory Boards, which members named by the owners or founders. All hospitals have to be licensed by Health Care Board, where license is issued for 5 years if minimum standards are met. Estonia has been very successful in re-organizing hospital sector to increase the efficiency and responsiveness of a system. In 1990, Estonia had about 120 hospitals with about 14 000 acute care beds. Since then, the number of hospitals and the number of beds have fallen dramatically. By 1998 the number of acute care beds had fallen to about 8600 and by 2003 103 Kunst, A. E., Leinsalu, M., Kasmel, A, & Habicht, 1. (2002). Social inequalities in health in Estonia. Main Report. Tallinn:Estonian Ministry of Social Affairs. 104 The regulation covering the whole health care service provision in Estonia (previous legislation from mid- nineties with updates available until this) from primary care to hospital services. In addition, the patient- provider relations are regulated by Obligations Act and provider-purchaser by insurance regulations and contracts. Good Practice in Health Financing, p. 262 to about 6000 acute beds. The reduction of acute beds has been related to the establishment of hospital licensing system where small hospitals providing predominantly long-term care lost their acute care status and were turned into nursing homes. Other hospitals have been turned into primary care centers providing out-patient care. In terms of hospital closure, the access to care measured by acute care hospital admissions, has remained same (see figure 13). In recent years, reduction in the number of acute beds has been due to hospital mergers. The average length of stay has reduced form 15 days in 14 days in 1990 to 6.4 days in 2003, but the bed occupancy rate remains pretty low - 68% in 2003 compared to 77% in 1994. Figure 12. Number of hospitals and acute care hospital admissions, in 1985-2003. 140 - 25 VI c o 'iii - 20 .!!! E 'tI III - 15 iii 0 .... 0 'ii. ... .. C GI 80 • l 10 VI .. ,g ~ 60 .Q E ::::I .. GI Z 40 .Q E ::::I 20 t +--------- -'-~'. ----r---+--"+-~-_j ----"--j--- +-~f L 0 Z ~~~OO~O~NMV~~~OO~O~NMV oooooooooo~~~~~~~~~~ooooo ~~~~~~~~~~~~~~~ooooo ~~~MMMMMMMMMMMMNNNNN ~ Number of hospitals _ Number of acute care hospital -Acute care hospital admissions per 100 Source: Ministry of Social Affairs 2006, www.sm.ee One of the most critical aspects of Estonian health system is the human resource planning, and the quantity and quality of health care professionals. The entire health workforce is independent and not part of public service. This means that doctors and nurses are free to provide their services on contractual basis to health care providers (e.g. hospitals). From one hand this has provided motivations for professionals, but from other hand made the long term planning more difficult. Good Practice in Health Financing. p. 263 Figure 13. Number of doctors and nurses per 100 000 inhabitants, in 1998-2004. ~~~L~-- .,,~ ~~ 200 --- ~ 00 m 0 .-< N M V m m 0 0 0 0 0 m m 0 0 0 0 0 .-< .-< N N N N N __ Number of doctors per 100 000 inhabitants --- Number of nurses per 100 000 inhabitants Source: Ministry of social Affairs 2006, www.sm.ee When health care reforms started in the early 1990s, it was assumed that there was an oversupply of doctors, particularly in certain specialties. At the same time, there was - and still is - a shortage of nursing personnel and an uneven distribution of specialist services around the country (see figure 12). After independence, underinvestment in health facilities and human resources was a major source of cost savings, resulting in relatively low salaries and poor morale among doctors and nurses until last years. During the last years more attention has been paid to salary increases and this has improved the motivation of medical staff. More recently, the prospect of the free movement of medical professionals within the European Union has put further pressure on this part of the health system. This has been reflected in the high number (for more than 300 doctors and around 150 nurses in 2004) of the documents prepared by health care board to work outside in Estonia, however the number for both nurses and doctors have declined four times for 2006. In addition the perception of professionals have chanced toward working short term aboard and not so much moving for several years. At the same time the outflow has been around 4% of doctors and 1% of nurses over the period 2004-2005. 3.4. Pharmaceuticals The pharmaceutical sector was reformed during the 1990s with the aims of establishing drug regulatory authorities, creating a modem legislative framework, introducing a system for reimbursing drugs and privatizing pharmaceutical services (e.g. pharmacies). During the eighties the choice of the pharmaceuticals was lagging behind of the evidence based practice by then to twenty years and only a few patients had the access to contemporary medication, but instead of medical need, due to their hierarchical position. The new scheme of medicines reimbursement was introduced in 1993 and since then the principles of reimbursement Good Practice in Health Financing. p. 264 category is determined by the severity of disease, efficacy of medication and social status (ability to pay) ofthe patient. Whereas the lack of effective medication was the main issue until 1992, the increase of medicines cost has become a major driving force for reforms nowadays (see figure 14). The increase of medicines cost has three main reasons. Firstly, the amount of medication used in Estonia has increased (budget numbers over three times during the nineties). Secondly, older drugs are replaced by medication, which is more effective, but more expensive. And third, new pharmaceuticals have been introduced for the treatment of diseases, which had not been used for treatment before or the existing treatments was not available in Estonia. 105 Figure 14. Cumulative increase in EHIF's pharmaceutical and health care services expenditures in 1993-2006. 1900 + 1700---------------------------------------- cO' 1500 ,_ 0 ~ -- OJ," 1300 III"" la I!! en I en 1100 u ... -+ c ...... 900 - ' ; III ,~ I!! 7001 - - - - .... :::J S~ :::J'a 500 - - - - - - - - - - - - EC I :::J OJ 300 -,- v~ OJ 100~~--------,-----~----------------~ ~ ~ 00 ~ ~ ~ 0 0 ~ 0 N 0 M 0 0 ~ V * * ~ ~ ~ ~ 0 0 0 000 0 ~ ~ ~ N N N N N 0 0 N N -*- Pharmaceuticals __ Health care services Source: EHIF 2006, www.haigekassa.ee *Forecastfor years 2005 and 2006 The last two qualitative changes have increased the coverage with medicines and allow the prescribing doctor (both family doctor and specialist) to use more effective and safe medicines. In volume terms, the amount of prescription pharmaceuticals used per capita has doubled during the nineties (but still not at the level of developed EU countries). This was increased to the level of 314 DDDIl OOO/day in 1994, and continued to increase to the level of 574 in 1999 and 754 in 2004. In monetary terms from 1999 to 2004 the pharmaceutical market has increased 1.26 times. Examples of important changes in the patterns of medicine use in Estonia include ulcer diseases, hypertension, depression and inflammatory joint diseases, in which cases the use of medicines is targeting more the chronic conditions in need of continuous medication. 105 Kiivet R, Harro J.(Ed.). 2002. Health in Estonia 1991-2000. Tartu: University ofTartu Good Practice in Health Financing. p. 265 Over the years the consumption of medicines has increased on average, but in regard to the different pharmaceuticals used the practice has changed as well. This has been influenced both by reducing to utilization of older drugs, but also increasing the availability of medicines not available in eighties. Looking only to the treatment of hypertension (see figure 15) we can observe that there is move towards evidence based prescription where ACE inhibitors are used more widely and has reached almost to the European average level, in case of other evidence based pharmaceuticals utilization has at least doubled. The changes in general was driven by opening the market to new medicines, targeted reimbursement system through positive list applied by health insurance system, sharing information on evidence based practice with service providers and pUblic. Figure 15. Utilization of cardiovascular medicines by groups in 1994-2005 250 200 > III 'tI ~ :g, 150 ...... o o o pj 100 ...... Q ~ Q 50 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 o Antihypertensives • Diuretics o Beta bloc king agents o Calcium channel blockers o ACE inhibitors • Angiotensin II antagonists • HMG CoA inhibitors D Fibrates - - - - - - _.. ---_._--- Source: State Agency of Medicines, Annual statistics on medicines 2006, www.sam.ee Availability of medicines with known quality and proven efficacy in the pharmaceutical market has been thus achieved and access for the public to prescription medicines is supported by the functioning reimbursement system. Under the Ministry of Social Affairs the State Agenoy of Medicines is fully responsible for the control of all pharmaceutical activities, including veterinary products - for registration, quality control and other activities. The reimbursement policies have been leaded by the Ministry, where until 2003 the main driver of the new initiatives was EHIF and since 2003 the Ministry with its special policy unit for medicines. For the prescription medicines the reimbursement goes up to either 90% or 100% for medicines assigned to lists of more serious diseases. Over-the-counter pharmaceuticals, vitamins and mineral supplements, herbal remedies, etc. are not reimbursed. On the other Good Practice in Health Financing, p. 266 hand, all phannaceuticals used in hospitals are free for the patients, similar to other inpatient services for insured persons. The first essential drug list for Estonia was developed in 1992 and since that time it was used as a guide for medicines donations. Later it was used by hospitals to develop their own medicines fonnularies and also to create the positive lists of the medicines reimbursement system for outpatient use of phannaceuticals. At hospital all the medicines are free for the patient as covered by the EHIF. 4. Refonns for Health Coverage Estonia has been going through extensive health care refonns since Estonian independence was regained in 1991. In Estonia the increase of the coverage with the insurance and health care services has been one of the implicit objectives through past fifteen years after independence. There have been two goals against what the refonns can be evaluated: first, to ensure access to high quality services, and second, to provide financial protection (both in tenns of providing insurance coverage with broad benefit package and protecting against financial risk from out-of-pocket payments). Figure16. Milestones of health sector reform in Estonia. Whole country Health Care Health Insurance covered with Establishment of Organization Act, Fund Act, Health Health effective health insurance licensing of Beginning of Hospital Master Service Insurance primary care system providers primary care reform Plan 2015 Organization Act Act svstem I I I I I I I ) 1992 1994 1997 2000 2001 2002 2003 Chronologically, Estonian health care refonns can be divided into four phases - early nineties; mid-nineties; late nineties and early twenty first century; and current system. In this chapter the health care refonns over the last 15 years are described chronologically and their impact to the health coverage in tenns ofthese goals is described. Main milestones of the refonns are given chronologically in figure 16. 4.1. Early Nineties - Establishing Social Health Insurance System to Ensure Sustainable Financing During the eighties, when Estonia was part of the Soviet Union, there was virtually universal coverage with health care services in tenns of breath of the coverage as everybody was eligible for health care services. In reality, the depth of coverage varied among population groups and some groups had better access to services. Although, services were developed Good Practice in Health Financing. p. 267 more in some areas (e.g. maternal and child health), but in others there was difference in use of modem technology or clinical methods for treatment there was difference compared to Western European countries. The first health care reform was started to be prepared even before Estonia became officially independent. Health financing reform was the first fundamental reform with aim to build up health insurance system to secure sustainable financing for health care sector. The intention of the health financing reform was to shift away from the centralized integrated state model of Semashko to the decentralized model of social health insurance to protect health funds trough earmarked tax and to enhance the system efficiency and responsiveness. The preparation of health financing reform began in late eighties when changes were foreseen and need to establish functioning health system emerged. The first Health Insurance Act was approved by Parliament even before the political independence was achieved. The professional associations (Estonian Medical Association representing majority of doctors) had a significant role in successful implementation of health insurance reform. They saw in insurance system the possibility to ensure sustainable funding for medical care in new economic environment. Thus, their support was very important in streamlining the reform. Another important factor of successful reform was the consensus and commitment of political parties to the system due to the regained independence and to the political stability between 1993 and 1995. The health insurance system was designed on regional basis where 22 non-competing sickness funds were established. This was one part of the plan to decentralize health care system. The sickness funds were organized by County or City Government. Government's authority was to approve the statute of sickness funds and rules of health care benefits calculations. The Ministry of Social Affairs had a supervisory role over the health insurance system. In order to enhance national coordination, the Government set up an Association of Sickness Funds. As mentioned above, one reason for establishing Bismarkian type of health insurance system was to secure certain revenue base for health care system. Another important rationale has been to relate health insurance closely to the labor market to give incentives to participate to formal labor market. Thus, health insurance tax was introduced with contribution rate of 13% paid fully by employers on employees salaries. There were other contribution rates for different kind of entrepreneurs. That 13% of contribution rate on salaries has remained same to this day, but due to changes in legal environment, these other types of entities are reorganized (e.g. farmers who are currently self-employed) and no separate taxation rates are used for them. At the beginning health insurance tax was a separate tax. In 1994 health insurance tax was incorporated to the social tax as earmarked share to the health care system. In early stages the regional sickness funds collected the health insurance tax and there was no central pooling or any risk-adjustment of funds. Therefore, more deprived areas had lower revenues and this affected adversely their access to care. This was recognized as one Good Practice in Health Financing, p. 268 important shortcoming of this kind of organization of health insurance system and one of the reasons to introduce administrative changes (see below). The established health insurance system was mandatory without opt-out possibility, but it was allowed to take supplementary private insurance to cover additional expenditures that are not covered by mandatory health insurance. At the beginning, the health insurance coverage was almost universal as most of population groups were eligible to health insurance, including persons paying health insurance contributions and their family members, pensioners and other groups. Out-of-pocket payments for health care services were almost non-existing during Soviet era. Formal out-of-pocket payments were not introduced at the beginning. In 1993, Estonia introduced the prescriPotion pharmaceuticals reimbursement system, which based on positive-list principle 06 and limited cost-sharing. The introduction of purchaser-provider split was seen as one tool to facilitate the downsizing the capacity of providers' network. Sickness funds were contracting to service providers to ensure necessary care to insured persons in their region. The payment to providers took place according to the price list approved by Ministry of Health (afterwards Ministry of Social Affairs). The pricelist was developed based on the German price list, where tariff proportions of different services were taken over and adjusted into local prices. Before 1995 co-payments were 0.38 US dollars per prescription for medicines in the positive list with co-insurance rate of 100% (both defined by disease groups and pharmaceutical names) and for popUlation groups as disabled persons, children under 3 years and aged over 70 years. Similar co-payments were applied for medicines with co- insurance rate of 90%, and slightly higher co-payment for prescription medicines with co- insurance rate of 50%. Such overall policy resulted with the small share of the private expenditures in early nineties and good access to the medicines. In conclusion of the first wave we may say that it sets a basis for the functioning health system by establishing sustainable and fully operational health financing system being inevitable precondition for the following reforms. Health insurance system was crucial in terms of ensuring the breadth and depth of coverage after Soviet system was collapsing and Estonia started to build up social security system. Important change while expanding the depth of coverage was the introduction of new reimbursement system for prescription medicines, which made medicines available with limited out-of-pocket payments. Another important change was the introduction of purchaser-provider split to have transparent contractual arrangements to achieve more efficient use of resources. 106 Positive list included a list of pharmaceuticals reimbursed at different levels of co-payments (with small deducible). For development ofthe list several characteristics were taken into account as the different groups of pharmaceuticals (starting from essential), age groups of the population, people with special health needs as disabled persons. Over the time the list has become more comprehensive. Good Practice in Health Financing, p. 269 4.2. Mid-Nineties - Decentralization of Providers' Network to Ensure Access to Modern and High Quality Health Care Services The second phase of reforms can be mainly related to the reorganization of providers' network. After re-independence and introduction of purchaser-provider split, health care organizations were gaining more autonomy than they had in Semashko type of system and this in turn gave more incentives to start restructuring the providers' network to increase system responsiveness and efficiency. In 1994, Health Service Organization Act came into force. By this legislation the health service planning function was largely delegated to the municipalities' level to decrease the central government role. Also, the providers' licensing system was enhanced, which was important pre-condition for decreasing hospital network capacity and to ensure quality. The licenses were provided by the Ministry (actually already in the 1991) for the 6 months to 5 years period were criteria's (e.g. attestation of professionals, safety, and agreement from municipality Government) were applied. In 1994 the detailed revision of all providers were performed leading to closure of providers not reaching the criteria. However, the introduced legislation was not very comprehensive while not supporting the supervision and accountability to be developed. In line with hospital sector reorganization primary health care reform was initiated. The first step in primary health care reform was to introduce the family medicine as separate medical specialty (in 1993) and to start post-graduate training in the University of Tartu, the only university providing training for future doctors. Re-specialization training was introduced already in 1991. Even the basic framework for family medicine was set up in Health Service Organization Act, more detailed regulation Family Practice Act was developed in 1994. Due to the non-favoring environment this act remained to the draft level and reform slowed down as there were no appropriate incentives. The primary care reform plan was re-initiated in 1997 by ministerial decree (see next chapter). There were administrative changes going on also in health insurance system. Based on the first years of experience, it was clear that fully decentralized sickness funds system is too fragmented and there is need for central coordination. Another weakness was the fragmented revenue collection without central pooling and risk-adjustments, which in turn make for widening inequalities between regions. Therefore, the Central Sickness Fund was established in 1994 to strengthen central functions such as planning, redistribution of revenues between regions to provide regional equity, and control of financial resources. After this change, regional sickness funds were directly subordinated by the Central Sickness Fund and collected funds were pooled centrally and re-allocated to the regions on capitation basis. This also led to the reduction of regional sickness funds to 17 aiming for more efficient use of resources. Jointly with introduction of Central Sickness Fund, the State Health Insurance Council (SHIC) was established. SHIC consisted of 15 members - one representative of Ministry of Social Affairs, one representative of State Social Insurance Board, one county doctor, one municipal doctor, one representative of providers, five representatives of employers' unions and five representatives of insured's unions. The mandate of members was for three years and they were nominated by representing organizations. SHIC main responsibilities were to Good Practice in Health Financing, p. 270 approve the state health insurance budget and to develop price list for health care services. The role of SHIC was mainly advisory. Concurrently with establishment of SHIC, also Regional Health Insurance Councils were established to each regional sickness fund. Even their role was advisory, they had remarkable influence on contracting process. In addition, there were made some changes in eligibility criteria for health insurance, by defining more precisely the groups eligible for health insurance without contributions. As a result some population groups in working-age population remained uninsured 107. The motivation behind this change was to give clear incentives to participate in legal labor market and to decrease the share of informal payments, which were a serious concern due to the ongoing privatization and other radical changes in economic environment. Another important amendment of Health Insurance Act was that Sickness Funds can take financial responsibilities only in terms of available resources. During the first years this provision was not so important, as there were plenty of resources available and it was more question of providers' capacity to provide enough services. But it was expected, that in the near future the providers capacity raises and health insurance system may not be able to cover all expenses. This principle - balanced revenues and expenditures - has been very strictly followed by fund's managers since 1994 and only once (in 1999 due to economic crisis, which decreased the expected revenues of health insurance system) running year expenditures were higher than revenues and deficit was covered by using sickness fund reserves. During the nineties this balanced budget principle was mainly followed due to the fund's own willingness in the health sector to show its independence from state budget and ability to function autonomously. Of course the general fiscal policy has always supported the balanced budget in all sectors from early nineties. Currently this is supported by strict regulation and general conservative fiscal policy, which disapproves deficit in public sector. In 1995, patient co-payments for primary care and specialist visits were introduced. Introduced co-payments were very low (5 Estonian kroons, which is less than half a dollar) and wide range of population groups were exempted as children, pensioners. The main objectives of introducing the co-payments were to decrease informal payments, which were common during Soviet era, and to rationalize health care service utilization by eliminating unnecessary visits. However, this was politically very sensitive decision as people were used to free medical care and their willingness, but also ability to pay for medical care was very low. In addition, after 1995 the cost sharing for pharmaceuticals was increased by raising the co-payments and adding ceilings on benefits covered by EHIF. The increase of pharmaceuticals' co-payments continued, being modest in the mid-nineties, but more rapid in late nineties and in current century. The second wave of the reforms ensured further development of the service provision by enhancement of primary health care system by initiating family medicine reform and by separation of ambulance services from health insurance system. Latter ensured access to ambulance care to the whole population. In addition, to ensure the quality of the provider network the clear licensing system was created that resulted to the closure of the providers 107 Data on coverage from this period are not available. Good Practice in Health Financing, p. 271 not reaching to quality standards measured. As a result, the number of hospitals decreased enabling more efficient use of resources and securing the service quality. 4.3. Late Nineties and Early This Century - Re-Centralization to Enhance the Depth of Coverage After re-independence, Estonian health care system was widely decentralized, where most of functions were moved to municipality and county level. The weakness of implemented refonns was the lack of preparation, in tenns of staff training, accountability procedures and guidelines for policy sustainability. In health insurance system some re-centralization was going on already in 1994, when Central Sickness Fund was established. But it was evident that the providers' network restructuring needs some enforcement. This was partly related to the decentralization of providers related planning functions to the municipalities' level who failed on this where mainly protecting the interests of local providers rather than targeting system level efficiency and accountability. Also, municipalities were very fragmented administrative units lacking revenue base and competences. Therefore, it became clear that some functions should be re-centralized and the legal status of providers has to be clearly established, and Ministry of Social Affairs has to take the stewards role in planning providers' network. The important milestone of the third refonn phase was the development of the Hospital Master Plan 2015 to make projections about future hospital capacity commissioned by Ministry of Social Affairs. This was a necessary accelerator for continuing hospital sector refonn and to set clear targets on that (see in detail in chapter 3). Until 1998, primary care was mostly provided in polyclinics and ambulatories, owned by the municipalities, and by a few private providers. In 1998, some primary care planning functions were re-centralized from municipality level to the county level. To foster the primary care refonns, Ministry of Social Affairs introduced the primary care refonn plan in 1997 with main objective to cover whole population with family physicians services to 2003. Ensuring access to primary care has been seen as important precondition for centralizing specialist care and downsizing hospital network capacity. All Estonian inhabitants, insured and uninsured persons, should be registered to the family physician. New mix of different payment methods for family physicians was introduced (see chapter 3). To give additional incentives for retraining special fee (about US$80) to doctors who were trained as family medicine specialist and certified as family physicians was introduced. Overall the coverage (currently universal coverage) with the ambulance services has been increasing over the years as entitlement broadened to whole population and at the same time the quality of the services increased, even with the smaller number of teams available. This has been related to the overall development of the health care sector where centralization of the hospital services and creating new possibilities to have high quality services with doctors and nurses available at the primary care level. Good Practice in Health Financing, p. 272 In the meantime there were made some administrative changes in health insurance system. In 1999, the social tax collection function was removed to Tax Agency and collected tax is transferred to the health insurance fund. This allowed health insurance fund to concentrate to the purchasing function as due to the changes in the legislative environment the there were no right for insurance fund to control tax payments and the function worked through tax offices. This change has made the collection of social tax more effective (in terms of organizational structures in public sector) and allowed EHIF to focus more on the purchasing function. In addition there was observed increase of tax revenues collection generally after such merging. Over the decade, as Estonia's own legal system developed, the legal status of sickness funds became ambiguously defined, having some features of public independent legal person and some of a state agency related to the Ministry of Social Affairs. In 2001 EHIF was established (through special legislation) as public independent legal body with seven regional departments. This fundamental change enabled also to clarify the roles of central and regional departments. At that time also the legal status of service providers were defined more precisely and clarified the relations between purchaser and providers. In 2003, seven regional departments were merged to four departments as a natural step of centralization, each covering 200 000 to 500 000 insured persons. In 2002, new Health Insurance Act was introduced, which clarified the regulation of all aspects of the health insurance system, including benefits, reimbursement lists and levels for health services and drugs, maximum levels of cost sharing for insured people and contractual relationships between the EHIF and providers (current system described in detail in chapter 2). Most important feature of the third wave of health coverage reforms has been the arranging the overall regulative framework for health financing system and providers network. In terms of breadth of coverage there were changes made in eligibility criteria for health insurance and some population groups were excluded. However, the share of uninsured population has not increased due to increasing employment rates in the country. The third phase can be characterized with growing out-of-pocket expenditures for pharmaceuticals and dental care. Latter was excluded from benefit package of adult population and replaced with low yearly monetary compensation. In addition, maximum ceilings for patient co- payments were set for specialist visits and hospital stays keeping FP visits free of charge. However, latter applies only for insured population and uninsured have to cove all costs out- of-pocket if not emergency care. 4.4. Further Directions and Challenges The most fundamental changes to build system with wide coverage in Estonian health care system were made during the early stages of reform in early nineties. The further incremental arrangements were made to support the primary health care and hospital sector reforms, and to strengthen the purchasing function ofEHIF. Currently, there is no need for Good Practice in Health Financing. p. 273 fundamental changes, but rather some fine tuning of a system to improve the system performance. One continuous discussion obj ect is the relatively low level of health expenditures of GD P. This has raised the question, which could be the potential sources for additional funding. In long term the financial sustainability of health system financing is a big concern as aging population will adversely affect the contributors and non-contributors share in health insurance system. In terms of public financing the broadening of health insurance revenue base (taxing other incomes in addition to salaries) is one alternative considered. Second alternative is to urge local municipalities to increase their financing level by increasing their responsibilities, e.g. health care services for uninsured. Both alternatives need strong political commitment and therefore it is doubtful that in the near future radical changes will be made. The only decision that has been made in 2006 is to increase the state's contribution rate, which concerns 4% of insured persons. However, this gives only marginal budget increase (less than 2%) for insurance fund as related to minimum and not average salary in country. Another discussion object is to increase private financing by making more favorable conditions for private insurance or by increasing out-of-pocket payments. Latter have been rising over the time, which has been driven by the intention to rationalize health care services and pharmaceuticals utilization to exercise macro level cost-containment on public funds. Therefore, when rising out-of-pocket payments, its impact on different social groups should be evaluated, where current evidence shows that access to medicines especially for those with chronic conditions might be the highest problem. Even currently some groups face high out-of-pocket expenditures and the possibility to introduce targeted exemptions could be considered. In addition, supply side measures as prescription budgets, active feedback to doctors, rational prescribing training programs have been discussed, but without strong commitment to implement as there has been high yearly increase of the funding and therefore the pressure to contain costs is low. Most vulnerable population group in terms of high out-of-pocket health expenditures are uninsured. Currently there is a policy that eligibility to insurance for the working-age population should be related to the active labor market policy measures and health insurance is seen as one incentive to become active in labor market. This has been driven from the Government strong policy to reduce the unemployment where all sectors are actively involved, including health sector. Therefore, it is doubtful that universal coverage will be introduced soon even the topic is higher in discussions than ever before. In terms of service delivery system the biggest challenges for the coming years would be fine tuning through the system. The strengths of the developed delivery system are the strong family medicine centered primary health care that provides coverage with wide range of the first level services without co-payments and only with minimum waiting time. This is complemented with the ambulance services available out of the working hours and weekends, even slight overuse of this service can be counted still as the weakness. This is counted as one fourth of the visits, where large share of such visits without need, followed Good Practice in Health Financing, p. 274 by cases treatable by family physicians or just transportation 108. Partially this is observed as increased activities when primary health care centers closed in evenings or weekends - for which now 2417 primary care call line was opened in late 2005. The difficulty would be to make the system more patients centered and to coordinate the care at primary care level with development of additional nursing and rehabilitation services. As an outcome this would enable to use available resources more efficiently and to ensure better access and quality of care. The continuity of care would be other challenge to have close to patient out-patient care and high quality hospital services without long queues. All this depends on large extent on human resources available as undersupply of nurses and external pressures due to migration (out of the sector and to neighboring countries) have been persisted over the last years. 5. Health Coverage Reforms - Lessons for Other Countries Over the last fifteen years health care coverage has improved in Estonia where health financing reforms have played key role as described in the previous chapters. In the following chapter key enabling economic factors, political system and institutional arrangements are briefly described as the more detailed overview is given in previous chapters. The rest of the chapter includes main lessons learned to improve health care coverage in Estonia. Estonian health financing reform in terms of establishing health insurance system, which ensures nearly universal coverage, has been successful. The society's commitment for radical changes, political consensus, and medical profession's support have been the most important enabling factors of health financing reform. In late eighties and early nineties the whole population was very open to changes and there was great willingness to move away from Soviet system to the market economy. Even if the public attention to changes in health system was low, the establishment of health insurance system was rather favored. Politicians, as well general public, were open to radical changes and in terms of establishing social health insurance system, there was a political consensus. Another vital factor for health insurance system development has been the general economic developmene o9 , which has assured the continuous growth of health insurance revenues and has supported the restructuring and up-scaling of health system. These factors have guaranteed stability and continuity for the longer period, which has been supported to build up institutionally sound social health insurance system in Estonia. Medical profession support in early stages of reform has been crucial. Medical profession saw on new health insurance system a sustainable way to finance health care as earmarked tax was seen as a certain yearly revenue source compared to budget allocations in Soviet times. Health insurance contribution rate (13% on salaries) was set by taking examples from other social health insurance systems, but calculated in a way to bring more revenues that was expected to be the actual need in early nineties. This assured enough resources available 108 First level emergency care organization. National Audit Office of Estonia, Tallinn 2004 (in Estonian) 109 GOP growth in real terms has been 7.1 in 2003,8.1 in 2004, and 10.5 in 2005. Good Practice in Health Financing, p. 275 for the first years of the changes, which secured the medical profession's satisfaction with new system. As Estonia was coming from Semashko type of centralized health care system, there was a very strong incentive for decentralization and during the first years of refonns health financing system as well as providers' network were decentralized. While the capacity to conduct all necessary functions varied considerably between regions, this lead to the re- centralization after some years of experiences. Already in 1994, the Central Sickness Fund was established to supervise over regional sickness funds. At the same time, the service delivery planning function was decentralized and the role of Ministry of Social Affairs decreased remarkable. After some years, when it was clear that regions are not able to organize and plan service delivery without central oversight, the Ministry had to start to enhance the steward's role. Over the last years the role of Ministry has been increasing, but still there is no clear vision of the steward's role and balancing the vision setting and direct control functions. Latter is especially important in the environment where both, service providers and medical profession, are not directly under the Government. Key lessons of Estonian health coverage refonns to low- and middle-income counties are summarized as follows. 1. Health system revenue collection should be in line with country's general fiscal policy and take into account labor market policies and trends. In Estonia health insurance system is closely related to the labor market as health insurance contributions are mainly made on salaries. At the same time the system has high degree of the solidarity as children, pensioners and some special groups (e.g. disabled persons; those in maternal leave) are insured and no contributions are made behalf of them. For the working-age population the eligibility criteria for health insurance are set in a way to give strong incentives to participate in a fonnallabor market. Therefore, a share of working-age population remains uninsured. The rationale behind that has been to support overall labor market policies, where earlier in nineties the objective has been to decrease the infonnallabor market and currently over last years the focus is to increase labor market participation. Such collection arrangement, related to overall Government policies, has been supporting privatization, labor market development and decreasing unemployment, but has also ensured stable resources basis for the health insurance system until now and allowed Estonian Health Insurance Fund to provide wide benefit package. On the other hand, in long run, when economy develops and population is ageing, only salary related contributions could become a threat to the financial sustainability of a health insurance system and therefore options to broaden the revenue base should be considered when system is designed. Thus, even the system has ensured sustainable funding in Estonia over the last 15 years its relations to the fiscal policies and labor market patterns should be continuously evaluated. 2. Fonnalization of the labor market and economy as part of the decreasing corruption has positive impact also to the health sector. Infonnal payments (in tenns of money of services) were common in Semashko type of health care system until early nineties in Estonia. When health care refonns started in early nineties, infonnal payments in health care were even Good Practice in Health Financing, p. 276 more serious concern. This was one of the reasons why co-payments for primary and specialist care were introduced in the middle of nineties to formalize the payments. Even with some drawbacks this has led to the common understanding from both patient and medical profession side and has been stipulated in clear regulation some years ago. However, lately the informal payments have been becoming an issue again, especially in some specialties. At the same time the patients and their relatives are not asserting publicly because of that, which shows that there is need for patient empowerment by informing them on their rights and by developing channels for informing about bribery cases. This has leaded to the conclusion that the formalization of economy or clear regulation is not enough to avoid invisible barriers to access to care due to informal payments and there is need to empower patient to raise the issue and build processes to tackle corruption. 3. Relying mainly on the public sources can ensure protection against financial risk but out- of-pocket payments play important and increasing role in the health financing. Estonian health financing system relies mostly on public sources, which enables to secure financial protection for majority of population. However, over the time out-of-pocket expenditures have been increasing as country and its health system is developing. For some extent this is related to the overall increase of incomes and changing consumption patterns. On the other hand out-of-pocket expenditures are used to ration the health care utilization and to optimize the use of scarce public funds. Still for some groups increasing out-of-pocket expenditures are a serious access barrier for necessary care, mainly for prescription medicines that have be proven to be difficult to control. Thus, even out-of-pocket payments are important source of revenues for the health sector in most of the countries, their impact on access to care by different population groups should be evaluated and monitored and if necessary, income or health condition related targeted exemptions should be introduced. 4. Strong and transparent health insurance system may lead to the de-prioritization from the health by other financial sources, thus there is need to engage different levels of the Government to the designing health financing system. When health insurance system was introduced in Estonia, it was expected that Central Government and local municipalities will continuously finance health care at the level they had done before. In practice they started to fund less health care sector as other sectors development has been higher priority. Currently, the majority of public funds come through health insurance system. Therefore, countries planning to introduce mandatory health insurance system have to keep in mind that for Central Government and local municipalities this may give signals to prioritize away form health sector and this may adversely affect long term sustainability of financing system. In addition, building on different revenue sources can ensure higher level of sustainability in the transition economies or where the demographic changes are taking place. 5. There is always need to adjust the roles and functions through the system in the changing environment to improve performance. This has taken place in relatively short time period in Estonia. The mandatory social health insurance system was built up as a decentralized one, having non-competing sickness funds at each region, which collected the health insurance tax and purchased services to their enrollees in the region. After one year it was obvious that having a population less than one and half million, such an organization is administratively non-effective. In addition, it was clear that lack of risk-adjustment mechanisms will lead to Good Practice in Health Financing, p. 277 the widening disparities between regions, especially due to different economic development and drawback in the early transition period. Therefore, Central Sickness Fund was established and regionally collected funds were started to pool centrally and thereafter allocated to the regions based on capitation. Hence, centralization on one hand has enabled to improve administrative efficiency, and on other hand supported central pooling of funds. Further centralization of some functions has continued, where client services and relations to insured persons have been developed further at regional basis and moved to be part of the internet based services (e.g. insured person can check the validity of its insurance using internet based banking services)). And even planning and overall monitoring is performed at central basis, the purchasing and service quality monitoring by trustee doctors is performed by regional branches of insurance fund. This type of functions need to be analyzed in each country and designed according to the health system with ability to apply structural changes when needed to improve performance of the insurance system. 6. Centralization has supported insurance system to develop stronger purchasing function and improve the efficiency of the resource use. Another important feature of centralization of health insurance system has been that it enabled to develop purchasing function, where one strong purchaser has much more market power than fragmented local level sickness funds over several providers. This has been also important enabling factor for restructuring providers' network, as there was need to merge and restructure hospital network and the EHIF's overall purchasing process has been one important facilitator of that. This process would have been more complicated having regional independent sickness funds, which all might have different local interest. 7. Development of the clear, transparent and obligations describing contracts has been basis to achieve cost containment, quality and access to care. As mentioned above, health insurance fund has been putting a lot of efforts on enhancement of purchasing function. One important component in addition to planning, need assessment and monitoring of that have been contracts, which contents has developed through the time to build basis for whole purchasing function. Even in some societies contracts are not always taken as serious agreements between parties, in Estonia these have been followed over time and thus constituted the basis for accountability. Currently, standard conditions of contracts are negotiated with providers' associations (not with single providers), which makes contracting process more efficient and transparent. Estonian experiences demonstrate that well contents contracts between insurer and providers with clear responsibilities on both sides are important tool for performance improvement in terms of access and quality of care. In addition, contracts have been very important cost-containment mechanism, as by contracts the minimum number of treatment cases and the upper limit of volume are agreed. 8. To achieve coverage with the services comprehensive providers' network should be developed. In Estonia this has taken place in all levels as family doctors centered primary care has been developed, ambulance services restructured, hospital sector restructured with development of high technology based services, and access to modem medicine provided. All this has taken place in sequenced way where health financing reforms have supported the development of the providers network and practicing new initiatives (as incentives for primary care). With the overall providers network development new balanced system has Good Practice in Health Financing, p. 278 been achieved providing coverage with wide range of services. Now the issues related to continuity of care, and integrating disease prevention and health promotion services into the current system has been raising. The on-going challenge has been that with step by step approach (e.g. providing new curricula's, staff training, new equipment and facilities, financial incentives) among different levels of care and networks the coverage can be significantly improved. 9. To support providers and health financing institutions to achieve coverage with services for population IT development has an important role. Estonian population is very much favoring IT technologies and this applies also for health sector. Earlier, in mid nineties, all bills from providers were transferred to health insurance fund on paper and further on magnetic data carriers. Currently, all health care providers are equipped with personal computers and electronic channels are used between EHIF and its partners (patients, employees, providers) to increase efficiency in data processing and faster outcomes. Today all transactions with health care providers and pharmacies are made trough electronic channels. In addition to more efficient transactions, electronic data transmitting enables to collect high quality data for detailed data analysis as benchmarking of providers and assessing service utilization by population groups. Good Practice in Health Financing, p. 279 Kyrgyz Republic Melitta Jakab Elina Manjieva Good Practice in Health Financing. p. 280 The purpose of this paper is to describe the reforms in the Kyrgyz Republic aiming to increase coverage of health care services through a comprehensive health system strengthening approach. The early transition period, with its economic decline and severe fiscal contraction, exposed similar problems in the Kyrgyz health system as in other transition economies of the FSU & CEE region. These problems included erosion in previously high levels of financial protection and coverage, inequitable distribution of public resources disproportionately favoring tertiary facilities in the capital city, inefficiently large service delivery sector, and quality problems. Over the past 15 years, Kyrgyzstan has been addressing these challenges by implementing a systemic approach to strengthening its health system including complex structural changes in health care financing and service delivery. This paper reviews the Kyrgyz health system reforms (Manas National Health care Reform Program 1996-2005) from the perspective of coverage. As the coverage problems and their causes were embedded in a broader web of performance problems and inadequacies of the inherited Soviet health system, a systemic approach was necessary. Improvement in coverage required (i) clarifying the role of the government and individuals in funding health services based on the actual fiscal space; and (ii) downsizing the inherited extensive service delivery infrastructure to achieve and re-channel efficiency gains from fixed costs to variable costs and reduce patient financial burden. The ensuing reforms encompassed the entire spectrum of health system with four main pillars: • Clarification of entitlements through the introduction of the State Guaranteed Benefit Package (SGBP); • Centralization of previously fragmented pooling arrangements in oblast level purchasing pools to allow greater cross-subsidization and enable rationalization across administrative boundaries; • Replacement of historical line-item budgeting linked to capacity with popUlation and output based provider payment mechanisms to give incentives for downsizing; • Restructuring of service delivery with strengthened primary health care and more efficient and streamlined hospital care. The Kyrgyz reforms are still under evaluation to understand their impact for financial protection. There is clear indication that the reforms are associated with declining financial burden of hospital care, especially for the poor, and increasing transparency. In addition, utilization of health care services has become more equal both for outpatient and for inpatient services. At the same time, seeking health services continues to place a sizeable financial burden on households nationwide due to the continued presence of large out-of- pocket payments. The paper is structured as follows. Section 1 provides an introduction and background to understand the economic, social and political context of the reforms. Section 2 overviews the health financing reforms in the country. Section 3 describes the system of service delivery. Section 4 synthesizes the linkages between the described reforms and coverage and presents some evidence for its impact. Section 5 concludes with lessons learnt. Good Practice in Health Financing, p. 281 1. Introduction and Background 110 1.1 Economic environment Kyrgyzstan is a former Soviet republic that became independent in 1991 following the dissolution of the USSR. It was one of the poorest republics in the Union and heavily depended on subsidies from Moscow. The loss of subsidies and the economic consequences of disintegration led to a painful transition process. Between 1991 and 1995, GDP declined by more than 50 percent. Although this was followed by strong economic growth in 1996 and 1997, it was not sufficient to make a positive impact on average standard of living of the majority of the population: GDP per capita measured in PPP is still below its level immediately prior to country's independence (Figure 1). Moreover, the impressive growth rates of 1996 and 1997 were followed by another downturn in 1998 largely attributed to the spillover effects of the Russian financial crisis. Since then, annual GDP growth rate has stabilized at around 5%. Figure 1: Key Economic Indicators, 1990 - 2004 GOP per capita, 1990-2004 G[P growth, 1990 - 2004 2~,------------------------------, 1M ... I ~ 10.0 co 2000 c I a, o .. '" e 1500 - c - - - - - - . - - -..-________- ___ ~ C co M B 1'5 ~o .: I~ -5.0 -:: 1000 ~ _____ ~._t.-II.__..__.I-I _____.--.. ca ! ::s :; c ~ c 0.. 500 I !! Q. 2.00% ii. 2.00% 1 - - - - - - - - - - - - _ _ _ _ j 0 Q. &. 0.00% +--~--_-_--__.__-~ '0 0.00% '0 Poorest 2nd 3rd 4th Richest .... Poorest 2nd 3rd 4th Richest ~ Source: HBS 2001, and KIHS 2004; WHO staff calculations Note: The population was grouped into five quintiles using consumption data and quintile data calculated by the National Statistical Committee. Each quintile group contains 20% of the population ranked from poorest to richest. Transparency The Single Payer system has improved transparency of the health system for the population by creating a clear system of benefits and entitlements through the SGBP and the co- payment policy. Previously, the lack of clarity about entitlements coupled with great pressure on providers to replace markedly declined public funding led to a wide-spread system of informal payments, in particular for hospital care. Most hospitalized patients had to pay at the time of hospitalization informally to physicians, for medicines and various other supplies. However, there was no system of what patients were required to contribute in-kind and in-cash and physicians exercised price-setting as well as social justice of how much to ask of whom. Various surveys indicate that patients found this system poorly transparent, made them vulnerable to financial demands at the time of illness, the cost of hospitalization was unpredictable, and not known in advance even at the time of hospitalization as expenditures added up during the course of a hospital stay. The reforms made progress in improving the transparency of entitlements although problems remain. The figure below shows the decomposition of out-of-pocket payments for hospitalization. The level of payment reflects the mean payment per hospitalization episode among those who were hospitalized in the public at least once in the year preceding the survey. The dark solid area is the level of expenditures made upon admission and for lab tests and these payments reflect a combination of formal and informal payment that cannot be separated based on this survey. The checkered area indicates payments for medicines, supplies, and personnel all of which constitute informal payment. The white area is the amount of payment made for food. Good Practice in Health Financing, p. 312 Figure 10: Mean Payment for Hospitalization Among Those Hospitalized in the Public Sector 2500 ~------------------------------------~ 2,072 1,995 2000 +--.-- 1500 o Food I!I Medicines, supplies, personnel i • Admission & lab 500 +.-11,-.-•.••-01-- Bishkek Late reform Early reform Source: HBS 200 I, KlHS 2004; WHO staff calculations Note: "Early refonn" includes Issyk-kul, Chui, Naryn and Talas; "Late refonn" includes ash, lalal-Abad, Batken. The checkered area (medicines, supplies, and personnel) is unequivocally infonnal payment as patients should receive these services free of charge against their co-payment. The dark blue area in 2000 is also largely infonnal payment, as there was no charge at that time for admission. In 2003, the dark blue area in early refonn oblasts if fonnal co-payment but in Bishkek and late refonn oblasts reflects a mix of co-payment and infonnal payment due to the varied timing of these policies. The following two main conclusions can be drawn from this graph: • The overall level of out-of-pocket payments made at the time of hospitalization grew at the slowest rate in early reform oblasts. This indicates some progress of the reforms in slowing the rate of growth in OOPs and the resultant burden on households. • Informal payment increased in areas where the single-payer reforms and co-payment was introduced the latest and reduced in early reform oblasts. It grew by 73% in the late reform oblasts of Osh, lalal-Abad, and Batken, and it grew by 24% in Bishkek. In contrast, informal payment reduced by 2% in the early reform oblasts ofIssyk-kul, Chui, Naryn and Talas where the reforms have had a longer period to take hold. As the evaluation above shows, there has been tangible improvement in depth of coverage, equity of access and transparency. However, there are still issues that need to be addressed before benefits of the reforms reach the entire popUlation. One of the key impediments to reaching this goal is lack of awareness among the vulnerable groups of their entitlements. Although the SGBP clearly regulates entitlements, many do not know that primary care is free or that insured population groups have a right for the Outpatient Drug Benefits Package. At the hospital level, many continue to pay informally despite publicly presented co- Good Practice in Health Financing, p. 313 payment schedules. Lack of awareness of rights is a particular problem among internal migrants who do not know that their health insurance benefits are portable and their enrollment with the Family Group Practice does not depend on their permanent place of residence (World Bank, 2005c). 4.4 Scaling-up The Kyrgyz reforms were piloted and the introduced in a phased manner. Although in many contexts, pilots do not get scaled up for nationwide implementation, the Kyrgyz experience is positive from this perspective. The reforms were introduced in two additional oblasts each year allowing covering the country between 2001 and 2005 completely. Strong political ownership was essential when taking forward politically sensitive unpopular reforms as resistance grew and side effects began to show. Nevertheless, the gradual implementation allowed learning-by-doing which is critical when completely new financing and institutional mechanisms are introduced. In addition, phased implementation also allowed sufficient attention from policy makers to those oblasts and providers that were next in line. 5. Lessons for other countries 5.1 Enabling factors The successes of the Kyrgyz health reform are in part due to focusing on a comprehensive approach rather than on isolated instruments or "magic bullets". The Kyrgyz reforms were not attempting to find the magic bullet to improve all health system objectives by implementing the "right" benefit package, or the "right" provider payment reform, etc. The reforms involved a coordinated package of changes in pooling, purchasing, design of benefit package and patient co-payments. A key element of this package was a carefully structured and well aligned financial incentive system both on the provider and patient side. In addition, reforms of health care financing were complemented with careful harmonization of the organization and structure of service delivery at both the macro and micro levels. This included restructuring of facilities and extension of provider autonomy. Complex reforms require careful sequencing of various reform steps. The Kyrgyz reforms stand out in this regard. First, it was implicitly acknowledged that without improvement of efficiency through downsizing, other health system objectives such as equity and quality cannot be improved given the fiscal constraints of the country. Through the combination of financing and service delivery reforms, Manas-I was extremely successful in achieving efficiency gains and paved the way for focusing on other health system objectives in the next phase. Second, the reform measures within health financing were also carefully sequenced. Changes in revenue collection and pooling were the initial steps to ensure optimal flow of public funds first. This was followed by introducing new purchasing arrangements that would change the incentives providers faced. These steps created transparency within the public sector. The benefit package/co-payments were the last step and created a transparent social contract between the popUlation and providers. Good Practice in Health Financing, p. 314 Paying attention to the institutional aspects of the reforms was important to ensure sustainable results. For example, progress on efficiency and quality would have been much more difficult had the MHIF not been formed as a parastatal agency in charge of purchasing, The Single Payer system has successfully adopted "strategic purchasing" into its day-to-day operations. This has included the introduction of output based payment mechanisms, cleverly structured incentives for referrals and exemptions, the Additional Drug Package and regular monitoring of quality. All these tools have provided explicit incentives for efficient and high quality provision of health care. This has led to documented improvements in efficiency of service delivery. If the MHIF had not been set up outside the core public sector, it would not be able to purchase services "strategically" and would have to return to the use of hierarchical line-item input control. This type of input-based budgeting was a major source of inefficiency in the system inherited from the USSR and could not have overcome without completely changing purchasing methods. Phased implementation and successful piloting were an effective implementation approach and helped build capacity and stakeholder support as well as allowed learning by doing. The Issyk-kul Intensive Demonstration Site provided an opportunity for the basic health reform model to be developed. Many of the early adjustments and fine- tuning of the reforms already took place in Issy-Kul. Also, Issy-Kul became a visible symbol of reform, and greatly helped the expansion of these reforms to other parts of the country, especially in the further expansion of the FGP model. Issyk-Kul continues to serve as a demonstration site for the country as the reforms here have been deepened over time. For example, the inpatient co-payment policy was first implemented in Issy-Kul and Chui (another oblast where the reforms have been ongoing for at least 5 years). The finding that the greatest and most sustained reduction in informal payment occurred in Issyk kul is likely to be due to this longer experience with the reforms which allowed building capacity and commitment to its principles. Strong collaboration of the development partners facilitated harmonized support for reform design and support. Donor presence in the health sector is strong in Kyrgyzstan. However, the experience from other high donor dependent countries indicates that donor support is often fragmented and overlapping, which means that often the health sector does not benefit from donors support as much as it should and there are few results to show at the end of a long period of donor support. This was not the case in Kyrgyzstan where donor collaboration is strong and more importantly, the MOH coordinated the donors. The MOH could be in the "driver's seat" because there was a clear sector strategy that facilitated a unified approach. 5.2 Key financing factors The decline in public expenditures experienced during the implementation of the reforms limited the reform effect on financial protection and transparency. Oblast funding began to decline in the wake of successful restructuring. Oblast level pooling and output based payment have achieved their intent: the Kyrgyz health system underwent significant downsizing and became slimmer over 1996-2003 with resultants gains in utility and maintenance costs. The rest of the public sector has not taken the same giant steps as Good Practice in Health Financing, p. 315 the health sector to move to output based payment of public providers. Declining staff and beds triggered the well-instilled response of the Soviet budgeting system: if inputs decline, so should public funds. This mechanistic response has effectively taken out savings from the health sector reducing provider incentives to embark on painful downsizing processes elsewhere. The introduction of co-payment also had a crowd-out effect on public spending. The period of accelerated decline in public funding, especially at oblast level, coincided with the introduction of co-payment. This seemed to create a crowd-out effect in budget negotiations and marginal resources were allocated to those sectors that did not have a chance to collect additional revenues through co-payment. The slow pace of reforms in the overall public finance system created a challenging environment for the health sector to operate in and limits its ability to achieve efficiency, quality and equity gains. The health sector gradually but firmly moved away from input-based line-item budgets and administrative control mechanisms and began to adopt performance management and output-based payment of public providers. These changes have achieved considerable successes in terms of improving transparency and efficiency of public management in the health sector. At the same time, these ideas remained revolutionary within the overall public system which continues to largely operate on the inherited principles of input based budgeting and strict line-items. This mismatch between the pace of the public finance reform and health finance reform has created significant problems for the health sector in the annual budget negotiation process. 5.3 Key political economy factors In recent times, the health sector has been subject to political pressures for eliminating co-payment without commensurate increase in public funding leading to the return of un-funded mandates. In the 2006 revision of the GBP, co-payment was eliminated for deliveries and pensioners above 70. This move was the result of political pressure from the new government and was not accompanied by clear and transparent calculations of costs and affordability of the policy. As a result, it is likely that commensurate funding is not available to compensate providers for the loss of official co-payment and this will mean the return of informal payment and deterioration of hard-fought improvement in transparency. The impact of this policy is currently under evaluation. 5.4 Replicability in low and middle-income countries The Kyrgyz reform in its entirety is particularly applicable in transition economies that face excess capacity and drastically reduced fiscal space. The main logic of the Kyrgyz reforms was to squeeze efficiency gains out of the system and return these for improving the depth of coverage. Good Practice in Health Financing, p. 316 REFERENCES Asymbekova, Gulnara. Unpublished Report. "Analysis of the Maternal Mortality for 1999 in the Kyrgyz Republic." Ministry of Health of the Kyrgyz Republic, Bishkek. Atun, Rifat. 2005. Evaluating Manas Health Sector Reform (1996-2005): Focus on Primary Health Care. Policy Research Papers. Bishkek: WHO-DflD Manas Health Policy Analysis Project. CIA. 2006. World Fact Book. https:Uwww.cia.gov/cialpublications/factbook/geos/kg.html International Crisis Group. 2001. Kyrgyzstan at Ten: Trouble in the "Island of Democracy". Asia Report. BishkekIBrussels: International Crisis Group. International Crisis Group. 2005. Kyrgyzstan: A Faltering State. Asia Report. BishkekIBrussels: International Crisis Group. Institute for War and Peace Reporting. 2003. "Bishkek Braced for Aksy Anniversary." News Report RCA No. 188, March 4. http://www.iwpr.netl?p=rca&s=f&o=177056&apc state=hruirca2003 Jakab, M. 2006. "The Impact of the Kyrgyz Single Payer Reforms on Financial Protection: Did the Poor Benefit?" Mimeo. PhD Dissertation. Harvard University, 2007. Kutzin, Joseph. 2000. "Towards Universal Health Care Coverage: A Goal-Oriented Framework for Policy Analysis." Working Paper, World Bank, Washington, DC. Kutzin, Joseph et al. 2001. Addressing the Informal Payments in Kyrgyz Hospitals. Policy Research Papers. Bishkek: WHO-DflD Health Policy Analysis Project. Kutzin, Joseph. 2002. Health Care in Central Asia. Policy Brief. Copenhagen: European Observatory on Health Care Systems. Kutzin, O'Dougherty, Jakab 2005. Fiscal Decentralization and Options for the Kyrgyz Health Financing System: Reflections on Three Options. Policy Papers. Bishkek: WHO- DflD Manas Health Policy Analysis Project. Meimanaliev, Tilek. 2003. Kyrgyz Model of Health Care System. Bishkek: Uchkun. Meimanaliev, Adilet-Sultan et al. 2005. Health Care Systems in Transition: Kyrgyzstan. Copenhagen: European Observatory on Health Care Systems. National Statistical Committee. Kyrgyz Integrated Household Budget Survey 2004. Bishkek. Good Practice in Health Financing, p. 317 National Statistical Committee. Household Budget Survey 2001. Bishkek. Oxford Policy Management. 2006. "Kyrgyz Republic: Public Financial Management Assessment." Final Report, 18 January. Oxford Policy Management, Oxford Available http://siteresources. worldbank. orglINTECAPUBEXPMANIResourceslPFMkyrgyzReportFin a1090806.pdf Purvis, George et al. 2005. Evaluating Manas Health Sector Reforms (1996-2005): Focus on Restructuring. Policy Papers. Bishkek: WHO-DflD Manas Health Policy Analysis Project. Republican Medical Information Center Database. Ministry of Health of the Kyrgyz Republic. Sargaldakova, Acelle et al. 2000. Health Care Systems in Transition: Kyrgyzstan. Copenhagen: European Observatory on Health Care Systems. United Nations Population Division. World Population Prospects: the 2004 Revision Population Database. World Bank. 2001. Poverty in the 1990s in the Kyrgyz Republic. Washington, DC: World Bank. World Bank. 2005 (a). Kyrgyz Republic Poverty Update: Profile ofLiving Standards in 2003. Washington, DC: World Bank. World Bank. 2005 (b). Kyrgyz Republic: Country Economic Memorandum. 2 vols. Washington, DC: World Bank. World Bank. 2005(c). Operationalizing the Health and Education MDGs in Central Asia: Kyrgyz Republic Health Education Case Studies. 2 vols. Washington, DC: World Bank. Wodd Bank. 2004. Millennium Development Goals for Health in Europe and Central Asia: Relevance and Policy Implications. Washington, DC: World Bank. World Bank. 2004(b). Public Expenditure Review. 2 vols. Washington, DC: World Bank. World Bank. 2006. World Development Indicators. Washington, DC: World Bank. WHO Health for All Database. WHO Global Burden of Disease Estimates, 2002. Good Practice in Health Financing, p. 318 Legal norms Kyrgyz Republic. Ministry of Health. 2006. Manas Taalimi Health Reform Strategy. Decree #98, February 25,2002. Law on the Basic Principles of the Budget, dated June 11, 1998. Law on Health Care Organizations, dated August 13, 2004. Law on Insurance Contributions for the State Social Insurance for 1998, dated January 17, 1998. Law on Insurance Contributions for the State Social Insurance for 1999, dated December 26, 1998. Law on Insurance Contributions for the State Social Insurance for 2000, dated January 18, 2000. Law on Insurance Contributions for the State Social Insurance for 2003, dated February 15, 2003. Law on Medical Insurance of the Citizens of the Kyrgyz Republic, dated October 18, 1999, with amendments from April 21, 2003, and July 15,2003. Good Practice in Health Financing, p. 319 Sri Lanka Ravi P. Rannan-Eliya Lankani Sikurajapathy Good Practice in Health Financing, p. 320 Abbreviations CWC Ceylon Workers Congress DHS Demographic and Health Survey GDP Gross Domestic Product ICHA International Classification for Health Accounts IHP Institute for Health Policy IMR Infant Mortality Rate LTTE Liberation Tigers of Tamil Eelam MCH Maternal and Child Health MOH Ministry of Health MOOH Medical Officer of Health OECD Organization for Economic Cooperation and Development PPP Purchasing Power Parity SHA System of Health Accounts SLFP Sri Lanka Freedom Party TB Tuberculosis TFR Total Fertility Rate UN United Nations UNP United National Party WB World Bank WHO World Health Organization Key Words Health systems, equity, health financing, Sri Lanka Responsibilities and Acknowledgments Any opinions expressed, or recommendations made in this report are those of the authors alone, and not necessarily those of the Institute for Health Policy or the World Bank. The editing of the first draft by Neluka Silva is gratefully acknowledged. Good Practice in Health Financing, p. 321 1. Background This section provides a brief introduction to Sri Lanka's history, economy, people, politics and health conditions, and main trends during the past fifty years. These factors are distinctive, and playa role in Sri Lanka's achievements. 1.1 Geography Sri Lanka is an island located close to the equator, with a land surface of 66,000 square kilometers. Relatively compact, it is mountainous in the center, with rivers flowing outward in all directions. The climate is tropical, and the island receives two separate monsoons, one in the southwest from April to June, and the second in the northeast from October to February. Rainfall is high, particularly in the southwestern comer. 1.2 Economic environment Sri Lanka's economy was historically based around agriculture, primarily rice cultivation, but trade has been a significant component for several centuries and has resulted in a society which has been more open to outside influences and interactions than most countries in Asia. Prior to British occupation of the island, which commenced th in the late-18 Century, large-scale irrigation agriculture and later export of spices accounted for most of government taxation and dictated key aspects of government organization, in particular, a high degree of state-led social organization. The British introduced coffee, then tea and rubber, cultivation, and a classic dualistic export economy emerged by the end of the 19th Century (Snodgrass, 1966). The export of cash crops led to relative prosperity and a trade surplus, and the taxation of crop exports provided a ready revenue source for the government. When Sri Lanka became independent in 1948, although the most of its people were involved in subsistence rice cultivation, the economy was highly trade dependent. Tea, rubber and coconut accounted for more than 95% of exports, and living standards were the highest in South Asia. Relative prosperity continued until the Korean War commodity boom in the 1950s, after which declining commodity prices and a failure to diversify exports led to economic stagnation and ever-tighter import controls and the pursuit of inward-oriented import-substitution policies (Bruton, 1992). Income levels stagnated and there were high levels of unemployment (>20%), which resulted in growing social tensions. These contributed in the years after 1970 to two Maoist insurgencies and an ethnic-based separatist conflict, which have presented major challenges for Sri Lanka's economy. Following a change in government in 1977, Sri Lanka became one of the first developing countries to embark on economic liberalization. This has been pursued ever since. Trade was liberalized, export taxes on cash crops removed, and the economy opened up, in return for which the country benefited from substantial Western aid inflows for more than a decade. These policies led to substantial improvement in economic growth (Table 1), and the economy has sustained an average of3-4% real per capita income growth ever since, despite the series of debilitating internal conflicts that started in the early 1970s. Growth has been led by export-oriented manufacturing, initially concentrated in garments, but now gradually diversifying. By the 1990s, more than 75% of Sri Lanka's exports were industrial products. In recent years, the continuing economic growth has pushed unemployment to less than 8% of the Good Practice in Health Financing, p. 322 workforce, raised per capita income in 2005 to more than US$1 ,000 per capita, and achieved modest reductions in the numbers living poverty (Table 2). More substantial reductions in poverty have not occurred, because recent economic growth has been associated with increasing income inequality, and living standards for the lowest income quintile have hardly changed. Although official development assistance remains significant, private foreign direct investment is now more important for growth, although not at levels comparable to other economies in South-East Asia. Table 1: Sri Lanka - Economic indicators 1930-2005 Year GDP per GDPper Revenue Expenditure External ODA capita capita (% GDP) (% GDP) Debt (% GDP) P990 US$) (1990 PPP$) (% GDP} 1930 180 945 -10.0 -10.0 -0 - 0 1950 273 935 16 20 3 0 1970 316 1,130 20 27 18 1.7 1990 577 1,935 22 31 72 5.7 1995 704 2,636 20 31 67 4.5 2000 844 3,626 17 27 55 0.4 2005 962 4,390 16 24 48 3.4 Source: Central Bank of Sri Lanka (2006), IHP databases, and estimates of pre-1950 GDP originally prepared by author for Rannan-Eliya and de Mel (1997). Table 2: Sri Lanka - Social indicators 1930-2005 Year Population Poverty head Poverty Literacy Infant Life (millions) count head count (%) mortality expectancy «PPP$l per «PPP$2 per rate da~) da~) 1930 5.3 n.a. n.a. n.a. 175 40 1950 7.7 n.a. n.a. 69 82 55 1970 12.5 n.a. n.a. 82 47 65 1990 16.3 n.a. n.a. 88 19 71 1995 17.3 2.5 31.3 90 16 72 2000 18.5 2.3 31.5 91 13 73 2005 19.6 2.3 22.7 92 11 73 Source: Central Bank of Sri Lanka (2006), IHP databases, Medical Statistician of MOH, and and de Silva (F orthcoming). A key element in the post-1977 economic liberalization was the removal of export taxes. This led to a collapse in government revenues, and caused a structural fiscal deficit that in the past decade has averaged 7-9% ofGDP (Table 1). Although government spending was also cut, this has not led to fiscal improvements, since concomitant tax policies led to continuing revenue reductions that kept pace with the spending reductions. Much of the pressure to cut taxes in recent years appears to have been ideologically driven by key donors, as fiscal realities in fact point to the need to increase taxation to achieve fiscal balance. Currently, taxation is predominantly from a mix of indirect taxes, including value-added taxes and excise taxes, with smaller contributions from import taxes. Direct income taxes on individuals contribute a small fraction of revenues. The fiscal deficit has resulted in a mounting public debt, constant pressure on the exchange rate, and the inability of the government to increase social expenditures or invest in needed physical infrastructure. As a consequence, government policy is now focused on raising taxes, given that there is no room for more substantial spending reductions. Good Practice in Health Financing. p. 323 1.3 Demography and health Sri Lanka's population is 20 million, and remains predominantly rural. For most of the past half century, Sri Lanka was notable for having no significant rural-urban migration, largely due to the good provision of social services in rural areas. However, with increasing industrialization, there has been a shift of population into urban areas in the past decade. The urban share of the population was 15% in 2001. Sri Lanka is a multi-ethnic, multi-linguistic and multi-religious society. Three quarters of the popUlation are Sinhala-speaking; most are Buddhist and the rest Catholic and Protestant Christians. The rest comprise of three distinct ethnic groups: the Tamil- speaking Sri Lankan Tamils (13%), Muslims (7%), and Tamil-speaking Indian (or Estate) Tamils (5%). The first is mostly Hindu, with a significant minority being Christian, and the latter is predominantly Hindu also. There are also small Eurasian Burgher, Malay and Indian communities «1 %). Sri Lanka's health indicators were worse than much of South Asia in the 1920s (Langford and Storey, 1993), but its health reforms rapidly reduced mortality rates from the 1930s (Langford, 1996). Following the end of World War II, mortality rates rapidly fell for a decade, before entering a slower but still rapid phase of decline, which continues. Prior to the 1990s, the bulk of mortality decline was due to substantial reductions in infant, child and maternal mortality. Life expectancy has continuously risen, but since the 1970s gains have been largely confined to women, and male life expectancy has stagnated (Table 3). Life expectancy is currently 71 years, and the infant mortality rate is less than 13 per thousand live births. The main drivers of these remarkable health gains have been policies that have ensured widespread easy access to medical services for the whole population, the emphasis on universalism, mass female education which has enabled women and mothers to make use of these services, and a continuous policy-driven process of behavioral change which has made Sri Lankans highly sensitive to illness and predisposed to make ready use of modern medical treatment when ill (Caldwell, 1986; Caldwell et al., 1989; De Silva et al., 2001; Rannan-Eliya, 2001; Rannan-Eliya, 2004). Table 3: Sri Lanka 1930-2003 - DemograEhic and health indicators Year Infant Life Life Maternal Total Population mortality expectancy expectancy mortality fertility growth rate rate at birth at birth rate rate (%) {female} {male} 1930 175 39 41 21 n.a. 1.4 1950 82 55 56 6 5.3 2.8 1970 47 67 64 2 4.2 2.2 1990 19.5 73 67 1 2.2 1.0 1995 16.5 75 68 <1 1.9 1.1 2000 13.3 76 70 <1 1.9 1.4 2003 11.2 77 71 <1 1.8 1.3 Source: Data kindly provided by Medical Statistician of MOH, Department of Census and Statistics, and De Silva (Forthcoming). Good Practice in Health Financing, p. 324 During the 1950s-70s, mortality reduction led to rapid population growth. Fertility rates began to drop after a lag, and the total fertility rate fell below replacement level by 1993. The TFR is now less than 1.9, and may drop as low as 1.5 according to some projections (De Silva, Forthcoming). Population growth, which was a major concern in the 1960s-70s, is giving way to concern about population ageing. The population size may stabilize at 22 millions by 2030, and decline thereafter, without substantial immigration. With low fertility rates and high life expectancy, Sri Lanka's population is becoming older. The numbers of elderly are rapidly increasing, whilst the numbers of children are falling. Sri Lanka will be the one of the most rapidly ageing societies in Asia in the next decades. This is already reflected in the age structure of the population, which is no longer pyramidal as in most developing countries (Figure 1). Fi ure 1: Sri Lanka - Po ramids 1991,2006,2026 and 2051 Sri Lanka, 1991 Sri lanka, 2006 90-94 75-79 60-64 45-49 30·34 15-19 0 .. 90 ... 75-79 60-64 45-49 30·34 15-19 0.. 10 10 10 10 Population (Of.) Population (%) Source: Authors computations using data from De Silva (Forthcoming) Sri Lanka's mortality transition is largely complete, and its mortality patterns increasingly resemble that of a developed country. There are few deaths from infectious diseases, such as cholera, measles, malaria, TB, etc., and mortality from communicable disease is declining. Increasingly, non-communicable diseases and accidents dominate mortality, and ischemic heart disease is the largest single cause of death ( Table 4. In fact, Sri Lanka faces growing epidemics of diabetes, ischemic heart disease and cerebrovascular disease, which affect particularly adult males. Good Practice in Health Financing, p. 325 Table 4: Leading causes of mortality, Sri Lanka 2001 Rank order Cause Percent of all deaths 1 Ischemic heart diseases 8.5 2 Remainder of disease of the nervous system 6.9 3 Other heart diseases 6.1 4 Intentional self-harm 4.2 5 Diseases of the liver 4.2 6 All other external causes 4.1 7 Chronic lower respiratory diseases 3.8 8 Hypertensive diseases 3.7 9 Remainder of malignant neoplasm 3.2 10 Cerebrovascular diseases 3.1 Source: Computed by authors from data kindly provided by Registrar Generals Department. Note: Deaths classified with no clear diagnosis are excluded, but accounted for 24.9% of all recorded deaths. The mortality trends are indicative of the underlying morbidity in the country, with a high prevalence of non-communicable disease in the adult population. However, reliable morbidity data are not available or routinely collected. Table 5 presents the morbidity profile of inpatients at government hospitals and of outpatients in a recent study of private clinic doctors, which can be considered representative of the morbidity profile seen by health care providers, but not necessarily of the overall disease burden. Table 5: Patient morbidity in Sri Lanka, inpatients and outpatients Inpatient morbidity Ou!gatient morbidity Cause % Problem % Traumatic injuries 16.7 Viral fever 15.6 Diseases of respiratory system 10.8 Asthma 6.3 Symptoms, signs and abnormal clinical and 7.6 Upper respiratory tract infection 5.2 laboratory findings not elsewhere classified Viral diseases 6.3 Hypertension 4.8 Disease of gastrointestinal tract 6.3 Respiratory infection 4.2 Direct and indirect obstetric causes 4.7 Gastritis 3.2 Diseases of the urinary system 4.1 Gastroenteritis 2.8 Intestinal infectious diseases 3.8 Lower respiratory tract infection 2.2 Diseases of musculoskeletal system and 3.6 Urinary tract infection 2.0 connective tissue Diseases of skin and subcutaneous tissue 3.5 Muscle pains 2.0 Source: Inpatient morbidity statistics from unpublished data for 2003 kindly provided by Medical Statistician, Ministry of Health. Outpatient morbidity statistics from Rannan-Eliya et al. (2003) 1.4 Government and politics Sri Lanka consists administratively of nine provinces (two of these were merged in 1987, but in late-2006 Sri Lanka's Supreme Court declared the merger illegal until a referendum of their voters is held), and 24 districts. The central government is responsible for national policy and other areas. In health and education, the responsibility for operation of facilities and services, except certain specialist and national ones, is devolved to the provincial governments. The provincial governments are themselves elected and accountable in theory to provincial legislatures, although in practice politics remains quite centralized. Good Practice in Health Financing. p. 326 Democracy has been the primary explanation for Sri Lanka's health achievements. Since 1931, Sri Lanka has maintained continuity of democratic government, with all governments elected through the ballot box, despite many challenges including an attempted military coup in the 1960s, two Maoist insurgencies in 1971 and 1987-89, and a separatist conflict since 1972, when the precursors of the LTTE terrorist organization first took up arms against the government. During this time, several developments have shaped the political system. First, initial electoral competition between a dominant conservative political establishment and Marxist-Trotskyite challengers, gave way in the 1950s to a two party model, in which two dominant political parties - the right-of-centre United National Party (UNP) and the left-of-centre Sri Lanka Freedom Party (SLFP) - competed for power. From 1956, incumbent governments have tended to lose elections, which has led to political leaders being extremely sensitive to voter concerns. In the case of health, this has encouraged bipartisan consensus on major features of health policy, such as an emphasis on universal access, no user fees, and continuing public sector dominance of healthcare delivery. Secondly, the British-inherited constitution was replaced in the 1970s by a presidential system, with an executive President directly elected by single-transferable vote, and a legislature elected through proportional representation. These changes make it harder for governments to introduce radical changes in policy where there is a strong pre-existing consensus, and make coalition politics the norm. Thirdly, as a result of international exhortations to solve the separatist conflict, extensive devolution of government to the provincial level was introduced in 1988, although it has failed to stop the conflict. 2. Health Financing and Coverage This section presents an overview of the financing of Sri Lanka's health system, key historical trends, and its performance in terms of equity and efficiency. 2.1 Health expenditures Sources offinancing Total expenditure on health in Sri Lanka was provisionally Rs. 100 billion (US$ 1 billion) during 2005 (Annex Table A.l), equivalent to 4.2% ofGDP (Institute for Health Policy, Forthcoming). There has been a slow increase in the level since the early 1990s, driven mostly by increased share from private financing. Government spending accounts for 46% of this, and private financing for the rest (Annex Table A.2). In per capita terms, expenditure in 2005 represented US$ 50 per capita at official exchange rates, and government spending was equivalent to US$ 23 per capita. Health services account for eight percent of government budgetary spending. Private financing is mostly out-of-pocket spending by households, with smaller contributions from employers and insurance. NGO spending is small. Good Practice in Health Financing, p. 327 Table 6: Trends in the composition of health expenditures, Sri Lanka 1953-2005 1953 1980 1990 2000 2005 Health expenditures (% GDP) Total 3.3 3.1 3.3 3.5 4.2 Government 2.1 1.7 1.7 1.7 1.9 Government as % of budget 8.4 4.1 5.6 6.5 7.8 Source of national expenditure (%) Public 62 57 46 50 46 Private 38 43 54 50 54 Of which out-of-pocket 33 41 42 43 48 Hospital composition by source of expenditures (%) Total 47 45 41 43 43 In public spending 72 70 78 69 71 In private spending 4 12 11 17 19 Hospital expenditure by source of financing (%) Public 96 87 93 80 76 Private 4 13 7 20 24 Non-hospital expenditure by source of financing (%) Public 33 31 17 19 23 Private 66 69 83 81 77 Source: Rannan-Eliya et al. (1997), and IHP Sri Lanka Health Accounts database. Estimates for 2005 are provisional and as of compilation in November 2006. Use of expenditures Health spending reflects the structure of what is a public hospital dominated healthcare system. Government expenditures have concentrated on hospitals since the health reforms of the 1930s, whose primary goals were increasing equity in access and improving risk protection, both of which required substantial increases in hospital coverage. Hospital spending accounted for about 70 percent of government recurrent spending in the 1950s, and the share has changed little since then (Table 6). Government hospitals have been the primary mode by which modern medical treatment has been made available to people in rural areas, and the prioritization of these facilities in budgetary spending has ensured that the health ministry was able to achieve sufficiently high levels of coverage that most services were available to all Sri Lankans. In contrast, most private spending is for outpatient care and for purchasing medicines, but the share of hospital spending in private outlays has increased in recent years. This is partly because of the expansion in provision of outpatient services by private hospitals, and partly due to the increased availability of private insurance. As a consequence, until recently more than 85 percent of hospital spending was by government, whilst more than 80 percent of non-hospital and outpatient care spending was financed privately. Benefit-incidence ofgovernment health expenditures Government health expenditures have reached the poor effectively since at least the 1950s, after the health reforms expanded government health services into rural areas. Estimates of the actual incidence of government health spending are only available from the late 1970s. Although these estimates are not strictly comparable, as they were computed by different authors using different methods, they suggest that the targeting Good Practice in Health Financing, p. 328 of government health spending was quite pro-poor in the late 1970s, then became less so during the 1980s-90s (Table 7). In 2003/04 the poorest quintile received 20 percent of government health spending, and the richest quintile 15 percent. Table 7: Incidence of public health expenditures, Sri Lanka 1979-2004 Year Share of government health Share of government health expenditure received by poorest expenditure received by richest household quintile (%) household quintile (%) 1979 30 9 1992 24 15 1996/97 22 18 2003/04 20 15 Source: Alailima and Mohideen (1983), and estimations by authors and Aparnaa Somanathan ofIHP. Government outpatient spending is the most pro-poor (27% went to the poorest quintile in 2003/04 versus 11 % to the richest quintile), and inpatient spending is more evenly distributed (18% versus 16%). Since Sri Lanka does not means test access to public services the main reasons for the pro-poor targeting of government health subsidies are (i) a highly dense network of health facilities which make government health services physically accessible to the poor; (ii) lack of user charges; and (iii) the voluntary opting out of the rich into the private sector (Rannan-Eliya, 2001). In effect, what Sri Lanka does is guarantee to its poor effective access to free health services, especially hospital care, and then relies on differentials in consumer quality in services to persuade the richer patients to voluntarily opt to use and pay for private provision. The role of these consumer differentials is discussed in more detail later. This approach resembles closely that seen in the other two Asian health systems where government-operated hospital services do effectively reach the poor - namely Malaysia and Hong Kong. Together with these cases, Sri Lanka differs as a result in the income gradients in use of public and private hospital services. Throughout Asia, the rich use private hospital services more than the poor, and this is generally true for public services, but in Sri Lanka and the other two cases, the gradient is reversed for public hospitals. Unfortunately, a full explanation as to why Sri Lanka, and other countries like it, are able to achieve this, when most countries don't, is sti11lacking, and requires intensive research. Unfortunately, it is unlikely that such research will take place in the near future, as understanding how the public sectors reach the poor in Sri Lanka, Malaysia and Hong Kong has generally not been a priority for most research funding agencies. Incidence offinancing The health financing system is close to progressive, and in comparative terms it does better than developed countries in Asia, but not as well as some countries such as Thailand and Philippines (Table 8). The burden of paying for the half of total health expenditures that come from general revenues falls mostly on the richer households. Indirect taxation is neither progressive nor regressive, but direct taxes are very progressive. On the other hand, spending from private sources is mostly out-of-pocket, and these payments are very progressive, as the rich are more likely to seek private care than the poor. For Sri Lanka to improve the progressivity of its healthcare financing, it would need to increase the share of direct taxation in overall government revenues, as Good Practice in Health Financing, p. 329 well as modify its system of indirect taxes so that they place more of a burden on goods and services used more by the rich than the poor. Table 8: Progressivity of health financing, Sri Lanka and selected Asian countries (using the Kakwani Index) Direct Indirect Social Private Direct Total taxes taxes insurance insurance payments Bangladesh (1999/00) 0.552 0.111 n.a. n.a. 0.219 0.214 China (2000) 0.152 0.040 0.235 n.a. -0.017 0.040 Hong Kong (1999/00) 0.386 0.119 n.a. 0.040 0.011 0.166 Japan (1998) 0.095 -0.223 -0.041 n.a. -0.269 -0.069 South Korea (2000) 0.268 0.038 -0.163 n.a. 0.012 -0.024 Kyrgyz Republic (2000) 0.074 -0.096 -0.034 n.a. 0.264 0.125 Nepal (1995/96) 0.144 0.114 n.a. n.a. 0.053 0.063 Philippines (1999) 0.381 0.002 0.205 0.120 0.139 0.163 Sri Lanka (1996/97) 0.569 -0.010 n.a. n.a. 0.069 0.085 Thailand (2000) 0.510 0.182 0.180 0.004 0.091 0.197 Note: Excerpted results of the Equitap study as published in Rannan-Eliya and Somanathan (2006). The Kakwani index is a numerical index of the distribution of payments in relation to ability to pay. It is calculated graphically by looking at the distribution curve of overall tax payments made by poor to rich households, and comparing this distribution with the distribution of overall consumption across the same households, with the index computed as twice the size of the area between the two curves. A positive number implies that the share of payments by richer households is greater than their share of overall consumption. A negative number implies the reverse. For further details of these methods see the World Bank's technical notes for quantitative techniques for health equity analysis at http://web.worldbank.orgIWBSITE/EXTERNALITOPICSIEXTHEALTHNUTRITIONANDPOPULATI ONIEXTP AH/O"contentMDK:20216933-menuPK:400482-pagePK: 148956-piPK:216618-theSitePK:4 00476,00.html. Protection against catastrophic risk Sri Lanka's health system performs very well in protecting the poor against catastrophic financial risks associated with illness. The Equitap study of equity in Asian health systems found that Sri Lanka is in the category of countries where few people are pushed into poverty as a result of medical spending (van Doorslaer et al., 2006), with only 0.3% of households pushed below the PPP$ 1.08 international poverty line as a result of health expenditure (Table 9). Good Practice in Health Financing, p. 330 Table 9: Proportion of population pushed below the PPP$ 1.08 poverty line as a result of household health spending, selected Asian countries Country Prepayment Postpayment Change in poverty headcount (%) head count (%) headcount (%) Bangladesh 22.5 26.3 3.8 India 31.1 34.8 3.7 China l3.7 16.2 2.6 Nepal 39.3 41.6 2.2 Vietnam 3.6 4.7 1.1 Indonesia 7.9 8.6 0.7 Philippines 15.8 16.4 0.6 Kyrgyz Republic 2.2 2.7 0.5 Sri Lanka 3.8 4.1 0.3 Thailand 2.1 2.3 0.2 Malaysia 1.0 1.1 0.1 Source: van Doorslaer et al. (2006) 2.2 Benefits package All government health services, with few exceptions, are available free to all citizens, including all inpatient, outpatient and community health services. Free services range from anti-retrovirals for HIV/AIDS patients to coronary bypass surgery. Access to all services is reinforced by a policy of permitting patients to visit any hospital in the country without restriction, and with no enforcement of a referral system. The government is able to do this because of a high level of technical efficiency in its delivery system which keeps costs low, and the implicit strategy of encouraging the richer patients not to burden the government health system by voluntarily opting to use private provision. The first exception to free care consists of family planning commodities. Condoms and oral contraceptive pills are made available through government primary care facilities at cost, although the prices are much lower than private sector alternatives. The second exception consists of private paying wards in less than a dozen government hospitals. These provide the same treatment as in the main hospital wards, but greater privacy. However, in practice these beds are under-used, and these beds account for less than 1% of all MOH inpatients. The third exception is the Sri Jayewardenapura General Hospital, which is an autonomous, tertiary care, 1,000 bed hospital that was constructed with Japanese development assistance. Probably for ideological reasons and poor economic analysis, the Japanese stipulated that user fees be charged at this facility. The hospital maintains three classes of wards, at which different levels of fees are charged based on a means test, although in no case are the fees sufficient to cover full costs. In practice, even though the hospital's subsidy per bed is greater than any other government hospital, it is unable to generate sufficient fees, and operates at a loss. Nevertheless, there is implicit rationing of care. This occurs in a number of ways. First, through internal purchasing controls and investment decisions the MOH can and does restrict the availability of services it considers too expensive. For example individual drugs or certain high technologies may not be purchased for government hospitals, or only purchased in limited numbers. Second, using administrative controls, the central ministry can restrict the supply of specific services to only certain government hospitals. This can be done by controls such as the placement of specialists, or through Good Practice in Health Financing, p. 331 the lists of drugs that are approved for different levels of hospital. Even basic equipment is rationed, e.g., most lower-level MOH facilities lack X-ray machines. However, although services may be restricted to certain facilities, all patients still have the right to travel to those facilities, and many typically do. Although these decisions often have some rational basis, most are taken implicitly and without public debate. Whilst this lacks transparency, it has the benefit that it prevents much public opposition. Third, it has been official policy that if medicines are not available in hospital stocks, that patients may be asked to self-purchase drugs from private pharmacies. This results in extensive self-purchasing by patients, as the medicines budget is inadequate, but there is evidence to indicate that MOH personnel protect the poor by using discretion to reserve limited drug stocks for the poor and by being more likely to encourage richer patients to self-purchase. 2.3 Financing and payment Public sector health spending is financed exclusively from general revenue taxation, with a small contribution from international development assistance «5%). There is no social insurance. Government health spending is mostly by the central government (62% of public), and provincial governments (36%), with small contributions from local governments at municipality and village level. However, financing for provincial and local government health budgets comes principally from the central government. Provincial governments have the authority to raise their own tax revenues, but owing to inherent economic limitations, they raise less than 5 percent of overall government tax revenues, and most of these are raised in the Western Province. This is because most provinces are essentially rural and so direct income taxation is limited, whilst the most buoyant domestic tax mechanism is value-added tax which can practically only be levied at the national level given the small size of the island. In addition, 50% of the country's economic output and an even greater share of its formal sector is located in the Western Province. There is a modest contribution from international donor assistance (4-6% of public financing). User charges and miscellaneous income account for the rest «2%). Private health services are funded principally from out-of-pocket spending by households. Private doctors either dispense and charge only for their medicines, or charge separately for consultation and for drugs (Rannan-Eliya, Jayawardhane, and Karunaratne, 2003). However, it should be noted that most private doctors are actually government medical officers who do private practice in their off-duty hours. Private hospitals typically charge by item for cost of services, except for the fees of attending physicians, who may bill the patient separately. One fifth of private financing is from employer spending on medical benefit schemes for their employees and on group medical insurance schemes, plus a smaller amount from individually purchased medical insurance (Institute for Health Policy, Forthcoming). Most of this spending pays for private hospital services, and benefits mostly the top income quintile. Typically, all these schemes reimburse the patient for their expenses, so patients must first pay out-of-pocket, and then claim reimbursement. The amounts reimbursed vary according to the specific rules of their insurance or employer policy, but most medical insurance in Sri Lanka is indemnity insurance, and will reimburse expenses up to a fixed maximum, beyond which no further payments are Good Practice in Health Financing, p. 332 made. The contribution of non-profit institutions to health care fmancing is about 1 percent (Annex Table A.2). 2.4 Development of health financing Before the health reforms of the 1930s, the government took a non-interventionist role in health, and health services and financing were treated as a private responsibility. Its involvements were limited to the funding of urban hospitals to look after European residents and its own civil and military personnel, and regulations to ensure medical services were provided to the important plantation workforce. However, during the reforms in the 1930s-40s, government expenditure on health increased substantially. By the mid-1950s, national health spending was about 3.2-3.5% ofGDP, of which the public share was about 60% (Rannan-Eliya and de Mel, 1997). From the early 1960s, spending fell as the government faced stringent fiscal constraints, and has remained in the range of 1.3-1.8% ofGDP (Figure 2). During these years, private spending has gradually increased its share of total financing to more than half. Fi ure 2: Government recurrent health s 3.0% , - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - , 2.5% + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 1 2.0% + - - - - - - - - - - - - - - - + \ r = - - - - - - - - - - - - - - - _ _ I 1.5%+---~~-~--~~._-------;.----------~ 1.0%+--~~---~~~-----------~~~-~---__I 0.5% + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - _ _ 1 0.0% + - - - - - , - - - - , - - - - - , - - - - , - - - - , - - - - - - ; - - - , . . - - - - . , . - - - ' 1920 1930 1940 1950 1960 1970 1980 1990 2000 Source: Adapted from Rannan-Eliya and de Mel (Rannan-Eliya and de Mel, 1997), with estimates for 1990-2005 derived by modifying estimates for central MOH and provincial council health department expenditures as compiled in IHP's Sri Lanka Health Accounts database to fit earlier definitions of recurrent departmental spending. 2.5 Equity - Health indicators and outcomes, and their distribution Health outcomes and inequalities Prior to the 19th Century, there is no evidence of significant differentials in health outcomes. However, as in Britain, urbanization and growth of the formal economy in the 19th Century led to deterioration in health conditions and significant health inequalities. In Colombo, the emergence of an urban proletariat led to the growth of a large poor quarter, in which people lived in crowded, unhygienic conditions, with no Good Practice in Health Financing, p. 333 basic sanitation, and were prey to disease epidemics. Dense population settlement in poor and unsanitary housing conditions was also found in the plantation sector, where indentured labor was imported from India. Two significant differentials in health status arose. First, health conditions in urban areas were worse than in rural areas. The urban middle-classes enjoyed better health than their poor neighbors, but their mortality rates were no lower than those in rural areas and, in fact, worse for infants. Secondly, the worst health was amongst poor urban and plantation sector workers (Meegama, 1986). Evidence for this comes from analysis of death registration data by ethnic group and district, which are reasonable proxies for socioeconomic status (Table 10). It was in the context of these large health disparities that the introduction of democracy had its impact. Full democracy did not occur until 1931, but municipal governments, which were elected by residents on the basis of a limited franchise, were introduced into most Sri Lankan towns during the 19th Century. These local governments intervened over several decades to enforce basic standards for sanitation, to provide maternal and child health services, and to construct public sewage and water systems. These interventions succeeded in reducing urban mortality rates, and by the early 1930s had eliminated the urban-rural differentials in mortality rates. By then, the major mortality differentials of note were between the malarial and non-malarial areas of the country. Table 10: Infant mortality rates in different social groups, 1920-22 Ethnic group and district Proxy for social group Males Females Sinhalese in Kalutara Rural villagers 114 101 Tamils in Nuwara Eliya Plantation workers 248 210 Tamils, Moor and Malays in Colombo Urban poor 341 320 Burghers and Eurasians in Colombo Middle class 144 158 Source: Computed from data of Registrar General of Ceylon as given in Meegama (1986). After full democratic government was established in 1931, electoral pressures led the government to expand the network of medical facilities in rural areas, and this expansion was most substantial in the previously under-served malarial districts (Rannan-Eliya and de Mel, 1997; Langford, 1996). This was in fact consistent with the original goals for introducing democracy to the island, and this democratic mechanism worked particularly well for encouraging expansion of rural health services, owing to the small size of electorates which encouraged a form of parish pump politics in which national politicians, some elected by as few as 5,000 voters (Wriggins, 1960), competed to ensure that the government built dispensaries and later hospitals in their home constituencies. Sri Lanka is in fact an example of how democratic politics can provide a means of accountability of government services to the poor (World Bank, 2003). This did not have an immediate health impact until World War II ended though. Between 1945 and 1952, mortality rates in Sri Lanka across all demographic groups halved and life expectancy increased 12 years. Although this was for many decades explained as being due to the introduction of DDT spraying against the malaria vector, the most recent assessments demonstrate that malaria control played only a minor role, as mortality reductions occurred in both malarial and non-malarial areas. The critical intervention was the expanded access to curative facilities in rural areas, plus the improved supply of antibiotics, other drugs and staff that became possible after 1945 (Langford, 1996). Electoral pressures ensured that the expansion of health facilities benefited all areas ofthe country, and in subsequent decades mortality rates were substantially reduced in both rural and urban areas, and in all districts of the country. Good Practice in Health Financing, p, 334 Table 11 shows this has substantially reduced and even reversed inter-district inequalities in health outcomes. Table 11: Infant mortality rate in selected districts of Sri Lanka, 1921-2000 Year Colombo Anuradhapura Kandy Hambantota 1921 161 366 196 293 1931 139 266 167 189 1951 94 69 89 62 1971 41 34 61 37 1991 26 21 23 6 2000 18 16 20 4 Source: Data from Registrar General's Department and Annual Health Bulletin ofMOH. With consolidation of the health system in the 1960s, physical access to basic health services within close proximity became the reality for almost all the population. High coverage has been the key to reducing mortality rates (Caldwell et al., 1989; De Silva et al., 2001; Langford, 1996; Rannan-Eliya, 2001), and has led to continuous improvement in health indicators for rich and poor. Continuing expansions in access at the margin primarily benefit the poor, once the richer households are served. This is illustrated by using asset indices to disaggregate data collected in the Sri Lanka DHS. Infant mortality rates declined in all income groups, and the absolute difference between the richest and poorest quintiles substantially narrowed during the 1990s (Figure 3). The role of universal access in this is demonstrated by trends in access to qualified medical care at childbirth. The major beneficiaries of marginal improvements in access to such care in the 1990s were the poorest quintile (Figure 4). Moreover, access of the poor to services is such that when demand is greater amongst the poor, actual uptake of services can be more than in rich, as shown by trends in use of modern contraception where the poorest quintile has higher use (Figure 5). Figure 3: Differentials in infant Figure 4: Differentials in medical mortality rate by asset quintile, Sri attendance at childbirth by asset quintile, Lanka 1987-2000 Sri Lanka 1987-2000 50 - - -----.---- 1_1987 _1993 20001 100% ~_1987 _1993 40 30 90% c,-~·-\lli.f-- 20 80% I 10 1 70% Poorest Second Middle Fourth Richest Poorest Second Middle Fourth Richest Good Practice in Health Financing. p. 335 Figure 5: Differentials in use of modern methods of contraception in currently married women, Sri Lanka 1987-2000 80% Iri_1987 _1993 112000 I I 60% 40% 20% 0% Poorest Second Middle Fourth Richest Targeting and equity implications of reforms Central to Sri Lanka's health reforms has been the concept of universalism and its link to citizenship. From their inception, Sri Lanka's health reforms were driven by the close connection between citizenship and political rights and government obligations to the people. The government does not explicitly target services to specific groups, and does not accommodate different systems of care to different groups. So much so, that pilot projects to test different approaches to healthcare delivery in small areas have not flourished. Access to healthcare is treated as a fundamental social right, and thus not subject to arbitration. The symbolic commitment to the principle of free universal care to all citizens is taken so seriously that, despite more than a quarter of a century of war against the LTTE (Liberation Tigers of Tamil Eelam) terrorist group, successive governments have consistently refused to restrict the right to free care that LTTE terrorists have by virtue of their citizenship. Hospitals in LTTE-controlled areas continue to be funded, supplied and staffed at government expense, and injured LTTE rebels continue to be treated in these facilities. Although this attitude may seem impractical given resource limitations, it has been a critical success factor. Firstly, gQvernment services continue to be used by and accountable to all in society, including the influential middle-classes and elite urban residents, who have remained political supporters of good quality government services. Furthermore, expansion has not been at the cost of reductions in clinical quality of services, although it has been at the cost of accepting lower levels of consumer quality. Moreover, with a universalist approach, once the rich and middle-income classes are provided for, marginal increases in provision inevitably favor the poor. There has been one defect to this approach. If individuals do not have citizenship or more importantly lack the vote, they fail to benefit. This link was most obvious in the case of the Tamil-speaking population ofIndian origin, who work on the plantations. As noted elsewhere, they were the first beneficiaries of state action designed to expand access to medical services, and by the 1940s their health indicators were better than for the rest of the rural population. However, just before independence, most were deprived of Sri Lankan citizenship and, thus voting rights as a result of legislative changes to deprive Indian-origin residents of Sri Lankan citizenship. After 1948, they had no electoral representation, and government left responsibility for their social service provision to the British and US private plantation companies. The results of this natural Good Practice in Health Financing, p. 336 experiment to compare private and public provision are quite evident. The private sector failed to match the improvements in public service provision that came about through government intervention in rural areas in the 1950s-70s, even though such efforts would have improved labor productivity. Health improvements in the estate population began to lag those of the rest of the population. By 1970 they had much worse mortality rates than the rest of the country, and health conditions were effectively stagnant. Then two important changes occurred. First, the government nationalized the plantations in the early 1970s, and thus indirectly became responsible for healthcare provision to this population, and second the CWC (Ceylon Workers Congress), the political party representing the estate workers,joined the government in 1978. Once in government, this party lobbied for enhanced state social service provision to its community and in the 1980s persuaded the government to restore first voting rights and then citizenship. The impact of the restoration of voting rights was immediate, and led to concerted government efforts to improve health services in the estates. In the 1990s, the plantations were effectively privatized, but at the urging of the CWC, the government agreed to nationalize the plantation healthcare facilities and to integrate them into the MOH network. The impact of these changes in citizenship and voting rights on health indicators in the estate population has been dramatic (Figure 6). Prior to the 1940s, government- legislated employer mandates were successful in eliminating and, eventually, reversing the mortality disadvantage of the estate popUlation. In the four decades that followed disenfranchisement, the community missed out on the national health reforms, and a significant disparity again appeared, with the IMR reaching almost double the national average. Within two years of the CWC joining government, rapid improvements in mortality began to take place, and thirty years later there are now good signs that the disparity in health outcomes will soon be eliminated. This performance in reducing these ethnic differentials compares favorably with that of countries such as the USA, which have conspicuously failed to narrow historical ethnic disparities in mortality. Figure 6: Trends in infant mortality rates in Nuwara Eliya district and Sri Lanka, 1920-2003 250 200 150 100 50 j-Nuwara-Eliya --Sri Lanka I o 1920 1930 1940 1950 1960 1970 1980 1990 Good Practice in Health Financing, p. 337 Note: The population of Nuwara Eliya district is predominantly Indian-origin estate Tamil, and so its IMR rate provides a good proxy for health conditions in the plantation community. 2.6 Efficiency There are two different types of efficiency - technical efficiency and allocational efficiency, and these can be considered either from a system or "macro" perspective, or a provider or "micro" perspective. Although Sri Lanka's health system does not meet conventional expectations, it can be considered to be highly efficient. Efficiency has been an important and critical element in Sri Lanka's success in extending coverage. It enabled it to use a limited health budget to reach the poor. In other countries, increasing access to services often leads to such increases in patient demand that the public sector must ration access to care in order to match available resources to apparent patient needs. However, Sri Lankan healthcare managers were denied this response, and were, instead, forced to pursue continuous gains in efficiency. This approach made it possible to financially sustain Sri Lanka's policies of universal access to health services (Rannan-Eliya, 2001; Rannan-Eliya and de Mel, 1997). In fact, Sri Lanka managed to expand access to health services whilst reducing government health spending as a share ofGDP after 1960 (Table) Technical efficiency at the macro-level From a macro-level perspective, the Sri Lankan health system is a global outlier. It spends less in absolute and relative terms than comparable low-income developing countries, but achieves health indicators better than some European countries, and does so by providing levels of access to medical services comparable to a developed country (Table 12). For five decades after 1950, Sri Lanka was a low-income developing country, with a per capita GDP less than US$350. Overall national health spending was only 3.0-3.6% ofGDP throughout, equivalent to less than US$10 per capita, well below the US$ 13 per capita cost of the minimum cost-effective package of services proposed by the World Bank (1993).114 Of this only half was public spending, which averaged US$4-6 per capita between 1950 and 1990. The bulk of Sri Lanka's health transition was achieved with national and government health spending being less in per capita terms than in the majority of Sub-Saharan African countries in 1990 (24 out of 30 spent more than US$ 5 per capita according to the W orId Development Report 1993). Table 12: Health services use and spending in Sri Lanka and comparable countries Country Year Physician visits per Inpatient admissions per Total expenditure on ca~ita ~er l:ear 100 ca~ita ~er l:ear health {% GDP} Bangladesh 2001 1 2 3.2 Egypt 1996 4 3 3.7 Indonesia 2001 1 1 4.2 Sri Lanka 2003· 5 22 3.6 Thailand 1993 2 8 3.6 Source: Estimated by authors from various sources. Technical efficiency at the micro-level 114 For the sake of comparison the US dollar per capita spending levels given in this and the following paragraphs are in constant 1990 US dollars, to pennit direct comparison with the spending estimates published in the World Bank World Development Report 1993 (World Bank, 1993). Good Practice in Health Financing, p. 338 The OEeD suggests three low-level indicators for assessing this (Hurst and Jee- Hughes, 2001) - unit costs, length of stay, and ratios of day cases to all surgery; Sri Lankan data are available for the first two. They show that Sri Lanka's public sector services are highly efficient in their use of available human and financial resources. Sri Lankan public hospitals deliver inpatient admissions and outpatient visits at a cost to per capita GDP ratio far lower than the average developing country, and in many instances at lower costs than any other country for which data are available (Table 1318). This is achieved by high patient throughput, reflected in bed-turnover rates and short average length of stay, and high levels of labor productivity with government doctors and nurses seeing, on average, more patients in both inpatient and outpatient settings than is the norm in other developing countries (Rannan-Eliya, 2001). Table 1318: Technical efficiency of public hospitals, Sri Lanka and selected countries Country Year Cost per Cost per Bed Average admission outpatient visit turnover length of stay (ratio of per (ratio of per rate -ALOS capita daily capita daily (days) GDP) GDP} Complex hospitals Sri Lanka 1997 7 1.0 65 5 Bangladesh 1997 26 0.8 47 11 Colombia 1978 25 0.8 38 7 Indonesia 1985 26 0.7 29 9 Jamaica 1985-86 40 1.5 35 8 China 1989 76 0.8 14 25 Basic and intermediate hospitals Sri Lanka 1997 5 0.1 57 3 Bangladesh 1997 14 0.5 77 4 Indonesia 1987 6 0.6 33 6 Jamaica 1985-86 18 1.1 32 8 Malawi 1987-88 17 0.4 47 9 China 1986 30 0.5 21 16 Source: Excerpted from Rannan-Eliya (2001) and Rannan-EIiya and Somanathan (2003) How these high levels of technical efficiency came about is not well understood. Yet it does appear that they are the result of incremental productivity improvements since the early 1950s or even earlier. Research on trends in productivity in public sector health services in developing countries has been almost non-existent (Hensher, 2001), but recent research by Sri Lankan researchers have suggested that such trends can be positive over long periods of time in many countries, and that the rates of improvement in Sri Lanka historically have been greater than in most (Rannan-Eliya, 2006). More specific empirical analysis of why Sri Lanka has performed so well in this area is lacking, but possible explanations include a strong public service ethos established in MOH by the 1950s, strong centralized control of budgets, inputs and operating procedures which minimized input prices and constantly forced health workers to meet increasing demand through efficiency savings instead of relying on more resources, and low administrative overheads associated with a civil-service run command-and-control management system (Hsiao and Associates, 2001). Good Practice in Health Financing, p. 339 Allocational efficiency Allocational efficiency refers to the correct allocation of available resources to different treatment or service interventions. This may refer to the allocation of expenditures by disease according to cost-effectiveness criteria, and or to the allocation of resources by service type. With respect to the former, the question is moot, as Sri Lankan health service managers have never allocated budgets by disease using cost-effectiveness criteria, and have not had the capacity to do so since budgets are not managed this way. With respect to the second, Sri Lanka has consistently followed a strategy of allocating the largest share of its budget to hospitals (75-85%), and within that to inpatient care (Table 14). Preventive and public health spending has averaged 25 percent or less of the budget, and during the past decade less than 12 percent. The hospital emphasis has been a long-standing feature of the system since the 1950s, and from a regional perspective is surpassed only by Hong Kong (Rannan-Eliya and de Mel, 1997). It is much higher than the 30-60% share seen in other Asian developing countries. Whilst this runs counter to standard prescriptions, in the Sri Lankan situation it made sense for two reasons. Firstly, a key goal of health policy, and one which clearly benefits the poor the most, has been protection against catastrophic risk, and a high hospital subsidy makes sense, considering that many patients are more able to pay for private outpatient services. Secondly, Sri Lanka has found that well-run government hospitals are an efficient way of delivering primary care, owing to economies of scale. Most government hospitals have only minimal capital investment and treat only simple illnesses, but are more cost- efficient than smaller outpatient facilities (Somanathan et al., 2000). Table 14: Proportion of MOH expenditures devoted to hospitals Year Recurrent expenditures Capital expenditures Total expenditures 1958 75% n.a. n.a. 1973 n.a. n.a. 65% 1986 77% 59% 75% 1991 78% 86% 80% 1996 74% n.a. n.a. 2005 n.a. n.a. 71 % Source: Derived by authors from various sources cited in Rannan-Eliya and de Mel (1997), and Sri Lanka Health Accounts database maintained at Institute for Health Policy. 3 Health Delivery System 3.1 Health services organization Sri Lankans benefit from extensive and organized health services, consisting of parallel public and private sectors. Public services are financed and provided in an integrated fashion by the Ministry of Health and eight provincial Departments of Health, and span the full range from preventive and basic primary care activities to complex hospital- provided tertiary care. The public sector network ranges from teaching hospitals with specialized services to small dispensaries that provide only outpatient services. Medical Officers of Health units (MOOHs) provide most preventive and public health services through teams of doctors, community midwives and others. Their organizational model was developed in the 1920s and then expanded during the 1930-40s expansion. Good Practice in Health Financing, p. 340 MOR facilities form an integrated network, with more than 1,000 institutions. The density of coverage is such that most Sri Lankans live within 2-3 kilometers of a public sector facility. Although there is a formal referral system with patients expected to use primary level services as the first point of contact, this is not enforced, so patients can seek care in the medical institution of their choice. The reason for this is related to the inequity in access that would arise, given the reality of variations in quality of service and the lack of organized general practitioner services, which might fully undertake the responsibility of acting as gatekeeper to accessing hospital services. Table 15: Provision of health service inputs and activities, Sri Lanka 2003 Resource Public sector Private sector Total Population ratio Physicians -7,000 -1,000 -8,000 0.4 per 1,000 Nurses -6,000 -5,000 -11,000 0.6 per 1,000 Hospital beds 59.262 -3,000 62,000 3.2 per 1,000 Inpatient admissions 4.1 million 0.3 million 4.4 million 22 per 100 Outpatient visits 45 million -50 million 95 million 4.8 per capita Source: Computed by authors from data provided by Medical Statistician, Ministry of Health and IHP databases. Table 16: Trends in treatment sources used by sick persons, Sri Lanka 1978-2004 Source of treatment 1978179 1981182 1986/87 1996/97 2003/04 Western government sector 42.6% 45.6% 44.1% 50.7% 43.5% Ayurvedic government sector 1.9% 2.2% 1.9% 2.0% 1.2% Western private sector 34.3% 34.2% 37.2% 38.1% 45.1% Ayurvedic private sector 16.1% 12.1% 12.9% 7.6% 5.0% Others 5.1% 6.0% 3.8% 1.7% 1.6% Source: Central Bank Consumer Finance Surveys based on published reports and analysis by authors. Note: The percentages are for those who reported falling ill during a 14 day reference period, and used any source of treatment. Western private includes private clinics, private hospitals and pharmacies. The largest part of private sector provision is ambulatory, with most outpatient services provided by government medical officers working in their off-duty hours. There are a smaller number (~1 ,000) of full-time private doctors, but they are concentrated in urban areas. This is supplemented by a private hospital sector which provides inpatient and tertiary services; this is concentrated in the Colombo district. The overall level of health services use in Sri Lanka is high in comparison with comparable countries, and the overall cost of the health system is low (Table 1318 ). The public sector dominates inpatient provision (>95%), but shares the outpatient load with the private sector (Table 16), whose contribution has been increasing in recent years (Table 16). 3.2 Growth of health service provision Before the 1930s, the supply of modern medical care was limited to government, missionary and private hospitals in urban areas, and provision for the plantation workforce by the estate companies. There was little in rural areas, as there was insufficient demand and purchasing power to make private sector investment feasible. The reforms starting in the 1930s changed this. Although there was significant construction of rural facilities in the 1930s, expansion was constrained by the difficult economic conditions of the time and later during World War II, and so was only enough to keep up with popUlation growth. After the end of the war and the recovery in the economic situation, the hospital-building program accelerated, and provision grew significantly faster than the popUlation. The number of government hospital beds Good Practice in Health Financing, p. 341 increased from 1.9 per 100 before 1945 to just over 3.1 by 1960 (Figure 7). Since then hospital expansion has continued, but only so as to keep up with population growth. Figure 7: Government hospital provision, Sri Lanka 1920-2000 (beds per 1000 o ulation 3.5 3.0 2.5 2.0 1.5 1.0 1920 1940 1960 1980 2000 Without the reforms, rural people would not have had equitable access to basic medical services. Today, although higher-level facilities and services are located only in urban areas, they are still accessible to the rural population, owing to relatively short distances in Sri Lanka between rural areas and towns and cheap public transport. Consequently, urban-rural differentials in access are minimal, with respect to availability of services. The differences that do remain relate more to quality differentials and the increased travel costs that rural people face in accessing tertiary care. 3.3 Medical education and regulation Medical education and licensing The public sector is responsible for physician training, and in the past decade the annual output of medical graduates has reached a thousand. The public sector universities operate medical schools. Admission is based strictly on marks in the school-leaving examinations, and tuition is free, as with all public sector university courses. The medical school curriculum is based on the British system, and is a five-year course. All medical graduates must complete an internship in the public sector before they receive a full license to practice. Placement in an internship is strictly according to merit, and Sri Lankans who have obtained medical degrees abroad are given last preference. The duration of the Internship is two years and is in government hospitals. For those who are low in the merit list, the placements are in rural areas, but refusal to undertake the allocated internship effectively debars a medical graduation from ever obtaining a license, so compliance is good. Further training to become a specialist requires both formal postgraduate examinations, as well as obtaining necessary expertise. This post- graduate training is only available in the public sector, including government-funded placements overseas, and is a major incentive for doctors to remain in the public sector. Good Practice in Health Financing, p. 342 Once a doctor receives a full license, he or she may practice indefinitely. The Sri Lanka Medical Council, which is a statutory body, regulates physicians and has the ultimate power to remove the license. However, this power is almost never exercised. Although the law does allow for cases to be brought against physicians on the grounds of professional negligence, this is rare owing to the difficulty of proving these cases in court, and lack of awareness by patients of their legal rights. Regulation of private sector Sri Lanka follows a laissez-faire policy of light regulation with respect to the private sector. Private doctors can practice as they wish, as long as they have the basic license. In addition, private hospitals are barely regulated, and government does not even maintain a requirement for these institutions to register centrally. The government also does not attempt to control prices in any way, although a statute to do so has been proposed by the health ministry for more than a decade, without it passing Parliament. The small private medical insurance market is not subject to specific regulation, although it is regulated as part of the overall insurance and financial services sector. 3.3 Pharmaceuticals and medical technology Pharmaceutical supply and regulation Sri Lanka's public sector initiated policies for the control and management of medicines in government hospitals as early as the 1950s, several decades before WHO adopted the concept of the rational use of drugs. These policies include the adoption in the 1950s of a national hospital formulary, which prescribes which drugs are approved for use in government hospitals, a policy of purchasing public sector drugs only through international tender and bulk purchasing, the use of only generic medicines in the public sector, and the adoption of a national essential drugs list. Most government drugs are purchased centrally and then distributed to medical facilities and the provinces. However, it should not be assumed that the existence of these policies implies a limited range of medicines in the public sector. The health ministry's essential drug list contains several thousand products, and MOH purchased almost 3,000 different drugs and medical items in 2005. Nevertheless, not all these medicines are available in all hospitals. For reasons of cost control, the number of drugs that lower level hospitals can stock is controlled through administrative procedures based on an assessment of how essential a drug is, and the more expensive drugs are restricted to tertiary facilities. The private sector may import any drug that is registered with national authorities. In practice, any drug that is registered in the USA and Europe will be registered in Sri Lanka, if the manufacturer files an application. Until 2002, prices of imported drugs were controlled by setting the retail price at a maximum of 165% of the import price. Medical technology There is no policy to control medical technology, other than a basic registration requirement. However, adoption and purchase of expensive high technologies are tightly controlled in the health ministry using managerial procedures. For example, the purchase of CAT scanners and MRI scanners was long delayed in the public sector, despite pressure from medical specialists, and only small numbers have been purchased. Good Practice in Health Financing, p. 343 The private sector may import any registered medical technology, but until recently the implications of this were limited, since tertiary medical care was effectively a public sector monopoly. However, in the past decade with growth in investor-financed tertiary private hospital facilities in Colombo, the private sector has begun to import some of the more expensive high technologies, such as MRI scanners. 3.4 Strengths and weaknesses The main strengths of Sri Lanka's system are its equity, high system efficiency, good health outcomes, and relatively low costs for government and households (Hsiao and Associates, 2001). It is able to provide effective protection to the poor against financial risk of illness, and is able to ensure that the poor are can access a high volume of basic medical services. The incidence of government spending is also not pro-rich as in many developing countries. In addition, the system of financing is progressive as taxes are relatively progressive, and out-of-pocket spending is more by the richer households. The weaknesses of the system are primarily the result of the under-funding that has arisen, because of the inability of the government to increase the budget owing to its failure to raise taxes sufficiently (Hsiao and Associates, 2001). Under-funding means that hospital services do not meet the needs and demands of richer households and has resulted in increasing use of private services in recent years. There is a recognized risk that if this trend continues, the continued shift of patients out of the public system may destabilize the health system as a whole, and undermine political support for government health services. The system has also failed to adopt modem methods for management and treatment of chronic, non-communicable diseases, such as ischemic heart disease. This is a growing challenge owing to the stagnation in male adult life expectancy. However, a reorientation is not feasible owing to lack of funding. A related problem is that as consumer expectations have increased with time, the health system is not able to respond to these non-health needs, such as better amenities in government hospitals and a more consumer-oriented approach by staff. Again, the organizational changes and flexibility that would be required as a response cannot currently be implemented until the financing gap is resolved. Good Practice in Health Financing, p. 344 4 Health Coverage Reforms 4.1 Sri Lanka - the early reformer The good and equitable health outcomes in Sri Lanka result from healthcare reforms that commenced after 1931, and are not an inevitable outcome of culture and history. The main, scaling-up phase lasted from 1931 to 1959. Prior to reform, health conditions in the island were no better and mortality rates were higher than the South Asian average (Langford and Storey, 1993). The subsequent health achievements of Sri Lanka are chiefly due to the role of its health services in reducing mortality and morbidity, and the success in expanding coverage of its health system. The scaling-up phase of reforms put in place all key features of the current Sri Lankan health system, including high levels of coverage. Subsequent developments have been incremental, but have not disturbed the core elements. Therefore, to understand how Sri Lanka reformed its health system to achieve high levels of coverage of the poor with health services, we must focus on the formative 1930-50s, and to understand why on the years before that. T a bl e 17 : Ch rono 02)' 0 f searmg-up pih ase 0 f Sn. L an k' as h eaIth re ~orms Year Event or Reform Government Inpatient IMR expenditure admissions {%GDPl /100 cal!ita 1858 Establishment of first government medical <0.5% <1.0 n.a. department 1931 Democracy - First national elections on 1.4% 3.3 158 basis of universal franchise and transfer of power to elected Sri Lankan leaders 1934-35 Ceylon Malaria Epidemic ravages economy 1.5% 4.3 263 of rural areas 1936-47 Expansion of government medical services 0.9-1.7% 5.5-7.9 120 into rural areas 1948 Commission on Social Services rejects 1.7% 9.3 101 introduction of social health insurance, recognizing that direct government hospital provision is a form of insurance 1951 Abolition of user fees at government 1.5% 11.0 82 hospitals 1959 End of increases in health budget 2.1% 14.0 58 4.2 The precursors to Sri Lanka's health reforms There are several historical reasons why Sri Lanka was able to and chose to embark on its health reforms. A history of state intervention The public financing of health services in Sri Lanka dates back at least 2,300 years. In the pre-modern era, there was a tradition of Sri Lankan kings establishing public hospitals at state expense, with funding from government revenues. Their motivation was the prospect of merit, which Buddhism taught would accrue to the builder, as well the importance that Buddhist values place on the alleviation of human suffering Good Practice in Health Financing, p. 345 (Rannan-Eliya and de Mel, 1997). This was in a context, where the Theravada strand of Buddhism, which is dominant in Sri Lanka, encouraged a close nexus between religion and state action. Unlike in pre-modem Europe, most hospitals were built in Sri Lanka not through private charitable action, but as state initiatives, with records dating the earliest such facilities to the 4th century B.C. (Uragoda, 1987). State financing of health services collapsed in the 13 th Century, with internal conflict and foreign invasion, and the resulting collapse of the system of public revenue collection. Although, the period from then until the end of the colonial period represents a clear discontinuity in social policy development, contemporary public attitudes in Sri Lanka, which regard the state as having primary responsibility for providing health care, echo these earlier traditions and find support in contemporary religious thought. Exposure to western medicine During the colonial era, the Dutch and British occupiers opened a few urban hospitals for the benefit of colonial officials and European residents. They were financed by a mix of user fees and general revenues, and were beyond the reach of the Sri Lankans who lived in rural areas. For them the only option was to obtain treatment from traditional doctors who, in accordance with Buddhist tradition, charged fees only for the dispensing of medicines, and not directly for diagnosis. Nevertheless, the colonial medical services were important in introducing the concepts of modem scientific medicine. Sri Lankan culture, much as in Japan and Kerala, proved receptive to the western biomedical model, as it blended easily into a context where individuals would look for signs of illness and doctors would treat with medicines (Caldwell et al., 1989). This ready cultural acceptance of scientific medicine partly explains the alacrity with which Sri Lanka's rural population later welcomed expansion of health services. In the 19th Century, the British occupation authorities made efforts to extend health services, by introducing in 1880 a .scheme for the plantation companies to provide basic medical services to their workers, to be financed by reimbursements of a cess on exports. This was motivated not by humanitarian reasons, but because of the importance of these workers to the plantation economy and pressure from the Indian government. Consequently, by the 1930s the estate workers enjoyed better health services than the rural population, and better health indicators. After 1931, these plantation services became a model for the rural population to aspire to. The colonial state and introduction of democracy The British occupation was marked by the development of the plantation economy. Plantation exports were easily taxed, and provided the authorities with a ready source of revenues. In 1848, the British defeated the last armed rebellion against occupation. This event was an important milestone. Firstly, the rebellion was essentially a peasant revolt, and marked the transition in Sri Lankan society between leadership by feudal elites to one based on the support of ordinary people. Secondly, the occupation authorities no longer faced internal threats requiring maintenance of a substantial military, and tax revenues were invested in building a physical and social infrastructure servicing the plantation sector, as well as an administrative structure that was more sophisticated and substantial than in the rest of South Asia. One indicator is that by the 1930s, Sri Lanka had a functioning vital registration system registering most births and deaths. Good Practice in Health Financing, p. 346 The relatively advanced state of institutions in the island, led the British in 1928 to embark on a major experiment in social engineering, by granting Sri Lankans self-rule on the basis of representative government elected through universal suffrage. This transferred power not to the local elites but directly to the majority of the population, and was opposed by the leading Sri Lankan politicians. Its radical nature is evident as the first elections on the basis of universal suffrage were held in the UK only in 1929. In introducing democratic government there was an explicit recognition that democratic accountability to voters would promote health conditions and reduce mortality: 'We have given serious consideration to the question of women's franchise. Apart from the familiar arguments in its favour, and the general principle of sex equality, we have been impressed by the high infantile mortality in the Island, and the needfor better housing, andfor the development of child welfare, midwifery and ante-natal services, all providing problems in the solution of which women's interest and help would be ofspecial value, It is true that though the position of women in the East has not, till recent years, been suitable for the exercise ofpolitical power, that position is rapidly changing and the demand for the vote was put to us by a large and representative deputation of Ceylonese ladies' (Government of Ceylon, 1928). The first legislature selected through universal franchise was elected in 1931, and British civil servants ceded all responsibility for domestic policy to Sri Lankan politicians until independence in 1948, when the new Sri Lankan government took over responsibility for foreign and security affairs. Competitive electoral politics during the following two decades were to drive all major changes in social policy, including the introduction of personal income tax (1932), the expansion of government health services to rural areas (1931-1940), the introduction of free education (1930s-40s), and the abolition of user fees for health services (1951). The lessons of market failure in the health sector The period leading up to 1931 created the preconditions for Sri Lanka's health reforms: (i) a population receptive to modem medical services, (ii) a model of healthcare provision in the plantations which the rural population could aspire to, (iii) a fiscal base adequate to support significant public expenditures, and (iv) a political mechanism to translate social preferences into actual policy. The missing element was the understanding by policymakers of the need for reform of the existing health system. This came about as a result of an unprecedented health crisis, the Ceylon Malaria Epidemic (Rannan-Eliya and de Mel, 1997). During 1934-35, owing to unusual climatic conditions, a major epidemic of malaria spread to all areas of the island, hitting most severely normally non-malarial areas, where the population had no natural immunity. Most of the population was infected, and a significant fraction died. When this hit, the politically conservative and newly elected Sri Lankan government left the response to market forces, primarily charitable action. This proved totally inadequate in face of the disaster, which followed a deep rural crisis induced by the impact of the global recession of the early 1930s. Other than direct morbidity, the epidemic economically devastated rural areas, as farmers were left too ill to cultivate their crops, and their family members were burdened by the Good Practice in Health Financing. p. 347 responsibility of caring for the sick, and thus prevented from attending to their normal responsibilities. In the absence of state intervention, the Marxist opposition parties of the time made much political capital by organizing missions to assist the affected rural poor. Although these were not that effective, they created political alarm in the elite. After the epidemic, there was an official government inquiry, which observed that the health crisis had forced rural households into poverty, and that relying on charitable and market actions was totally inadequate to deal with the challenge. It concluded that there was a need for direct state intervention through provision of hospitals, which could care and feed the sick, so as to help the affected families survive such events. Interestingly, this series of events where recognition of the financial impoverishment created by ill- health in rural areas led through official reports to increased political pressure to expand health services is almost identical to the sequence of events in Japan in the early 1930s which led the Japanese government to embark on the eventually successful effort to extend health insurance coverage to all its people (Hasegawa, 2005). In effect, Sri Lankan state and policy-makers realized very early that health was not just an individual matter, and that health policy was more than simply curing disease. At the time, there was anyway little effective treatment for malaria. They understood that ill health had economic implications and was linked to poverty, that the market would fail to provide effective health insurance against catastrophic risks, and the public sector had a crucial role to play in providing this. They also realized that failure to address the market failure would carry significant political risks in the new political environment where the poor had a vote. The recognition that direct public provision of hospital services was a form of social insurance would be later explicitly stated by the influential Commission on Social Services (1947). 4.3 Health system reforms Expansion of health services into rural areas 1931-51 From the early-1930s, the government launched significant efforts to expand the availability of free health services in rural areas, primarily through building and staffing of hospitals and dispensaries. The immediate impact of this expansion was obscured by financial constraints during World War II. When the economy returned to normal after 1945, there was an immediate increase in the quality of services at the new facilities as staffing and the supply of the new antibiotics improved. This fed directly into the jump in life expectancy that took place in the years after 1945 (Langford, 1996). The initial expansion of the service network chiefly involved the building of new health ministry facilities in places where there had been none. At first these were simple dispensaries and unsophisticated rural hospitals, but they served to bring most of the population within a short distance of some treatment point and encourage them to try modem medicine instead of traditional care. The construction program was promoted by the lobbying of the health ministry by individual members of parliament which, in effect, ensured that all electorates ended up with at some provision. This lobbying was biased in favor of rural areas, as rural electorates had fewer voters on average than urban electorates, and thus were over-represented in parliament. In addition, until the late 1940s, party political organization was weak, and most legislators competed for election not on the basis of a party platform, but their ability to bring back benefits to their constituents, such as hospitals, schools and roads (Wriggins, 1960). By 1945, the Good Practice in Health Financing, p. 348 health ministry was operating more than one thousand hospitals and dispensaries for a population of only seven million. Later, after the network was in place, further expansion shifted to upgrading and expansion of existing facilities, a process that continues today. Sri Lanka thus ended up by the 1950s with a highly dispersed facility network. The importance of this for reaching the poor is that econometric analyses of demand for health services in developing countries now find that distance is as important a factor as quality. By building so many facilities, the distance that the poor have to travel to obtain care was reduced, and a key barrier to access was removed. Abolition of user fees In common with most British colonies, government hospitals were charging user fees prior to the 1930s. However, there was a policy to provide a means-tested exemption to those considered poor, and the income limit for this was set so high that most rural patients did not pay. Yet in the new electoral scenario that emerged after 1931, even these limited fees were considered unreasonable, and user fees were abolished in 1951 by the UNP government, which was in power at the time. Subsequently, a Trotskyite finance minister did reintroduce them in 1971, but they caused a substantial and immediate reduction in demand, especially by the poor. They were soon abolished again in 1977 by the next UNP government. Consequently, in Sri Lanka user charges have never been a barrier to access by the poor, and the national policy of free care has been firmly supported by all major political parties, and in fact instigated by the most pro-market of them. The emphasis on hospitals The health reforms from the beginning placed a heavy emphasis on the role of hospitals in extending coverage. Unlike many developing countries today, Sri Lankan health planners recognized from the beginning not only the insurance function of a hospital, but also the fact that most illnesses could not be prevented through preventive programs, and thus needed to be treated through curative interventions. They thus believed that hospitals needed to play the lead role in combating illness, and allocated budgetary resources accordingly. The prioritization of hospitals in the health ministry budget was thus an important feature of official health policy from the 1930s, and represented a major change from previous British colonial policy that had focused on preventive, sanitation and quarantine measures for the rural population. Indigenization of medical department The health ministry that executed these reforms was distinctively different to other ministries. In the latter part of the British occupation, government departments became no longer the preserve of British civil servants, and Sri Lankans were progressively recruited into them. This process of indigenization had most advanced in the case of the medical department, as the British regarded it as the least important or prestigious. Not only were almost all its personnel Sri Lankan by 1930, but it was also the first department in which it was permitted for the most senior civil service position to be filled by a local. This departmental history was important. It imbued the ministry with a distinctive pro-poor attitude, as its personnel took pride in the fact that it was the first to be controlled by Sri Lankans, and this, in tum, encouraged an ethos in which the mission of the ministry was seen as "serving the masses". This contrasts with Good Practice in Health Financing, p. 349 neighboring India, where the Indian Medical Service was seen as an agent of the occupiers (Jeffery, 1988). Trade off of quality versus access As rural access to services improved in the 1940s, the health ministry was met with sustained surges in patient demand that were wholly unanticipated. For example, in 1948 total inpatient numbers increased by 22% year-on-year, and outpatients by 30%. Government hospitals were filled beyond their design capacity, and average bed- occupancy rates of more than 200 percent were common. In this situation, medical personnel and planners often pushed for measures to restrict demand, including closing hospitals when bed-occupancy breached the official limit of 200 percent, since admitting more patients would damage treatment quality. However, political pressures not to restrict access made it impossible for the health ministry to accede. In several instances, doctors were dismissed from service for implementing departmental rules on over-crowding, after patients who were refused admittance complained to their legislators, who then complained to the health minister. In this situation, an implicit trade-off was made to prioritize access to services over service quality. Although it had the perverse result of making overcrowding inevitable, it benefited the poor, since any measures that might have restricted demand would have affected the poor inevitably more than the rich. Productivity improvement The ambitious expansion of health services that was planned and achieved was expensive. After the mid-1950s, fiscal constraints made it impossible for the government to increase the health budget, and subsequently, health spending fell as a share of national income. Despite this, political leaders would not accept any reductions in services by the Health Ministry or any policies to restrict demand. Faced with these conflicting pressures, the Health Ministry responded by relentlessly searching for productivity increases and pushing its staff to work harder. To do this it had to rely largely on administrative and managerial measures, ranging from simple changes such as reducing the minimum distance allowed between hospital beds to changing the pensions regulations, so that doctors were compensated adequately for working longer hours. Over time this created an organizational culture that has promoted continuous productivity increases. Compulsory posting and dual medical practice An important reform, which was introduced prior to the 1930s, that has facilitated expansion of coverage to rural areas concerns private practice. When the health department was first established, medical officers were not permitted to engage in private business, consistent with general civil service regulations. However, the department discovered that it was difficult to recruit medical officers from the UK to work in Sri Lanka owing to the difficult working conditions and low pay, and it was decided to permit medical officers to supplement their official salary by doing private practice outside official work hours and outside government premises. This enabled the department to recruit and retain medical staff, and this policy was in effect in the 1930s, by when most new recruits were local graduates. It has supported expansion of coverage into rural areas, as the health ministry cannot afford to pay market wages to Good Practice in Health Financing, p. 350 persuade doctors, but doctors can substantially raise their incomes by private practice. In rural areas where usually the government medical officers are the only physicians, private practice can be lucrative. During 1970-77, when private practice was abolished, the distribution of government doctors to rural areas suffered, as did overall retention of medical officers in the Health Ministry. The Health Ministry has adopted another policy to directly improve availability of medical personnel in rural areas. This has been to rotate all junior doctors on a regular basis, and compulsorily post many to rural areas. This has been enforced by firing doctors who refuse to comply which is a significant disincentive, as outside the public sector junior doctors cannot obtain specialist training. This policy uses the carrot of specialist training and seniority, which are necessary for doctors to earn the highest private practice incomes, to encourage doctors to accept lower than market wages during the early part of their careers. 4.4 Evaluation of reforms Expansion of coverage Between 1931 and 1951, Sri Lanka expanded access to health services by using direct government provision and building of a highly dispersed health facility network in rural areas. The reforms fundamentally altered the health system. They changed it from one in which the urban rich used modem medical services and rural people relied on traditional providers, to one in which the whole popUlation had easy access to and used modem medical services. This involved not only increases in coverage, but also a change in health seeking behavior as rural Sri Lankans were persuaded to switch from traditional to modem medical care. So effective was the expansion in coverage that Sri Lanka was able to achieve by 1951 quantitative levels of health service access comparable to many middle-income developing countries today, and to substantially equalize use of modem medical treatment between rich and poor. In addition, this expansion of coverage involved not only access to primary care services, but also to general hospital services including inpatient treatment. The aggregate increase in population coverage was 200-300% during the twenty-year period (Table 18). Table 18: Expansion of health service coverage, Sri Lanka 1931-1951 1927 1931 1936 1946 1951 Inpatient admissionsllOO capita 3.7 3.3 6.0 7.5 11.0 Outpatient visits/capita 0.5 0.6 1.0 1.1 1.5 Government health expenditure (% GDP) 0.8 1.4 1.7 1.3 1.5 Note: Data derived from IHP databases and official statistics. Data are given for 1927 to provide an indication of the pre-reform trends. Other years selected based on availability of data, which are incomplete for the 1939-1945 time period owing to wartime censorship of government administrative data. There is an important aspect to this expansion in coverage. Since the initial reforms in the 1930s, a major implicit goal of healthcare provision has not been to improve health outcomes, but to protect households against the catastrophic economic impacts of severe illness. The reforms have not only increased access to services, but it has also extended insurance protection to the poor in the form of hospital services. This has been possible because the public sector has allocated the largest share of the budget to Good Practice in Health Financing, p. 351 hospital services, and has implicitly accepted the role of the health services in providing insurance against the catastrophic financial risks of illness. Institutional change To achieve this expansion in coverage, Sri Lanka transformed its government health ministry from one whose mission was to serve the workforce of the occupation authorities to one whose ethos was to serve the rural people. Moreover, the expansion created a management culture, which focused on productivity improvement as the means to expand coverage within limited resources. This was done without changing the existing civil service structure or regulations, and relied primarily on organizational ethos and culture. Health outcomes \ The expansion in coverage enabled the country to rapidly reap the advances in medical technology that had been made globally in the previous half century, and substantially reduce mortality in all areas and in all popUlation sub-groups. Not only were health indicators dramatically improved after 1945 as a result of the better access to medical treatment, but the evidence indicates that a process was started which substantially reduced health inequalities between urban and rural and rich and poor Sri Lankans. The health reforms started a process of rapid mortality decline that still continues and has enabled Sri Lanka to complete its mortality transition in less than half a century. Cost implications of scaling up and targeting The expansion in services during the scaling-up phase of the reforms involved substantial increases in government outlays in funding. However, the government was not able after the 1950s to continue to increase funding in line with the increases in patient demand that started to manifest from the early 1930s. At the end of the scaling- up phase, the reforms could easily have undermined themselves with the increase in demand so out-stripping the capacity for financing that the underlying principle of universalism in access would have had to be sacrificed. The fact that this did not happen can be attributed to two factors: (i) technical efficiency gains in the public sector that substituted substantially for the lack of additional funding, and (ii) the implicit targeting that was achieved by interaction with private sector supply. As already noted, more than half the increase in service volume was achieved through productivity increases. By the 1950s the government was able to start reducing health spending as a share ofGDP, whilst still continuing to expand and improve services. The other contribution to meeting the financing gap involved in committing to universal coverage has come from the dynamics of demand for public and private services. As noted, the health ministry has permitted dual practice by its medical officers since the th 19 Century, and takes a laissez-faire attitude to the private sector. Consequently, in all areas there is some private supply of medical services, and in rural areas these are almost all outpatient services provided on a fee-for-service basis by government medical officers. Even in urban areas, where there is significant private inpatient provision, the physicians attending to private inpatients are usually government specialists, or they are full-time private GPs, most of whom previously worked a Good Practice in Health Financing, p. 352 significant number of years in the public sector (Rannan-Eliya, Jayawardhane, and Karunaratne, 2003). Consequently, there is little difference in the technical quality of medical services in the public and private sectors. The only reason that patients choose to pay for private services, when they can, in theory, see the same doctors for free in government hospitals, is that in the private sector the hotel amenities are better, the facilities are less crowded and queues shorter, the opening times can be more convenient, patients can choose their doctor, and the doctors and staff will spend a little more time with the patient, and are perhaps more polite and courteous. In this context, what happens is that it is the richer patients, who have a higher opportunity cost of waiting, and who have a greater preference for the consumer aspects of quality, voluntarily choose to use the private sector. This is reflected in a strong pro-rich gradient in the use of private services, and a pro-poor gradient in the use of public services. However, this is only seen for outpatient services, and for inpatient services, which few Sri Lankans without insurance can afford to pay for out-of-pocket, the gradients are less obvious. The advantages of this situation is that it allows the government to effectively target its spending to the poor without sacrificing its explicit commitment to universalism, and that it relies on the rich revealing their own preferences by voluntarily opting out, instead of having to identify them directly which would be quite difficult. Politics and sustainability There is no evidence to suggest that Sri Lankan politicians are more altruistic and concerned for the poor than those of any other country. When the country was struck by an unprecedented health disaster in the form of a devastating malaria epidemic in 1934- 35, they mostly stood by and left the population to fend for itself. However, the existence of a political system where political power must be obtained though the ballot box and where the voters are not disinclined to throw-out incumbents, taught political leaders to be highly sensitive to the demands and preferences of the population, including the rural poor. This led the political leadership to embark on health reforms that invested significant public funds in expanding access to free health services, and subsequently to ensure that this system was maintained and improved. This sensitivity to social preferences has also been behind the prioritization of budgetary resources to hospitals. To this extent it can be said the current health care system is sustainable, as the political rationale still exists for governments to maintain the system, as long as the population favors it. At the same time, the political system has been effective in transmitting its concerns about financial costs to the health ministry, so overall government health expenditures in Sri Lanka have never been high by international standards. 4.5 Key conditions The key conditions for Sri Lanka's success in expanding access were as follows: Democracy - A competitive electoral process has ensured that the political elite must take into account the preferences and needs of the poor, and forced policy makers to build a health system accessible in all areas of the country. (i) Taxes - The early availability of a ready tax base in the form of export taxes on tea, rubber and coconut enabled the government to provide the public financing that was necessary to pay for service expansions for the rural poor. Good Practice in Health Financing, p. 353 (ii) A committed and efficient public sector - Sri Lanka still faced the challenge of how to expand services within limited budgetary resources. It was able to do this primarily because its centralized civil service management of health care services provided it with an efficient means of expanding services and seeking continuous productivity gains. (iii) The willingness to rely on a pragmatic approach to targeting services which involves adopting an explicit commitment of universal access, but implicitly allows the rich to opt-out and use private services thus allowing government spending to be focused on the poor. (iv) The existence of models of emulation - at various points, Sri Lankans were able to copy or emulate or aspire to introduce models of health care delivery. These provided them with a template for transformation, and reduced the need to experiment. This included the Western model of biomedicine introduced by the European occupiers, the provision of free health care to estate workers and civil servants, and the well-organized health ministry that attempted to follow a British model. 5 Lessons for Other Countries Sri Lanka offers many lessons to those concerned with expanding access to basic health services to the poor in the context of a developing country. 1. Democratic accountability can ensure that health systems are responsive to the needs of the poor For health systems to be responsive to the needs and preferences of the poor requires that the political system itself be responsive to their wishes. A democratic political system in which the poor are given voice through regular free and fair elections can be an effective mechanism to achieve this, although the scale and level of electoral representation are important mediating factors. Democratic accountability should not be confused with community participation or cost-sharing or political decentralization. Sri Lanka has had little success with community participation in local government or in the running of individual facilities, but the most basic decisions about the health system are not taken at the local level, but nationally, and that is where the poor have had a voice. In fact the most important issue for the poor, which is the redistribution of resources from the wealthiest parts of the country to the poorest, can only be resolved through central government action. Nor should it be thought that well-meaning technocratic intentions will permanently guarantee attention to the poor. Only when a political system has incentives to consider the interests of the poor will it actually do so in a sustained manner. Political considerations frequently override technical concerns, and Sri Lanka's history provides numerous examples of public choices being superior in welfare terms to the choices of experts. At the same time, for a health system to be able to act on the preferences of the poor it must not emasculate its capability to do so. Building and maintaining efficient systems for public revenue collection and public production of health services are integral to making democratic accountability real in poor countries. Good Practice in Health Financing, p. 354 2. Equity of access should be a priority goal of health systems Equity of access to services must be a priority goal if health gains are to reach the poor. All health systems face resource constraints. It is unavoidable that there will be actions to match supply and demand, but almost all rationing mechanisms have negative implications for the poor. Unless equity of access is the highest priority, choices about rationing will be made which, inevitably, negatively affect the poor. Sri Lanka's democratic system ensured that equity of access was the priority goal. This led to the disavowal of user fees and other financial barriers to accessing health services, the building of a highly dispersed government health care system which reduced non- financial barriers to access by reducing travel costs, the imposition of incentives for government providers to accept all patients regardless of cost and quality implications, and the outranking of concerns for rational referrals by the principle of free choice of providers. 3. Health systems must provide the poor with insurance against catastrophic illness Independent of the impact on health status, health systems make an important contribution to the welfare of the poor when they provide insurance against the economic costs of severe illness. This has been one of the most important objectives of Sri Lankan health policy, recognized as early as 1935. Its importance is reflected in the highest priority being given to inpatient services in government budgetary allocations. The economic costs of illness include not only the cost of medical treatment, but also the care and feeding of the patient, and the loss of income as household members must divert time from other activities to caring for the sick or substituting for their normal household contribution. Such insurance cannot be provided by charity or by private markets in any country, and only public action can provide this either through provision of free inpatient services or through some other method of social insurance. 4. Efficiency in health service production is more important than resource mobilization in overcoming resource constraints Equity of access in resource-constrained environments requires that health services be efficient producers. Sri Lanka achieved high levels of access to modem medical care not by rationalizing the services delivered, but through continuously achieving efficiency gains in its public sector of the order of 1-4% per annum. The ability to achieve efficiency gains was greater in Sri Lanka's case than the ability to mobilize additional resources. The potential for such efficiency gains in health systems should place serious doubts about assuming that resource constraints represent insuperable barriers to ensuring better health for the poor. Sri Lanka completed its health transition and achieved levels of access to health services similar to OEeD economies whilst spending less than 50% of the World Banks stipulated "minimum cost-effective package", and without resorting to user fees, community financing or insurance. 5. Pessimism about the relative inefficiency of public sector health service production is as unwarranted empirically, as it is theoretically Although Sri Lanka's experience is unusual, it should not be assumed that the public sector is inherently incapable of efficient delivery of services, nor is it necessarily the most inferior of possible alternatives. Sri Lanka's government hospital system has been Good Practice in Health Financing, p. 355 effective both in achieving high efficiency and in generating continuous efficiency gains. This was done without giving autonomy to individual hospitals, without changing civil service conditions of employment, despite the existence of strong health sector unions and without decentralization. More attention by researchers and by policy-makers needs to be given to understanding the determinants of efficient performance by public sector providers, and perhaps more than is currently given to maximizing efficiency of private sector provision. Sri Lanka (and the experience of several countries with similar health departments, such as Malaysia, Hong Kong and Mauritius) suggests that within the spectrum of public sector organization patterns, there remain unobserved and poorly understood possibilities for excellence in public sector delivery. Since Sri Lanka neither pays its health personnel well nor uses performance-related financial incentives, the evidence indicates that non-financial incentives and organizational culture are more important determinants of performance. 6. Cost-effectiveness of interventions and a disease-focused approach to allocational efficiency are irrational and inefficient guides to resource allocation and may lead to erroneous use of resources Sri Lanka never relied upon cost-effectiveness as a guide to allocating resources. Practically, allocating resources according to specific diseases and interventions would have been mostly impossible given that resources were budgeted by facility and input type, and that decisions about treatment are taken at the lowest level by physicians. Even if such decisions could be implemented, Sri Lanka never had the data or expertise to make the necessary calculations of cost-effectiveness. The Sri Lankan experience fundamentally contradicts the basic conclusions of most cost-effectiveness analysis. Sri Lanka, under the influence of its political process, has tended to allocate government budgets according to where they would have the greatest marginal welfare benefits in the context of social inequality and dual public and private markets. Inpatient care has been favored, not because of its health impact, but because of underlying welfare gains in terms of risk protection, and the inability of private markets to ensure adequate supply. Routine primary care receives a lower share of government budgetary resources, since many households are more likely to be able to finance adequate care themselves from private sources. 7. Use of differentials in consumer quality in a dual pUblic-private system can be a more effective mechanism of targeting health subsidies than explicit targeting Sri Lanka found that targeting any subsidy to the poor faces the informational constraint that governments lack sufficient information to accurately distinguish between the poor and rich. This is the same information constraint that prevents poor countries from relying on income taxation as the main source of general revenue. Consistent with the insight of the theory of optimal income taxation that tax and subsidy systems should be incentive compatible, a health system can efficiently target health subsidies by enforcing universal access and relying on differentials in consumer quality to persuade richer individuals to self-select private services. Such a system, as in Sri Lanka, is politically sustainable as long as sufficient people continue to use the public service. That this is a general attribute of the health system is demonstrated by similar outcomes in several other ex-British dependent territories with similar health systems, such as Jamaica, Barbados, Mauritius, Malaysia, Hong Kong and Ireland. Good Practice in Health Financing, p. 356 REFERENCES Alailima, Patricia J., and Faiz Mohideen. 1983. Health Sector Commodity Requirements and Expenditure Flows. Unpublished report. Colombo, Sri Lanka: National Planning Department. Bruton, Henry J. 1992. Sri Lanka and Malaysia. D. Lal and H. Myint, The Political Economy of Poverty, Equity and Growth. New York, USA: Oxford University Press. Caldwell, J.C. 1986. Routes to Low Mortality in Poor Countries. Population and Development Review 12 (2): 171-220. Caldwell, John, Indra Gajanayake, Pat Caldwell, and Indrani Peiris. 1989. Sensitization to illness and the risk of death: An explanation for Sri Lanka's approach to good health for all. Social Science and Medicine 28 (4):365-79. Central Bank of Sri Lanka. 2006. Annual Report 2005. Colombo, Sri Lanka: Central Bank of Sri Lanka. Commission on Social Services. 1947. Report of the Commission on Social Services. Sessional Papers. 1947: VII. Colombo: Ceylon Government Press. De Silva, M.W. Amarasiri, Ananda Wijekoon, Robert Hornik, and Jose Martines. 2001. Care seeking in Sri Lanka: one possible explanation for low childhood mortality. Social Science and Medicine 53:1363-1372. De Silva, W. Indralal. Forthcoming. A Population Projection of Sri Lanka for the New Millenium, 2001-2101: Trends and Implications. IHP Research Papers. Number 1. Colombo, Sri Lanka: Institute for Health Policy. Government of Ceylon. 1928. Report of the Special Commission on the Constitution. Colombo, Sri Lanka: H. Ross Cottle, Government Printer. Hasegawa, Toshihilko. 2005. Japan. In Social Health Insurance: Selected Case Studies from Asia and the Pacific. New Delhi, India: WHO Regional Office for South-East Asia and WHO Regional Office for Western Pacific Region. Hensher, Martin. 2001. Financing Health Systems through Efficiency Gains. CMH Working Paper Series. Paper No. WG3:2. Geneva, Switzerland: Commission on Macroeconomics and Health. Hsiao, William C, and Associates. 2001. A Preliminary Assessment of Sri Lanka's Health Sector and Steps Forward. Cambridge, MA, USA: Harvard University. Hurst, Jeremy, and Melissa Jee-Hughes. 2001. Performance measurement and performance management in OECD health systems. Labour Market and Social Policy Occasonal Paper. No. 47. Paris, France: OECD. Institute for Health Policy. Forthcoming. Sri Lanka National Health Expenditures, 1990-2005. Colombo, Sri Lanka: Institute for Health Policy. Good Practice in Health Financing, p. 357 Jeffery, Roger. 1988. The Politics of Health in India. Berkeley, USA: University of California. Langford, Christopher M. 1996. Reasons for the decline in mortality in Sri Lanka immediately after the Second World War: a re-examination of the evidence. Health Transition Review 6 (1):263-282. Langford, Christopher, and Pamela Storey. 1993. Sex Differentials in Mortality Early in the Twentieth Century: Sri Lanka and India Compared. Population and Development Review 19 (2):263-282. Meegama, A. 1986. The Mortality Transition in Sri Lanka. In Determinants of Mortality Change and Differentials in Developing Countries: The Five-Country Case Study Project. Population Studies, No. 94; ST/ESAISER.Al94, Pub. Order No. E.85.xIIIA. New York, USA: United Nations. Rannan-Eliya, Ravi P. 2001. Strategies for Improving the Health of the Poor - The Sri Lankan Experience. April 20, 2001. Colombo, Sri Lanka: Institute of Policy Studies Health Policy Programme. Rannan-Eliya, Ravi P. 2006. Productivity Change in Health Services: An Empirical Analysis and Exploration of Institutional Determinants. Unpublished paper prepared for publication by Global Development Network. Colombo, Sri Lanka: Institute for Health Policy. Rannan-Eliya, Ravi P., and Nishan de Mel. 1997. Resource Mobilization for the Health Sector in Sri Lanka. Data for Decision Making Publication. Boston, MA, USA: Harvard School of Public Health. Rannan-Eliya, Ravi P., Prashanthi Jayawardhane, and Leela Karunaratne. 2003. Private primary care practitioners in Sri Lanka. In Health Policy Research in Asia: Guiding Reforms and Building Capacity, edited by A. S. Yazbeck and D. H. Peters. Washington, DC, USA: World Bank. Rannan-Eliya, Ravi P., and Aparnaa Somanathan. 2003. The Bangladesh Health Facility Efficiency Study. In Health Policy Research in Asia: Guiding Reforms and Building Capacity, edited by A. S. Yazbeck and D. H. Peters. Washington, DC, USA: World Bank. Rannan-Eliya, Ravindra P. 2004. Towards a Model of Endogenous Mortality Decline: The Dynamic Role of Learning and Productivity in Health Systems. A thesis submitted to the Faculty of the Harvard School of Public Health in partial fulfillment of the requirements for the Degree of Doctor of Public Health, Department of Population and International Health, School of Public Health, Harvard University, Boston, MA, USA. Rannan-Eliya, Ravindra, and Aparnaa Somanathan. 2006. Equity in Health and Health Care Systems in Asia. In The Elgar Companion to Health Economics, edited by A. M. Jones. Cheltenham, UK: Edward Elgar Publishing Limited. Good PraCtice in Health Financing, p. 358 Snodgrass, Donald R. 1966. Ceylon: An Export Economy in Transition. Homewood, IL: Richard D. Irwin. Somanathan, Apamaa, Kara Hanson, Tamara Dorabawila, and Bilesha Perera. 2000. Operating Efficiency in Public Sector Health Facilities in Sri Lanka: Measurement and Institutional Determinants of Performance. Small Applied Research Paper No. 12. August 2000. Bethesda, MD, USA: Partnerships for Health Reform, Abt Associates Inc. Uragoda, C. G. 1987. A History of Medicine in Sri Lanka. Colombo: Sri Lanka Medical Association. van Doorslaer, Eddy, Owen O'Donnell, Ravi P. Rannan-Eliya, Apamaa Somanathan, Shiva Raj Adhikari, Charu C. Garg, Deni Harbianto, Alejandro N. Herrin, Mohammed Nazmul Huq, Shamsia Ibragimova, Anup Karan, Chiu Wan Ng, Badri Raj Pande, Rachel Racelis, Sihai Tao, Keith Tin, Kanjana Tisayaticom, Laksono Trisnantoro, Chitpranee Visasvid, and Yuxin Zhao. 2006. Effect of payments for health care on poverty estimates in 11 countries in Asia: an analysis of household survey data. Lancet (9544): 1357-1364. World Bank. 1993. World Development Report: Investing in Health. New York, USA: Oxford University Press. World Bank. 2003. World Development Report 2004: Making Services Work For Poor People. New York, USA: Oxford University Press. Wriggins, W. Howard. 1960. Ceylon: Dilemmas ofa New Nation. Princeton, NJ: Princeton University Press. Good Practice in Health Financing, p. 359 Annex Table A.1: Total health eXl!enditures, Sri Lanka 1993-2005 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Total expenditures on health Rs. million 17,263 20,346 24,372 28,485 31,964 38,224 42,694 48,457 55,724 62,828 72,108 86,893 100,115 US$ million 358 412 476 515 542 592 607 640 624 657 747 859 996 Ratios Share ofGDP (%) 3.5% 3.5% 3.6% 3.7% 3.6% 3.8% 3.9% 3.9% 4.0% 4.0% 4.1% 4.3% 4.2% Per capita (US$) 21 24 27 29 31 33 33 35 33 35 39 44 51 POEulation ~millionl 16.3 16.4 16.6 16.9 17.1 17.3 17.5 17.7 17.9 18.2 18.5 18.7 19.0 Source: IHP Sri Lanka Health Accounts database. Annex Table A.2: Health eXl!enditures b~ source, Sri Lanka 1993-2005 (% oftotal~ 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 General government (%) 44 45 47 47 47 49 48 49 47 47 45 47 47 Central government 27 26 30 30 31 33 32 33 31 31 30 32 31 Provincial government 14 16 15 15 14 14 14 15 14 14 13 13 15 Local government 2 2 2 2 2 2 2 2 2 2 2 2 2 Private sector (%) 56 55 53 53 53 51 52 51 53 53 55 53 53 Household out-of-pocket 51 51 48 48 49 48 47 46 48 49 50 49 49 Private insurance 1 1 1 1 1 1 1 1 1 1 1 1 2 Employers 3 3 3 4 2 1 3 2 3 2 3 1 1 Non-Erofit institutions 1 1 1 1 1 1 1 1 1 1 1 1 1 Source: IHP Sri Lanka Health Accounts database. Annex Table A.3: Health expenditures by provider, Sri Lanka 1993-2005 (% of total) 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Hospitals 33% 34% 37% 36% 37% 38% 38% 38% 37% 39% 39% 42% 43% Nursing and residential facilities Ambulatory care providers 32% 31% 29% 30% 28% 26% 27% 27% 27% 26% 27% 26% 27% Retailers of medical goods 22% 22% 20% 20% 21% 21% 21% 20% 22% 23% 22% 21% 20% Providers of public health services 7% 6% 6% 5% 6% 6% 4% 4% 4% 3% 3% 3% 4% General health administration 4% 4% 5% 6% 5% 7% 8% 8% 8% 6% 5% 6% 4% Others 2% 2% 2% 2% 2% 2% 2% 2% 2% 2% 3% 2% 2% Rest of the world 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Source: IHP Sri Lanka Health Accounts database. 359 Good Practice in Health Financing. p. 360 Thailand Suwit Wibulpolpraset Suriwan Thaiprayoon 360 Good Practice in Health Financing, p. 361 Background Thailand is a middle come country located in Southeast Asia. It is the only Southeast Asian country never to have been colonized. The country covers an area of 513,115 square kilometers, and is divided into geographical four regions: the central, the northern, the southern, and the northeastern regions. The administrative system consists of three main categories; central, provincial, and local administrations. There are 76 provinces, 876 districts, 81 sub-district, 7255 tambons (communes), and 68,839 villages (Wibulpolprasert,2004). 1) Economic Environment The Thai economy has shifted from an agricultural self sufficient low-income economy towards a manufacturing based export-led middle-income economy. Since 1965, economic (GDP) growth has averaged 7.8 annually, with a 10.5 percent annual growth rate during the period 1986-1990. This growth rate had led to Thailand being considered a middle-income country. By mid-1997, it was a period of economic crisis for Thailand where economic growth dramatically declined; -1.7 and -10.8 in 1997 and 1998 respectively (Wibulpolprasert, 2004). The baht currency was instantly devalued by 50 percent. Undoubtedly, the crisis affected household income. The GDP per capita (Purchasing Power Parity) decreased from $US 6,372 in 1997 to $US 5,757. The percentage of people living under poverty has climbed from 18.8% in 1998 to 21.3% in 2000.In response to the economic crisis, the Thai government sought assistance from the International Monetary Fund (IMF). Thus, the external debt to GDP ratio dramatically increased - from 49.9 in 1996 to 93.9 percent ofGDP in 1998 (NESDB,2004). However, with a free enterprise economy and pro-investment policies, the Thai economy has recovered from an economic crisis in 2002 with a growth rate of 5.4% (NESDB, 2004). This also led to a decline in the proportion of people living under the poverty line, from 21.3% in 2000 to 9.8% in 2002 (UNDP, 2004). The Thai economy has continued to recover with a GDP growth rate of 6.8 and 6.4 percent in 2003 and 2004 respectively before slightly declining to 4.2 percent in 2005. Nevertheless, GDP per capita (Purchasing Power Parity) has gradually increased from US $ 7,483 in 2003 to US$ 8,300 in 2005(NESDB, 2005). Thailand's total external debt has also fallen to 30% ofGDP, accounted for US$ 50.9 billion at the end of2005 (World Bank, 2006). The trend in seeking international financial assistance has also declined. In contrast to being a recipient, Thailand has recently become an emerging donor providing assistance to neighbor countries. In fiscal year 2003, Thai official development assistance (ODA) was US$ 167 million accounted for 0.13 percent of its Gross National Income (GNI). Over 90 percent of Thai ODA was in support of infrastructure projects in Cambodia, Lao PDR, and Myanmar whilst the remainder of Thai ODA was in the form of technical assistance (Ministry of Foreign Affair &United Nations, 2005). 361 Good Practice in Health Financing, p. 362 2) Epidemiological and Social Environment According to major health indices, the health status of Thai people has significantly improved. Life expectancy at birth gradually increased from 67 years in 1995 to 70 years for men and from 71 to 74 years for women (WHO, 2005). The maternal mortality rate (MMR) had declined from 374 per 100,000 live births in 1962 to 15 per 100,000 live births in 2005. In addition, the infant mortality rate (IMR) had declined from 84 per 1,000 live births in 1964 to 20 per 1,000 live births in 2005 resulted from an improvement of immunization programs and maternal and child health services. Family size (member per household) has fallen from 5.6 in 1960 to 4.4 in 1990 and 3.9 in 2000 (Bureau of Policy and Strategy, 2006). Presently, Thailand is experiencing the demographic transition. The population growth of the country tends to decrease resulting from a success in family planning campaigns. The contraceptive prevalence rate has increased from 14.4% in 1970 to 79.2 % in 2001. The condom use campaign against HIV/AIDS transmission in 1989 also had some effect on birth control. As a consequence, the population growth rate has dropped from 3.2% in 1970 to 0.7% in 2005. It has been projected to decline to less than 0.5 % in 2020. This has affected the quantity and age structure of the population. The proportion of people below 15 years has dropped while the proportion of working age and elderly has been rising. The base of Thailand's population pyramid has become narrow as shown in Figure 1. Figure 1 : Population Pyramids '---'-- - - - - - Thailand,1990 I Thailand.200! 10"" 10"" 80-84 80-84 60-&< 6i}64 & " 4""" : 4Q-44 2tr24 20-24 (}4 (}4 15 10 10 15 15 10 10 15 PopulatIon (Of.) Population (%) Thailand, 202Cl 100+ 80-84 (/) 10 Non-poor household not children aged less than 12 pensioners, and their employees establishment eligible for the MWS, years, the disabled, dependants (spouse, community leader and veteran, and monks parents, and children) health volunteer family 3 Financing Tripartite 1.5% of payroll Household 500 baht+tax 3.1 Source of funds General tax revenue General tax revenue each 1000 baht reduce to 1% since 1999) Proportional 3.2 Payment mechanism Global budget Fee-for-service Capitation reimbursement among levels of care Maternity emergency 3.3 Co-payment No Yes, at inpatient care servIces No at private hospitals if beyond ceiling budget 4 Benefit Services 4.1 Ambulatory Services Only public designated Public only Public and private Public (MOPH) 4.2 Inpatient services Public only Public and private Public and private Public (MOPH) emergency only) 4.3 Choice of provider Referral line Free choices Contracted hospitals or Referral line its network, registration required 4.4 Condition included Comprehensive package Comprehensive package Non-work related injuries Comprehensive package illness, iniuries 4.5 Condition excluded 15 conditions NO 15 conditions 15 conditions 4.6 Maternal benefit Yes Yes Yes Yes 4.7 Annual physical check up No Yes No Yes 4.8 Prevention and Health Health education, Promotion Very limited No immunization Yes Private beds, special Private beds, special Private beds, special 4.9 Service not covered nurse, Special nurse nurse, nurse, and eye glasses and eye glasses Source: ModIfied from Tangcharoensathlen et aI, 2002 3. Overview of Health Delivery System The Thai health system has been developed for centuries from self-reliance to the current health care system that mostly depends on western medicine, The Ministry of Public Health (MOPH) has invested in the infrastructure of health care units in every district and sub-district resulting in hospitals and health centers gradually established in all areas of the country during the 1900s. In 2003, there were 9765 health centers covering all sub- districts; 725 community hospitals (30- 60 beds) , covering 91.2% of all districts;70 general hospitals (100-150 beds) and 57 military hospitals, covering 100 % of provincial 368 Good Practice in Health Financing, p. 369 level; 25 regional hospitals; 59 specialized hospitals; and 9 medical school hospitals. These public health facilities provided 99,590 beds, with an occupancy rate of71.8 percent. In the private sector, there were 436 private hospitals and 14,953 private clinics providing 34,863 beds which accounted for 26% of the total beds in the country (Wibulpolprasert,2004). In total, 19,956 medical doctors, 59,131 nurses, 6,003 dentists, 10,054 pharmacists and 30,633 rural health workers completed their training in 1999. The major sources for educating and training these qualified health personnel are public universities (Wibulpolprasert, 2002). At the same time, Praboromarachanok Institute of Health Manpower Development, an agency under the Ministry of Public Health plays a crucial role in producing nurses and health workers. Nevertheless, deployment of health worker to rural areas, the Northeast in particular, still challenges human resource management in the Thai context. The distribution of health resources in Thailand was shown in Table 2 (Bureau of Policy and Strategy, 2006). The health care system in Thailand is currently based on a public and private mixed system of health care financing and health service provision. The MOPH acting as the principal public agency plays a crucial role in providing health care services nationwide while the private sector participates in a pluralistic health service system. The main functions of the MOPH include health promotion, prevention, and medical services delivered through the primary, secondary, and tertiary health care system. Basic health care including disease prevention, health promotion, and simple treatment is provided at a primary care level through health centers. General, regional, and university hospitals, at the same time, provide secondary and tertiary care while community hospitals located at districts or sub-districts focus mainly on health prevention, promotion, and curative care. At the village level, village health volunteers who have been trained by health personnel provide primary health care services which covered health promotion, disease prevention and rehabilitative care without service charges (Wibulpolprasert, 2004). Presently, there are 816,555 village health volunteers covering all villages throughout the country. These volunteers are not remunerated, but receive fringe benefits from the Ministry of Public Health (Department of Health Support Service, 2006) The Existing Challenges of the Health Systems Inequities in Access to Healthcare Inequality in accessing health services among Thai people in urban and rural areas is obvious. Urban residents have a greater opportunity in accessing health facilities with doctors than the rural residents do. The poor, a most vulnerable group also faced financial barriers to access health care services. Pannarunothai and Mills (1997) found that there were inequities in the Thai health care financing system. The poor were likely to pay more for their own medical costs than the rich. They also reported an inequitable pattern of household health expenditure by income quintile and per capita. The underprivileged spent 8.17% of health care services relative to their household incomes while the privileged groups spent 1.27 % (NSO, 2003 ) 369 Good Practice in Health Financing, p. 370 Inequities in Health Status Like most developing countries, infants born in urban areas have a much higher chance of survival than rural infants. Thus, the infant mortality rate (IMR) is a good health indicator to show the differences in health status of various groups of the population in Thailand. IMR in rural areas is about 1.9 times greater than that in the urban areas (Wibu1po1prasert,2004). Inequities in Bearing Health Cost The burden of health expenditure does not reflect the ability of Thai people to pay. According to a comparison of health expenditure proportion in each income group, the poor pay for their health expenditure more than the rich. Moreover, under the Universal Healthcare Coverage Scheme, provincial hospitals favored the rich rather than the poor. On the contrary, the poor received more benefits than the rich when they utilized health care at health centers and community hospitals (Wibu1po1prasert, 2004). Inequities in Resource Allocation One of the most obvious is inequity in resource allocation. Despite a high proportion of health resources per capita, the differences between regions in terms of human resources, the number of beds and health facilities has been revealed as illustrated in Table 4. This situation clearly reflects the equities in allocating resources between geographical areas. Table 4 Distribution of Health Resources (Resource to Population Ratio) by Region, 2004 The Type and Region Bangkok Central The North The South Northeast Nationwide Beds 1:224 1:390 1:503 1:501 1:749 1:469 Doctors 1:879 1:3,134 1:4,534 1:3,982 1:7,466 1:3,305 Dentists 1 : 5,583 1 : 15,775 1: 16,039 1 : 15,620 1 : 24,699 1:15,143 Pharmacists 1 : 4,632 1 : 6,819 1 : 9,037 1 : 8292 1 : 13,032 1 : 8,432 Professional Nurses 1 : 289 1 : 593 1 : 684 1 : 659 1 : 1,045 1 : 652 Technical Nurses 1 : 3,250 1 : 2,402 1 : 3,228 1 : 2,481 1 : 4,141 1 : 3,085 Source: Bureau ofPohcy and Strategy, 2005 In addition, there are vast differences in the diffusion of medical and health technologies. The number of such medical technologies has rapidly increased (Figure 3). The allocation of the health care budget also reflects inequities in health care. With respect to allocation and diffusion of health care budget per capita, the overall budget allocation for the Northeast is lower than those for the other regions. In contrast, the Central region and Bangkok and its vicinity obtain the highest. 370 Good Practice in Health Financing, p. 371 Health system efficiency The problems of the health service in technical efficiency were revealed in a variety of ways. The curative service system was considered to be inefficient due to the over- utilization of drugs, unnecessary admissions to hospital, the excessive lengths of stay and excessive use of diagnostic tests. In addition, the Thai health service system focused mainly on curative care, which is much less efficient compared to promotive and preventive care. Moreover, health personnel, particularly doctors were trained with a focus on disease treatment rather than disease prevention. Undoubtedly, Thailand spent more than 250 billion baht (6.59 million US$) (5.6% ofGPD in 1996) annually on health expenditures, with a 10 percent increase yearly (Bureau of Policy and Strategy, 2006). The national health expenditure gradually increased from 3.82 percent ofGDP in 1980 to 6.05 percent of GDP in 2003 with government budgets of 61.6 percent and private payments of 38.4 percent. In contrast to Malaysia and Sri Lanka, their expenditures on health were less than 4 percent of their GDP, but their health situations have improved with reference to their major health indices (World Health Report, 2006). The overall resources allocated to health care also have a tendency to increase. The Ministry of Public Health budget as a percentage of the total national health budget has risen from 6.7 percent in 2001 to 7.6 percent in 2004 due to the government policy on universal coverage (Bureau of Policy and Strategy, 2006). 4. Description of health coverage reforms The existing Thai health system is a result of a long evolution of reforms, which took place in specific socio- political contexts. Based on the changes in the system in each era, the evolution of the health system in Thailand has been divided into four eras. 4.1 The Era of modernization of medicine (1888-1960) Before 1888, prior to the establishment of Siriraj Hospital, people were solely responsible for their own health depending mostly on home remedies and traditional medicines. The role of the government in providing health care or medical welfare for the general public was relatively limited. The idea of an organized health care system and public health were not socially feasible under a traditional system of knowledge and social organization. The first era started from era of King Rama III when Dr. Dan Beach Bradley, an American Christian missionary, brought the first smallpox vaccination into Thailand in 1882. The real start of modernization, together with overall modernization of the country to cope with the colonization movement, was in the reign of King Rama V. This was when the first modem medical schools were built and the first provincial dispensary was established in 1888 under a user-fee system. After Siriraj Hospital was established, other specialized hospitals, mainly situated in Bangkok were built to provide health and medical care. To finance these hospitals, the 371 Good Practice in Health Financing, p. 372 government subsidized them. In the provinces, on the other hand as the government did not fund the hospitals themselves, health care financing in part relied on donations from the local people. In this way, hospitals were established in every region nationwide to improve accessibility to modern medicine. Nevertheless, the health care infrastructure was not well developed. Thus accessibility was not greatly improved. In 1942, there were only 15 provincial hospitals and 343 health centers. To support modernization, the Ministry of Public Health (MOPH) was established in 1942 (Jongudomsuk, 2005) to prevent epidemic diseases and to provide medical care. Afterwards, health care infrastructure continually expanded. 4.2 The Era of initial expansion (1961-1976) The second era started in 1961 and was the initial phase of the National Economic and Development Plan, supported by the World Bank. There was rapid expansion of infrastructure, including modern provincial hospitals and rural health and midwifery centers. Besides this, there was a regional hospital in each region to act as a referral center for provincial hospitals. However, a shortage of medical doctors working at health centers was a major constraint. In particular, the distribution of health care providers among urban and rural areas was imbalanced. After 1972, the private health care providers began to increase their roles in servicing the people although this was restricted to urban areas. This seemed to widen the gap of mal-distribution of heath personnel among urban and rural areas. In this period, the public health infrastructure was able to provide a health service to 15 percent of the population only while 51 percent of them still sought care from private providers and traditional healers (Wibulpolprasert, 2004). 4.3 The Era ofPrimary Health Care and Health For All and the rapid expansion of rural health care systems (1977-2001) The first political movement against the military government in 1973 resulted in temporary democracy. This political changes together with the global Health For All movements resulted in rapid expansion of rural district hospitals and recruitment of Village Health Volunteers and many community-financing schemes in health, for example the village drug fund, the sanitary fund, and the nutrition fund. For five years, from 1982-1987, while the economic growth was low, the public health budget was shifted from the urban provincial hospitals towards the rural district hospitals and health centers. During this era, trained village health volunteers and village health communicators were used as the main tool in the primary health care strategy. Their roles were to assist health personnel in delivering basic curative care. Health volunteers somewhat improved the coverage of basic health care services to villagers. However, the main objective of primary health care was not truly achieved. According to the primary health care approach, communities and individuals must be actively involved in improving their own health. Health volunteers, on the one hand were expected to be an initiative to increase community and individual involvement in health development by promoting their participation in health activities. Unfortunately, the role of health volunteers was passive 372 Good Practice in Health Financing, p. 373 in health development because health personnel failed to understand the role and potential of health volunteers (Jongudomsuk, 2005). With respect to the issue of accessibility to health care, the poor continued to face the problem of access to care due to financial barriers. This problem became a major issue of concern, which caught the attention of the government. Consequently, the Medical Welfare Scheme (MWS), with the aim of targeting the poor was first established in 1975. It was funded through general taxation. Eligible persons in the MWS consisted of the poor whose incomes were less than 1,000 baht per month (NESDB, 2005). They were offered medical services from public health facilities at registered primary health care providers without user charges. By 1997, MWS beneficiaries were allowed to seek care at district hospitals as the first contact. In the case of bypassing this system, beneficiaries had to pay for health services out of their own pockets (Pannarunothai, 2002b). Civil servants and their family members were covered by the Civil Servant Medical Benefit Scheme (CSMBS), which is non-contributory. The benefit package of this scheme included: outpatient and inpatient care with private rooms, emergency services and drug expenses. Beneficiaries were allowed access to both public and private hospitals, but the cost incurred at private hospitals was not fully reimbursed. The CSBMS was funded by general revenue through the Ministry of Finances' Comptroller General Department. Due to the nature of the fee-for-service payment method and the inefficient monitoring system of the scheme, the expenditures of the CSBMS dramatically increased in the past decade and led to the highest expenditure per capita in 2003 (IHPP, 2005). 4.4 The Era ofHeath Care Financing Reform and expansion of Health Insurance coverage (2001- present) The Social Security Act was promulgated and started to cover the private employees since 1991. The Social Security Scheme, under the capitation payment system, was started in 1992. It was financed by the contribution of employers, employees and the government. To receive medical benefits, insured workers were required to register with the main contracting hospital, either public or private with more than 100 beds. The Welfare Health Services for the indigents was expanded to cover the elderly, the children under 12, and the social contributors. The Health Card Schemes was also greatly expanded. These health insurance schemes together covered around 80 percent of the population as of year 1998. 5. The 2001 Universal Coverage of Health Insurance Policy 5.1 Setting the Agenda The 1996-1997 political reform movements with the promulgation of the new "People Constitution" in 1997, has resulted in strong political movements towards public interest policies. The policy elites in the Ministry of Public Health, with the connection to the leading political party, the Thai Rak Thai Party from their past engagement in the social movements against the military government, were able to push the "universal coverage of 373 Good Practice in Health Financing, p. 374 health insurance" on to the political agenda. This has become one of the main populist policies for the campaign in the first general election under the new constitution in early 2001. The leading Thai Rak Thai party designed the motto of"30 Baht (0.79US$) 122 treat all diseases" to represent their universal health insurance policy. The proposed policy then led the party to winning the election. 5.2 Policy formulation processes and social mobilization Immediately after the landslide victory, about 40.6% of the popular vote (TRT, 2005) in the election and the formation of the new government, the policy was reformulated based on existing evidences and experiences. The policy was immediately implemented in six provinces, which has tested the new systems of financial reform under the World Bank support, two months after the new government was set up. It was expanded to cover all provinces in one year. Evidences from previous researches as well as further synthesis of new information were used extensively in the implementation. For example the information from previous costing researches and the information of the health care utilization behavior from the Health and Welfare Survey, were used to calculate the capitation figure l23 . Strong social movements to support the National Health Security Bill has been so successful that it was the first bill that more than 50,000 people proposed the draft bill to the parliament. This is possible under the new constitution. This movement has made it possible for the civil societyl24 to gain significant and influential seats on the Parliamentary Commission to consider this bill. This has resulted in significant and influential seats for civil society organizations on the National Health Security Committee, the top authority in decision making for the Universal Coverage of Health Insurance System. 5.3 The Content of the policy and the essential reform component 5.3.1 The coverage and benefit schemes The Universal Healthcare Coverage Scheme aims at reducing national health expenditures and out-of pocket payments by establishing a tax-based financing system and paying providers on a basis of the number of registered population 125. All Thai citizens are entitled to access to qualityl26 health care and a single standard of benefit package. The benefit package includes a set of health interventions that has been stipulated in a contract between the purchaser and the provider at all levels of health services. It has been classified into three components: the curative package; the high cost care; and the promotive and preventive package. The curative package covers ambulatory and hospitalisation services with some exclusion, such as cosmetic surgery, infertility 122 Exchange rate 38 baht per 1US$ a co-payment per visit for all essential benefit package provided 123 1659 baht (43.66US$) per head in 2006 124 Civil society refer to active individuals and groups from voluntary associations and informal networks such as the Rural Doctor Society 125 Demand-based allocation 126 Quality Assurance of health care services has been assessed by the Hospital Accreditation (HA) procedure. 374 Good Practice in Health Financing, p. 375 treatment, and the provision of private room including board. For the high cost care, the UCS has adopted a similar package to that provided by the SSS in order to standardize packages across the scheme to minimize the health care service inequity. Thus there are substantial expensive interventions offered as shown in Table 5. The preventive package of the UCS is focused on health promotion and disease prevention which covers immunization, annual physical check ups, premarital counseling, and voluntary HIV testing, antenatal care and family planning services. All contract provider networks including public and private providers are bound to provide these services to registered beneficiaries (NHSO, 2001) Table 5 Inclusion and exclusion list of expensive health care interventions of the ues Inclusive list Exclusive list Chemotherapy for specific cancers Renal replacement for patients with Radiation therapy for specific cancers end stage renal disease ** Open heart surgery including prosthetic Organ transplantation cardiac valve replacement Cosmetic surgery Percutaneous Transluminal Coronary Infertility treatment Angioplasty (PTCA) Coronary Artery Bypass Grating (CAB G) Stent for treatment of Atherosclerotic Vessels Prosthetic hip replacement therapy Prosthetic shoulder replacement therapy Neurosurgery, e.g. craniotomy Antifungal treatments for cryptococcal meningitis Antiretroviral treatment for patients * Living with HIV/AIDS Source: NHSO, 2004 * It has been included in October 2003 ** It has been currently piloting 5.3.2 The financial reforms towards more equity and efficiency The Health Systems Research Institute (2001) has reviewed the strengths and weaknesses of the capitation payment under the SSS and Diagnostic Related Groups (DRGs) payment of the MWS and suggested two main payment mechanisms for two main split budgets: outpatient and inpatient services as depicted in Figure 3. Capitation payment has been used for ambulatory care whereas case payment, DRG with global budget was used for inpatient care. Inclusive and exclusive capitations have been proposed as provider payment methods for the UCS. It is notable that the inclusive capitation covers the cost of health promotion, preventive and curative cares including primary, secondary, and tertiary care while the exclusive 375 Good Practice in Health Financing, p. 376 capitation covers the cost of health promotion and preventive and primary health cares. It also demands collection of funds to be managed by provinces as inpatient costs incurred are paid in a basis of weighted allocations under a national budget (Pitayarangsarit et aI, 2004). Figure 3. Financing mechanism in a province during 2001-2002 National office for Health insurance I OP 574 J i .t 1,052 bahts / cap. / . I Hospltas year IP 303 P&P 175 Primary Inclusive Exclusive Primary Area Health care Capitation care Board Units Units ~ OP+IPp ~P+IP+P& I r I OP+P&P ~ OP i DRG c global ~ I Hospitals I j budget for IP Emergency cost Emergency cost High cost High cost Contingency Contingency fund fund Capitabon National office for Health insurance Source: Banchuin 2002 In regard to resource allocations from the central government to the provinces, budgets were shifted from a historical supply-based to demand -based allocation, according to the number of their registered population. In fact, UC beneficiaries have freedom to choose their preferred providers. However, a registration information system was not well developed. Thus, they have been assigned to the providers close to their communities. In 200 I, during the first year of implementation, salaries of staff were included in the capitation budget. In fiscal year 2002, due to a financial problem, the salary budget for staff was then excluded. Initially, the MOPH allowed provinces to choose either inclusive or exclusive capitation (split outpatient and inpatient budgets) to pay providers within their provinces. However, due to the disincentive to pay providers for high-cost care and delays in case referrals resulting from the inclusive capitation employed, the MOPH decided to choose a single system of exclusive capitation for all of the provinces in fiscal year of 2003 (Pitayarangsarit et aI, 2004). 376 Good Practice in Health Financing, p. 377 5.3.3 The primary care based systems Primary care provider units (PCUs) have been determined to be the gatekeepers to provide care for UCS beneficiaries. As gatekeepers, PCUs are expected to better provide continuous and comprehensive care with a holistic approach to people in catchment areas. Based on the services provided, health facilities under the UCS can be classified into three groups (NHSO, 2001). 1. Contracting unit for primary care (CUP); CUPs are the health centres providing curative, promote, preventive, and rehabilitative services as ambulatory care, home care, and community care 2. Contracting unit for secondary care (CUS); CUSs can be community hospitals, regional hospitals, general hospitals, and university hospitals including those public hospitals outside MOPH and private hospitals which are responsible for inpatient services 3. Contracting unit for tertiary care (CUT); CUTs are the organization providing expensive care and specialized care with high technology. They can be regional hospitals, university hospitals or specialized health institutes All contracting unit for primary care are required to have at least one PCU for every 10,000 registered and the prerequisites for primary care provider to be a main contractor are outlined in Box 1. Box 1 Minimum Requirement of CUP Inputs and Structure 1. One facility for 10,000 - personnel work in CUP more than 75% Population their working time 2. Facility is located close to the responsible - available laboratory system for the Pop. (Transportation time< 30 min.) investigation 3. Adequate health personnel - available vehicle for the referral - Physician 1: 10,000-20,000 Provision of Services - Dentist 1 : 20,000-40,000 - service available at least 56 hours/week - Pharmacist 1: 20,000-30,000 - be able to provide comprehensive care - Registered nurse 1 : 5,000 - be able to provide in-house service and - Health personnel 1: 1,250 community based services Source: NHSO,2001 377 Good Practice in Health Financing, p. 378 5.4 The Difficulties and Negative Implications 5.4.1 Increase in Workload The ues has increased the demand for health services, as it has reduced the financial barrier to access to health care services. This increase in demand had implications for health care personnel's workload. More than 70 percent of health personnel claimed that their workload increased as a result of the UCS. Comparing service utilization rates between 2001 (before the UCS) and 2003 (after the UCS) revealed that utilization of public health care service increased by 25 percent for outpatient care and hospitalization rate was increased by 9 percent (Srithamrongsawat and Torwatanakitkul, 2004). The Suan Dusit Poll (2004) interviewed 4417 health workers in 11 provinces and also reported that daily workloads were a 50 percent increase. This significantly contributed to doctors' resignation. In comparison with 2001, the number of doctors who resigned from the public sector in 2002 has doubled. Their reasons were also stated as being overworked and poor payment in the public sector (Srithamrongsawat and Torwatanakitkul, 2004) as depicted in Figure 4. Figure 4 Number of doctors resigned from public sector during 1995 - 2003 _. . _........................._ _.... 700 .. .... 600 Economic 500 400 300 200 UC implementation 100 o+---~----~--~----~--~----~--~----~--~ 1995 1996 1997 1998 1999 2000 2001 2002 2003 Source: Srithamrongsawat and Torwatanakitkul (2004) 5.4.2 Adverse effects of resource allocation reform A per capita budget, which included staff salary and non-salary budget to provinces, has created a conflict. The new budget allocation, shifting from a historical supply-based to a need or demand - based allocation, aimed to allocate equitable budgets. Due to a mal- distribution of health personnel, it seemed to create adverse effects. The affluent provinces that had a high number of health personnel received lower budgets due to a small registered population, while the provinces in the Northeast and district hospitals with a low number of health personnel received higher budgets due to a large registered 378 Good Practice in Health Financing, p. 379 population when using a single capitation rate in 2002 (Srithamraongsawat & Torwatanakitkul, 2004). However, the situation reversed once salaries were excluded from the capitation rate in 2003, but resulted in a gap of inequity of budget allocation when salary was taken into account. The capitation rate favored the affluent provinces and provincial hospitals. They received higher budgets. Nevertheless, there were some provincial hospitals, which continue to be affected by the financial constraints after salaries were excluded. The less affluent provinces, on the one hand, although receiving lower budget, a number of district hospitals experiencing financial constraints in the Central and Southern regions declined; except a number of district hospitals in northeastern region gradually increased (Srithamraongsawat & Torwatanakitkul, 2004). 5.4.3 Financial impacts The UCS was criticized by healthcare providers for being under-financed; particularly for inpatient care since a capitation payment of 1,202 bath (31.63 US$) 127 (the real budget allocated was 1052 (27.68 US$) as 150.40 baht (3.96US$) has been set aside) per registered person per year paid was not adequate (Leesmidt et aI, 2005). The rate was based on the 1996 service utilization rate that has not taken into account the aging popUlation and an increase in admission between year 1996 and 2001. When using the 2001 service utilization rate and unit cost of services provided in the same years it appears that per capita budgets for fiscal year 2002 was under estimated by 212 baht (5.58 US$). Besides, the capitation rate has not taken into account the aging population in the first year of implementation (Srithamraongsawat & Torwatanakitkul, 2004). Kittikanya (2004) also reported that more than 30 percent of public hospitals under the UCS were in financial trouble, with an accumulated debt of 1.365 billion baht (35.9 million US$). In particular, districts hospitals in the North and Northeast were severely affected. The percentage of district hospitals with financial problems in the Northeast significantly increased from 3 percent in 2002 to 27 percent in 2003. The costs from the rapid expansion of Primary Care Units according to budget gained in the first year could partly explain the problem of these district hospitals when they received lower budgets in the following year. The number of hospitals with financials constraints and receiving extra financial support from the Contingency Fund was shown in Figure 5. 127 Exchange rate at 38 baht per 1 US$ 379 Good Practice in Health Financing, p. 380 Figure 5 Number of hospitals with financial constraints and receiving financial support from the contingency fund during 2002-2004. 120~----------------"------------"-------------~ 100+--------------------------------4 i_North SO+-------------------------------~ II1II Central 60+-------------------------~ ~-1~ ~ 0 Northeast 40+-----------------~ __~~~ ~-1~ o South 20 o GeneraVRegional hospitals District hospitals (N=720) (N=94) Source: MoPH, 2002-4 5.5 The Evidences ofSuccesses 5.5.1 The great decrease a/the uninsured The percentage of population coverage showed that the ues has achieved universal coverage. After the first year of rapid implementation (2001-2002), the number of the uninsured has dramatically decreased from 20% of total population in 1998 to 5% of total population in 2003. The result of the Health and Welfare Survey (HWS) conducted in 2003 showed that the ues covered 47.7 million people (74.7%) (Vas avid et aI, 2004), as shown in Figure 6. Figure 6 Population covered by health insurance schemes, 2003 60.0 •••...• -- ..................... _.• - •.••.•.••...••.•.••.• - •.• - •.• 47.7 SO.o . . . . . . . . . . . . . . . . . . . . . . . (~5%) . . . . . . . . . . g ~ 40.0 fg 3(J.O :E 20.0 .. ' ............. . i 5.7 6.1 10.0 .. (9..,) ..... (!oo/?)...... IT ... . (2%) o.o.L........I_i.--- CSHlS sss PHI uc lkJinsured Source: Vasavid et al (2004) 380 Good Practice in Health Financing, p. 381 5.5.2 Out-of- pocket payments reduced ue beneficiaries would be effectively covered if their own out-of-pocket payments were reduced. Visavid et al (2004) reported that the household expenditures were decreased after the ue implementation. In 2002, the household savings on health care services was approximately 9,650 million baht (253.95 million US$). The year later, savings gradually increased to12, 726 million baht (334.89 million US$) among the previously uninsured and currently insured by the ues. The evidence from three low-income provinces suggested that the ues would ease the burden on household expenditures only ifUe beneficiaries seek care from public facilities operated under the ues (Suraratdecha et aI, 2005). 5.5.3 eonsumer satisfaction The survey concerning perspectives of the ues among ues beneficiaries was conducted in 13 provinces 128 with a sample size of 6294 beneficiaries. It was found that more than 90 percent of the sample satisfied with the quality services provided as shown in Table 6. Table 6. Satisfaction of UC Beneficiaries Increase Issue Satisfaction 2003 2004 TheUeS 92.9 92.9 Quality of services Physician 92.9 92.9 - Nurse 89.4 91.2 1.8 Other personnel 89.5 91.9 2.4 Medicines 83.3 86.6 3.3 Medical equipments 85.8 90.4 4.6 Source: ABACK Polls, 2004 128 Bangkok, Ayutthaya, Prajerbkirikhan, Samutprakarn" Prae, Kampaengpet, Chieng Rai, Nongkhai, Srisaket, Nakhonrajasima, Trang, and Nakhonsithamaraj 381 Good Practice in Health Financing, p. 382 5.5.4 The increasing benefit package Despite improvements to the benefits package, at the end of 2003, antiretroviral medicines for HIV/AIDS treatment had not been included in the benefits package. In addition, comprehensive services for the maternal and child health have also been included in the UCS preventive package. 5.5.5 The knowledge generation and management The key strategy of health sector reform in Thailand was "the Triangle that Moves the Mountain" (Wasi, 2000). The Mountain refers to a big and very difficult problem that is usually unmovable, emerging from several factors such as political, socio-economic, cultural, environmental, and so forth, conducive to health deterioration. Figure 7: Triangle that Moves the Mountain 1. Creation of rtltnnt knowledge 2. SocialllloHlllfnt 3. Political inl'olHlllfnt Source: Wasi (2000) The Triangle, depicted in Figure 7, is the interaction between (1) creation of relevant knowledge, (2) social movement, and (3) political involvement. Knowledge has been created through research and helped policy makers and civil society network find innovative ways to restructure health system. The 'Mountain' cannot be moved without the political involvement since politicians influence resource allocation and utilization and have a significant role to promulgate laws. Through mutual cooperation, this 'Triangle' is able to move the Mountain (Wasi, 2000). In the process of health system reform in Thailand, the triangle has been applied to restructuring and reforming the Thai health system, a hierarchy and sophisticated system as if it was a mountain. The National Health System Reform Committee (NHSRC) chaired by the prime minister appointed four sub-committees comprised of researchers, the civil network, policymakers, and mass media representatives to design a new health system and to draft the National Health Act (Poolcharoen, 2004). 5.5.6 The strong public involvement and support The UCS was formulated through civil society involvement and political support. A network of civic groups established a campaign for the UC policy. These groups began drafting the National Health Insurance Act. With more than 50,000 signatories, in March 2001 the draft was submitted to Parliament for its consideration (Chuengsathiansup, 2004). 382 Good Practice in Health Financing, p. 383 6. Lessons for Other Countries Thailand is one of the few countries which have struggled to achieve universal coverage of health care policy during an economic recession. Attempts to achieve universal coverage have been made for a long period of time but they have sped up in the past couple of years. The policy was rapidly implemented and national coverage was achieved within one year. As a lower middle income country, the 2001 policy of Universal Coverage of Health Insurance in Thailand is clearly a very bold political decision. At the outset, two World Bank staff proposed an analytical paper criticizing its likely failure due to inadequate funding. After five years of implementation, it has evolved to be a system with strong social involvement and ownership, and political commitment. It has affected not only financial reform, but also reform in other components of the health care system. It is an example of the evidence-informed decision, in real life. Thus, it would be very beneficial as a case study for lessons learnt for policy makers and analysts. 6.1 The strategy of "Triangle that moves the mountain" The linkages among the three main strategies that form the triangle is the key to the success of many social reform in Thailand. These three strategies include "Knowledge Generation, management and sharing", the "Social movements and supports", and the "Political movements, involvement and commitment". 6.2 The Tipping Point Many factors contributed to tipping the situation towards the universal coverage. These include the "Stickiness" of the issue. The motto "30 Baht treats all disease" has been so popular that it sticks into the mind of every Thai. The existence and the roles of the "Marvens", the "Connectors", and the "Salesman" contributed greatly to building up the political and social movement. The conducive political, social and economical environments allow the issue to be accepted and also fund mobilized to support the implementation. 6.3 The Dos and Don't (1) Dos: These include, in addition to the strategies of the "Triangle that moves the mountain", the gradual but determined and progressive reform, the flexibility of implementation, the decentralization, the transparency and accountability, the appropriate involvement of the private providers, and the adequate investment in the infrastructures for services and management. (2) Don'ts: The sound policy directions of the Universal Coverage Scheme were undermined by the circumstances of its hasty introduction. Health infrastructure was not well enough developed to perform its new functions adequately. Undoubtedly, the policy faced a variety of obstacles. Health personnel had limited capabilities to perform their new roles 383 Good Practice in Health Financing, p. 384 and functions. The direction of the scheme was misinterpreted due to the lack of policy communication. Financial sustainability became a serious concern for the DeS. Having been funded by tax revenue, the stability of the budget was uncertain. A co-payment of 30 baht did not reflect the marginal cost of interventions, but it helped encourage people to utilize health care services at an affordable cost. Adversely, the demand for health services was increased and attributable to hospital workloads which influenced physicians to resign. Moreover, the problems of being under-financed and a less than ideal equality of medical services resulted from an inappropriate plan for resource allocation. Hence, it should be avoiding too rapid expansion of the depth of the coverage without adequate consideration on the financial and health services burden, too rapid, aggressive, and non flexible reform, too aggressive social marketing which can create over- expectation and demand among the people, underestimating the effect of the growth of the private sector particularly those that cater foreign patients, and the use of mainly financial incentives to solve the problem of human resource for health. 384 Good Practice in Health Financing, p. 385 ANNEX - Table 9. Comparison of characteristics of health insurance schemes in Thailand, 2002 Characteristic CSBMS SSS UCS 1 Scheme Nature 1.1 Beneficiaries Fringe benefit Compulsory Social welfare 1.2 Model Public reimbursement Public contracted Public contracted Government employees and 1.3 Covered population their degendants Private sector People not covered by Employee> 1 CSBMS or SSS 2 Benefit Package 2.1 Ambulatory services Public only Registered public and private Registered public and private 2.2 Inpatient services Public Registered ]:lublic and private Registered public and private 2.3 Choice of provider Free of choice Registration required Registration required 2.4 Conditions included Comprehensive package Non work related illness Comprehensive.package 2.5 Conditions excluded No 15 conditions 12 conditions 2.6 Maternity benefits Yes Yes Yes 2.7 Annual physical check - ups Yes No Yes 2.8 Prevention and health Health education, ipromotion Yes immunization Yes 2.9 Services not covered Special nurse Private bed and special nurse Private bed and special nurse 3 Financing 3.1 Source of funds General tax Tripartite, 1% of payroll each General tax National Health Security 3.2 Financing body Ministry of Finance Social Security Office Office Capitation for OP;DRG for 3.3 Payment mechanism fee-for-service(>2000 baht) Capitation( 1500 bah!) II'(1,202baht) Yes: at public and private 3.4 Co-payment hospital Maternity, emergency Yes, 30 baht per visit , IP private clinic only for life- threatening care if beyond ceiling 3.5 Per capita tax subsidy 2106 baht 519 baht 1275 baht Source: Bureau ofPohcy and Strategy, 2005 385 Good Practice in Health Financing, p. 386 REFERENCES Banchuin, C. (2002). New Health Insurance Policy in Thailand [On1ine].Avai1ab1e from:http://www .adb.org/Documents/Events/2002/Socia1Protectionibanchuin~aper. pdf# search='socia1%20hea1th%20insurance%20in%20thai1and. [Accessed August 10,2005]. Bureau of Policy and Planning, Ministry of Public Health (2002). Burden of disease and irijuries in Thailand. Bangkok: Printing House of The War Veterans Organization of Thailand Under Royal Patronage of His Majesty the King. Bureau of Policy and Strategy, Ministry of Public Health. (2005). Health Personnel Resource in Thailand (in Thai) [Online]. Available from: http://203.157.19.1911per8- 47.doc [Accessed July 25,2006]. Bureau of Policy and Strategy, Ministry of Public Health(2006). Health Policy in Thailand. Nonthaburi: Bureau of Policy and Strategy, Ministry of Public Health. Chuengsathiansup, K. (2004). Deliberative Action: Civil Society and Health System Reform in Thailand. Nonthaburi: Health Systems Research Institute. Donaldson, D., Pannarunothai, S., &Tangcharoensathien, V. (1999). Health Financing in Thailand Summary Review and Proposed Reforms. Nonthaburi: Health Systems Research Institute. Gajeena, A. (2004). Thailand: Country Report. The 2nd ASEAN & Japan High Level Officials Meeting on Caring Society,30 August-2 September 2004. Hayama and Yokohama, Japan. Health Systems Research Institute. (2001). Proposal to Achieve the Universal of Healthcare (in Thai). Nonthaburi: Health Systems Research Institute. International Health Policy Program (2005). Universal coverage: what has been achieved, current and fUture challenges. The Meeting of High Officials of the National Health Security Office, Bangkok, 5 Apri12005, Bangkok: Thailand. International Labour Organization (2004). A Technical Note to the Government: Financing Universal Health Care in Thailand. Geneva: The International Labour Office. Jongudomsuk, P. (2005). Health Care System in Thailand: Reforms towards Health Promotion. 6th Global Conference on Health Promotion, Bangkok, 12-15 August 2005. Bangkok: Thailand. Kittikanya, C. (2004). Bangkok Post Economic Review 2004: Health [Online]. Available from: http://www.bangkokpost.net/ecoreviewye2004/health.html [Accessed September 9,2006]. 386 Good Practice in Health Financing. p. 387 Leesmidt, V., Pitayarangsarit, S., Jeegungwal, N., Piravej, N., & Bums, M. (2005). Assessment of the purchasingfunctions in the Thai Universal Coverage scheme for health care [Online]. Available from: http://www.ihpp.thaigov.netJpubli- report!Assessment%200f%20the20purchasing.doc. [ Accessed July, 18,2006] Mill, A., Bennett, S., Siriwanarangsun, P. & Tangcharoensathien, V. (2000). The response of providers to capitation payment: a case-study from Thailand. Health Policy, 51, 163-180. Ministry of Foreign Affairs of Thailand& United Nations Country Team in Thailand (2005). Global Partnership for Development: Thailand's Contribution to Millennium Development Development Goal 8. Keen Publishing (Thailand): Bangkok National Economic and Social Development Board (2004). Summary of Economic Growth Rate[Online]. Available from: http://www.nesdb.go.th/econSocial/macro/dev5.htm. [Accessed July 10,2006]. National Economic and Social Development Board (2005). Thailand's Official Poverty Lines [Online] . Available from: http://www.nscb.gov.ph/poverty/conference/papers 17_Thai%official%20poverty.pdf. [Accessed July 16,2006]. National Economic and Social Development Board (2006). Executive Summary Thai Economic Performance in Q4 and Outlookfor 2006 [Online]. Available from: http://www.nesdb.go.th/econSociallmacro/Outlook_dataiecon_outlook_ q4_48/03 _ Execut ive-Q4-2005.pdf. [Accessed July 20 , 2006] National Health Security Office (2001). The Universal Healthcare Coverage Scheme .. [Online]. Available from: http://www.nhso.go.th/30bahtEnglish/index.htm. [Accessed July 26,2006]. National Health Security Office and ABAC-KSC Internet Poll Research Center (2004). The Perspectives ofUC Members and Healthcare Providers to UC Program (In Thai). Nonthaburi: National Health Security Office. Pannarunothai, S. (2002). Medical Welfare Scheme: Financing and targeting the poor. In : Pramualratana, P.& Wibulpolprasert, S. eds. Health Insurance Systems in Thailand Nonthaburi: Health Systems Research Institute, 62-78. Pannarunothai, S. & Mills, A. (1997). The Poor Pay More: Health-Related Inequality in Thailand. Social Science & Medicine, 44(12), 1781-1790. Sriratanaban, J. (2002). Civil Servant Medical Benefit Scheme: Unregulated fee-for- service and cost escalation. In : Pramualratana, P.& Wibulpolprasert, S. eds. Health Insurance Systems in Thailand Nonthaburi: Health Systems Research Institute, 43-50. 387 Good Practice in Health Financing, p. 388 Pannarunothai, S. (2002). Medical Welfare Scheme: Financing and targeting the poor. In : Pramualratana, P.& Wibulpolprasert, S. eds. Health Insurance Systems in Thailand. Nonthaburi: Health Systems Research Institute, 62-78. Pitayaransarit,S., Jongudomsak, P., Sakulpanich, T., Singhapan, S.,& Homhual, P. (2004). Policy formulation process. In: Tangcharoensathien, V. & Jongudomsak, P. eds. From PoliCY to Implementation: Historical Events during 2001-2004 of Universal Coverage in Thailand. Nonthaburi : National Health Security Office. Phoolcharoen, W. (2004). Quantum Leap: The Reform of Thailand's Health System. Nonthaburi : Health System Research Institute. Sriratanaban, J. (2002). Civil Servant Medical Benefit Scheme: Unregulated fee-for- service and cost escalation. In : Pramualratana, P.& Wibulpolprasert, S. eds. Health Insurance Systems in Thailand. Nonthaburi: Health Systems Research Institute, 43-50. Srithamrongsawat,S.,& Torwatanakitkul, S. (2004). Implications of the universal coverage scheme on health service delivery system in Thailand. In: Tangcharoensathien, V. & Jongudomsak, P. eds. From Policy to Implementation: Historical Events during 2001-2004 of Universal Coverage in Thailand. Nonthaburi : National Health Security Office. Suraratdecha, C., Saithanu S., &Tangcharoensathien, V. (2005). Is universal coverage a solution for disparities in health care? Findings from three low-income provinces in Thailand. Health Policy, 73, 272-284. Tangcharoensathien, V. & Pitayarangsarit, S. (2003). Health Insurance Systems in Thailand: Major Research Questions (in Thai). Journal ofHealth Science ,12, 159-68. Tangcharoensathien, V., Srithamrongsawat & Pitayarangsarit, S. (2002). Overview of Health Insurance Systems. In: Pramualratana, P.& Wibulpolprasert, S. eds. Health Insurance Systems in Thailand. Nonthaburi: Health Systems Research Institute, 28-38. The World Bank (2006). Thailand Economic Monitor [Online]. Available from: http://siteresources.worldbank.org/INTTHAILAND/Resources/Economic- Monitor/2006apri1.pdf [Accessed July 21,2006]. United Nation Development Program (2000). Thailand Country Assessment. A New Poverty-Focused Development Path [Online]. Available from: http://www.undp.org/povertyreport/countryprofiles/thailandl.html [Accessed July 21,2006]. Vasavid, c., Tisayaticom, K., Patcharanarumol, W., & Tangcharoensathien, V. (2004). Impact of universal health care coverage on the Thai households. In: Tangcharoensathien 388 Good Practice in Health Financing, p. 389 ,V.& Jongudomsak, P. eds. From Policy to Implementation: Historical Events During 2001-2004 of Universal Coverage in Thailand. Nonthaburi: National Health Security Office, 129-149. Wasi, P. (2000). "Triangle That Moves The Mountain" and Health System Reform Movement in Thailand. Human Resources for Health Development Journal (HRDJ), 4 (2), 106-110. Wibulpolprasert, S., ed. (2004). Thailand Health Profile 2001-2004. Nonthaburi : Ministry of Public Health. World Health Organization (2006). The World Health Report 2006: working together for health France: Raphaell Crettaz 389 Good Practice in Health Financing, p. 390 Tunisia Afra Chokri Achouri Hedi 390 Good Practice in Health Financing, p. 391 Abbreviations and Acronyms ATCE : Agence Tunisien de Communication Tunisian Agency for External Communications Exterieure BCG : Bacille Calmette-Guerin Bacille Calmette-Guerin BM : Banque Mondiale The World Bank BTS : Banque Nationale de Solidarite National Solidarity Bank CHU : Centre hospitalo-universitaire University Hospital Center CNAM : Caisse Nationale d'assurance maladie National Health Insurance Fund CNPV : Centre National de Pharmacovigilance National Center for Pharmacovigilance CNSS : Caisse Nationale de Securite Sociale National Social Security Fund CSB : Centre de soins et de sante de base Primary Health Care Center DEP : Direction des Etudes et de la Planification Research and Planning Direction DPT3 : Three doses of combined diphtheria- Three doses of combined diphtheria-pertussis- pertussis-tetanus vaccine tetanus vaccine DTH : Direction de la Tutelle des H6pitaux Hospital Direction EPS : Etablissement de Sante Publique Public Health Establishment FMI : Fonds Monetaire International International Monetary Fund FSE : Fonds National d'Emploi National Employment Fund FSN : Fonds National de Solidarite National Solidarity Fund HC : H6pitaux de Circonscription District Hospitals HR : H6pitaux Regionaux Regional Hospitals INS : Institut National de la Statistique National Institute of Statistics IPT : Institut Pasteur de Tunis The Pasteur Institute - Tunis ISF : Indice Synthetique de Fecondite Total Fertility Rate LNCM: Laboratoire National de Contr6le des National Laboratory for Drug Control Medicaments MASSTE: Ministere des Affaires Sociales et de Ministry of Social Affairs and Solidarity and of Solidarites et des Tunisiens a l'etranger Tunisians abroad MENA: Middle East and North Africa Middle East and North Africa MSP : Ministere de la Sante Publique Ministry of Public Health OPV3 : Three doses of oral poliovaccine Three doses of oral poliovaccine PAS: Plan d' Ajustement Structure I Structural Adjustment Plan PCT : Pharmacie Centrale de la Tunisie The Tunisian Central Medical Store PIB : Produit Interieur Brut Gross Domestic Product PNAFA: Programme National d'aide aux Familles National Aid Program for Families in Need N ecessiteuses PT : Population Totale Total Population SMIG : Salaire Minimum Interprofessionnel Garanti Minimum wage SSP: Structures Sanitaires Publiques Public Health Facilities TBM : Taux Brut de Mortalite Crude Mortality Rate TBMI : Taux Brut de Mortalite Infantile Crude Infant Mortality Rate TBN : Taux Brut de Natalite Crude Birth Rate TCN : Taux de Croissance Naturelle Natural Growth Rate TPU : Taux de Population Urbaine Urban Population Rate 391 Good Practice in Health Financing, p. 392 Introduction The current report is realized within the framework of a World Bank study ["Good Practice" in Expanding Health Care Coverage: Lessons from Reforms in Low and Middle Income Countries]. Its principal goal is to provide the readers with a general understanding of health care coverage in Tunisia. Founded on a contextual analysis of the system's evolution and adaptation to its environment, this work intends to document this system's performance and to predict, based on current policy, future transformations. This occurs during a period marked by emerging challenges to meet the needs of the population and by the country's socio- economic, demographic, and epidemiologic transitions. The current situation calls for an in-depth study based on the population's health status, current policies and programs, as well as health sector reforms, including health insurance coverage, Tunisia's health status is among the best in Africa and in the MENA region, and is superior to that of other countries' with similar income levels, This is largely attributable to a combination of various factors, notably economic performance and poverty alleviation, as well as social development within all of its components. Yet, disparities between urban and rural areas as well as between different socio-economic groups persist, although not to the same extent as in other countries. The population's health status is significantly better in urban areas than in rural areas, However, this difference is also to a lesser extent than in other countries of the MENA region, especially in terms of infant mortality and life expectancy. Tunisia's positive record of accomplishments in health demands a comprehensive analysis of the health system, in order to understand the extent to which the country has successfully assured an acceptable level of health coverage for all of its population, even its vulnerable groups. On one hand, the emphasis will be placed on measuring the capacity of Tunisia' health system to offer services which improve the health status of the population. On the other, the performance of the health sector will be highlighted, This will be accomplished through an analysis of the key determinants affecting the current health coverage of the Tunisian population, the availability of human, financial, and material resources, and the access to health services. After describing and thoroughly analyzing the current situation and the process by which it evolved, for both the public and private sectors, our work explores potential future developments. The implementation of the mandatory social health insurance policy is the principal such development which introduces a new paradigm in the entire health system. Since this chapter developed within the framework of a multi-country study, we have been committed to follow, as much as possible, the terms of reference and an outline established beforehand by the World Bank. Consequently, we apologize to our readers for certain unavoidable repetitions and redundancies. 392 Good Practice in Health Financing, p. 393 Our report consists of three main sections. The first section provides an insight to the transitioning socio-politic and economic context. The second describes in detail the evolution of health care coverage in Tunisia during the last three decades, focusing on the components of the health system and the health insurance mechanism in relation to health provision, financing and regulatory mechanisms. The last part will put in perspective the latest reforms of the health sector and of health insurance coverage and will present efforts for evaluating these reforms. 1- The Economic Environment Tunisia is situated in the north of the African continent, in the Mediterranean basin, within the Maghreb regional collaboration. The country's total surface area is 163,610 km2. Its population, according to the most recent census from 2004, reached 9,910,872 million inhabitants. Tunisia has been continuously improving its human development indicators, as evidenced by the sustained increase of GDP per capita, the reduction in poverty rates, and the continuous amelioration of the population's living conditions and wellbeing. In 2004, with an average GDP growth rate of 5.5% during the past thirty years, the Tunisian economy is the most dynamic south of the Mediterranean, albeit slowed down due to the structural adjustments undertaken in the early 1980's. Besides its climate, its proximity to Europe, its socio-political stability, and its relatively highly qualified population's, Tunisia owns significant economic assets in spite of its modest natural resources (oil and phosphates). These assets, together with a stable macro-economic environment and a sound administration, have allowed Tunisia to continuously increase its GDP per capita (2336.5 US$ in 2004, compared to 1 502.9 en 1990 (see Table 1)). Due to this macro-economic management, the inflation rate has been maintained at level between 2 and 3% per year during the past ten years. Table1 : Tunisia's Economic Indicators IIndicators 1990 1995 2000 2004 iQDP per capita (constant 2000 US$) 1502,921 1655,289 2035,696 ~336,479 PDP per capita, PPP (current international $\ 3705,988 4681,146 6251,551 7767,597 pDP growth (annual %) 7,951 2,320 4,671 5,843 1P0verty gap at $1 a day (PPP) (%) .. .. 0,500 .. Poverty gap at $2 a day (PPP) (%) .. .. 1,328 .. lRevenue, excluding grants (% of GDP) 30,705 30,089 29,203 29,676 iTax revenue (% of GDP) 19,953 20,534 21,279 20,664 ~xternal debt (% of GDP) 59,6 67,8 ~id (% of GNI) 3,294 0,440 1,203 1,216 iUnemployment, total (% of total labor force) .. .. 15,6 13,9 mflation, consumer prices (annual %) 6,545 6,244 2,929 3,570 Source: WBI (World Bank IndIcators) 2005 Since the early 1990's, Tunisia introduced numerous reforms which allowed it to maintain an annual GDP growth rate around or above 5% and a control over their inflation rate during the decade spanning between 1996 and 2005. Additionally, these 393 Good Practice in Health Financing, p. 394 reforms supported significant growth of the private sector which contributed to investments, exports, job creation, and the subsequent reduction in the unemployment rate. Additionally, Tunisia concluded commercial agreements with about 60 countries, of which some involve preferential trade agreements I29 . Tunisia signed important trade agreements, including the WTO Agreement, the Bilateral Agreement with the European Union, and the Agreement with the League of Arab States. The liberalization of services, including health care, carries high stakes for Tunisia. 1.1. Government sponsored benefits Government sponsored benefits represent one of the most important instruments for the success of the development strategy's social dimension, through their decisive role in reinforcing mutual aid among various social categories, in supporting the advancement of human resources, and in achieving the indispensable balance between the urgent need for economic efficacy imperatives and the demand for social well-being. Table 2 : Government sponsored benefits 1986 1998 2001 Government sponsored benefits (millions dinars) 1,330 4,352 5,581 Percent of Government sponsored benefits which are 44,1% 50,1% 51,6% part of the state budget Government sponsored benefits per household and per 81 194 221 month (dinars) Source: TunIsian Agency for External CommunIcations, 2002 Tunisia's social policy includes social assistance programs which provide numerous benefits for both rural and urban populations. These programs include health care, schooling, housing, and employment benefits for youth and indigent populations, as well as financial assistance for the purchase of food and other basic products. The latter benefit is proven to be a costly method (0.5-1 % of GDP) for combating poverty, given that the entire population can access it (Ghali S. 2004). Improving living conditions and alleviating poverty have always been a central priority for Tunisia's public authorities. The government commits more than half of its budget to the social sectors, accounting for 19% ofGDP. The government's investments in social programs have been growing constantly and, between 1996 and 2005, the amount of Government sponsored benefits doubled. This resulted in 275 dinars of additional monthly revenue per household. The government commits financial and in-kind assistance to certain groups within the framework of its aid and social assistance programs. These programs can be categorized according to their permanent or temporary status, or according to their target populations 129 The Ministry of Trade website contains a list of the majority of Tunisia's trade agreements (Ministry of Trade, Information on-line. Available at: http://www.infocommerce.gov.tnlindexfr.htm.) 394 Good Practice in Health Financing. p. 395 and qualified social categories. Several programs, initiated by MASSTE and the Tunisian Union for Social Solidarity, can be placed into four categories: • Aid and social assistance programs • Programs for job creation and support and diversification of revenue sources • Programs to improve living conditions • Social protection and integration programs Besides government support, there are numerous programs for job creation and support and diversification of revenue sources. Some specific examples are: • The National Solidarity Funds; created in 1993. • The National Employment Funds; created in 2000. • The Tunisian Solidarity Bank; created in 1997. To promote revenue generation, non-governmental organizations have set up micro-credit programs to help rural populations finance investments, especially in agriculture and in arts and crafts. 1.2- Poverty In Tunisia, financial poverty is captured through the National Statistics Institute's national survey130 (for household budget and consumption) which is conducted every five years. Table 4 (below) provides information on trends in the poverty line, the poor population and poverty rates in urban and rural areas since 1975 (PNUD 2004). Poverty reduction was most perceptible beginning in the second half of the 1990's, alongside an increase in the rate of economic growth (an average of 5% in the 1995-2004 period) and more efficient and targeted policies and programs in the fight against poverty. For the period between 1980 and 2000, we notice: - The reduction by a third in the poverty rate: This conclusion is also valid for 2004 and 2005, since the poverty rate is estimated at 4% and 3.9%, respectively. - A significant reduction in financial poverty in rural areas. - Poverty is an urban phenomenon. 130 Eight surveys have been carried out since independence; 1967, 1975, 1980, 1985,1990, 1995, 2000,2005. The results of the last survey have not yet been published. 395 Good Practice in Health Financing, p. 396 Table 3 : Trends in the poverty line and poverty rates by area (1975-2000 Amounts in Tunisian Dinars 1975 1980 1985 1990 1995 2000 Poverty line Current Dinars Urban Areas 87 120 190 278 362 428 Rural Areas 43 60 95 139 181 221 Entire Country 64 91 147 222 292 351 Constant Dinars Urban Areas 272 267 269 278 273 276 Rural Areas 135 134 134 139 137 142 Entire Country 200 203 207 222 220 226 Poverty Rate Urban Areas 26,5 % 11,8 % 8,4% 7,3 % 7,1 % 4,9% Rural Areas 18,0 % 14,1 % 7,0% 5,8% 4,9% 2,9% Entire Country 22,0 % 12,9 % 7,7 % 6,7 % 6,2 % 4,2 % Source: INS Surveys (1975-2000); (PNUD 2004) Concerning the non-monetary aspects of poverty, significant progress has been made in terms of health and education, as evidences through Tunisia's latest social indicators. 2- Social, Demographic, and Epidemiological Environment In this section, we will analyze the Tunisia's triple transition (socio-economic, demographic, and health-related), based on the evolution of indicators and their potential effects on the health system. 2.1- Socio-economic Transition We noticed the following principal social transformations: • Urbanization: the population living in urban zones has rapidly increased from 40.1 % in 1996, to 64.1 % in 2004, and is projected to reach 67% by 2015. • Lifestyle: In particular, one may notice a tendency towards a sedentary lifestyle and an evolution towards the alimentation patters characteristic to populations from richer countries. • Tobacco and other addictive behaviors: The global tobacco prevalence is around 30%, exceeding 50% in males. Some other behaviors which impact health without yet representing a major public health problem should be carefully recognized and monitored. Alcoholism and drug abuse are two such behaviors. • Changes in family structure: The Tunisian family is becoming increasingly mononuclear, composed of the parents and their dependent children. 2.2- Demographic Transition Tunisia has unquestionably followed a unique transition pattern, in contrast with the other countries in the region. The abolition of polygamy immediately after the country's independence in 1956, as well as the related family planning policy and the health system's significant progress, provide a comprehensive explanation for the transition model observed in Tunisia. 396 Good Practice in Health Financing, p. 397 • The country's total population has grown by a factor of2.6 since independence, from 3,780 million in 1956 to 9,910 million in 2004 131 ; it is projected to further increase to 12,028 million inhabitants in 2020. The natural growth of the population has been significantly reduced: 3.2% in 1966,2.35 % (1984-1994) and 1.21 % (1994-2004). According to the Population Division of the United Nations (2004), this rate will be 0.87% for the period between 2020 and 2025. • The population's age structure has also been drastically transformed during in the recent past years: Tunisia's population will probably age rapidly (see the three population pyramids below). In 2020, according to United Nations Population Division projections (2004) 132, the number of those older than 60 years will near that of those less than 9 years old, thereby increasing to 12.5% (a dramatic difference from 9.5% in 2004 and 6.7% in 1984). In addition, it is worth mentioning that the transition from a high-mortality and high- fertility situation to one characterized by low-mortality and low-fertility is not occurring at the same time nor at the same pace observed in other high, middle, or low income countries. Source: INS Census (2004) 131 Data from United Nations Population Division, World Population Prospects The 2004 Revision 132 Population Database: http://esa.un.org/unpp/index.asp. 397 Good Practice in Health Financing, p. 398 Tunisia, 1990 Tunisia, 2005 90-94 90-94 75-79 75-79 60-64 60-64 QI QI CI CI < 45-49 < 45-49 30-34 30-34 15-19 15-19 0-4 0-4 15 10 505 10 15 15 10 5 0 5 10 15 Population (%) Population (%) - _.. _ - . - - - Tunisia, 2020 90-94 75-79 1/1 60-64 Q) ~ 45-49 30-34 15-19 0-4 15 10 5 0 5 10 15 Population (%) 398 Good Practice in Health Financing, p. 399 • The decrease in infant mortality resulted in an increase of life expectancy at birth, which has changed from 37 years in the late 1940's to 52 years in the late 1960's and to 73 years in 2004 (71.4 in males and 75.3 in females). Between 1956 and 2004, the Tunisian population has gained exactly 26.3 years in life expectancy at birth (Figure 1). The infant mortality rate (IMR), which was close to 200 per thousand in 1956, currently reaches 20.6 per thousand. According to the National Institute of Statistics (INS) projections, the IMR will further decrease to 10.0 per thousand in 2020 and to 8.0 per thousand in 2030. This spectacular fall in infant mortality is attributable to a close interaction of three principal strategies: • The promulgation, since August 1956, ofa law that prohibited polygamy. • The full-scale implementation, since the early 1960s, offamily planning policy and of fertility reduction. • The integration, since 1960, of maternal and child health and of reproductive health into the most important national programs: monitoring of pregnancies, immunizations, post natal care, and birth spacing. These strategies have evolved since and have been updated periodically, within an economic environment characterized by the fight against poverty, increased access to mandatory education for females, and a continuous improvement of health care access, in particular regarding childhood infectious diseases. • The continuous drop in fertility is one of the most remarkable aspects of Tunisia's demographic transition. Starting from a level close to that of less developed countries, Tunisia's fertility rate was 2.02 in 2004, nearing that of developed countries, where average fertility rate is 1.57. • The crude mortality rate is 16.8 per thousand in 2004, compared to 50 per thousand 45 years ago. It is predicted to decrease to 15 per thousand towards the period between 2025 and 2030. Thus, because of the slow mortality reduction process post-1950, the consequent rapid decrease in mortality and, in particular, the reduction in fertility, Tunisia can be classified among other countries which have followed a model of delayed demographic transition. Such countries include: South Korea, Hong Kong, Sri Lanka, and China (Omrane A.R. 1983). 399 Good Practice in Health Financing, p. 400 ___________________________ Table ~~ 4: Trends fl dd em02rap hie indo . T .. ____________________________ ~~ Indicators 1956 1966 1975 1984 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 1 TP (millions) 3.780 4.533 5.588 6.966 8.154 8.318 8.490 8.657 8.815 8.958 9.089 9.215 9.333 9.456 9.564 9.670 9.782 9.840 9.9101 TPU% 40.1 47.5 52.8 58.0 58.6 59.6 60.0 61.7 61.30 61.60 61.90 62.2 62.5 62.8 63.1 63.4 63.7 64.1 J TeN par an (%) 3.50 3.10 2.70 2.60 2.40 2.00 2.00 2.00 1.70 1.60 1.50 1.40 1.3 1.3 1.1 1.1 1.1 1.09 1.08 i TBN (%0) 50.0 44.0 36.6 32.4 25.2 24.9 24.9 24.0 22.7 20.8 19.7 18.9 17.9 16.9 17.1 16.9 16.7 17.1 16.8, TBM (%0) 25.0 15.0 10.0 6.5 5.6 5.6 5.6 5.7 5.5 5.8 5.5 5.6 5.6 5.6 5.6 5.5 5.8 6.1 6.0 I TBMI (%0) 200 120 76.9 51.4 41.0 31.8 32.0 30.0 27.6 26.6 26.2 25.0 20.6 1 ISF 7.2 7.15 5.8 4.7 3.5 .. 3.2 .. 2.9 2.67 2.51 2.38 2.23 2.09 2.08 2.05 2.0 2.1 2.021 Source: MSP/DEP et INS PT Total Population TPU Urban Population Rate TCN Natural Growth Rate TBN Crude Fertility Rate TBM Crude Mortality Rate TBMI = Crude Infant Mortality Rate ISF Total Fertility Rate 400 Good Practice in Health Financing, p. 401 2.3. Epidemiologic Transition A multi-faceted epidemiologic transition followed Tunisia's demographic transition: health- related, of course, but also social, economic, and cultural. This transition is inadequately documented in the absence of a reliable and sustainable information system. Tunisia's epidemiologic profile is no longer the same as the one which prevailed until the late 1980's. Indeed, we notice: • The decline - occasionally the eradication - of the «traditional» infectious diseases (malaria, schistosomiasis, trachoma, tuberculosis, infectious diarrhea) and of early childhood ones (poliomyelitis, neonatal tetanus, diphtheria). In 2002, non-communicable diseases accounted for 79.7% of mortality and 70.8% of morbidity (Table 5) • The emergence of chronic and degenerative non-communicable diseases, with multifactorial etiology and high fees to access health care services. • The aggravation of road traffic accidents and their consequences in terms of mortality and morbidity. Table 5: Global Burden of Diseases 2002 Estimates Burden of Diseases by Disease Burden of Diseases by Disease Categories: Deaths only Categories: DALY s (Estimated (Estimated total deaths ['000] total DAL Y s ['000] by cause, 2002) by cause, 2002) Value % Value % Communicable, maternal, perinatal & 5.3 9.50% 217 14.10% nutritional conditions Noncommunicable 44.5 79.70% 1093 70.80% diseases Injuries 6 10.80% 234 15.20% Total 55.8 100.00% 1544 100.00% Source: WHO (World Health OrgamzatlOn) 2004 WHO estimates (2004) of the top five mortality sources for 2002 finds cadiovascular disease in the first place, followed by malignant neoplasms, injuries, digestive diseases, and respiratory diseases. On the other hand, neuropsychiatric conditions precede cardiovascular diseases, and occupy the second place when morbidity and disability-adjusted life years are included in the calculations. 401 Good Practice in Health Financing, p. 402 T a bl e 6 : T op F' lseases c auses lve 0 fB urden D' Top 5 Burden of Diseases: Deaths only Top 5 Burden of Disease: DALYs (Estimated total deaths ['000] by cause, 2002) (Estimated total DALYs ['000] by cause, 2002) GBD Cause Value GBD Cause Value Cardiovascular diseases 26 Neuropsychiatric conditions 268 Malignant neoplasms 5.5 Cardiovascular diseases 237 Unintentional injuries 5.4 Unintentional injuries 211 Digestive diseases 3.3 Sense organ diseases 171 Re~iratory diseases 2.9 Digestive diseases 88 Source: WHO (World Health OrganIzation) 2004. Additionally, Ben H. Romdhane's 2002 study shows a disease profile similar to those of western countries: • Hypertension prevalence varies between 15 and 21 % (Canada: 14%). • Diabetes prevalence is predicted to change from 3.5% to 10% during the next 20 years (France: 16.2%, Canada: 5%). • A preponderance of tobacco related cancers in men, with lung cancer on the top of the list. According to cancer records, its incidence (per 100,000 population) is 27.9117.5 (North/South, respectively); it remains, nonetheless, much lower than that of industrialized nations, such as France (55.1) or Canada (82.5). - The prevalence rate of mental illnesses (particularly depression) is similar to those reported in the international literature, with a global prevalence of major depression of 8.2% and of schizophrenia of 0.57%. - The number of disabled individuals has increased by a factor of 2.5 since 1975; the global prevalence of disabilities in individuals older than 60 years is four times higher than that of the general population (28% vs. 6.5%). 3- The social and political context • Internationally, Tunisia has been dedicated to reinforce its integration in the Arabic, Muslim, and African environment; belonging to both the Maghreb and the Mediterranean region. It has signed agreements with the World Trade Organization - Organization Mondiale du Commerce (OMC). Additionally, it is the first country in the MENA region to have reached Associate Agreements with the European Union (AAEU) in 1995. Moreover, Tunisia remains the principal leader of initiatives for the revitalization of the Arab Maghreb Union. • Tunisia's economic and social progresses have been supported through the coordination of policies and reforms and a competent public administration. First of all, the pursuit of practical macro-economic policies has created an environment conducive to the growth and development of the private sector. Second of all, the constant efforts of sector-wide reforms pursued since the early 1990' s, as well as the determination of integrating Tunisia in the global economy, have created favorable conditions for growth in productivity and for an increase of economic competition. • In terms of development, Tunisia's activities have promoted the emergence and expansion of a large middle class, who is particularly encouraged to participate in the development process. 402 Good Practice in Health Financing, p. 403 • Regarding social development, Tunisia has documented rapid and sustained progress. Apart from indicators related to the development of education and training, health indicators have also significantly improved. According to a recent MENA report regarding Governance (BM 2003), Tunisia, in comparison with the average lower-middle income countries, is ranked markedly higher for the quality of its public administration. This attribute translates itself into the administration's capacity to formulate and implement sound economic policies as well as into the credibility of the institutions that govern the relations between the State and its citizens. 4- Health financing and health insurance Health financing sources are often grouped into three major categories offinancing agencies: the State, the Social Security Funds (Caisses de Securite Sociale (CSS)), and the households. To that list, one must add the following: • Donors and other international organizations which contribute in diverse ways to health financing. Although their contribution has never surpassed 1% of total health care expenditures, it was not integrated as a component of financing. • Medical services • In-house medical services provided by companies and occupational health also contribute about 1%. These various financing mechanisms are further explained in the following paragraphs. 4.1. Health care coverage in Tunisia The health insurance system includes a multitude of schemes, put together over time according to the needs of certain professional categories. The majority of the Tunisian population benefits from health care coverage (99%), through various means: mandatory social security schemes (66%) or free medical services (8% - free of charge, 25% reduced charges). Additionally, supplementary insurance schemes (group insurance and mutual insurance companies) have developed in response to the insufficiencies inherent to health insurance schemes of the social security fund. Figure 2: Coverage rates by health insurance scheme .8% ii66% 025% D1% I_ F.... of charge 0 Reduced Charge 0 Other • $ociallnsurance 403 Good Practice in Health Financing, p. 404 4.1.1- Health care benefits packages: The concept of benefit packages has not yet been fully incorporated into the health care and the health insurance systems. Generally speaking, patients in need of care have access to all health care categories, without the limitation of a particular benefit package. The only exception lies in the reimbursement option ofCNRPS's mandatory scheme which limits the coverage of chronic illness (from a predetermined list) and of surgical interventions. Nevertheless, all reimbursement schemes stipulate an annual expenditure cap (limit) per household. 4.1.2- Various health insurance mechanisms: The health insurance mechanisms vary according to social category and to sources of financing. • Preventive and public health activities, individual or collective, are guaranteed at no cost for the entire population. The same model applies to those with illnesses which qualify them to take part in research studies. Individual costs for preventive services provided by the private sector are relegated to the respective households. • Certain professional categories (including their families) benefit of free health care in primary health care centers (soins de sante primaires (SSP)) and other special schemes, in addition to their social security membership. Additionally, in certain situations, they also benefit from their mutual insurance membership which permits access to services from private providers. These categories are: the military; national security forces; customs agents; health care professionals. The members of the resistance movement (participants in Tunisia's liberation movement) and their dependants also benefit from free health care services through the SSP and military hospitals. • According to the health care laws currently in effect 133 , the State guarantees free or subsidized health care services to low-incomes groups through two public medical aid schemes 134: Free health care scheme: This category, defined according to the poverty line, includes target families ofa long-term government assistance program. Ruling Nr. 98-1812 describes the conditions and allocation and distribution mechanisms of the free health care card. It also specifies that these free services pertain only families who are eligible candidates or current beneficiaries of a long-term government assistance program. The free health care card, for a period of 5 years, is issued in function of set national limits and regional quotas. Beneficiaries of reduced fees or charges: Fee reduction cards are granted according to annual household income accounting for family size, and should not exceed: • The minimum wage (salaire minimum interprofessionnel garanti (SMIG)) for families of2 members ofless. • One and a halftimes the SMIG if the family includes between 3 and 5 members. • Twice the SMIG if the family includes over 5 members. 133Article 35 of law Nr. 91-63 July 29th, 1991, the public health organization law. 134 In 1998, there was a refonn of this health assistance scheme. It approved the revision of health care card eligibility and distribution, mainly to better target the needy families counted by MASSTE, and of reduced service fees for persons with limited incomes, which cannot enroll in one of the social security funds. 404 Good Practice in Health Financing, p. 405 In addition to the conditions mentioned above, current affiliates to social security are not eligible to receive fee reductions. The MASSTE social agents will work in the field to verify the lawful implementation of eligibility conditions. The fee reduction card is issued for a period of 5 years and must be validated annually through the deposit of 10 TND. They are subject to payment of the inclusive contribution for each medical visit. The total number of fee reduction cards is determined through regional quotas. In order to better target the intervention and the necessary support to families in need, the MASSTE has established a national poverty database. This database includes socio-economic information for the families who are beneficiaries of the National Program of Aid to Poor Families (Programme National d'aide aux Familles Necessiteuses (PNAFN)). This program includes tools to monitor socio-economic status, which are an important resource for various social partners who much choose beneficiaries for various aid programs (long-term, temporary, free or reduced charges for health care). It is also useful in the selection of beneficiaries for income generation activities within the context of regional development programs and of job creation assistance mechanisms. Within the framework of the two medical aid programs (free health care and reduced health care charges beneficiaries), the poor population's access to health services is not subject to any restriction or rationing. However, the budget limitations of public hospitals and drug shortages can be considered as an implicit form of service rationing. In addition, the reform of this scheme, introduced in 1998, has contributed to the expansions of social coverage under formal insurance schemes, particularly by appealing to independent individuals, workers in the agricultural and informal sectors. In combination with the other measures, this program has improved the real social coverage rate which reached 85.5% in 2004 and is estimated to further rise to 90% in 2006 (see Table7). • Social security schemes are available for employees and employers, both of which are mandated to participate. The funds are pooled within two principal funds: the National Social Security Fund (la Caisse Nationale de Securite Sociale (CNSS)) and the National Pension and Social Protection Fund (la Caisse Nationale de Retraite et de Prevoyance Sociale (CNRPS)). These national funds provide various allowances for pensions, family services, social protection, and for industrial accidents and occupational health. The CNSS covers private sector workers, while CNRPS covers employees of the State, local public unions and other public institutions. The coverage rates vary by sector: a constant 100% for State employees (CNRPS), it continues to increase within the private sector (CNSS) due to several factors: The growth of legal allowances for those socially insured and their dependents which reached over 1,943 million dinars in 2003, or the equivalent of 30.8% of government sponsored benefits and 6% ofGDP. The extension of coverage to social categories which were previously not reached by the social security scheme (homemakers and construction workers), to those unable to participate in their respective schemes (small-scale fishermen, farmers, and artisans), to craftsmen and to artists. 405 Good Practice in Health Financing, p. 406 Table 7: CSS insurance coverage rates (% of eligible population) 1987 1997 1998 2001 2004 2006 Total, private sector (affiliated with CNSS) 41.6 74.6 77.0 Non-agricultural workers 66.3 94.0 96.3 Agricultural workers 16.9 41.7 42.8 IndeQendents 9.2 48.5 50.22 Total, public sector (affiliated with CNRPS) 100 100 100 100 100 100 Total, eligible beneficiaries 54.6 80.7 82.6 84.2 85.5 90 . . .. Source: MInIstry of socIal and sohdanty affaIrs and of TunlSlans abroad (MASSTE). The health care benefits covered, which vary by fund organizations, can be summarized as follows: .:. The CNRPS offers two schemes: ~ A mandatory scheme with two mutually exclusive options: ./ The first allows beneficiaries with chronic illnesses and surgical interventions to access various public and private providers . ./ The second, a "health care card," covers all the costs within the SSP which are related to the Ministry of Public Health, in the same manner for both the ill as well as for the beneficiaries of reduced fees. ~ An optional scheme which supplements the reimbursement schemes for common illnesses . •:. The CNSS offers its beneficiaries, regardless to which scheme the head of the household belongs to, two means of illness coverage, based on in-kind benefits: ~ within the SSP, the same benefits for those ill as well as for those who benefit from reduced fees. ~ within its polyclinics, which only offer ambulatory care, these depend on the patient's co payment at the point of service. Besides the health care services covered by these schemes, the two Social Security Funds finance completely or partially: ~ Expenditures for special services, intensive care covered through special agreements with public and private providers. ~ International health services. • Contracts with private insurance companies are made, depending on the insurance company, by private and public sector enterprises for risk pooling to the benefit of their employees. • The mutual insurance companies and associations of private and public sector employees, offer several social services including coverage of health care costs. Certain mutual insurance companies, membership is mandatory. Membership in a mutual insurance company or in an insurance group is accomplished in parallel to the mandatory affiliation of both the employee and the employer to the CSS. They usually adopt a fee-for-service reimbursement scheme and limit their allowances per beneficiary per year. 406 Good Practice in Health Financing, p. 407 • In the complete absence of insurance, the patient bears the entire health care cost out of pocket, irrespective of whether the care was provided in the private or public sector. 4.2- Health expenditure indicators Tunisia commits almost 5.5% of GDP towards health expenditures, compared to the average 8% dedicate by European Union members. During the period between 1980 and 2004, total health care expenditures have increased by a factor of 15.2, with an average annual increase of 12%. In this manner, health care expenditures have risen from 143 to 2,170 MTND (Table 8). However, a closer analysis of this growth shows that it was 7.3% during the period between 1980 and 1990, subsequently reaching 9.9% between 1990 and 2000, only to become constant at 9.4% from 2000 to 2004. Table 8: Health expenditures, as part of GDP and as expenditure per caJ!ita {1980-2004). 1980 1985 1990 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Health 143, 290, 578, 938, 1160, 1258, 1238, 1367, 1489, 1644, 1673, 1824, 2170, expenditure 0 3 0 0 6 3 0 0 5 0 0 6 0 s (MTND) Expenditure s as % of 3,2 4,2 5,3 5,5 5,5 5,4 5,5 5,5 5,6 5,8 5,8 5,6 5,6 GDP Expenditure 105, per capita ---- 39 72 114,5 123,4 132,7 144,5 155,8 169,9 170,4 184,5 217,5 5 (TND) Source: MSPIDEP Similarly, the % ofGDP committed to health has grown from 3.2% in 1980 to 5.6% in 2004, and health expenditures per capita have grown from 39 TND in 1985 to 217.5 TND in 2004, which translates into an increase by a factor of 5.6 and an average annual growth of 9.5% (Table 8). 4.3- Trends and consequences of public and private health expenditures in Tunisia Health financing has largely been supported by the State budget in addition to the social security funds until the end of the 1980' s. During this period, the majority of public funds dedicated to health was around 65%. The financial crisis during the second half of the 1980s and the subsequent adjustment programs have significantly reduced the State's contribution which was replaced primarily by increased household expenditures and, later on, by a reinforcement of social security contributions. For the period between 1995 and 2004, the funds for health financing were equally assured through public funds (from the State and Social Security) and private funds (direct payment from households and supplementary health insurance). 407 Good Practice in Health Financing, p. 408 Table 9: Health expenditure indicators - Tunisia (1995-2004) 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Public expenditures 49.5 52.1 47.7 49.9 51.6 48.5 51.0 50.0 47.5 48.7 (PubExp) % Total Expenditures (TE) Private expenditures 50.5 47.9 52.3 50.l 48.4 51.5 49.0 50.0 52.5 51.3 (PrivExp) % TE PubExp % of total public 7.l 7.2 6.7 7.l 7.6 6.9 7.9 7.6 7.7 7.7 expenditures PubExp financed by 0.8 0.8 0.8 0.8 0.7 0.7 0.7 0.2 0.4 nJa external resources as % of TE Household health 92.5 100 100 92.7 83.0 81.7 82.5 83.0 82.9 82.9 expenditures % PrivExp Social security health 28.8 31.9 24.5 24.3 24.l 26.7 22.9 22.9 23.5 23.3 expenditures % PubExp PubExp financed through 70.8 70.7 69.9 70.2 70.5 70.5 70.6 general taxation as % Total Health Expenditures (THE) PubExp financed through 54.1 59.9 general taxation as % of total public expenditures Total health expenditures III 117 III 116 122 114 120 126 141 163 per capita at the current exchange rate ($ USA) Total health expenditures 270 286 302 320 344 367 403 413 per capita in international dollar. Source: MSP/DEP 4.4- Trends in health expenditures by financing agents The evolution of health expenditures according to the contribution of each financial agents for the period between 1980 and 2004 is shown in Figure 2, knowing that private insurance and mutual insurance is less known, therefore remaining at a low level and often being confused with other household expenditures. Only CNS estimates from 2000 (Arfa C. et Achour N. 2004) have allowed the analysis of health expenditures. The developments in this area lead us to arrive to the following conclusions: 4.4.1- Public expenditures • Financing through taxation (State budget): from 1980 to 2004, an average annual growth of 8.1 % has been recorded for State expenditures, however, subsequently this role in financing has decreased. • Financing through contributions to the Social Security Fund: from 1980 to 2004, an average annual increase of 13.6% has been recorded for these expenditures, with a significant increase relative to total expenditures. 4.4.2- Private expenditures Private financing sources includes individual expenditures, whether or not they are reimbursed by private or mutual insurance companies, and occupational health and other expenditures for curative services provided by companies. It is also noteworthy that household continue to bear the majority of these expenditures. From 1980 to 2004, we noted 408 Good Practice in Health Financing, p. 409 an increase of 14.3% in the average annual growth of private expenditures, as well as a significant increase from 36.6% in 1980 to 51.3% in 2004. From 1980 to the present, the financing of health expenditures has been marked by a net decrease in public financing and a significant increase in the direct contribution of households, due to diminishing incomes and to high prices of health care services and health commodities. Faced with inequity risks through a reduction of financial access to health and in hopes of improved sharing of the health sector's financial burden, the public powers have initiated reforms in the social securities health insurance schemes. Figure 3: Trends in the relative contribution 01 health financing agents (1980-2004) 100~~--~--~~~--~~~~~~--~--~--~~~~ 80~~~~~~~~~~~~~~~~~~~~~~~~~ 6°'==.lifiiR.ii5~ tJri 40 20~~~~"~~~~~~~~~~ O+*~~~~~~~~~~*+~~~~~~~~~~~ ~~~~~~~~~~~~~~*~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~ -.o+-State ...... CSS -'-Households ~ Mutual Ins. Compo Since 1990, the recorded changes have been ultimately confirmed by National Health Accounts developed for the years 1995, 1997, and 2000. Therefore, some explanations for these developments can be put forth: • The Structural Adjustment Program resulted in the reduction of the State's social expenditures, in particular of those devoted to health. Consequently, the SSP were confronted with financial difficulties which limited the availability of medical commodities. In this situation, patients often had to resort to the direct purchase of drugs and other necessary medical commodities. Moreover, copayments were raised several times in order to increase the financial resources available for SSP. • The CNRPS reimbursement rates and their annual caps were frozen at their 1980 levels in order to increase overall financial stability, while the fees and prices of medical commodities have not ceased to increase. Consequently, copayments have continued to increase with time. • Health care costs have continued to rise under the double burden of price increases and of the demographic transition and have been affected by budgetary constraints of public health care facilities and of health insurance organizations; the combination of these factors has contributed the sharp rise in households' direct costs in contribution to health care financing. This tendency reduced access to health care and led many households into a vicious circle of poverty caused by illness. Consequently, it should become a candidate for health policy reform. 409 Good Practice in Health Financing, p. 410 5- Payment and financing 5.1- Sources of financing The current financing system, characterized by a diversity of financing sources and payment options, varies by provider category. In general, government financing is set aside for the SSP. Social security financing also benefits the SSP. Whereas financing of private sector provision primarily remains the households' responsibility. 5.2- Public sector financing The SSP which are under the auspice of the MSP are financed by the state, social security, and households. Hospital reforms have contributed to a revision of the share of the burden held by these three agents. These reforms introduced new payment options through social security funds for hospital providers. • The share of social security funds has increased from 11 % to 22.2% between 1990 and 2003 (Figure 3). This rise was obtained due to: An initiative for priority investments in certain programs. The establishments of a billing system for hospital and ambulatory care, by virtue of annual agreements between MSP and MASSTE, implemented gradually since 1996 and limited to university and regional hospitals. • The share of households has also undergone a slight increase from 10% in 1990 to 14.2% in 2003. This is explained by: Step-wise modifications for copayment amounts for public entities, initiated in 1991, 1993,1994, and 1998. Patient contribution to costs incurred from diagnostic and treatment services (10% of the share applied to paying patients) for those socially insured in 1994 and for beneficiaries of reduced tariffs in 1998. An increase in tariffs of paying patients, initiated successively in 1991, 1993, and 1996. A revision of the fees associated with classification of medical services in 1996. The State's subvention to SSP and MSP has increased in value, but its share in public sector financing has decreased by 15.4%. Source: MSPIDEP Since 1990, the relative decrease in State subsidies to the public sector was greatly compensated by the CSS (73%) and by households (27%). Public funds remained the principal means of public sector financing, since the share of private funds for private sector financing has not surpassed 14.2%. 410 Good Practice in Health Financing, p. 411 5.3- Private sector financing Households bear the greatest share of financing. Insurance groups and mutual insurance companies, in addition to their low coverage, only take responsibility of a small share of these costs. Social security funds are responsible for health care costs within the private sector within the current regulatory conditions: CNRPS reimbursement schemes and CSS agreements. Table 10 (below) provides an overview of health care provider payment options according to the health care coverage scheme. Differences in beneficiaries' access to care originate from this situation, highly dependent on the coverage type. The benefit of free health care and reduced charges for these services is only provided by the MSP's SSP. Access to CNSS polyclinics is reserved to members of this fund. CNSS' socially insured cannot have coverage benefits for services provided within the private system (except for extreme cases). This may explain the resort to group insurance which allows coverage of health expenditures in the private sector. CNRPS's socially insured have the option to seek care in the private sector. In addition, they can join professional mutual insurance companies which allow them to access this type of care as a supplement to CNRPS' mandatory coverage. This fragmentation in health insurance coverage has arisen from numerous actions taken to improve the population's access to health care. These actions have resulted in the successful reduction of inequity among various categories of the population, of financial difficulties felt by social security funs (especially CNRPS), and of the budgetary disequilibrium for the MSP's SSP. In this context, the reform of social security health insurance schemes are initiated in order to harmonize the access to health care for the insured population, and therefore for the Tunisian majority. 411 Good Practice in Health Financing, p. 412 Table 10: - ---- Health --- --------- ----- -- ----~- dfi -----------------~ Affiliated Government CNSS CNRPS Supplementary insurance agency Assistance Scheme General Mandatory Mandatory Scheme Optional Scheme Free Reduced Group Insurance Mutual scheme and Scheme (reimbursement) (reimbursement) Health Fees insurance other schemes (health care Care companies cardL Financing 4.75% for the 1% 1% employers + 2.5% employers + Complete lOTND Monthly Contribution method general employers + 1% employees 4.0% employees exemption per year premium from rates vary from scheme, and 1% 4% to 7%, 1% to 7%. For less for the employees including both employees, the other schemes employers and rates are not employees higher than 2%. Types of Public and Public Public and private Public and private Public Public Public and Public and covered polyclinics of private private health care the CNSS providers Types of No cap in the No cap in the - Reimbursement of Reimbursement of All care A cap on total A cap on total coverage public sector, public sector, 80% of medical fees 80% of medical fees (ambulatory , expenditures and expenditures and depending on depending on and of all medical and of all medical hospitalization, some service- some service- availability. availability. services at regular services at regular emergencies) specific caps specific caps fees (year 1982) fees (year 1982) received III the (surgery, dental, (surgery, dental, - Complete public sector, drugs) drugs) reimbursement for - Annual cap for depending on pharmaceutical reimbursement of availability. commodities, pharmaceutical without a cap. commodities (200 TD per year) Source: Adapted from the Study of the Health Sector (BM, 2006) 412 Good Practice in Health Financing. p. 413 Table 11 : Health °d t thods bv t f - _ . ----0- Public Sector Para statal Sector Private Sector CNSS Military hospitals Polyclinics Without any Fee for service Payment of public fees Payment of specific fees coverage Free health care 0 N/A N/A N/A Reduced Fees Copayment CNSS Negotiated fees for the agreed services Copayment - Patient pays a - Patient pays a copayment copayment ~ CNRPS ~ - Billing to the funds of charges - Billing to the funds of Eo-< Mandatory - Negotiated fees the agreed services Fee in university and regional specific conventional ~ ~ for service for long-term illness and hospitals N/A ~ J. Scheme charges ~ surgical interventions - Negotiated fees the agreed ~ 0 services - Negotiated fees the U agreed services Optional - Negotiated fees the agreed services Fee for service (reimbursement) N/A Fee for service Scheme - Fee for service (reimbursement) (reimbursement) Group Insurance Fee for service and Mutual Fee for service (reimbursement) Fee for service (reimbursement) N/A (reimbursement) insurance companies 413 Good Practice in Health Financing, p. 414 5.4- Reimbursement procedures Payment options and procedures vary according to the health care schemes offered by social security funds and to the provider (Table 11) . •:. The reimbursement of the fees paid by patients to their public or private providers. This procedure is applied by CNRPS within the framework of the mandatory scheme- reimbursement option - and of the optional scheme, by group insurance, and by mutual insurance organizations . •:. Direct payment of public or private providers According to a series of special agreements of intensive care financing, the two CSS pay directly to the provider the entirety of health care costs, according to the tariffs in place at that time . •:. Payment in-kind for certain medical commodities Under certain conditions, CSS are required to provide their beneficiaries, in a social action framework, drugs, prosthetics, and other commodities necessary for the patient's complete medical care and attention . •:. An annual lump sum payment to the State's Treasury Historically, this was the first mechanism for the contribution of the two social security funds towards health care costs of socially insured, for services received in public health facilities. This assured sufficient resources in the Treasury for the allocation of finances for the budget ofMSP's SSP. This payment remains in effect and has been increased on several occasions, as recently as 2006 . •:. Billing procedures for the health care costs of those insured within the two social security funds This procedure has become the predominant means of university and regional hospital budget financing. It is based on the following principles: - Each year, a fixed amount is allocated for billing services and operations by the two social security funds. This is in addition to the amount that the funds do not cover, but which is provided by the health facilities involved. This annual amount is shared by the MSP among the various structures in question (university and regional hospitals). This distribution is assigned though forecasted resources of operating budget of each of the facilities in question. Service fees are negotiated with the funds' and are described as follows: • A single fee for outpatient services covering the medical visit, any additional medical tests and recommended drugs. • Fees for medical services in day hospitals, defined according to pathology and/or therapeutic procedures. • Hospitalization charges, fixed per duration of stay - regardless of the duration - for various medical and surgical specialties . •:. CNSS subsidization, in grants, of the polyclinics it owns . •:. In their encounters with the MSP's SSP, and outside CNRPS's reimbursement schemes, the socially insured from both of the social security funds are subject for copayments for outpatient consultations, for any additional procedures, and for hospitalization. The CNSS socially insured observe similar regulations when utilizing polyclinic services. 414 Good Practice in Health Financing, p. 415 Table 12: Provision for CNRPS affiliates Health Care Card Reimbursement System Reimbursement System Coverage of health services provided by Coverage for surgical interventions and Coverage of common illnesses Provision government-related hospitals (for health long terms illnesses (Common illnesses care card holders) excluded) General • CopaymentDO 1,500 D for CSB 80% of the charge set in the classification • private: 80% without surpassing 3.500 D Consultation • Copayment 2,000 D for HC of medical services in public and private • public: 90% of the official tariff facilities Medical Specialist • Copayment 3D for HR 80% of the charge set in the classification • private: 80% without surpassing 5.000 D Consultation • Copayment 4,500 D for CHU of medical services in public and private • public: 90% of the official tariff facilities Medical Services In public hospitals, the patient pays 20%, not to Long-term illness or surgery: 100% of • private: 80% official tariffs exceed 30 D official tariffs • public: 90% of the official tariff with a cap of SOD per beneficiary per year for both Laboratory In public hospitals, the patient pays 20%, not to 100% of official tariffs • private: 80% official tariffs services exceed 30 D • public: 90% of the official tariff with a cap of SOD per beneficiary per year for both Midwife services In public hospitals, the patient pays 20%, not to 100% of official tariffs • private: 80% official tariffs exceed 30 D • public: 90% of the official tariff Speech therapy In public hospitals, the patient pays 20%, not to 100% of official tariffs : 80% official tariffs services exceed 30 D • : 90% of the official tariff Dental surgery In public hospitals, the patient pays 20%, not to 100% of official tariffs • private: 80% official tariffs services exceed 30 D • pUblic: 90% of the official tariff - Copayment for consultation • copayment 1,500 D for CSB with a cap of 70D per beneficiary per year for both • copayment 2,000 D for HC provision types. • copayment 3D for HR DforCHU 135 In the public sector, copayments vary by facility type. 415 Good Practice in Health Financing, p. 416 Table 13: Provision for CNSS affiliates Provision Provision covera2e provided by SSP and CNSS polyclinics General consultation • Copayment of 1,SOOD for CSB • Copayment of 2,000D for HC Medical specialist • Copayment of 3D for HR consultation • Copayment of 4.S00 D for CHU General practitioner visit in a Copayment onD CNSS polyclinic Medical specialist visit in a Copayment of 4D CNSS polyclinic Medical services Patient pays: - In public hospitals: 20% of the current public sector tariff, no to exceed 30D - In polyclinics: 20% of the current Qublic sector tariff Patient pays: - In public hospitals: 20% of the current public sector tariff, no to exceed Laboratory services 30D - In polyclinics: 20% of the current public sector tariff. Patient pays: - In public hospitals: 20% of the current public sector tariff, no to exceed Mid-wife services 30D - In polyclinics: 2 D. Patient pays: - In public hospitals: 20% of the current public sector tariff, no to exceed Speech therapy services 30D - In polyclinics: 20% of the current public sector tariff.. Patient pays: - In public hospitals: 20% of the current public sector tariff, no to exceed Dental surgery services 30D - In polyclinics: 10% of the current public sector tariff. In CNSS polyclinics, the patient pays: - Long-term illness: IS % of charges. Drugs - Common illness: 20% of charges. In the MSP's SSPs drugs are provided without any patient contribution. 6- Other considerations related to health care financing: .:. Following an INS household survey, health expenditures have risen the quickest among all other household expenditures. Health expenditures have doubled between 1975 and 2000, from 5.4% to 10% of total expenditures. By income, health expenditures are equal to 4% for the poorest households with a monthly income less than 600 DT and to almost 6% for households with monthly incomes ranging between 200 and 2,400 DT. 416 Good Practice in Health Financing, p. 417 Source: INS .:. According to information from INS' household survey (Table 15), the private sector has absorbed a significant portion of household expenditures (87.3%). This can be explained by the subsidization of health care services offered in public and parastatal facilities. Table 15: Household ex enditures b health facili Public Facilities Private Facilities 9,9% 87,3% 2,8% 100,0% Source: Arfa C. and Achour N. (2004) .:. Household health expenditures represent 55% of total "hygiene and health care" expenditures. 79% of these expenditures are devoted to drugs and ambulatory medical consultations. T a ble 16 : H ouse h 0 Id expen dOltures b,y serVIce cat egory (2000) Amount Percentage TND Pharmaceutical commodities 33.3 47% ~bulatory Care 22.7 32% ~ospitalization 9.8 14% Additional consultations 3.7 5% Other 1.4 2% Total 70.9 100% Source: INS (Institut National de la Statistique) 2000 . •:. Figure 4 shows the progress in annual health care expenditures per capita, according to several household surveys. These expenditures, in current dinars, has grown by a factor of 16 between 1975 and 2000, from 4.4 TND to 79 TND. On the other hand, in 1990 constant dinars, the DPA have increased from 38.5 TND in 1995 to 48.2 TND in 2000. 417 Good Practice in Health Financing, p. 418 Graphique n04 : Depense par personne par an en soins medicaux 100~-----------------------------, 7- Health care provision system Health care provision is characterized by: The coexistence of two health care provider sectors, dominated by the public sector which has 87.5% of hospital beds and more than 55% of medical personnel; A predominance of tertiary health care services in the public sector; A concentration of health care facilities, both public and private, in the eastern coastal areas; A rapid growth in private health care provision, especially for hospitalization and for a basic level of medical services: the bed capacity has doubled throughout the past 5 years. Health care provision is assured by public, parastatal and private facilities. Table 17 : Health care provision in Tunisia (for 2004) Facility T~e Public Parastatal Private -CHU (22) - CNSS polyclinics (06) - Clinics -HR (32) - Military hospitals (03) Multidisciplinary (49) -HC (118) - Hospitals of the national security Mono-disciplinary (32) - CSSB (2067) force (01) - Dialysis centers (99) - Private physician practice (4641) Source: MSP * 0: number 7.1- Public Sector Trends The public sector is comprised of facilities which are directly subordinate to the Ministry of Public Health and of parastatal institutions. (i) Provision of care from the Ministry of Public Health is organized at three levels: Primary health care centers (2,067 in 2004 serving 4,500 persons), the entry point in the public sector, form a decentralized network which responds to the population's most immediate needs; Rural natal care centers and district hospitals (118 hospitals with 2,613 beds,); the latter offer medical care, a maternity ward, and a basic level of medical services and technology; Regional hospitals (34 hospitals with 5,479 beds) are located in the region's administrative center, and represent the first referral point for specialized medical servIces; 418 Good Practice in Health Financing, p. 419 University hospitals and their related centers (22 facilities with 8,590 beds) are found in cities which have a medical school (Tunis, Sousse, Monastir, and Sfax). Table 18: Public Health Facilities (MSP) SSP 1970 1980 1989 1992 1996 1998 2000 2002 2004 Nr. beds 8 22 21 20 22 22 22 22 CHU --- 7723 7659 7752 7987 8590 Nr. beds 12 20 24 24 29 31 32 32 34 Regional hospitals ---- 509 5360 5578 5379 5479 Nr. beds 54 54 98 109 111 105 118* 118 118 District hospitals* --- 2664 2793 2640 2647 2613 Primary health care Nr. 435 765 1841 1922 2008 2220 2067 centers Nr. beds 82 144 154 160 158 172 172 174 Total public hospitals 15407 15812 15970 16013 16659 16682 Source: MSP * Including autonomous maternities. (ii) The parastatal sector has the following components: Health facilities that belong to other ministries: 3 military hospitals and a national security forces hospital. Six ambulatory care facilities, or polyclinics, which are located in large cities and are reserved for CNSS affiliates. Nine health care facilities in-house medical services for the personnel of certain companies. 7.2- Private sector trends Medical practice has always been organized around the individual physician's practice, hospitals being a rather recent development. In effect, with the exception of St. Augustin's clinic, which was created in 1930, the majority of hospital institutions have been created since the 1970's, especially in the 1990s. In 2004,81 clinics with 2,379 beds were registered, or 12.5% of the national hospital capacity. The greatest progress has been noted in hemodialysis centers, whose numbers have increased by a factor of 5.5 between 1990 and 2004. Table 19 : Private health facilities in Tunisia Year 1990 2004 Facilities Number Beds Number Beds Clinics 33 1142 81 2379 Multidisciplinary 25 1060 49 2175 Mono-disciQlinary 8 82 32 204 Dialysis centers 18 205 99 986 Source: MSP 7.3- Increase in the number of hospital beds The reorganization of public and private sector roles is evident in Tunisia. The private sector's rapid development is visualized by the sharp rise in number of beds from 1992 on (Table 20). • The public hospitalization capacity has only slightly increased (especially in special services centers and university hospitals), from almost 15,000 beds in 1985 to 15,792 in 1996, and about 17,269 in 2004. 419 Good Practice in Health Financing, p. 420 • In the same period, the capacity of the private sector has doubled, from 974 beds to 2379 beds. • Overall, the two sectors' combined proceedings has resulted in the stabilization of the total average number of beds per 1000 population to a level close to 2 (2004) Table 20: Number of public and private hospital beds (1985-2004 Hospital beds 1985 1989 1992 1996 1997 1998 2002 2004 Public sector 15000 15400 15845 15792 15943 16013 16182 17269 Private sector 974 950 1638 1974 1894 1944 2000 2379 Total 15974 16350 17483 19766 17837 17957 18 182 19648 % Public 94% 94% 91% 89% 89% 89% 89% 88% Source: MSPIDEP and DTH 7.4- The evolution of the health sector's employment structure Along the 1960s, Tunisia's density of medical personnel has declined, especially with the departure of foreign physicians after the country's independence. It has not attained its levels from 1956 unti11970. Since, the ratio has constantly increased: in 1985, there were only 4 physicians per 10,000 popUlation; since, it has almost doubled. From 1981 to 2004, the number of physicians has increased by a factor of4.7, from 2,078 to 9,805 (Table 21). This growth is marked by: • A much faster increase in specialist physicians whose numbers increased by a factor of 6.9 compared to 3.7 for general practitioners. • A must faster growth in the private sector: 674 in 1981,2773 in 1995, and 4641 in 2004. Table 21 : The physician workforce (1981-2004) General Practioners Medical Specialists Total lYear Public !private Total Public !private Total 1981 547 380 927 857 294 1 151 2078 1985 925 670 1 595 947 357 1304 2899 1991 1524 959 2483 1 110 799 1909 4392 1995 1 767 1468 3235 1425 1 305 2730 5965 1998 2022 1 526 3548 1 735 1507 3242 6790 2002 1 983 1620 3603 2164 1667 3831 7434 2003 2552 1620 4667 2115 1677 3297 7469 2004 2737 2635 5372 2427 2006 4433 9805 Source: MSPIDEP Tunisia has also invested greatly in paramedical personnel. The number of nurses and of midwives has increased at a rate similar to that of physicians, especially due to the widespread patterns of recruitment. The paramedical personnel workforce employed in the private sector has increased almost tenfold between 1998 and 2002, from 500 to 5,000. Private clinics are mandated to respect the regulations for having 0.3 or 0.4 nurses per bed, depending on the type of care needed. In reality, one finds about 0.8 paramedical personnel and 0.3 physicians per bed. 420 Good Practice in Health Financing, p. 421 7.5- Human resources development by category The analysis of the development of the professional health workforce for various 5-year periods leads to the following conclusions (Figure 5): Between 1981 and 1986, the highest growth rates were recorded: 8% for paramedical staff, 10% for pharmacists and dentists, and 14% for physicians. The public sector's need for health professionals and the introduction of certification opportunities the labor market both explain this high value of the average growth rate. For the other periods, the growth rate remains below 10%. For physicians, growth rates dropped from 14% (1981-86) to 5% (1986-91). In contrast, given that the competitive admission to nursing schools depends solely on the Ministry of Public Health and that this department subsequently employs the majority of graduates, this category of the health workforce fluctuates in function of employment availabilities and needs in public facilities. For example, between 1991 and 2004, a growth rate of only 4% was recorded, with a decrease to 1% between 1991 and 1996, particularly due to the rise in unemployment observed in that period. Figure 6:Average growth rates for SA'ilA~al of health care pr()tessllonal$~t' Since the second half of the 1990s, Tunisia has increasingly been confronted by unemployment in certain health professional categories, especially for general practitioners, technicians, and sometimes even in some medical specialties. Unemployment is less of a problem for pharmacists and dentists. 421 Good Practice in Health Financing, p. 422 Table 22: Medical Personnel in Tunisia (1990-2004 Health professionals ;zfJlJll .1'U\~ :~1~~.,i;!:, iiA,'~~"?; .~ ~tJlI# Physicians 1800 3450 4424 5965 7444 9805 Dentists 320 525 809 1038 1315 1889 ,Pharmacists 700 1120 1240 1499 1951 2069 Paramedical staff* 13570 20300 23743 25874 27392 29584 Population per 1 3200 2110 1825 1500 1284 1013 physician Population per 1 340 346 340 336 paramedical staff . . .. * TechnICIans, nurses, mIdwIves, and health auxIlIanes. Source: MSP/DEP 8- Other components of the health provision system 8.1- Training of health professionals The medical education system is comprised of: - 4 Medical schools - 1 Pharmacy school - 1 Dentistry school - 4 Colleges of Science and Health Technology, for Medical Technicians (17 sections for mid- wives, physiotherapists, hygienists, laboratory technicians) 19 Professional Schools of Public Health for training of nursing staff. 8.2- Procurement and distribution of pharmaceuticals The pharmaceutical sector is characterized by collaboration between public and private actors: • The sector is regulated and supervised by the MSP and by certain public institutions on which it depends: the National Phannaceutical Control Laboratory (Le Laboratoire National de Controle des Medicaments (LNCM)), the National Phannacovigilance Center (Le Centre National de Pharmacovigilance (CNPV)), the Tunisians Central Medical Store (Pharmacie Centrale de Tunisie (peT», and the Pasteur Institute in Tunis (l'Institut Pasteur de Tunis (IPT»). • During the past fifteen years, the local pharmaceutical industry has seen significant growth. From 1987 to 2001, consumption of locally produced phannaceuticals has grown from 8% to almost 45%. • The total pharmaceuticals consumption in Tunisia is estimated to be about 400 MTND for 2001, which accounts for less than 0.1 % of the global market. The pharmaceutical consumption per capita has risen to 40% between 1995 and 2001, from 30 to almost 42 TND. • The Tunisian Central Medical Store is charged with pharmaceutical distribution and, through a network of private wholesale dealers, it provides coverage for the country's regIOns. • The distribution of pharmaceuticals to patients is ensured by the SSP through a dense network of private pharmacies. • The PCT is the only institutions authorized to import drugs and vaccines, acting as a central purchasing agency which assures the coverage of the entire country's needs. 422 Good Practice in Health Financing, p. 423 8.3 The role of medical technology The evolution of medical technology is illustrated in Table 23. On one hand, MSP regulations establish the list of medical equipment that needs approval prior to their installation. On the other, these regulations gather the data needed to indicate regional need for medical technology. Table 23 : Description of lar !!e-scale medical equipment (1995-2004) Equipment 1995 1997 2003 2004 Public Private Public Private Public Private Public Private MRI(IRM) 1 0 2 0 3 5 4 6 Scanner 7 20 8 29 15 54 20 54 Lithotripteur 3 3 3 5 3 12 3 13 Telecobaltotherapy 1 4 2 6 4 5 4 5 Angiography Scanner nla nla 2 9 6 10 6 10 Cardio-vascular 4 1 7 4 9 10 10 10 Catheterism Cardiopulmonary 4 nla 4 6 6 12 6 12 bypass Source: MSP Health Report 9- Health sector regulation Health policy formulation and health care system regulation are assured by: • The State: government and ministries. • Professional health organizations • The social security funds The regulation tools are put in practice in health care provision characterized by recent efforts to integrate quality assurance. 9.1- Health care provision regulation Health care provision policies address productivity factors, professional practice, and financial remuneration for goods and services. During the past thirty years, they have quickly diversified. The current legislation credits the entire progress of health care provision, public or private to the "health card", renewable at least every five years . •:. Public sector infrastructure Infrastructure development is strongly integrated in the five-year plan and the annual budget. Indeed, "needs" are defined within the five-year development frameworks that decide infrastructure and equipment projects. The development process is highly associated with the regions which express their needs to the Ministry of Public Health and to the Ministry of Economic Development (both in charge of planning), as well as to the Ministry of Finance within the annual budget framework. .:. Private sector infrastructure Private provision of ambulatory health services and hospitalization is not subject to regulations regarding geographic location, but to standard norms regarding buildings, technical installations and equipment. Private dialysis centers are the only private health care 423 Good Practice in Health Financing, p. 424 facilities whose creation was subject to a preliminary authorization based on the regional government's indicators of the population's needs. The distribution of pharmacists is regulated by numerus clausus (limited number or quota) . •:. Equipment and technology The public health organization law of defines large-scale equipment whose installation is subject to preliminary authorization. The list of such equipments is defined by a joint decree of the Ministries of Finance, Trade, and Public Health. The MSP defines the population's need indicators for each type of equipment. .:. Drugs and other pharmaceutical commodities Drugs and pharmaceutical commodities are subject to one of the strictest regulations which covers their procurement (State monopoly), manufacturing, distribution, and commercialization. Drugs distributed in Tunisia are listed under their generic names, both in the public and the private sectors. A strategic sector with social implications, the pharmaceutical sector is strictly limited in regards to the retail price level: pharmaceutical costs are officially recognized and compensated for some, imported by the peT, the sole procurement agent. .:. Regulation of the medical demography Access to medical education, limited by the admission quotas of the 4 national medical schools, has been constant since 1994 at around 700 per year, in spite of a large increase in the number of bachelor degree holders. Access to medical specialization studies is regulated by annual residency competitions for which the number of available positions is limited according to MSPs annual budget. 9.2- Demand regulation Demand is regulated through user fees. In the public sector, copayments have been in effect since the beginning of the 1980s for access to health facilities, and have been regularly increased. Besides their quality of "moderating" health care consumptions, they are also believed to direct patient referral according to various health care needs: copayments are highest in tertiary care facilities and lowest in primary care facilities. All reimbursement schemes, regardless of the insuring institution, have two regulation strategies: user fees and an annual expenditure cap, for which the amounts vary by scheme and by insurance policy. 9.3- Quality assurance and control Since the beginning of the 1990s, the quality improvement of health care has become an important component in the management of the various health facilities. The actions taken to improve the quality of health care include: The development of a national strategy for continuous quality improvement, currently in its early implementation stages. Awareness-raising and encouragement of practical education of quality management A system of quality assurance for pharmaceuticals Quality control of laboratory tests The execution and acknowledgement of patient satisfaction survey results. 9.4- Other types of regulation 424 Good Practice in Health Financing, p. 425 A « best practice» guide. Contracting, initially of intensive care services for socially insured, and subsequently expanded to routine care in regional and university hospitals under the auspice of the MSP. Medical inspections and controls. 9.5- Current regulation trends The latest trends in the health system's regulation, which bring together all stakeholders, are reflected in the framework of health insurance reform implementation. In particular, they address contracting and the promotion of medical control procedures. Health insurance reform is based on a clear distinction between the basic mandatory scheme, administered by CNAM, and the supplementary schemes, open to private health insurance. The latter must only insure those health care services truly supplementary to those of the basic scheme in order to avoid the duplication in coverage. Contracting between the CNAM and providers should be the only mechanism to govern the relationship between insurers and providers. These agreements allow the definition of specific regulations for medical practice and the setting of health care fees. Medical control procedures monitor the quality of the health care services delivered to the insured based on national and international standards. 10- Health information system The information system is one ofthe Tunisian health system's weaknesses, in spite of the various attempts for its establishment and its implementation. In its present state, the information system is composed of sub-systems of management for the various components of the health care system: Data from the oldest component, the system of statistical information collection for health facility activities, are collected by the central administration's statistics division. The national office of family and population has instated an efficient information system, regularly updated according to family planning, birth rate regulation and reproductive health mission and activities. During the 1980s, the accent has been placed on basic health care services. Collection methods and data management guidelines have been put into place in various domains. From the 1990s to today, the hospital system has been the principal interest, sustained by two World Bank projects, of which one component has been the implementation of a computerized information management system used by the Center of Computer Science of the Ministry of Public Health (Centre d'Infonnatique du Ministere de la Sante Publique (CIMSP)). To this day, this system only offers classic management tools. Moreover, it does not integrate patients' health care information or data regarding human resources management. The information system's current inefficiencies, attributed to its fragmentation in the absence of concrete integration and to novel challenges of the hospital sector, have driven public powers to request a comprehensive study. The study's objectives are to conceive new basis for development and an implementation strategy for an information system centered on the patient. This study has been realized and has been submitted to the government within the framework of investment proposals for the next development plan (2007-2011). 11- Expenditures, services, and outcomes In the absence of quantitative information linking expenditures to services and to outcomes, worsened by a decline in expenditures and revenues, we can only provide general information based on comparisons with other countries. 425 Good Practice in Health Financing, p. 426 .:. Tunisia is a middle-income country at an average development level. The level of health care expenditures (5.6%) is moderate and comparable with other Middle East and North African countries (MENA). Although the annual per capita expenditures have been less than $150 per capita, Tunisia has achieved the greatest successes, especially in terms of health: • Among the most pertinent health-related accomplishments is the significant decrease in the infant mortality rate. • Since the end of the 1960s, the Tunisian population's life expectancy at birth has increased by 21.8 years, from 51.5 in 1966 to 73.3 in 2004. During this period, women have gained 23.7 years, while men gained 20.8. (Figure 1) During these 44 years, Tunisia's life expectancy at birth has always been noticeably higher than that of other countries in the MENA region (BM 2006), a trend which continues to this day, brining Tunisia's health status closer to that ofOECD countries. Table 24: Life expectancy and infant mortality (1960-2004), Tunisia and the MENA regIOn 1960 1980 2000 2004 Indicators MENA Tunisia MENA Tunisia MENA Tunisia MENA Tunisia Life expectancy 47 48,6 58 62,4 68 72,1 68 73,3 at birth Infant mortality per 1000 live 262 160 138 60 47 25 45 20,6 births ... Sources: WDI (World Development IndIcators) 2004; INS 2005; UNPD (Umted NatIOns PopulatIOn DIVIsIOn) 2005 .:. The analysis of health care coverage, displayed in Table 24, show that Tunisia has achieved significant progress, which allows it to hold a reputable position among other middle-income countries and other MENA region countries . •:. The indicators for health care coverage in terms of primary health care services have consistently improved (Table 25), as indicated by the percentage of the population (95%) who has access to them, to pre-natal care (92%), and to skilled personnel for birth and child care, 90% and 96%, respectively. Similarly, immunization policy has led to more than 98% of Tunisian children under-one to be vaccinated, particularly for diphtheria, tetanus, pertussis, poliomyelitis, measles, and for BCG. Vaccine coverage rates reach similar levels to OECD countries. Table 25: CoveraKe withprimary health care services indicators (2005) Population with access to local health services, total (%) 95 Contraceptive. prevalence rate (%) 62 Antenatal care coverage (%) 92* Births attended by skilled health personnel 90* Infants attended by trained personnel (%) 96** One year-olds immunized in 2005 with BCGJ%l 98 Oneyear-olds immunized in 2005 with DPT (%) 98 One year-olds immunized in 2005 with OPV (%) 98 One year-olds immunized in 2005 with measles vaccine (%) 96 One year-olds immunized in 2005 with Hepatitis B vaccine (%) 97 Pregnant women immunized with two or more doses of tetanus toxoid (%) 88 * Year 2004/ ** Year 2000 Source: WHO: http://www.emro.who.intlemrinfo/index.asp?Ctry=tun 426 Good Practice in Health Financing. p. 427 Table 26: Health outcomes for Tunisia and similar countries Upper Lower Middle East Middle middle middle & North Tunisia Indicators income income income Africa 19HfaM 199811084 I_HIM 1998 1884 1990 2804 Reduce child mortality Immunization, measles (% of children ages 12-23 months) 79.7 91.2 87 85.4 85.9 86.5 83.3 92 93 95, Mortality rate, infant (per 1,000 live births) 33 23 46 33 44 31 60 44 41 211 Mortality rate, under-5 (per 1,000) 41 28 62 42 58 39 81 55 52 251 Improve maternal health Births attended by skilled health staff (% of total) .. 94.7 .. 85.7 .. 87.3 45.6 72.3 83 (1998) 92* Maternal mortality ratio (modeled estimate, per 100,000 live births) .. .. .. .. .. .. .. .. 69 (1993) 48* Combat HIV/AIDS, malaria, and other diseases Incidence of tuberculosis (per 100,000 people) 69 113.6 134 114.9 120.8 114.7 66.2 53.9 33.1 22 Prevalence of HI V, total (% of population ages 15-49) .. 2 .. 0 .. J .. 0 .. 0 Tuberculosis cases detected under DOTS (%) .. 72.3 .. 68.4 .. 69.5 .. 50.2 .. 95.5 Life Expectancy, Fertility and Literacy rate Fertility rate, total (births per woman) 2.6 1.9 2.7 2.2 2.7 2.1 4.8 3 3.5 2 Life expectancy at birth, total (years) 69.2 69.2 67.3 70.2 67.6 70 64.3 69.3 70.3 73.3 Literacy rate, adult total (% of people ages 15 and above) 91.5 93.6 78.1 89 80.7 89.8 51.8 72.1 59.1 74.3 Source: WBl (World Bank Indicators) 2006 (www.ddp-ext.worldbank.org); * MSP (Ministry of Public Health) 2005 427 Good Practice in Health Financing, p. 428 12- Equity and access to health care .:. Due to the current health care coverage schemes, access to health care is within the reach of the entire population regardless of the household' financial capacity. These schemes are for: - the social protection with or without supplementary coverage, in place since the end of the 1950s and the beginning of the 1960s; - the medical assistance of the population that is poor or that has limited revenues, and is not covered by the mandatory health insurance schemes. However, in spite of near universal coverage of the Tunisian population by one of the schemes, household expenditures for health care remain very high. This situation is the source of inequity and one of the principal drivers for health insurance reform. Indeed, Tunisia is confronted by two categories of problems that explain this high proportion of household expenditures: The under-financing in the public sector, which prevents it from providing the goods and services needed to all its users. The low reimbursement rates and the CNRPS' out-dated reimbursement schemes. The World Health Report (WHO 2000) illustrates the inequity by the financial contribution indicator, similar to that of other middle-income countries of the MENA region (Table 27). Table 27 : Inequity of financial contribution indicators, Tunisia and similar co untries Inequity of financial Country Rank contribution indicators Jordan 49 0,958 Iraq 56 0,952 Tunisia 108 0,925 Iran 112 0,923 Egypt 125 0,915 Morocco 125 0,915 Syrian Arab Republic 142 0,904 Source: WHO World Health Report (2000) .:. The expansion of public provision has improved access to health care, notably through the development of the primary health sector, based on a dense network of primary health centers throughout the entire country connected to a network of district hospitals which provide primary health, pediatric and obstetric services. A large majority of the population (90%) can access a primary health care center by less than an hour of walking . •:. Along the same line, access to specialists' services has gradually improved: the public hospital network, which offers various specialist services, is comprised of regional and university hospitals. The government, through various financial incentives encourages access to specialist doctors, especially in underserved areas. Meanwhile, significant disparities remain between urban and rural areas, in terms of availability of health services: hospital provision and specialized physicians continue to be concentrated in large urban centers (Tunis - the capital, Sousse, Monastir - central Tunisia, and Sfax - southern Tunisia) .:. The inequality in the improvement of health indicators coincides with geographic disparities in terms of socio-economic development. Additionally, it affects vulnerable groups, such as women, children, adolescents, and the elderly. 428 Good Practice in Health Financing, p. 429 The currently available statistical information reveals: Infant mortality rates that are much higher in rural than in urban areas, sometimes even double. Maternal and neonatal mortality, maternal and infant anemia, diarrhea and respiratory infections in those under-5 consistently show higher rates in rural or periurban underserved areas, as well as in the western and southern regions. The health indicators published by WHO (2006), although reporting 1998 data for Tunisia (Table 28), confirm significant differences according to residency area (rural or urban) as well as to the mother's education level. 429 Good Practice in Health Financing, p. 430 Table 28: Health inequalities in Tunisia and in similar countries Proportion of under-S children with delayed growth Under-S mortality rate (per 1000 live births) (%) Country Year Rural (1) Urban Ratio (2) I ILowest IHighest (1) / (2) (1) (2) Ratio (1) /(2) I Rural (1) Urban (2) IRatio ILowest IHighest (1) / (2) (1) (2) Ratio (1) /(2) Source: World Health Statistics WHO 2006. Measles immunization coverage among 1-year-olds Births attended by skilled health personnel (%) Country Source: World Health Statistics WHO 2006. 430 Good Practice in Health Financing, p. 431 13- Evaluation of the health system's strengths and weaknesses 13.1- Health system strengths .:. The main strength of the Tunisian health system lies in its accessibility which acts on there levels: The implementation and responsiveness of the health insurance system, covering almost the entire Tunisian population since the late 1950s and the early 1960s. The creation and expansion of a modern public health infrastructure that is regularly maintained and updated. The national education of various categories of health care professionals, since the early 1960's . •:. The introduction of a primary health care strategy, based on WHO recommendations, responding the population's basic health care needs, including family planning and maternal and child health . •:. The coexistence of a public and private health provision system, which largely contributed to the population's overall health status improvement. 13.2- Health system weaknesses The growing needs of the population and of its socioeconomic status, medical technology innovations, and the scarcity of financial resources allocated to this sector, are all proof of certain weaknesses in the health care system: .:. In terms of health care provision: regional disparities between the east and the west, in terms of infrastructure, human resources, and large-scale medical equipment. .:. In terms of health care financing, we identify the following insufficiencies: The under-financing of the public sector, particularly of primary health care. Under the current financing of health care consumption, direct payments represent a large proportion of household expenditures . •:. In terms of governance, the system suffers from the following: The absence of an organized evaluation plan, especially for the quality of the provision of health care services. The weaknesses of the information systems as well as its fragmentation represent the major barrier to any evaluation attempts. The weakness of referral as a part of medical practices, of health care and of management. .:. In terms of continuous medical education, the current efforts remain insufficient in terms of their general organization and of the means allocated to it. 13.3- The health system during the reform At the moment of the design of the health insurance reform, before the start of implementation efforts, the Tunisian health care system can be described as follows: .:. In terms of health care provision Health care provision assured predominantly in a dense and decentralized public sector and in a private sector of which the ambulatory component was gradually decentralizing and the hospital one rapidly increasing. A concentration of health care providers in the large urban centers, needing sustained efforts of re-equilibration. Health care services which remained primarily under the direction of the public sector. •:. In terms of financing Two financing methods coexisted: the Bismarck mandatory health insurance and the Beveridge system targeting the poor and those with limited incomes. 431 Good Practice in Health Financing, p. 432 Historically, the latter has been the predominant model that has been gradually replaced by the former, especially since 1995. Financing from public funds is almost completely reserved to the public health care provision sector while provision in the private sector are mostly financed directly from households. The heavy burden of direct financing from households for health expenditures (51.3%) resulted in inequity of health care access and, to a certain extent, the impoverishment of certain vulnerable populations. From a grant-based financing of health care provision, the system moved gradually towards fee for service: more and more, "the money follows the ill patient." A new sharing of health care related financial expenditures is being designed to reduce those of the government and of households in order to replace them by a contribution towards a mandatory social health insurance scheme. The CNAM increased competition among health care providers and health care services, based on quality and price. More and more, it will position itself like an "informed consumer" of health care providers for its members . •:. In terms of resource mobilization Tunisia has developed important capacity in terms of training health care professionals of various categories. Nevertheless, the health sector's evolution needs the integration of new occupations, especially in management, engineering, and computer science. Newly certified health professionals undergo a longer than usual unemployment period. Certain paramedical personnel and general practitioners are particularly affected. In spite of a rather high medical density (1 physician per 1000 population), disparities continue to persist. Specialist physicians are scarce in numbers, especially in Tunisia's western regions. The development of medical technology has encouraged public powers to introduce a "health card for large-scale medical equipment" as a regulatory mechanism for investment in this type of equipment. At the same time, the rapid development of the private hospital sector has promoted a significant growth in capital investment, without any evaluation of the utilization of such large-scale medical equipment. Due to current regulations, drug management is satisfactory in the entire sector. Meanwhile, public health facilities, strained by the above-mentioned under- financing, are not able to fill all the prescriptions received . •:. In terms of general administration and governance Tunisia has introduced a strong legal and regulatory arsenal for the administration of the health sector, especially for its policy component. Meanwhile, the regulation of health care provision in the private sector remains weak because of the near- absence of an operational arm, such as public financing. Increasingly, private sector health professionals tend to promote strategic attitudes, allowing them to better manage social negotiations. The public sector continues to be characterized by an extreme centralization of decision-making, both in terms of investment and of investment of human resources. In spite of current regulations promoting autonomy of the hospital sector, strategic and operational decisions regarding hospitals continue to be made externally . •:. Stakeholders and powers involved Apart from the various ministerial departments involved, in particular those related to health, to social affairs (under the auspice of the CSS and the health insurance), and to 432 Good Practice in Health Financing, p. 433 financing, there are three main actors which playa role in the health insurance reform process, each with diverging interests and often conflicting goals. • The central labor union: the General Union of Tunisian Laborers (l'Union Generale des Travailleurs Tunisiens (UGTT)) proposes the following: The ''upgrade'' of the public sector in order to meet the same standards as the private sector, rather than competition between the two. Behind this idea, we find fear of health services privatization and the subsequent risks regarding job security in the public sector. Resistance against an increase in the employees' contributions to the health insurance schemes, in fear of subsequent unrestrained increases, and the call for a direct government contribution to mandatory health insurance schemes. The expansion of the benefits packages under the new health insurance system, without increasing contribution rates or copayments. The expansion of third-party payers both for payment authorization for the ill and for promoting a reduction in reimbursement systems. • Professional organizations of private providers: comprised of various orders and unions of private sector professionals, they develop the following principal themes: Preference for the reimbursement of charges through third-party payers. Resistance against « managed care» as an organization based on the family doctor, prioritizing the notion of patient's freedom of choice. Resistance against the movements regarding the control of the medical practice and prescriptions, unveiling the notion of freedom for the prescription of drugs. Militating for the reduction of the insured's health care management procedures. • Work-based organizations: the Tunisian Union for Industry, Trade, and Crafts (l'Union Tunisienne de l'Industrie, du Commerce et de l'Artisanat (UTICCA)), and the Tunisian Union for Agriculture and Fisheries (l'Union Tunisienne de I' Agriculture et de la Peche (UTAP)), which promote the following: The influence of enterprises' social responsibility in matters of national and international competition and employment. Job insecurity in the agricultural sectors, resulting in a weak contributory capacity for laborers and enterprises. 13.4- Reasons behind key health sector reforms The actions taken at the core of the Tunisian health care system can be categorized into three overlapping stages based on the shifting needs of the sector, of its national and international environment, and of the resulting constraints . •:. An extensive growth in health care provision, based on the development of the public health care provision network, financed primarily by the government, which also ensured the training and the recruitment of the large majority of health professionals. The principal reason for this stages' implementation was the burden of infectious diseases, of malnutrition, of maternal mortality and morbidity . •:. A stage of intensive growth of the public sector, in search of greater efficiency through better allocation of available resources and through higher quality of service provision. The public powers have implemented refonns in management of the SSP in order to improved their management perfonnance, especially since the government could not continue to sustain all of the sector's needs as it had previously done in the past. 433 Good Practice in Health Financing, p. 434 .:. A stage, which continues to this day, of exploration with pUblic-private partnerships in tenns of provision and financing, in order to provide for the population's growing needs under the double burden of the demographic and the epidemiologic transition. In the context of the epidemiologic transition, the public sector limited in its ability to respond to the population's needs, the private sector's rapid development, and the high burden of health expenditures shouldered by households, public powers have been driven to reexamine the partitioning of health expenditures and to offer to those insured the choice between public and private providers. Certain underlying societal values, especially equity, have been translated into the promotion for access to health care during the different stages mentioned above. The quality of health care have been valued differently by various categories of the population, of which the better- off appeal more and more to the private sector, which is much more responsive to the services they offer. Economic efficiency remains a great constraint of health care system management and not a concept widely adopted by health professionals. 14 - Timeline and description of the principal measures taken for the reform The system refonns (from 1980 to 2004) have involved the provision and financing of health care services . •:. In regards to health care provision, we identify three major periods: From 1980 to 1990: the development of primary health care, in continuation of the extensive growth stage described above. From 1985 to 1999: the university hospital sector marked by the adoption of new management methods beginning with the creation of the "public health establishment." During this period, a legal arsenal was introduced to incorporate the private hospital sector, which was until then regulated by provisions from the end of the 19 th century. A series of regulations were promoted for the governance of private paramedical practice. From 1990 to 2004: regional hospitals have benefited, within the framework ofa sector-wide project, from a revision of their organization and management, associated to the renewal and the modernization of their medical equipment. .:. In tenns of financing, various paths have been followed: The care of the poor and of those with limited income has been the goal of various revisions, regarding better targeting of eligible populations. The introduction, since 1983, of copayments and their integration into the resources of health care institutions. Health care financing mechanisms in public hospitals through social security have undergone several developments: o The government's contribution to the treasury for the financing of the SSP has been the primary means introduced which remains in place even to this day. o Agreements for intensive care services have been reached with some SSP since the end of the 1980s, and have been expanded since to private facilities in a movement to reduce the number of foreign health care services to which the popUlation resorted to and to build national capacity. o Since 1996, a billing system has been gradually introduced and expanded. In addition to some adjustments in the classification of medical service and medical tariffs and fees, the private sector has not been a target of specific measures regarding financing of its service provision . •:. Measures regarding the general organization of the sector have been undertaken since the 1990s. The principal ones are based on: 434 Good Practice in Health Financing, p. 435 The national health information system (Le systeme national d'information sanitaire (SNIS)). o The adopting of a national strategy for the SNIS and its development from existing sub-systems o The identification of SNIS development strategies regarding unavailable information: surveys, registries, death certificates '" Organization measures o The revision of the central organization and administration o The revision of the mission, objectives and organization of regional administration o The revision of public hospital management Quality assurance of service provision has been the goal of the national strategy that is currently being implemented. 15- Evaluation of the reforms The various reform strategies have contributed to the extension of health care coverage and to the improvement of health care access. However, under the impact of the socio-economic, demographic, and epidemiologic transitions, these strategies and actions have generated three categories of problems: (i) inefficiencies which have led to management reforms, centered on public hospitals, (ii) the need to mobilize additional financial resources for these institutions, and (iii) the need for a revision of financial cost-sharing between various institutional sources of financing. Moreover, these reforms have not been developed in a global manner, generating fragmentation in the system. It is difficult to evaluate the impact of such a reform on the sector's performance. Additionally, the weaknesses of the information system pose a major obstacle, both for the reform's implementation and its evaluation. In spite of these limitations, certain evaluation elements can be proposed: .:. The Tunisian health care system ensures a reasonable coverage of its population's health care needs, but its eastern and western regions remain underserved, especially in regards to medical specialists. Re-balancing attempts, through to the introduction of financial incentives, have not resulted in the desired outcomes . •:. The private health care provision sector has rapidly evolved due to a generally favorable environment and to some additional support. It currently remains a tributary of private financing by households. The health insurance reform allows the reduction of this limitation. •:. The various health care financing, undertaken in the public sector, have only minimally improved the financial situation of public hospitals, which remain markedly under- financed. Nonetheless, they allowed the introduction of new mechanisms, permitting resource mobilization from the social security funds that gradually replace those allocated by the government. .:. The reform strategies for management in public institutions have successfully resulted in a modernization of management procedures. However, they have not reached the desired objectives, especially in terms of the hospitals' greater autonomy, due to several delays for the revision of the various management levels in the health sector. •:. The general regulation of the sector, especially for its private component, remains weak, in spite of the numerous regulatory texts governing it. .:. Finally, the sector has a strong need to improve its capacity for mid and long-term strategic planning in order to actively anticipate future trends. In this regard, two domains should be given priority: the health sectors response to the epidemiologic transition and to 435 Good Practice in Health Financing, p. 436 the sustainability of health care financing. Indeed, the health sector is once again confronted by new challenges, such as: A transformation of the population pyramid, marked by an increase in the proportion of the elderly, thereby inducing a demand for expensive, tertiary care; The promotion of financial accessibility to health care through the mobilization of public financial resources through a more equitable cost-sharing mechanism between the government and social security. A global tendency towards to liberalization of all of the economic sectors' activities, including the service sector, with the anticipated impact on employment, on the role of the public sector and its resources, and on access to health care. 16- Health insurance reform in Tunisia In the following section, we look at the different aspects of Tunisia's health insurance reforms (its purpose, objectives, the introduction process, lessons learned and the associated obstacles) 16.1- The purpose of the health insurance reform While the health insurance system has allowed to ensure a reasonable coverage and to maintain access to health care, there remain certain insufficiencies: .:. The rapid increase in health care expenditures between 1985 and 2004 and the increasing proportion of these expenditures shouldered by households . •:. The fragmentation of the health insurance system has generated several inequalities in terms of coverage and contribution rates . •:. The low level of reimbursement rates for health care provision charges within the CNRPS schemes. In addition to these intrinsic insufficiencies, we would like to mention the influence of private sector physicians. Given that their revenue depends on the solvency of their clients, they lobby in favor of the extension of health insurance coverage to the private sector. 16.2- Goals and objectives of the reform The decision to reform the health insurance of social security schemes was announced on May 1S\ 1995. It has been analyzed in a preliminary study, which was presented to the government on February 16 th 1996. The principles of this large-scale reform has been announced: The implementation of a basic scheme, mandatory and unifies, assuring the necessary health care coverage and managed solely by the social security funds. The introduction of a complementary insurance scheme, insuring health care services complementary to those of the basic scheme, managed by insurance companies and mutual health insurance organizations (in the future, by CSS). The organization of relations between financial institutions and public and private health care providers, based on predetermined agreements. The control of health expenditures so that the reform does not impose too heavy a burden on public finances, while maintaining the competition among enterprises in Tunisia's open economy. The decision to combine the management of social security's health insurance regime within a single agency - the National Health Insurance Fund (CNAM) - was taken later and was implemented in July 2004 by law nr. 2004-71 of August 2nd, 2004. This law mandated the creation of a new health insurance system. 436 Good Practice in Health Financing, p. 437 Negotiations with various partners have been since in the working in order to reach agreements, implementation decrees, and the approval of a framework agreement to regulate the interaction between CNAM and private providers. 16.3- Evaluation of the reform The reforms remains in an abstract, theoretical phase and cannot be evaluated based on its strengths and potential risks (see Table 25) given the current situation. The reform should, however, noticeably improve the existing situation, beginning with the safeguarding of the health systems' current accomplishments. 16.4- Scalability of the reform A conclusive decision has not yet been taken in regards to planning the implementation of the health insurance reform, however, the following have been proposed: • The implementation of a reimbursement system for common illnesses for those interested, with reimbursement caps. • The continuation, for those who are not enrolled in a reimbursement scheme, of care in a public health facility. • The introduction of a process to transfer long-illness costs to third-party payers • The continuation of coverage for hospital care expenses as it is currently handled and its gradual expansion to the private sector for certain illnesses or for medical care needing the specialization that cannot be supported by the public sector These propositions are, of course, minor compared to overall priorities in terms of payment methods, health care procedures and general tariffs. 16.5- Key conditions for the reform's success • The underestimation of the actors' strategic capabilities results in a lengthy reform process, drifting away from the intended objectives. • The proper implementation of a mandatory social health insurance system depends on the contribution rate. Studies on this subject have been strongly recommended, however, in practice, only one has been carried out, permitting the estimation of the contribution rate of the new mandatory scheme proposed by the reform . •:. Capacity building in management of CNAM and other agencies involved . •:. The completeness of the legal arsenal necessary for the implementation of the reform, and in particular: The contents of the benefits package The reimbursement rate and annual cap Sector-wide conventions The stakeholders' frame of reference, standards, and norms, in terms of quality assurance of services and cost-control .:. The development of a clear communication strategy known by all partners, to avoid the unnecessary misunderstandings and disputes. Despite the above-mentioned factors, we continue to believe that a principal condition for the success of the health insurance reform is the promotion of the public health care provision sector, as well as the perceptible increase of well-managed resources. This should serve as the reference model for the health system, in order to it to play its full role in service provision, training, and medical research. The economic and socio-cultural environment could bring about many unforeseen threats to the sustainability of a health care system with such a significant private component. 437 Good Practice in Health Financing, p. 438 Table 29: Stakeholder analysis for the health insurance reform Stakeholder Actions BeneJiJs Risks Challenges - Access to higher quality health - Potential for subsequent - Does the ditribution of health care services increases in contribution rates. professionals offer a true - Access to public and private - The increase in copayments. choice between health Increased providers? providers - Additional non-regulated The socially contribution rates for - Reduction of the uncovered payments - Undeclaration of employees insured health insurance medical expenditures and their revenues. (social security) - Impact on vested rights and - Better monitoring of personal benefits health status (individual health files~ - Reduction in complementary Increased insurance fees. - Vested rights Competition in the national and Employer contribution rates for - Better access to health care for - Increased social costs international market health insurance emQLoYet:s (p~odu~tivi!y) - Limitation of the choice of - Many new administrative physician? procedures - fees: amounts and revision - Continuing medical education Private providers Greater access to Predicted demand and solvency of methods - Health care networks and of goods and public financing the insured - delay in payment ofCNAM fees physician referrals services - monitoriong and evaluation - Litigation management with systems theCNAM - transparency of revenues - A reduction in resources Public providers A reduction in the number of - Degradation of quality Under-financing of the public Competition with the of goods and patients which lightens the sector and mis-matching of its private sector - Migration of competent services workload capacity and reach professionals towards the private sector 438 Good Practice in Health Financing, p. 439 Stakeholder Actions Benefits Risks Challenf,!es - Employment impact of the sector's health professionals - Coordination between CNSS and CNRPS for the collection - Increase in administration costs of contribution, the CNAM Management of a single being responsible of its own Social Security Increase in resources mandatory scheme by a single - Financial equilibrium resource management agency - The introduction of various versions of the basic scheme - A Financial and Medical Information System - Socio-political consequences - Optional enrollment in Group insurance The drastic reduction in the supplementary insurance and mutual health Coverage limitation turnover of « illness» has schemes insurance (only supplementary) significant appeal - Potential competitions with companies CNAM - A dichotomized health system: - Continue to assure - A reduction in health public for the poor and private health care to the poor expenditures The Government for those who can afford it - Governance capacity. and to low- income - Better information about the groups - Training capacity of the public private sector sector 439 Good Practice in Health Financing, p. 440 16.6- Lessons drawn from the reform process • To lead the analysis within a multi-disciplinary and autonomous framework, having a deep understanding of the national health system, yet avoiding that the initiation of the reform is driven by only one of the parties involved and involving health system actors from various categories; if necessary, appealing to international expertise. • To anticipate the stakeholders' strategies and any potential resistance to change. • To profoundly analyze all potential consequences of the financing reform on the other components of the system. • To appeal to political experts who must fully play their role as decision-makers regardless of party divisions and of multiple alternatives. • To estimate the missing information through timely studies in order to allow for informed decision-making. • To broaden the knowledge of international experiences in health care reform, • In order to avoid mistakes and to mitigate negative repercussions of any increased in the contribution rate, it is useful to aim to reach the ideal rate gradually. In Tunisia, the ideal contribution rate is 6.75% which will be reached during a three-year period. This could be prolonged to a five-year period for low-income populations (Table 30): Table 30 : Review of contribution rates for the health insurance scheme Current Rates Target Rates Variance Employer I Employee Employer Employee Employer Employee CNRPS - Mandatory 1 1 4 2.75 +3 +1.75 - Mandatory & 2.5 4 4 2.75 +1.5 -1.25 op_tional CNSS -RSNA 3.43 1.32 3.98 2.77 +0.55 +1.45 -RSA 0.68 0.23 3.98 2.77 + 3.3 +2.45 -RSAA 1.52 0.76 4 2.75 +2.48 + 1.99 RNS 3.04 6.75 + 3.71 Pensioners CNRPS Mandatory 1 3 +2 CNRPS Mandatory & 3 3 +0 optional 0 3 +3 CNSS Source: Khaled M. K. 2001 16.7- Implementation challenges Three key questions continue to delay the implementation of the health insurance reform. These continue to be the subject of stakeholder negotiations. • Payment methods Fee for service has been retained as the payment method for ambulatory care. This called for a revision of the general classification for medical services (realized in 2006) and for the fixation of standards fees for each of them (currently in negotiation). 440 Good Practice in Health Financing. p. 441 For the coverage of fees, two options are foreseen, which continue to be discussed among various stakeholders: (i) CNAM reimbursement of fees paid by the insured, or (ii) third- party payers, the insured therefore only contributing the copayment amount. For hospital care, the current system promotes a single inclusive fee, dependant on the public sectors costs and, a previously determined fixed fee for pathologic and specialized services. The CNAM will pay health facilities directly for these inclusive tariffs. This method assumes the availability of calculations for various combinations of price and quantity and the preliminary identification of the medical fees included therein. • Health care networks In line with the objective of cost-control, the formation of a health care network allows supervision and guidance of a patient's care-seeking behavior. Negotiations continue on this subject, especially in regards to the role of general practitioners as gatekeepers to specialized services. Meanwhile, medical organizations, especially medical specialists, do not favor this proposed system, as they fear the dichotomization of the medical profession and the over-consumption of health services. • Other questions remain unanswered. In this context, we mention the most significant ones: Referrals as a method of medical control; The management of drug procurement according to the financial levels adopted in their classifications (vital, essential, non-essential); The substitution, by the pharmacist, of drugs prescribed by the physician; The dental care benefits package Despite the capacity of defining the desired direction for reform and of integrating it within the social protection strategy, the architects of the reform realized that it will need many negotiations and political discussions to reach the desired consensus for it to be coherent on various medical, socio-economic, and political aspects. This situation will require the management of ideological differences, contradictory approaches of the preservation of the systems gains, and conflicting stakeholder interests. 441 Good Practice in Health Financing, p. 442 Conclusions The first part of this document traces Tunisia's economic and social performance since the end of the 1980s. For a long time, Tunisia has adopted an approach based on the indivisibility of the economic and social dimensions of development. Numerous reforms and programs that aim to improve the performance of Tunisian companies have accompanied the country's road towards liberalization and integration in the global economy. They do so by encouraging foreign investment, and diversifying the Tunisian economy, with an increasing role in social sectors. The social policy implemented during the last decade has efficiently contributed to the general improvement of the quality of life and to poverty reduction. The second part of this document, dedicated to the analysis of the Tunisian health system and health insurance, allows us to confirm the system's overall positive performance and the expansion of health care coverage . •:. During the past three decades, Tunisia has introduced a health care system which covers almost its entire population and which compares favorably with that of other middle-income countries. This health care system has developed in a transitional context, where the demographic, socio-economic, and health status of the population has profoundly changed. Therefore, this system is confronted by multiple challenges arisen from this triple transition, demanding the expansion of health care coverage that was accomplished through an increase in the number the health workforce, in health facilities, in modem medical equipment, and in health care coverage. Consequently, health care expenditures have surpassed 5% since the second half of the last decade . •:. The health status of the Tunisian population continues to improve, with the development of preventive and curative health care services and the decrease in birth rate. Tunisia demonstrates better health outcomes than other average middle-income countries, in particular in terms of life expectancy, infant mortality, malnutrition, and reproductive health . •:. The public sector is the main health care service provider, supplying about 85% of total hospital beds and more than 55% of medical personnel. .:. The private sector has developed rapidly: its bed capacity has doubled in the past ten years and the number of medical examinations neighbors that of the public sector. •:. In terms of health financing, the extension of the health care coverage has been accompanied, during the last sixteen years, by a growth in household contribution, a decreased in government contribution, and a slight increase in the social health insurance contribution. In addition, the description of Tunisia's health policy and of the reforms adopted at the end of the 1980s presents the country's inclination for the evolution of its health care coverage . •:. Until the 1980s, the health care system has evolved based on the colonial tradition of a hospital-centered health infrastructure, concentrated in large urban areas. Public powers have waited for an opportunity to improve health care provision and coverage, by prioritizing: 442 Good Practice in Health Financing, p. 443 Preventive programs (individual and collective) in the fight against illness, financed completely by the government. This allowed the eradication of certain infectious diseases and the significant reduction in the incidence of others. The development of geographic coverage of the population through primary health facilities, financially accessible for the entire population. The local training of physicians and paramedical personnel, combined with a policy friendly towards the foreign health workforce in order to meet the needs of the population. The gradual development of a hospital infrastructure with the capacity for specialized care. The early introduction of a mandatory health insurance for a rather large section of the population (civil servants and formal sector employees), which was gradually extended to additional groups. The existence of the private. health care provider sector, which efficiently assisted in the public sector's efforts . •:. Since 1990, the public powers have entered a new era of the health sector's development, based on a strategy which includes: A continuation of the primary health development through a provision consolidation program. An improvement in hospital care through a reform regarding structural and institutional aspects of university hospitals. A legislative reform regarding private health care provision. In terms of pUblic-private partnerships, we note two principal periods: .:. One period when the provision of health care was almost entirely public, the private ambulatory sector coexisted and provided services to only a limited section of the population. This situation prevailed until the end of the 1980s . •:. Since the end of the 1980s, there has been a rapid development of the private health care provision sector, in terms of number of providers and of health care categories. Private hospitalization, especially in surgery, has strongly developed, making available another form of hospital provision. One of the current challenges of the health insurance reform is the development of efficient public-private partnerships. The extensive growth of the private sector, associated with the quantitative and qualitative change in health care demand, has contributed to a significant growth in household health expenditures, increasingly depended on as a direct financing method for the health care system. Aware of the concerns generated by this situation, public powers have reformed the health insurance schemes in order to reduce the burden carried by households. In the current context of the health systems evolutions and of the reforms which it undergoes, regulation is one of the fundamental pillars of good governance, as it faces serious challenges and constraints. On the one hand, these arise from the system's progress following a decade of reforms. On the other, they come about from the evolution of its macro-economic and social environments. 443 Good Practice in Health Financing, p. 444 The following points summarize the factors playing a key role in the successful expansion of health care coverage and the lessons that one can learn from the Tunisian expenence . •:. The fundamental policies for human resource development across a modem social policy: "Tunisia's main natural resource are the Tunisians." .:. A public sector, government-regulated, planned in detail according to community health priorities and compliant with the standardization of its geographic expansion as well as with its missions, mandated by legal texts . •:. A near-universal "health insurance:" all Tunisians can access health care services in the public sector, through social security (Bismarkian) or government-managed schemes (Beveridgian). This coverage has been regularly maintained and updated in order to ensure its adaptation to the socio-economic conditions of the population it serves . •:. The regulated training of human resources in the health sector, in place since the 1960s, first concentrated in the capital. Since this has been decentralized, especially for the training of nursing personnel. .:. A strong commitment towards primary health care policy, associated to the gradual development of more intensive and more costly hospital care, influenced by the regulated introduction of new medical technologies . •:. The government's complete control of the drug policy: procurement, distribution, price, and, more recently, manufacturing and quality control. .:. The development of information systems represents a growing need for the health sector, for which health care providers and financing institutions, such as CNAM, must be prepared. The classic data collection and analysis techniques must be enhanced by the integration of studies to anticipate trends in health care demand and provision, and the subsequent expenditures and quality concerns . •:. The dimensions of the large-scale mandatory health insurance reform must not be under-estimated when implemented. Meanwhile, the reform is complicated and comes with high financial risks. Therefore, it is vital to pay special attention to a number of sub-systems that must be consequently adapted. We provide the following priorities: The development of public hospitals' management capacity and the promotion of participatory management which wi111ead to an improvement in hospital performance and their capacity to produce the necessary information for cost calculations and quality-driven resource allocation and management. The adaptation of public health facilities' leadership to the new system's objectives in order to improve the quality of the health services provided. The current framework combines the strong position of public functions applied to health professionals, the centralized decision-making, and the utilization of a global budget in order to assure the delivery of quality health care services. Therefore, the current system only provides little or no incentives for quality assurance of health care services. 444 Good Practice in Health Financing, p. 445 The promotion of effective regulation of the activities of health care professionals who own a private practice and the strengthening of capacity and medical control competency. It is often useful, given that investment in health, a durable pubic good, is one of the promoting factors for economic growth. The era where health expenditures were seen as an uneconomic investment has passed. 445 Good Practice in Health Financing, p, 446 REFERENCES ATCE (Agence Tunisien de Communication Exterieure) 2002. "Principaux indicateurs economiques et sociaux de la Tunisie 1978-2002." Publication de ATCE. Arfa C. et Achour, N. 2004. "National Health Accounts Tunisia 2000." Ministere de la Sante publique. http://who.int/nha. Ben Romdhane H. et al. 2002. "Transition epidemiologique et transition alimentaire et nutritionnelle en Tunisie." Options mediterraneennes, serie B : Etudes et recherches N° 41. BM (Banque Mondiale) 2006. "Etude du secteur de la sante en Tunisie." Departement du developpement humain, region Moyen-Orient et Afrique du Nord. BM (Banque Mondiale) 2004. "Project Performance Assessment Report: Republic of Tunisia." Report nO 31017: http://www-ds.worldbank.org/servletIWDSContentServer. BM (Banque Mondiale) 2003. "Vers une meilleure gouvernance au Moyen-Orient et en Afrique du Nord: Ameliorer l'inc1usivite et la responsabilisation." Rapport sur Ie developpement au Moyen-Orient et en Afrique du Nord, Washington, D.C BM (Banque Mondiale) 2003. "Strategie de developpement touristique en Tunisie." Washington D.C. BM (Banque Mondiale) 2000. "Forsuivre l'integration it l'economie mondiale et perenniser Ie progres economique et sociale de la Tunisie. " Revue Sociale et Structurelle. BCT (Banque Centrale de la Tunisie) 2004. "Rapport annueI2003". Publication de 1a BCT Tunisie. FMI (Fonds Monetaire International) 1999. "Rapport experimental du FMI sur l'observation des normes et des codes: Tunisie". http://www .imf. orglexternallnp/rosc/tun/fre/. Gha1i, S. 2004. "The Tunisian Path to Development: 1961-2001, A case study from Scaling Up Poverty Reduction: A Global Learning Process and Conference." Shanghai, 25-27 May 2004. http://www.worldbank.org/wbi/reducingpoverty/case-Tunisia-Path- Development.html. Hsairi M., NacefT, Achour N., et Zouari B. 2002. "Economie et Sante: Evaluation et Strategies de mise en oeuvre des interventions. " Publication de l'Institut National de la Sante. INS (lnstitut National de la Statistique) 1980-2000. "Enquetes nationales sur Ie budget et la consommation des menages." Serie de publication. 446 Good Practice in Health Financing, p. 447 INS (Institut National de la Statistique) 1966-2004. "Recensement general des menages et de la population, caracteristiques economiques." Serie de publications. MSP (Ministere de la sante publique) 1976-1996. "Serie des bulletins epidemiologiques." Tunisie. MSP (Ministere de la Sante Publique) 1986. " 1956-1986 Trente ans au service de la sante: un engagement, une ethique." Tunis: MSP. UNPD 0 2006. "World Mortality Report 2005." Department of Economic and Social Affairs, New York. PNUD (Programme des Nations Unies for Ie Developpement) 2004. " Strategie du phenomene de la pauvrete en Tunisie. " Etudes et rapports sur Ie developpement, Juillet 2004. Omrane AR. 1983. "The Epidemiologic transition A Theory. A Preliminary Update." Journal of Tropical Pediatric, vol 9: 305-317. UNPD (United Nation of Population Division) 1999. "World urbanization prospects: the 1999 revision.". United Nations Publication, New York, 260 p. UNPD (United Nation of Population Division) 2001. "Human Development Report." Ed. De Boek, Brussels, 275 P. WBI (World Bank Institute). 2004. "Public Health in the Middle East and North Africa: Meeting the Challenges of the Twenty-First Century." Learning Resource Series Publication 29163, Washington, DC World Bank. 2006. World Development Report: Equity and Development. Washington, DC. WB (World Bank) 2004. "Making Deeper Trade Integration Work for Growth And Jobs." Republic of Tunisia, Development Policy Review. Report No. 29847-TN. Social and Economic Development Group Middle East and North Africa Region 447 Good Practice in Health Financing, p. 448 Viet Nam Bjorn Ekman Sarah Bales 448 Good Practice in Health Financing, p. 449 Acronyms and Abbreviations ADB Asian Devel<:lQment Bank BMI Boc!Y Mass Index CHS Commune Health Station CPV Communist Party of Viet Nam DALY Disability Adjusted Life Years DAV Drug Administration of Viet Nam DHC District Health Center DOTS Directly Observed Treatment Shortcourse DPT Diphtheria, Pertussis, Tetanus DRG Diagnostic Related Group EAPR East-Asia Pacific Region EPI Expanded Program of Immunization GBD Global Burden of Disease GDP Gross Domestic Product GNI Gross National Income GSO General Statistics Office HIPC Highly Indebted Poor Country HIV/AIDS Human Immunodeficiency Virus/Acquired Immune Deficiency Virus HCFP Health Care Funds for the Poor IEC Information Education Communication ILO International Labor Office IMF International Monetary Fund MoH,MOH Ministry of Health NHA National Health Accounts ODA Overseas Development Assistance PHC Primary Health Care PPP Purchasing Power Parity TB Tuberculosis UNDP United Nations Development Programme USD United States Dollar VHW Village Health Worker VND Vietnamese Dong VNHS Viet Nam National Health Survey VSS Viet Nam Social Security WDI World DevelClpment Indicators WHO World Health Organization WHR World Health Report WHS World Health Survey WTO World Trade Organization 449 Good Practice in Health Financing, p. 450 Local currency unit: Vietnamese Dong Exchange rate as of November, 2006: USD 1 = VND 16,080 (24/10/2006) 450 Good Practice in Health Financing, p. 451 Introduction In the past two decades or so, Viet Nam has introduced a series of health care and financing reforms to increase coverage of service delivery and health prepayment. This report describes these reform efforts focusing on their impact on efficiency and equity in health care utilization and financing. Many of the reform efforts have been aimed at trying to increase coverage, utilization, and financial protection for the poor and other vulnerable groups. Hence, an underlying "theme" of this report is health reform efforts to try to increase health care coverage for the poor and the extent to which these have been successful. An important purpose of the report is to document experiences that are of interest to other countries that are in the process of implementing similar reforms. Given that Viet Nam is a low-income country with health outcomes that surpass those of similar countries, the experiences of Viet Nam might be of principal interest to other low-income countries. In line with many of the other policy areas, the Vietnamese government (or more specifically, the Communist Party of Viet Nam, CPV) has set very ambitious policy aims for the health sector. One particular such aim of relevance for this study is the goal of universal health insurance by 2010. Although this aim may be difficult to achieve, at least within the target date (indeed, the most recent CPV congress does not mention a set date for completion), it provides the general framework within which other reform efforts are being developed and implemented. As this report will describe, the goal is intended to be reached by means of various types of health insurance programs, including social health insurance, targeted health insurance coverage for the poor, and voluntary insurance schemes. In the case of Viet Nam, it is not possible to identify one or two particular health care coverage or financing reforms - "The Reform". Rather, Viet Nam has introduced a number of "reforms" in different health care financing areas, including resource mobilization (cost-sharing and user-fees), pooling of resources (social health insurance and health cards for the poor), and purchasing (payment ceilings, capitation, and case mix systems). These reforms have often taken the form of central government decisions or CPV resolutions. In addition, Viet Nam has recently initiated decentralization and management reforms, one specific goal of which is to make public hospitals more financially autonomous. The scope and implementation of these and other measures vary significantly as some have been in place at the national level for a long time, while others have been piloted in local settings only and still others are only at the initial or planning stage. Moreover, all reforms have not been documented to a similar extent. These and other factors will affect the extent to which one can say something about these experiences and their replicability. This report draws on what is currently available in terms of reports and evaluations of the identified health coverage reforms. Most of the documentation is publicly available while some are internal reports of the Ministry of Health (MOH) or other agencies. Most of the reference material is written in the English language or is a translation from Vietnamese 451 Good Practice in Health Financing, p, 452 documents. In tenns of data, the report makes use of the following data sources: World Development Indicators 2006, World Health Statistics 2006, national household surveys (various years), and National Health Accounts (NHA) and other official national health statistics. This report is outlined as follows. The next section of the report gives an overview of the current situation in Viet Nam in tenns of the economy, demography and epidemiology, and politics and governance. Section 3 describes the health financing system of the country. The first part of the section discusses the various programs that have been implemented while the second part describes the health financing situation based on the most recent versions of the NHA. The next section describes the current health delivery system in Viet Nam both in tenns of the infrastructure and with regard to the regulatory environment. Section 5 of the report describes the various health financing refonns that have been launched in the past two decades. The main focus of the discussion is on resource mobilization, pooling of resources, and purchasing of services. In addition, based on the available evidence, an assessment of the implications and impacts of the refonn efforts is made. The final section of the report discusses the replicability of the Vietnamese health coverage refonns and the main lessons that may be of interest to other countries in a similar situation as Viet Nam. The main messages of this report are that a low-income country with a difficult past has managed to increase coverage in health financing and services over the past two decades. Although many challenges remain, the increase in health insurance coverage has had a positive impact on those covered in tenns of reduced spending on care, increased utilization of services, improved health outcomes, and reduced income poverty. In tenns of health service coverage rates, Viet Nam has managed to increase immunization coverage rates from less than 50 percent of the population to well over 90 percent over the period. Moreover, significantly more people are reached by other services, including maternal and child health and disease prevention interventions. In combination with rapid economic growth, these improvements in coverage rates have substantially improved health outcomes, reduced the disease burden from communicable diseases, and contributed to the rapid poverty reductions over the past 20 years. More recently, concerted efforts have been made to also increase coverage to previously excluded and vulnerable population groups, such as the poor, ethnic minorities, and people in disadvantaged communes. Preliminary evidence from these efforts suggests that those targeted by these interventions have experienced increased utilization of services and reduced out-of-pocket spending for health care. 452 Good Practice in Health Financing, p. 453 Background Description of Viet Nam Economic environment Since the end of the American War in 1975 and the reunification of the country, Viet Nam pursued a socialist economic policy with government control in the production of most goods and services. In the immediate post-war period, Viet Nam was one of the poorest countries in the world with an estimated gross domestic product (GDP) per capita of US$130 per year. Poor economic performance, stagnant school enrollment rates, and widespread malnutrition characterized developments during this period. By the mid- 1980s, the Communist Party realized that a change was necessary for the country to be able to sustain its population in terms of food security and improved welfare. The government's response to the dismal developments was the adoption of an economic reform program referred to as Doi Moi ("renovation"). The reform program encompassed a series of policy changes that were implemented during the following years and which continue to this date. The most important changes involved the agricultural and trade sectors that saw wide reaching liberalizations. In the health sector, a far reaching deregulation of the health care system was introduced in the early 1990s. Among other things, the changes meant that health staffwere allowed to operate private clinics and that private entrepreneurs could sell drugs in private outlets. Moreover, user fees for health care were introduced as a means to raise revenues. Finally, a social health insurance program was initiated, the further implementation of which is currently taking place by means of various measures to increase coverage and improve efficiency. Table 2.1 shows the development of a set of key economic indicators for various years since 1990. As a direct result of the economic reform program, the country has had one of the fastest growing economies in the world; annual overall economic growth has been between 6-8 percent on average in the past decade. In the early part of the period economic growth was below that of other groups of countries with which Viet Nam can be compared, either geographically or economically. Since then, overall GDP growth has been comparable with that of other fast growing countries in the region and well above that of other low-income countries and even some lower middle-income countries. Growth rates have been particularly strong in per capita terms due to Viet Nam's reduction in fertility rates and subsequently limited population growth rates (see below). The most recent estimate of gross national income (GNI) per capita is $540 (current; WDI 2006) making Viet Nam a low-income country that is rapidly moving into the group of lower middle-income countries. The changes in per capita income over this period suggests that the effect of the policies in terms of economic stability has provided Viet Nam with a national income per capita in purchasing power terms of around US$2,700 per year. In contrast to the early years of the period, this is now above the average for the group of low-income countries. 453 Good Practice in Health Financing, p. 454 Table 0.1: Economic indicators, 1990-2004 Indicator 1990 1995 2000 2001 2002 2003 2004 GOP per capita (constant 2000 US$) 226.86 305.24 396.99 419.18 443.66 471.02 501.99 GOP per capita, PPP 989.34 1,418.72 2,011.74 2,169.96 2,336.38 2,518.06 2,744.7 (current international $) 8 GOP growth (annual %) 5.10 9.54 6.79 6.89 7.08 7.34 7.69 1 2 3 Poverty gap at $1 a day (PPP) (%) n.a. 18 9 0.5 0.5 n.a. n.a. 1 2 Poverty head count ratio at n.a. 58 37 n.a. 28.90 n.a. <20 national poverty line (% of population) Revenue, excluding grants n.a. n.a. n.a. n.a. n.a. n.a. 22 (% of GOP) Tax revenue (% of GOP) n.a. n.a. n.a. n.a. n.a. n.a. 17 External debt (% of GOP) n.a. n.a. 38.60 37.90 34.90 34.10 34 Aid (% of GNI) 3.12 4.08 5.47 4.50 3.70 4.52 4 Unemployment, total n.a. n.a. 2.30 2.80 2.10 2.30 2 (% of total labor force) Inflation, consumer prices (annual %) n.a. n.a. -1.71 -0.43 3.83 3.10 7.80 Source: WDI 2006; IMF Article IV Consultations, PRGF Reports; National Survey Data. 1) Data for 1993; 2) Data for 1998; 3) Data for 2002. Furthennore, as a result of the rapid economic growth Viet Nam has seen its rates of income poverty fall sharply over the same period. Although estimates and definitions vary and are surrounded by some degree of uncertainty, income poverty fell from around 58 percent in 1993 to 29 percent in 2002 (World Bank 2003, p. i). The most recent estimates suggest that poverty has fallen even further to less than 20 percent of the popUlation in 2004 (GSO 2004). Extreme poverty is relatively limited in Viet Nam and remaining absolute poverty is predominantly found among certain regions in difficult circumstances (flood prone areas and mountainous regions) and among ethnic minorities (WHO 2003), who make up approximately 13 percent of the population of the country. Income distribution remains an issue in Viet N am with the 20 percent most well off receiving around 45 percent of available income compared with only around 7 percent for the poorest one fifth of the population (WDI 2006; not shown). In tenns of government revenue, recent estimates suggest that Viet Nam has a revenue- to-GDP ratio of around 22 percent, which is expected to increase slightly in the coming years (IMF 2005; not shown). The overall budget deficit is around 4.5 percent ofGDP, suggesting that the overall fiscal situation of the country is largely sustainable. Furthennore, national accounts data show that Viet Nam has a tax revenue-to-GDP ratio of around 17 percent, the main sources of which are income and profit taxes (ibid.; not shown). Viet Nam has managed to limit its external debt over this period. In particular, Viet Nam is not a HIPC-country (Highly Indebted Poor Country) that can benefit from any such debt cancellation initiatives. Recent estimates show that total external debt amounts to around 33 percent ofGDP and that debt servicing is less than 10 percent of total exports of goods and services (ibid.). An important issue for its external relations is the coming accession to the World Trade Organization (WTO), which is expected to happen before 454 Good Practice in Health Financing, p. 455 the end of the year 2006. It is anticipated that the health sector in general and the pharmaceutical market in particular will be affected by Viet Nam's accession to the WTO. Among other things, drugs prices are expected to fall significantly with possibly far reaching consequences for both access to medicine and to health financing more generally considering the fact that pharmaceuticals make up half of all health spending. Likewise, Viet Nam has limited its dependence on overseas development assistance (ODA). In the early reform period, assistance was provided by the countries of the former Soviet-bloc and Sweden, a large share of which was directed toward the health sector. Since the late 1990s, when relations with the rest of the world were normalized, inflows of aid, including loans from development banks, have increased to reach around 5 percent of GNI, which is considerably more than the regional average although substantially less than, for example, several countries in Sub-Saharan Africa. Demographic, epidemiologic, and social environment Viet Nam covers an area of approximately 330,000 square kilometers (207,000 sq. miles) with a population of around 84 million resulting in some 255 inhabitants per square kilometer. The population consists of several different ethnic groups with the largest being the Kinh, accounting for around 87 percent of the population. The remaining share consists of some 50 different ethnic minority groups with distinctive ethno-linguistic and cultural characteristics (WHO 2003). Around a quarter of the population lives in urban areas, although with migration from the country side to the larger cities, the ratio of rural to urban population is falling. Viet Nam has pursued a relatively strict family and population policy over the past two decades that has resulted in a fertility rate barely above replacement level. Projections suggest that the population growth rate will be reduced even further to less than one percent after 2025 (WDI 2006). Table 2.2 presents selected population, demographic, and social indicators for various years since 1990. Life expectancy at birth was considerably higher in Viet Nam than in other countries at similar income levels well before the mid- 1980s when the economic reform program was initiated. 455 Good Practice in Health Financing, p, 456 Table 0.2: Selected population and social indicators, 1990-2004 Indicator 1990 1995 2000 2001 2002 2003 2004 Population, total (million) 66,2 72.9 78.5 79.5 80.4 81.3 82.2 Population ages 0-14 (% of total) 38.92 37.02 33.46 32.67 31.87 31.08 30.30 Population ages 15-64 (% of total) 56.13 57.96 61.14 61.89 62.67 63.46 64.24 Population ages 65 and above (% of total) 4.95 5.03 5.40 5.44 5.46 5.46 5.45 Median Population Age 20.20 21.30 23.10 n.a. n.a. n.a. n.a. Rural population (% of total population) 79.74 77.79 75.68 75.20 74.72 74.24 73.76 Population growth (annual %) 2.18 1.80 1.29 1.23 1.16 1.10 1.04 Age dependency ratio 0.78 0.73 0.64 0.62 0.60 0.58 0.56 (dependents to working-age population) Life expectancy at birth, female (years) 66.84 69.42 71.52 n.a. 72.20 72.50 72.80 Life expectancy at birth, male (years) 62.82 64.96 66.70 n.a. 67.30 67.60 67.90 Life expectancy at birth, total (years) 64.78 67.14 69.05 n.a. 69.69 69.99 70.29 Birth rate, crude (per 1,000 people) 28.78 23.76 19.43 n.a. 18.50 18.30 18.00 Death rate, crude (per 1,000 people) 7.26 6.68 5.80 n.a. 6.20 6.10 6.10 Fertility rate, total (births per woman) 3.62 2.67 1.90 n.a. 1.87 1.80 1.80 Mortality rate, infant (per 1,000 live births) 38.00 32.00 23.00 n.a. n.a. 19.00 17.40 Mortality rate, under-5 (per 1,000) 53.00 44.00 30.00 n.a. n.a. 23.00 23.20 Mortality rate, adult, female 153.15 n.a. n.a. n.a. 130.06 n.a. n.a. (per 1,000 female adults) Mortality rate, adult, male 215.02 n.a. n.a. n.a. 181.67 n.a. n.a. (per 1,000 male adults) Maternal mortality ratio n.a. n.a. 130.00 n.a. n.a. n.a. n.a. (modeled estimate, per 100,000 live births) Literacy rate, adult female n.a. n.a. n.a. n.a. n.a. n.a. 86.92 (% of females ages 15 and above) Literacy rate, adult male n.a. n.a. n.a. n.a. n.a. n.a. 93.92 (% of males ages 15 and above) Literacy rate, adult total n.a. n.a. n.a. n.a. n.a. n.a. 90.28 (% of people aQes 15 and above) Source: WDI 2006. One likely factor for the relatively favorable health outcomes in the country is the comparatively high level of education as reflected in the high literacy rates for both men and women. The literacy rates of Viet Nam, also for previous years not shown in the table, are significantly higher than those of other low- and even some middle-income countries. Figures 1 - 3 show population pyramids for the years 1990, 2005, and 2020 (estimated), respectively. The figures show the effect of the government's family and population policy, one prominent feature of which is the active encouragement of a maximum of two children per family. 456 Good Practice in Health Financing, p. 457 Figure 9: Population pyramid, Viet Nam 1990 Vietnam, 1990 --'--'---l I 90-94 I 75-79 Q) 60-64 I C) « 45-49 I 30-34 I 15-19 I 0-4 20 15 10 5 0 5 10 15 20 Population (%) Source: United Nations Population Division, World Population Prospects, The 2004 Revision Population Database. Figure 10: Population pyramid, Viet Nam 2005 Vietnam, 2005 90-94 75-79 Q) 60-64 C') < 45-49 30-34 15-19 0-4 15 10 5 0 5 10 15 Population (%) Source: United Nations Population Division, World Population Prospects, The 2004 Revision Population Database. 457 Good Practice in Health Financing, p. 458 Figure 11: Population pyramid, Viet Nam 2020 (estimated) Vietnam, 2020 90-94 75-79 f/) 60-64 CI) C) « 45-49 30-34 15-19 0-4 15 10 5 0 5 10 15 Population (%) Source: United Nations Population Division, World Population Prospects, The 2004 Revision Population Database. Although the projections indicate that the total population will grow to around 125 million by 2025, it is plausible that the country will be able to sustain its population with social services. The demographic transition of an aging population will, however, require that the social delivery system continuous to adapt to the needs of elderly people. Viet Nam has made significant progress in its disease priority programs. For example, it has managed to reduce the incidence of and death from malaria dramatically in the past few years, and moreover, effectively managed to reduce the incidence of polio, neonatal tetanus, and leprosy to levels where they no longer are regarded as threats to the health status of the population. The estimated global burden of disease (GBD) for the main disease categories is shown in table 2.3. With respect to morbidity patterns, Viet Nam has undergone an epidemiological transition along with the economic and social changes over the past two decades or so. In 1986, communicable diseases accounted for more than half of the entire estimated disease burden with non-communicable diseases making up almost the full balance (World Bank 2001). By the end of the 1990s, non-communicable diseases accounted for 65 percent of the main causes of illness with communicable diseases now making up around 27 percent, and accidents and injuries accounting for the rest. It is expected that accidents and injuries will continue to increase, not least as a result of road traffic incidents. 458 Good Practice in Health Financing, p. 459 Table 0.3: Estimated Burden of Disease, Viet Nam, 2002 Top 5 Burden of Diseases: Deaths only (Estimated total deaths ['000] by cause GBD Cause Value #1 Cardiovascular diseases 160.0 #2 Infectious and parasitic diseases 74.6 #3 Malignant neoplasms 64.1 #4 Respiratory diseases 51.3 #5 Unintentional injuries 36.1 Burden of Diseases by Disease Categories: Deaths only (Estimated total deaths ['000] by cause and relative share Value % I. Communicable, maternal, perinatal & nutritional conditions 126.2 24.5% II. Noncommunicable diseases 341.1 66.1% III. Injuries 48.5 9.4% Total 515.8 100.0% Top 5 Burden of Disease: DALYs (Estimated total DALYs ['000] by cause GBD Cause Value #1 Infectious and parasitic diseases 2337 #2 Neuropsychiatric conditions 2195 #3 Cardiovascular diseases 1389 #4 Unintentional injuries 1389 #5 Sense organ diseases 809 Burden of Diseases by Disease Categories: DALYs (Estimated total DALYs ['000] by cause Value % I. Communicable, maternal, perinatal & nutritional conditions 4305 32.2% II. Noncommunicable diseases 7334 54.9% III. Injuries 1721 12.9% Total 13360 100.0% Source: World Health OrgamzatIOn, Department of Measurement and Health InformatIOn, December 2004; http://www.who.intihealthinfo/bodestimates/eniindex.html Among other things, table 2.3 indicates the dual burden of disease that Viet Nam is facing with increasing morbidity and mortality from non-communicable diseases and injuries, on the one hand, and a persistent burden of disease from communicable diseases. As to specific causes of morbidity and mortality, it is estimated that the top ten causes of hospital admissions were: Pneumonia, Acute pharyngitis and acute tonsillitis, Acute bronchitis and acute, bronchiolitis, Influenza, Transport accident, Diarrhoea and gastroenteritis of presumed infectious origin, Essential primary hypertension, Gastritis and duodenitis, Diseases of appendix, Intracranial injury (MoH information system). Similarly, the main causes of deaths at hospital were: Intracranial injury, HIV, Pneumonia, Intracerebral haemorrhage, Transport accident, Acute myocardial infarction, Stroke (not specified as haemorrhage or infarction), Heart failure, Respiratory tuberculosis, Intentional self harm (suicide). It should be noted, however, that there is considerable dearth of information on this issue as these deaths only account of around 5 percent of all deaths. Political and governance environment Viet Nam is a one-party socialist republic with a president as chief of state and a prime minister as the head of government. The legislative branch consists of a national assembly with representatives from all 64 provinces of the country (comprising 611 459 Good Practice in Health Financing, p. 460 districts/provincial towns and 10,600 communes/wards) that elects the president from among its members. The cabinet is appointed by the president and ratified by the national assembly. The National Assembly is assuming a gradually more prominent role in policy debates and decisions, including in the social sectors. Given the country's relatively high level of decentralization, this is an important factor in terms of implementation of policies and possible adjustments. Formally, the country is governed by People's Committees at various levels of the administration, including the village level. These committees have far reaching authority over mobilization and allocation of resources at their respective level of administration. Furthermore, while it is fair to say that Viet Nam is a strongly centrally planned society - in contrast to implementation - focusing on production targets, attempts at active village level participation in decision making processes have been on-going for many years. As mentioned, in terms of implementation, Viet Nam is a highly decentralized country, even in an international perspective (World Bank 2004, ch. 8). For example, spending at sub-national level has increased steadily since the early 1990s and now constitutes around 45 percent of state budget expenditures. An important factor in the decentralization process is the extent of inter-provincial transfers that are determined by the share of locally raised revenues that each province can retain and the grant received by the central government. As resource availability varies considerably among provinces, some provinces are net contributors while most are net recipients of such grants. In general, the grants are directed toward a set of expenditure areas that include social and economic activities. A concern has been raised that the allocation formula is not sufficiently commensurate with local needs and resources mobilization abilities (World Bank 2003a). In particular, moving toward allocations based on development outcomes, including equity concerns, would most likely improve the incentive structure of the provincial and local level authorities. This is a particular concern as the health sector often suffers from being a low priority when cuts need to be made due to shortages of funds. A further issue of decentralization is the level of autonomy among administrative and service delivery units, including health care providers such as hospitals and clinics. The main point of such transfer of authority over resources is to improve efficiency of the operations of these units. However, concern has been raised that the absence of transparency, accountability, and sound monitoring might lead to inappropriate use of resources and rent seeking behavior among staff and managers of the spending units (ibid.). This is particularly the case when decisions about investments in expensive equipment are made without any higher level coordination and control. Continuity of care between different levels of the system is also threatened under the strong decentralization push currently underway. As noted above, Viet Nam has largely relied on a planning process that looks almost exclusively at the attainment of certain pre-defined production targets. Recently, the government has initiated a process that would take development goals into more consideration (Ministry of Planning 2005). Although this is a much more analytically demanding planning process, it would be expected that the country would thereby 460 Good Practice in Health Financing, p. 461 manage to sustain or even improve upon its recent record in economic development and poverty reduction. 461 Good Practice in Health Financing, p. 462 Overview of Health Financing and Service Coverage inViet Nam Trends in health financing Viet Nam's health financing indicators show a continuing increase in per capita health expenditures, but a relatively constant share ofGDP spent on health (table 3.1). Private spending accounts for about three fourths of all spending. Among health expenditures, out-of-pocket accounts for the largest share, although this share has seen a decline as health insurance has increased its share. Starting in 2003 with the establishment of the health care fund for the poor (HCFP) and in 2005 with free health care for children under age six, the government has greatly increased its spending commitments on health care. These new entitlements to free health care constitutes an impressive increase in government funding of health care not yet reflected in the figures in the tables below. Viet Nam is not overly dependent on ODA for health care spending in general with only 2.6 percent of total health spending coming from external sources, although the share of external funding for preventive care is somewhat higher (NHA). Table 0.4: Per capita total expenditure on health (international USD), Viet Nam 1999-2003 1999 1 2000 1 2001 1 2002 1 2003 Per capita total expenditure on 111 129 145 143 164 health at international dollar rate Total expenditure on health billion USD at international 8.6 10.1 11.5 11.5 13.3 dollar rate Total expenditure on health as a 4.9 5.3 5.5 5.1 5.4 share ofGDP General government health expenditure as a share of total 6.7 6.0 6.7 5.1 5.6 government expenditure Share of total health expenditures that are out-of- pocket 58.2 62.7 59.2 58.1 53.6 Social insurance as a share of general government expenditure on health 9.5 10.5 13.7 15.8 16.6 Private health expenditures as a share of total health 67.3 72.0 70.8 71.9 72.2 expenditures Public health expenditures as a share of total health 32.7 28.0 29.2 28.1 27.8 expenditures External resources for health (% 3.4 2.7 2.8 3.5 2.6 of total expenditure on health) Source: WHR 2006. Table 3.2 shows the share of health expenditures by type of service and table 3.3 indicates the sources of funds by type of service. The bulk of health spending is spent on self-treatment, that is, over the counter purchase of both prescription and non-prescription drugs not as part of a medical consultation with a relatively stable high share of around 42 percent of total health spending from 2000 to 2003 (Vietnamese NHA). This spending is primarily at private pharmacies, and is not covered by health insurance or exemption 462 Good Practice in Health Financing, p. 463 programs. Self-treatment is the primary choice for treatment among both the poor and the rich (Chang and Trivedi 2003), Spending on inpatient care accounts for about 27 percent of all spending on health in Viet Nam, mostly at state-run hospitals funded by user fees, government subsidies and health insurance reimbursements. Outpatient care accounts for about 14 percent of total health spending half of which comes from the state budget. Primary health care (PHC) and school health programs account for 7.5 percent of all spending and are funded largely by employers. Table 0.5: Percent of health expenditures dedicated to different types of care 1999 I 2000 1 2001 I 2002 1 2003 Inpatient 30.1 29.5 26.7 26.4 27.0 Outpatient 14.2 12.0 16.4 13.7 13.5 Self-medication 38.1 43.5 41.7 42.0 42.1 PHC and school health 4.8 4.9 6.3 7.7 7.5 Preventive and public health 10.5 7.9 6.6 7.7 7.4 Other 2.3 2.2 2.3 2.4 2.5 Total 100.0 100.0 100.0 100.0 100.0 Source: NHA 1998-2003. MOH, StatIstIcal PublIshmg House 2006. Preventive health programs have seen a declining share in total spending to reach 7.4 percent of total spending, three fourths of which comes from the state budget and one fourth from external assistance. Government administration and regulation of the health sector, research and training, account for a very small share of the totaL Table 0.6: Sources of funding for different components of health care, 2003, percent Health State budget I insurance I User fees I ODA I Other I Total Inpatient 27.2 15.2 37.5 1.4 18.8 100 Outpatient 49.6 19.8 10.1 1.3 19.3 100 Self-medication 0.0 0.0 100.0 0.0 0.0 100 PHC and school health 18.3 0.0 18.2 1.0 62.5 100 Preventive and public health 73.4 0.0 2.2 24.4 0.0 100 Other 67.3 10.8 0.0 6.1 15.8 100 Total 22.5 7.0 55.1 2.6 12.8 100 Source: NHA 1998-2003. MOH, StatIstIcal PublIshmg House 2006. Health financing programs As Viet Nam undergoes its transition to a market economy, the government continues to place a strong emphasis on preventive care, maintaining state budgetary and ODA funding of national target programs and other preventive health measures. The government also continues to invest in the grassroots health network to implement these programs along with primary and maternal and child health care. However, in recent years, financing of curative health care in Viet Nam has increasingly relied on mobilizing resources from society (xa hoi hoa or socialization), primarily through user fees at government facilities, expansion of the private sector and a contributory social health 463 Good Practice in Health Financing, p. 464 insurance scheme. The share of expenditures of state-run health facilities covered by the government budget has declined, from 62.7 percent in 1998 to 49.6 percent in 2003 (Huong et al. 2006). User fees and to a lesser extent health insurance have taken up an increasing share, ensuring funding to maintain and improve the quality of health care facilities. Social health insurance was introduced in 1992 as a solution to help mobilize resources and create a more appropriate mechanism for payment of user fees (table 3.4). The Vietnamese Communist Party and Government have decided that health insurance should be the primary mechanism for financing health care and have set the challenging goal noted above of reaching universal health insurance coverage in the near future. At the same time, social health insurance has gradually become the main mechanism for implementing subsidies to users in efforts to increase government health spending to promote health equity goals and overcome the negative consequences of user fees. Figures from VHI indicate growing coverage of the population by health insurance, especially among the poor who are benefiting from government assistance to purchase health insurance. Coverage by a prepayment mechanism is even higher if other government health financing programs are included. Table 3.4: Overall health insurance coverage rate and type of program, Viet Nam 1998-2004 Health insurance holders 1998 1999 2000 2001 2002 2003 2004 percent population covered by health insurance 12.7 13.4 13.4 15.8 16.5 20.3 27.7 percent health insurance for the poor - 0.6 1.1 1.9 2.1 4.6 4.8 percent health insurance for the non-poor 12.7 12.8 12.3 13.9 14.4 15.7 22.9 .. Source: 1998-2002 Health Insurance StatIstIcal Yearbooks. 2003-2004 proVIsIOnal figures from VIet Nam Social Insurance Agency as presented in Huong, P.B. et al. 2006. The section below provides a brief description of the various public and private health financing mechanisms for curative care currently available in Viet Nam. Compulsory social health insurance As noted above, compulsory social health insurance was begun in 1992. There are two groups of beneficiaries, those who have made contributions to health insurance and those whose contributions were paid by different agencies of the government. The contributory compulsory social health insurance covers all Vietnamese employees with a contract of 3 or more months as well as pensioners and people on disability. Despite the compulsory nature of social health insurance for these groups, compliance is particularly low for people in the informal private sector. Even among people making contributions for health insurance contributions, there is some evidence ofunderreporting income in order to pay lower amounts into social insurance (Castel (2005) on social insurance, but relevant also for health insurance). 464 Good Practice in Health Financing, p. 465 A large number of groups in the population are eligible for compulsory social health insurance paid through the government budget if they are not yet covered through their employment. The first group is those who have served the government or provided meritorious service to the country including such groups as elected officials, political beneficiaries who served in the revolution or wars protecting the nation, family members of current military or police officers. The second group includes parts of the population considered vulnerable for which the government finances social insurance contributions including the elderly aged 90 and older, social assistance beneficiaries (such as the disabled, elderly without family to care for them, orphans), foreign students on scholarships from the Government of Viet Nam. In 2003, the Government put in place the HCFP in order to ensure financing of health care services for the poor, ethnic minorities in disadvantaged provinces and residents of remote communes (under Program 135). Provinces were allowed to choose whether to implement health care for the poor through purchase of health insurance or through direct reimbursement of facilities and services used by the poor. When Decree 63 on health insurance was issued in 2005, the direct reimbursement option was discontinued and it became compulsory for the province to purchase health insurance for these target groups with funding from the Central budget. Voluntary health insurance Voluntary social health insurance is legislated to cover individuals with a need for health insurance who are not covered under the compulsory scheme, such as the self-employed or non-working population, or for those with compulsory health insurance who want to contribute to voluntary health insurance as a supplementary insurance to their compulsory coverage. The voluntary part of the social health insurance program is intended to be provided to groups of people rather than individuals so as to create a risk pool, reducing the risk of adverse selection. Currently most of the beneficiaries covered under voluntary health insurance are school pupils and students as it has been relatively easy to enroll an adequate share of pupils in a school to meet the minimum coverage rate required by insurance regulations. Other options for risk pools include communes or various mass organizations and associations however it has been more difficult to enroll sufficient members, and even then the members tend to be those who have more health problems and therefore pose a higher risk to the insurance scheme. In addition to social health insurance, several private and state insurance companies offer health insurance, usually as an add-on to life insurance products including policies for pupils and students as well as for older working people to cover hospitalization, surgery, and emergency transport. In 2003, according to the latest NHA, private health insurance reimbursements to facilities constituted about one third of all health insurance reimbursements (MOH 2006). 465 Good Practice in Health Financing, p. 466 Other government sponsored programs In Decision 139 of October 2002 on the HCFP a provision was made to not only cover health care costs for the poor and other target groups of the policy, but, in principle, to also provide assistance to the non-poor if they face unexpected and catastrophic health expenditures although implementation has been weak. According to the Law on Protection and Care of Children, starting in 2005 all Vietnamese children who are not yet a full 72 months of age are eligible for free health care at all government health facilities with facility services reimbursed directly from the State budget. In 2006 the Prime Minister requested the MOH of Health and Vietnam Social Security (VSS) to study how to convert this scheme into health insurance coverage for children under age six. Benefit packages of health financing programs Compulsory and voluntary health insurance pay for outpatient and inpatient diagnosis and treatment at government and private facilities that have signed a contract with the social insurance agency agreeing to the reimbursement rates set for public facilities. The package is quite generous and in principle should cover the costs of: 1. Consultations, diagnosis, treatment and rehabilitation during the time of treatment at the health facility 2. Lab tests, diagnostic imaging and other diagnostic techniques 3. Medicines in the list determined by the MOH 4. Blood, transfusions. 5. Medical procedures and surgery 6. Use of materials, medical equipment and treatment bed 7. Prenatal exams and assistance at delivery For the poor, political beneficiaries, people living or working in mountainous, remote areas, transport costs for referrals from district to higher level hospitals are also covered by health insurance. Health insurance does not reimburse fees for the following health problems or treatments because of moral hazard issues, the elective nature of the intervention, or because coverage of those health problems already exists in other government programs: l. Treatment ofleprosy, medicines for treatment ofTB, malaria, schizophrenia, epilepsy, and other diseases already covered with government funded programs 2. Diagnosis and treatment of HIVIA IDS except when HIV tests are part of protocols for treatment of other diseases or in the case the individual was infected with HIV through their work, syphilis and gonorrhea 466 Good Practice in Health Financing, p. 467 3. Immunizations, nursing care, early diagnosis of pregnancy, health checkups, family planning services and treatment for sterility 4. Plastic surgery, prosthetics, false teeth, eyeglasses, hearing aids 5. Occupational disease and labor accidents, war-related accidents 6. Treatment costs in case of attempted suicide, self-inflicted harm, drug addiction or health problems associated with illegal activities behavior 7. Health assessments for legal reasons 8. Consultations, treatments, rehabilitation or birth at home Patients may have to make co-payments in the amount of the additional cost incurred beyond what it would have cost for basic services at an appropriate level of care in government facilities in the following cases: • Special requests for services such as selecting their own practitioner, type of bed, select the facility they want to use • Bypassing primary care facility without a referral • Seeking treatment at facility that doesn't have a contract with health insurance, • Seeking care at a private facility • Seeking treatment overseas For children under age 6, the benefits include consultations, diagnosis and treatment at all primary care public facilities (including commune health stations (CHS), district hospital) and if formally referred then treatment at provincial and central level facilities. In case of an emergency, the family can seek treatment for the child at any level of facility without a referral. In case the family of a child requests to use certain high tech equipment, the family must pay the additional fees. When this policy is converted to health insurance coverage for children under age six, it is likely that the benefits will be the same as other insured groups. Financing and payment Contribution levels and financing sources Compulsory insurance contribution rates are set at a modest level of 3 percent of contractual salary and basic allowances, pension, social insurance payments, scholarship or minimum wage depending on which entitlement group the insured individual belongs to. For the employed, workers pay I percent and their employers pay 2 percent of salary. For retirees and people receiving social insurance benefits, the contributions are paid by VSS. For the other groups the government budget pays the contributions. For the poor and the elderly aged 90 and older, the contribution is a fixed amount of3.l USD per person per year with the contribution paid from the state budget. Voluntary social health insurance contributions range between 1.9 and 10.0 USD depending on the locality and the type of risk pool (for example, a school, an association, or a commune) and are paid by individuals and their households. 467 Good Practice in Health Financing, p. 468 For children under age six, the government has allocated an amount of 4.7 USD per child per year, but if the total amount required exceeded this amount, funds from the reserve budget line of the national budget would be used to pay the excess. Stipulations of the HCFP require that provinces allocate 4.4 USD per eligible beneficiary, with 75 percent paid in from the State budget and most of that used to purchase health insurance. The remaining amount, if it can be mobilized, can be used to pay for health care of non- eligible groups when faced with catastrophic health care costs. In the past, because of strict ceilings imposed on hospitals for how much they were allowed to be reimbursed, and despite the low contributions there has been a surplus in the health insurance fund. Since Decree 63 on revised health insurance regulations was passed in 2005, there is some concern about the solvency of the social insurance fund because low contribution levels were maintained, while the benefit package was expanded to include over 200 additional medicines and a large number of expensive high tech services. In 2005, VSS estimated that health insurance reimbursements exceeded total contributions to health insurance by 8.6 million USD, an amount equivalent to about 4.3% of total health insurance contributions. Currently this is being financed by accumulated reserves, but VSS is warning that these reserves will be exhausted in 2007 and it is a top priority now to find a solution for this issue either through increasing contribution rates or finding a way to better control costs. In order to gain a sense of their adequacy, one can compare contributions to per capita curative care spending. In 2003 according to NHA estimates, if one only includes health spending at government hospitals and curative care facilities per capita annual spending was 7.1 USD. If one were to include private curative care including self-treatment, annual per capita spending was 20.3 USD (NHA 2003). There is some concern that with the generous package, coverage of a higher and higher share of people for whom low contributions have been paid (poor, elderly and soon children under age six), and with adverse selection among the voluntarily insured, there may be major problems with balancing the insurance fund. In addition, the cost of health care is increasing rapidly with the development of new technology and drugs, which will require increasing contribution amounts periodically to keep up. Reimbursement procedures used to pay providers Reimbursement from health insurance as well as the direct reimbursement of services for children under six and the poor not yet covered by health insurance is on a fee-for-service basis. The amount reimbursed to facilities is based on the partial user fee schedule charged to uninsured patients, although in some cases the amount reimbursed by health insurance is less than the amount facilities charge to the uninsured. Drugs and disposables are reimbursed at the price paid by the facility to purchase them. The "major drug list" is the list of drugs reimbursed by health insurance. It was updated in 2005 and now includes 646 modem and 91 traditional medicines. There are regulations on certain drugs that require a committee to approve use before they can be used, and the 468 Good Practice in Health Financing, p. 469 drug list also specifies which level of hospital (district, province, central) is allowed to use each type of medicine. A list of procedures with prices that are reimbursed by health insurance was developed in 1995, and appended with a price list in 2006. Private facilities that sign contracts with the health insurance agency are reimbursed at the same rates as government facilities. Inherent in the fee-for-service payment mechanism are incentives for over prescription of certain health services and medications. In fact, a recent study found evidence that the inpatient admission rates and the average length of stay were higher for better off individuals and for the insured (Sepehri et aL 2005). In order to overcome this provider moral hazard problem (or provider induced demand), Decree 63 called for development of alternative payment mechanisms, including capitation and casemix options (such as DRG). In 2004, WHO negotiated an experiment at selected districts in Hai Phong Province. Initial results of the experiment were favorable both in terms of maintaining quality of services, solvency of the health insurance fund and satisfaction by the providers and patients. However, in 2005 when new health insurance regulations came into effect covering a wider range of services, especially high tech services, the health insurance agency would not allow the capitation amount to be increased to cover the higher cost of services provided so providers are not willing to continue with the experiment. The Department of Planning and Finance of the MOH is currently undertaking a costing study as part of the process of developing a casemix payment system to be pilot tested in 2007. It is hoped that the casemix system will improve cost-effectiveness of health spending by reducing use of unnecessary services and drugs. This will also contribute greatly to ensuring solvency of the health insurance fund. Measures to ensure solvency of the insurance fund At present, the health insurance scheme only reimburses facilities the partial user fees that are charged to patients paying out of pocket. The government continues to subsidize health facilities by paying salaries, investing in equipment and paying part of operating costs. If the government allows facilities to charge full cost for health care, the cost to patients and the insurance fund would increase substantially. A recent hospital costing study examining costs of treating five different medical conditions at provincial hospitals found that the amount currently charged covers from 39 percent of full costs for pneumonia in children to 71 percent of full costs for intracranial injury (Lieu et al. 2005). Currently the MOH intends to continue subsidizing health care, but would like to shift from payment to providers to payment through users although it has not yet identified the appropriate mechanism for this purpose. In the hospital costing study by Lieu et al. (2005) drugs accounted for from 30 to 70 percent of the full cost of services. Cost of diagnostic procedures (depreciation, labor and materials) were not found to be a high share of costs of care yet, although with the development of high tech services, their share may increase in the future. The major drug list and list of service user fees that are reimbursable by health insurance in principle 469 Good Practice in Health Financing, p. 470 could be used as cost control measures. However, in establishing the lists of drugs and services, little attention was paid to the criterion of cost-effectiveness and the benefit package is in fact extremely generous and not currently an effective tool for controlling costs. Another measure to ensure solvency of the health insurance fund is imposing ceilings on benefits. Individual medical services less than 7 million dong are reimbursed by insurance, but the amount exceeding 7 million dong will be reimbursed different amounts for different beneficiaries. Co-payments are often put in place to avoid overuse of health services (moral hazard). However, in 2005, Decree 63 eliminated all co-payments in an effort to make health insurance more attractive to potential participants. The problem with ceilings on individuals and co-payments is that they do not provide incentives to practitioners to provide services in a cost-effective manner. In health care it is usually the practitioners who determine the amounts and types of services to be provided, not the consumers. In order to avoid overspending of the health insurance fund under the fee-for-service reimbursement mechanism, VSS has imposed caps on the amount each first contact facility will be reimbursed. The insurance fund available for reimbursements at each hospital, in principle, is based on the number of insured and the average contribution amount in the province. However, an amount is subtracted from the insurance fund at each facility for payment to referral hospitals. The amount subtracted is calculated as 110% of the amount actually reimbursed to referral facilities in the previous year. In principle there are no caps imposed on services provided by referral hospitals and therefore no incentives for cost-effective services at this level. Inherent in this payment mechanism are strong disincentives for first contact hospitals not to refer patients. Although some expenditures in excess of the cap may be reimbursed by health insurance, there is a risk that they will not or that reimbursement will be severely delayed. To avoid overspending hospitals create their own controls on services provided to the insured, which often leads to differential services provided to the insured and those paying user fees. In some cases, the hospitals ask the insured patients to "voluntarily" agree to paying the differential between the amount health insurance reimburses and the amount that the hospital would normally collect from fee-paying patients which goes against the health insurance regulations but is not stopped because of sympathy with the difficulties of staying within the budget of the health insurance fund. Equity: Health indicators and outcomes, and their distribution Several studies have found inequitable utilization of health care services across living standards groups with the poor less likely to seek care at all, and when they do, to seek care at lower levels of facilities. These studies have also shown differentials in the burden of health care expenditures with the poor having higher burdens (Lieu et al. 2005; Knowles et al. 2003; Minh et al. 2004; Chien et al. 2005; Dong 2002). 470 Good Practice in Health Financing, p. 471 Earlier efforts of the Vietnamese government to improve health equity and access to health care by children under age six involved unfunded exemptions at government hospitals and provision of health insurance to a small share of the poor for which targeting was problematic. The Viet Nam National Health Survey was implemented in 2001-02, covering the period immediately prior to the approval of the HCFP policy (Decision 139). Survey results found low insurance coverage among the poor, ethnic minorities and people living in mountainous regions. In addition, fee waivers or health insurance reimbursements were obtained for only 31 percent of inpatient visits for people eligible for Decision 139 benefits compared to 66 percent among children under age six and 74 percent of people covered by health insurance (MOH 2003). The negative consequences and inequities in access to curative health care and financial burdens for the poor have been documented in numerous studies (Knowles et al. 2003; Minh et al. 2004; Chien et al. 2005; Dong 2002, and Wagstaff and van Doorslaer 2003). Among other things, Wagstaff and van Doorslaer find that out-of-pocket expenditures for care contribute to increasing the number of absolute poor by 3.4 percentage points in the end of the 1990s. The HCFP policy was developed based on this existing evidence base of inequities in health outcomes, access to health care and differential financial burden. Evidence of the strong impact of health insurance on access to health care helped to orient the policy towards subsidizing users through health insurance although originally the policy also allowed direct government reimbursement to facilities for services used by the poor. Implementation of the HCFP policy began in 2003 and an impact evaluation is planned with results expected by mid 2007. The policy on free health care for children under age six only began implementation in 2005. In the process of implementing these policies, administrative reports from the provinces have been compiled to identify shortcomings and difficulties in implementation. In addition, two small-scale studies involving fieldwork have been commissioned explicitly to look at the impact of the HCFP. Although these studies only examine a small geographic area, the findings lead to optimism about the positive impact of the HCFP on health equity. Targeting of specific populations The government's obj ective in targeting of provision of assistance to cover user fees is not based simply on need (i.e. health care for the poor, health care for the old elderly), but includes also a political rationale (people who have performed meritorious service to the country, non-poor ethnic minorities), concern about the health of future generations (health care for children under age 6). One of the early complaints from the provinces about implementation of Decision 139 was the perceived unfairness of providing assistance to some better off people who happened to be part of the target group, i.e. residents of disadvantaged communes or ethnic minorities in disadvantaged provinces. In addition, it was felt that there were many near poor ethnic minorities living in provinces other than the 12 targeted disadvantaged provinces that also deserved assistance but were not targeted. In the first revision of the 471 Good Practice in Health Financing, p. 472 HCFP policy in 2006, one of the main changes was to reduce leakages to better off groups, and reduce under coverage among near poor ethnic minorities. Lack of concrete criteria for selecting beneficiaries to implement assistance for catastrophic health expenditures for people who are not eligible for free health care cards or health insurance has hindered its implementation. This lack of concreteness also makes it difficult to evaluate whether provinces are using these funds appropriately. According to administrative records of implementation, only about one half of provinces have actually implemented this provision of the policy. Regarding leakages and under coverage of targeting of Decision 139, the MOH is planning a study to analyze VHLSS and administrative data to assess this issue, results should be available by mid 2007. 472 Good Practice in Health Financing. p. 473 Overview of Health Delivery System in Viet Nam Supply and organization of the health care system The health care system in Viet Nam is organized in a four-tiered configuration (World Bank 2001). At the upper-most level is the MOH, which, along with the People's Committees at the various levels of the administration is responsible for formulating and executing the health policy and programs of the country. In addition to the MoH there are several separate units and institutes that function outside of the Ministry in areas such as training, research, and policy development in especially important areas, including family planning. In addition, there are 31 central level hospitals providing primarily preventive care services along with training, research, and supervision of some of the preventive care target programs. The second level in the health system configuration consists of the Provincial Health Bureaus. These are relatively influential entities given Viet Nam's far-reaching decentralization. The Provincial Health Bureaus are part of the provincial governments under the Provincial People's Committees and are charged with the overall planning of health services and programs at the provincial level. Immediately below the Provincial Health Bureaus are the District Health Centers (DHCs). In 2006 as part of the decentralization process, the DHCs are being split into three entities with distinct responsibilities: District health offices, district hospitals and district preventive health centers. District hospitals and preventive health centers are for the most part still under the supervision and management of the provincial health bureau and ensure the curative and preventive care of the population in the district. The district health office under the District People's Committee has responsibility for government administration regarding overall management of the CHS (see below), food safety and inspections, and disease prevention. Finally, at the lowest level of the health care system are the CHS. These are mainly responsible for providing primary health care and implementing the national health target programs. In addition to the CHSs there exists a system of village health workers (VHW) that provide a minimum level of health care, including information-education- communication (lEC) and first aid, to the inhabitants of villages in remote areas. In terms of infrastructure, Viet Nam has around 817 hospitals with a total of 98,000 hospital beds. 65 percent of the hospitals are district hospitals, 24 percent are provincial hospitals, and around 2 percent are central hospitals under the responsibility of the MoH. In addition, there are some 78 non-MoH hospitals that belong to other ministries and government units. In 2006 there were a total of 41 private hospitals, a number that is expected to increase rapidly in the coming years as a result of further deregulation and the government's intentions to further diversify service provision in the country. As to the organization of health care and programs, most primary health care is delivered at the CHSs, although much such care is also provided higher up in the system like district hospitals and even provincial hospitals. The primary health care network in Viet 473 Good Practice in Health Financing, p. 474 Nam has a long history. Up until the mid-1980s, most care was provided by the local agricultural work brigades. This grass-roots system formed the backbone of the service delivery system, but with the introduction of the economic reforms at this time, these providers of primary health care were largely dismantled. Recognizing the difficulties that followed, in 1994 the government assumed the responsibility of paying the salaries of all health workers. An important feature of health care delivery in Viet Nam is the national health programs that are implemented by means of vertical structures from the central to the local levels. These programs are basically free of charge or with very low fees for the general population. Currently the MOH is negotiating a revised list of national target programs in health to include: HIV/AIDS, community mental health (mainly schizophrenia and epilepsy), TB, leprosy, malaria, dengue fever, EPI, malnutrition, food safety, reproductive health, military-civilian medical collaboration, school health. In the CHS a large range of maternal and child health services are also provided including prenatal care, deliveries and health care for small children. Coverage of these primary health care and preventive programs is nationwide although evidence from a recent national health survey (VNHS 2001/02) indicates that the implementation of the programs is not being done uniformly across provinces and districts. As previously noted, in addition to the public facilities, Viet Nam has a growing private health sector which provides outpatient care, traditional medicine services, and sales of pharmaceuticals. Along with improved ability to pay for health care and current policy reforms, the size and scope of the private health care sector is expected to increase. There is, however, a significant informational deficiency regarding the size and scope of the private health care sector as private facilities are under the jurisdiction of the provincial health bureaus and a large share of these providers is not licensed. There is also some difficulty in separating out public and private facilities as a large number of public doctors also operate private practices and many public hospitals are setting up semi- private wings providing more "hotel-like" services in order to increase revenues under the policy of promoting autonomization of hospitals. The most recent data on health care utilization suggest that private provision of care may be as high as 30 percent of all care with large variations across provinces and with particularly high levels in the urban areas (MoH 2003). Statistical overview of the health system Using recent international health systems statistics, this part presents a set of key health systems indicators for Viet Nam, focusing on human resources, coverage of services, and some related health outcomes. In terms of non-financial inputs of the health care system, table 4.1 shows inputs of human resources for health care. 474 Good Practice in Health Financing, p. 475 Table 0.7: Selected health systems indicators - Human resources for health, Viet Nam, various years* Physicians Nurses Midwives Pharmacists Total Number Density Number Density Number Density Number Density Density per 1000 (2003) per 1000 per 1000 per 1000 per 1000 42327 0.53 44539 0.56 14662 0.19 5977 0.08 1.3* Source: WDI 2006, WHS 2006. * See WHO World Health Statistics for details. * World Bank, Learning Initiative on Human Resources for Health. At around 1.3 government health workers per 1,000 inhabitants, Viet Nam is at par with that of other low-income countries. Along with continued professional training, however, it is expected that the staff of the health system will become gradually more qualified in the coming years. In addition to these formal health care provider groups, there are many alternative (or traditional) providers of health care in Viet Nam that constitute a relatively large and well recognized part of the health care delivery system. In addition, in 2003, Viet Nam had 23 hospital beds per 10,000 inhabitants compared with 39 for the WPR (not shown). Although there are many more details hidden behind the average numbers, it would seem that Viet Nam has been able to develop a relatively well staffed health delivery system with considerable physical infrastructure, including hospitals and clinics at lower levels. While the country compares favorably in terms of per capita availability of health care facilities, there is significant variation in quality of services and in access and utilization of care by region. For example, while people in the most remote and mountainous areas have access to primary health care facilities, utilization in these regions is limited for a number of reasons, including poor quality of care, restrictions in physical access, financial obstacles to using services, and possibly cultural factors that circumscribe utilization. Furthermore, CHSs are perceived as only implementing target health programs even though their mandate is to provide also curative care services as part of primary health care. There is a concern with the level of utilization ofCHSs as shown by recent estimates of the average number of visits of around 12 consultations per day (World Bank 2001). Reasons for the relatively low rates of utilization at this level of care include the availability of alternative providers of care, staffing irregularities, and perceptions oflow quality of care. Table 4.2 presents a set of selected health service coverage indicators for Viet Nam and some health outcomes variables that are considered to depend on the healthcare system. 475 Good Practice in Health Financing, p. 476 Table 0.8: Selected health care coverage and health outcomes indicators, 1990-2004. Indicator 1990 1995 2000 2001 2002 2003 2004 Births attended by skilled health staff (% of total) n.a. n.a. 69.60 n.a. 85.00 n.a. 90.00 Immunization, OPT (% of children ages 12-23 months) 88.00 93.00 96.00 96.00 75.00 99.00 96.00 Improved water source (% of population with access) 72.00 n.a. n.a. n.a. 73.00 n.a. 80.00 Improved water source, rural 67.00 n.a. n.a. n.a. 67.00 n.a. n.a. (% of rural population with access) Improved water source, urban 93.00 n.a. n.a. n.a. 93.00 n.a. n.a. (% of urban population with access) Improved sanitation facilities 22.00 n.a. n.a. n.a. 41.00 n.a. 32.00 (% of population with access) Malnutrition prevalence, height for age n.a. n.a. 36.50 n.a. n.a. n.a. n.a. (% of children under 5) Malnutrition prevalence, weight for age n.a. n.a. 33.80 n.a. n.a. 28.40 n.a. (% of children under 5) Diarrhea treatment (% of children under 5 receiving n.a. n.a. n.a. n.a. 39.00 n.a. n.a. oral rehydration and continued feeding) Prevalence of HIV, total (% of population ages 15-49) n.a. n.a. n.a. 0.30 n.a. 0.40 n.a. Malaria prevention, use of insecticide-treated bed nets n.a. n.a. 15.80 n.a. n.a. n.a. n.a. (% of under-5 population) Incidence of tuberculosis (per 100,000 people) 202.17 n.a. n.a. n.a. n.a. n.a. 176.48 Tuberculosis cases detected under DOTS (%) n.a. 29.61 82.06 83.40 86.87 85.38 88.81 ARI prevalence (% of children under 5) n.a. n.a. 9.30 n.a. n.a. n.a. n.a. ARI treatment (% of children under 5 n.a. n.a. 60.40 n.a. 71.00 n.a. n.a. taken to a health provider) Contraceptive prevalence (% of women ages 15-49) n.a. n.a. n.a. n.a. 78.50 n.a. n.a. Source: WDI 2006. Among other things, the table indicates that immunization coverage in Viet Nam is relatively high. Since the mid-1980s coverage has increased from less than 55 percent to more than 95 percent coverage rate (World Bank: 2001). Although immunization rates may be particularly high, other indicators of health service coverage also display similar increases. These improvements have contributed to the achievements in bringing down morbidity and mortality rates in Viet Nam. Regulatory framework As noted above, the MOH is the key regulator in the health sector. The Ministry is responsible for the development of national policies and for monitoring and supervision of the national programs of public health care. An important policy document is the Health Sector Master Plan to the Year 2010 and Vision to the Year 2020 that has recently been developed by the Department of Planning and Finance of the MoH in collaboration with other sector departments. Among other things, the government envisions a 476 Good Practice in Health Financing, p. 477 continued emphasis on the grass-roots level while simultaneously developing more advanced medical technologies. In addition, the Ministry is responsible for the development and drafting of central government decisions that relate to the health sector and that are subsequently promulgated along with specific guidelines to the provinces for implementation. The private sector is regulated by the Ordinance on Private Practice of Medicine and Pharmacy first passed in 1993 and subsequently updated in 2003, and the decree detailing the provisions for its implementation. The Ordinance stipulates the procedures for licensing and regulates the professional requirements for practitioners applying for a private practice license. Specifically, the regulations specify ten different types of private health facilities, four types of private traditional medicine facilities, and four types of private pharmaceutical facilities. Licenses for private providers including private pharmacies are issued by the Provincial Health Bureaus after formal procedures of testing and screening applicants. The regulatory environment regarding health insurance is currently undergoing rapid transformations. Currently, health insurance is regulated according to Decree 63 passed in 2005, and the Viet Nam Social Security (VSS) has the overall responsibility of the implementation, including collection of contributions, fund management, and reimbursements to providers, of the various programs. With technical assistance provided by WHO, a law on health insurance is being developed (ILO 2006; WHO 2004) and it is expected to be passed by the National Assembly in 2007. As a reflection of the importance that health insurance is obtaining as a health financing instrument, a specific department for health insurance was recently formed within the MOH. The new department is formally charged with developing legislation and strategies for attaining universal health insurance coverage and is expected to take active part in the supervision and monitoring of the various programs that are being planned for and that are currently in place. Additional components of the health delivery system Together with the Ministry of Education, the MOH is responsible for the education and training of all health care staff at national and regional health training facilities. Viet Nam has several medical universities that train health personnel, including specialists in most medical areas. In line with its overall responsibility for the health sector, the MoH is also charged with the production of most health statistics and information. VSS is responsible for the collection and compilation of health insurance data, which is expected to be of greater importance as insurance coverage is provided to larger shares of the population. As was noted above, the government of Viet Nam is planning for the health sector to become more advanced, not least technically, to be able to provide services also for export purposes. Recent information suggest that central level hospitals are becoming increasingly more technically sophisticated and to date many relatively demanding medical procedures can be performed in Viet Nam. While the medical technology improvements will most likely play an important role in enhancing medical care quality, 477 Good Practice in Health Financing, p. 478 it is not clear the extent to which this process is beneficial for the achievement of the major public health goals of the country. For example, the main pediatrics hospital in the country's largest city can today perform highly advanced neo-natal care with the possibility of saving relatively early pre-term newborns. Access and utilization of such services is likely to be limited, however, for many poor households living in remote areas of the country, and for the sti11large number of mothers giving birth at home. The pharmaceutical sector has a prominent position within the Vietnamese health care system. For example, in 2003 half of all health spending is on medicines (MOH 2006) by a wide range of providers only partly under the systematic control and regular supervision of the relevant authorities. While legislation and institutions are in place to regulate the sale of pharmaceutical products, Viet Nam is experiencing serious problems with antibiotic resistance, partly as a result of irrational use of drugs due to poor supervision and monitoring. Formally, the responsibility of inspection and control of the pharmaceutical sector lies with the Drug Administration of Viet Nam (DAV) and the Department of Therapy of the MoH. An indication of the difficulty in enforcing a policy of rational use of drugs is the fact that self-medication (i.e. the buying of drugs at public or private drugs vendors without a prescription) is used in as much as 73 percent of all cases of reported illness (MoH 2003). 478 Good Practice in Health Financing, p. 479 Description of Health Coverage Reforms in Viet Nam Background and rationale for reforms in Viet Nam The health system in 1986, prior to the series of socio-economic reforms of Doi Moi, was described in great detail in a historical review of Viet Nam's Health System published in 2001 (MOH, 2001), on which this section strongly relies. The guiding principle of the health system pre-reform was "health for all". At the grassroots level, an extensive network of CHSs and inter-communal polyclinics provided primary health care to the population. Government hospitals were available to treat people requiring a higher level of care. Despite the wide coverage, the quality of health care suffered from a lack of resources. From a severe shortage of trained health workers at independence and very short-term basic training, the medical training of Vietnamese health workers improved as the system developed and especially as the number, size and scope of medical schools improved. Whereas in 1945 there were 180,000 persons per doctor; by 1985 the ratio had fallen to only 3,137. Nevertheless, low access to developments in medical science for training medical workers and low salaries were important causes of low quality care. Materials and medicines were distributed through the health service but were generally insufficient, especially in CHSs. Short supply of western drugs was a major constraint to quality of care in the pre-reform period. The health sector relied heavily on traditional medicine for treatment of common ailments and the limited modem drug distribution was tightly controlled. Supervision of health care services through inspections and supervisory visits from higher levels were infrequent and weak due to limited availability of vehicles, fuel shortages and absence of telephones in addition to technical limitations. During the pre-reform period, only 3-4 percent of the state budget was allocated to health care, insufficient to cover the recurrent costs of government health facilities. Few funds were available to upgrade health facilities or provide medical equipment. Only some 8 percent of the perceived need for medical equipment was met. The basic operating costs of facilities above the commune level were covered solely through the government budget. The financing of CHSs depended on revenue from the agricultural cooperatives. Foreign assistance from the former Soviet Union and other socialist countries dried up abruptly at the end of the 1980s. Because of the high dependency on foreign aid and substantial government budget deficit at the time, this had a severe negative impact on health financing in Viet Nam. Many health facilities lacked sufficient funds even to maintain routine activities or pay salaries. Much of the medical equipment was obsolete or broken. The already inadequate supply of drugs worsened at all levels. The government was no longer able to subsidize current expenditure for all health facilities and there was an urgent need to mobilize alternative sources of finance if the health sector was to function effectively. 479 Good Practice in Health Financing, p. 480 Chronology of reforms The major economic and social reforms began with the Sixth Communist Party Congress in 1986. The economy has gradually made the transition from a centrally planned to a socialist oriented market economy. International trade and exchange of information have increased. Dramatic economic growth and rapid improvements in living standards have been the result. However, gaps between the rich and poor and between better off and disadvantaged regions have increased both in terms of economic development and in terms of access to quality health care. A series of health system financing reforms have been put in place first, to overcome resource constraints of the health system, and more recently to attempt to overcome the gaps between the rich and the poor, the advantaged and disadvantaged regions in order to ensure health for all. This section describes the main health financing reforms since Doi Moi began in 1986. A general overview of the chronology of the main health sector reforms in Viet N am is provided in table 5.1. Table 0.9: Chronology of health reforms, 1989-2006 Year Reform measure 1989 • Fee for service • Private practice • Pharmaceutical sale 1992 • Health insurance • Legal document on private practice 1994 • Upgrading of the CHS system • Salary for the CHS staff • Health Policy Unit at MoH 1996 • National drug policy • Policy for social mobilization (Xa hoi hoa) of health care • National target programs in health 1997 • Provincial health system model 1998 • Hunger Eradication, Poverty Reduction Program: Health Cards for the Poor 1999 • Redefine goals of health sector reform 2003 • Health Care Funds for the Poor (HCFP) 2004 • Implement free health care for children under age 6 2005 • New health insurance regulations deepening coverage 2006 • Revision of Health Care Funds for the poor • Drafting health insurance law Source: Adapted from Adams (2005) The replacement of cooperative agriculture with family based production in the Doi Moi reforms meant that the commune people's committees lost revenue from which to finance social services. The real income of CHS workers declined, their salary payments could be delayed by months, and their morale plummeted. 480 Good Practice in Health Financing. p. 481 Starting in 1987, as part of the market oriented reforms, private health and pharmaceutical sectors were officially sanctioned. As a result, Viet Nam saw a rapid increase in private medical practice and retail pharmacy outlets, although primarily in better off areas and cities. Most private practices were very small scale and a large number were run by moonlighting government health workers. Greater freedom given to state pharmaceutical production and import companies and an increasing number of private pharmaceutical distributors and producers resulted in a rapid rise in the volume and diversity of pharmaceuticals on the market. As the service at CHSs deteriorated, attendance at these facilities fell to low levels. People, including in the rural areas, resorted increasingly to over the counter drug purchases or private consultations for basic medical care because of the greater convenience and user friendliness despite its greater expense. Also in 1987, in response to the dire state of many CHSs, the government decided that a defined number (typically three) ofCHS staffper commune should have their salaries subsidized from the provincial budget. However, by the early 1990s only a third of CHSs were being subsidized in this way and in 1994 the Prime Minister signed Decision 58/TTg which designated CHS staff as government workers and allocated funds for their salaries to be paid from provincial budgets. This salary support has greatly improved the income and morale of CHS workers and it rescued the public primary health care system of the country. Partial service fees and charges for drugs and diagnostic investigations were introduced officially in public hospitals and other health facilities in 1989. The revenue was retained by the collecting facilities and was to be used to improve the service delivered. Certain population groups (such as children under six years, the poor, people living in remote or mountainous areas) were to be granted partial or total exemption from charges by the individual health facility, but this policy was inadequately funded and proved difficult to implement effectively or fairly. As a result user charges have deterred people from seeking public health care, especially the rural poor and especially for care at more expensive, higher level hospitals. While there has been no official policy for CHSs to charge service fees, many of them in fact do levy a small fee and these facilities may earn substantial revenues from the sale of pharmaceuticals. Health insurance was introduced in Viet Nam in 1992 with a view to reducing the cost burden on the government health budget and increasing population access to health care. There was an initial surge in the numbers of people with compulsory, employment related coverage and voluntary student insurance. The goal of universal health insurance by the year 2010 to achieve equity of cover and increase the element of cost sharing of health service financing has been stated in official documents. However, attempts to extend voluntary health insurance to the general population have not met expectations and, in fact, have led to serious problems of adverse selection. In 1998 and again in 2005, health insurance regulations were revised, gradually eliminating differences in the benefit packages between types of insurance and increasing the package of services and drugs reimbursable through health insurance. Cost control measures put in place by the social 481 Good Practice in Health Financing, p. 482 security administration in charge of health insurance have forced some hospitals to under provide health services to ensure that their costs do not exceed their allocation of health insurance funds each year. Public health spending increased rapidly in real terms in the 1990s. In 1994 the MOH initiated a program to upgrade medical equipment in public hospitals. By 2000 central and provincial hospitals had most of the necessary equipment and of district hospitals 90 percent had X-ray machines, 70 percent had ultrasound machines and 92 percent had ambulances. The Communist party and government issued guidelines for the consolidation and strengthening of the basic rural health network. External assistance helped to finance the rehabilitation of CHSs and DHCs, focusing on disadvantaged areas. By 2002 98 percent of communes had a functioning CHS and the goal is for every CHS to have a doctor by 2010. Although the government continues to increase its contributions to the health sector, it is strongly pushing a policy of social mobilization (cost sharing) in order to mobilize additional funds for the health sector. An important component of this policy is the policy on semi-autonomous public hospitals laid out in Decree JOin 2002 and revised in 2006 with Decree 43. These facilities would move from state budget funding to a self- financing status with decentralized powers of management. The purpose is to create incentives and conditions for public hospitals to operate more efficiently. There is, however, a concern that this policy wi11lead to commercialization of the Vietnamese hospital service, with perverse incentives to overuse investigations, drugs and services to increase revenue and to concentrate on money making activities to the detriment of other essential, but less lucrative, aspects of health care. Household health care costs increased rapidly after the reforms and became a particularly heavy burden on poor rural households. Ensuring equity in access to, and utilization of, health care has always been a high priority of both the public health service and the Communist Party in Viet Nam. Therefore in the mid-1990s, a policy of providing government subsidized health insurance cards for the poor was put into effect to overcome the financial barriers to health care that resulted from the user fee policy. Insufficient funds meant that not all the poor were covered and the contribution per card was inadequate to cover the cost of health services. To overcome these weaknesses, Decision J39/2002/QD-TTg provided for all provinces and cities to establish a HCFP with a major subsidy from the central state budget. The target beneficiaries include all poor people, as well as residents of disadvantaged communes and ethnic minorities in disadvantaged provinces. The amount allocated to the fund per poor person was increased substantially and the provinces were allowed to use it either to purchase health insurance or to reimburse hospitals directly for health care services used by targeted beneficiaries. In 2005 the Law on Protection and Care of Children came into effect mandating free health care for children under six years of age. State budget funds were allocated and the policy is being implemented as a direct reimbursement scheme from the state budget to state facilities providing health care to children under age six. Work is beginning on a policy of health care for the elderly. 482 Good Practice in Health Financing. p. 483 Evaluation of health coverage refonns As a result of Viet Nam's impressive achievements in reducing poverty and improving health and other outcomes among the population, the country has attracted considerable interest from the international community, including health policy researchers and analysts. This interest has resulted in a significant amount of studies on several key outcomes also of relevance for this report. However, in tenns of rigorous quantitative analysis of the causal effects of the policy interventions, the evidence base is limited in scope with only a handful of studies that have used modem impact evaluation methods. This sub-section reviews the most important contributions to date of the impact of the health coverage refonns of Viet Nam. Health insurance in Viet Nam has been shown to have important impacts on health outcomes, access to health care and on out-of-pocket expenditures, in particular for low income individuals. Four recent studies have looked at the impact of the compulsory social health insurance program. First, Wagstaff and Pradhan (2005) applied quantitative techniques, including double-difference and matching, to look at the impact of the government's social health insurance program on four outcomes: health, utilization, out- of-pocket spending, and poverty (non-medical spending). Using panel data from the 1990s, they found that Viet Nam's social health insurance has had a positive impact on height for age and weight for age of young school children and body mass index (BMI) of adults. This study also found that for young children, health insurance increased use of primary care facilities and reduced self-medication. Among older children and adults, health insurance led to an increase in the use of hospital inpatient and outpatient care. Finally, the authors found that health insurance led to a reduction in annual out-of-pocket expenditures on health and an increase in non-medical household consumption. Second, using econometric modeling on the same data as the previous study, Sepehri et al. (2006) also found a positive effect of the compulsory social health insurance program on out-of-pocket spending. Specifically, after controlling for unobserved heterogeneity, social health insurance reduced out-of-pocket expenditures from 16 to 18 percent with the reduction more pronounced for low income individuals. Third, Wagstaff (2005) analyzes the effect on income, medical spending, and household consumption of a negative health shock. The author found, among other things, that this type of health insurance does provide financial protection against the adverse effects of a health shock. And finally, Chang and Trivedi (2003) also found a strong negative insurance effect on use of self-medication as the insured were significantly more likely to seek a medical consultation in case of illness. The application of these evaluation techniques on household survey data, including panel data, provides strong evidence of a positive and causal impact of this type of insurance on key policy outcomes. With regard to the voluntary component of the Vietnamese health insurance program, Jowett et al. (2003) use a special survey to collect individual level infonnation on health care use and spending in three provinces in Viet Nam in the late 1990s. Among other 483 Good Practice in Health Financing, p, 484 findings, the authors report a 200 percent reduction in the average out-of-pocket spending for outpatient care. As to the effect of voluntary health insurance on utilization, Jowett et al. (2004) found that, using the same data source and controlling for potential selection bias, voluntary health insurance in Viet Nam led to higher use of outpatient facilities and public providers away from self-treatment and private providers, effects that were particularly strong at lower income levels. The limited scope of the information basis in these two studies somewhat compromises the strength of the contribution to the evidence base, but nonetheless, they do suggest that also the voluntary component of the Vietnamese government health insurance system has had a positive impact on relevant outcomes. As discussed above, the Vietnamese government introduced special programs to provide health insurance coverage to several target groups in the late 1990s. The impact of these health card initiatives was evaluated by Nguyen (2003). The author applies first difference techniques coupled with propensity score matching on national household survey data from 2002 to analyze the impact on health care utilization and out-of-pocket spending. The results indicate, however, only very limited or no impact of the health care program on these outcomes. A recent report discusses several reasons for the apparent absence of an impact from these efforts to target the poor and other socially excluded groups, including incomplete implementation of the program, low levels of coverage, and finally, the need to improve the operation of the scheme so as to provide an effective protection mechanism (World Bank 2003a). To date there has been no formal evaluations of the policy of free health care to children under 6. A recent report, however, discusses this policy based on a case study of one central level children's hospital in Hanoi (UNDP 2005). Among other things, the report notes the increased patient load that the hospital has experienced as a result of the policy. While it is too early to know for sure, the study discusses the possible effect on quality of care that the increase in utilization may have. In addition, future evaluations of this policy might usefully look at the benefit incidence of utilization of child care that may change to the advantage of the urban rich as a result of the policy. Small studies designed to evaluate Decision 139 in particular provinces or regions have shown promising results, although they will require confirmation with a nation-wide study to see whether the impact is localized or widespread. Households interviewed in a WHO study in the provinces of Bac Giang and Hai Duong provinces reported increased utilization of health services after implementation of the HCFP, especially for inpatient care (Axelson et al. 2005). The study also found a significant increase in health care seeking at the CHS for the first contact among HCFP beneficiaries. While these preliminary findings of this important program are encouraging, it is safe to say that problems do exist also with the implementation of this program. Quality of primary health care, especially in remote areas remains poor. Health insurance contributions are too low to cover the cost of the now expanded package of services promised to the insured leading to dangerous deficits that threaten the solvency of the health insurance fund. Incentives for both public and private health care facilities to over 484 Good Practice in Health Financing, p. 485 prescribe medicines or expensive diagnostic services have led to a rapid escalation in health expenditures and waste. The lack of adequate supervision and quality control systems have exacerbated the overprovision of drugs and services and delayed further improvements in quality of care. Main conditions for success During the trial and error process of health financing policymaking in Viet Nam, certain conditions for success have become evident. Providers require an adequate level of funding in order to continue providing health services. Although the State continues to be a primary financing source for Government health services, the collection of user fees has helped to fill the gap left by inadequate state financing in order to maintain the existence and attempt to improve the quality of health care services. Inadequate reimbursement of services by health insurance has led to under provision of services or unofficial co- payments by the insured. Actually knowing what full costs are is a precondition to ensuring that they are fully covered by existing financing sources. Studies are only now being undertaken to find out the full cost of providing health services at state facilities. Implementation of assistance programs to the poor of other target groups requires both clear targeting and adequate financing. The initial user fee policy called for exemptions or reductions for the poor and children under age 6. This policy failed because the services provided were not adequately reimbursed to the facility and targeting was unclear. A new policy to ensure health care for the poor ensured a small amount of funding to purchase health insurance for the poorest of the poor, but did not provide clear criteria for targeting these people, and the contributions were inadequate to cover the cost of health care services. The policy for the poor and for children under age six has since further evolved so that a higher level of funding is available and the criteria for targeting are very clear. Monitoring and evaluation of impact of health financing policy implementation has been important in orienting revisions or development of policies to counteract the negative impacts. Findings from evaluations of the impact of user fees on access to care and the heavy burden of health expenditures on the poor was instrumental in convincing relevant government bodies of the importance of a policy to assist the poor. Administrative monitoring of this policy, although it was not comprehensive and did not rely on rigorous research process, was still adequate to show that earlier policies were not providing adequate coverage and provided evidence for the development of the HCFP to ensure coverage for all poor and several near poor groups. Assessment of remaining health system strengths and weaknesses This section has described the main health service and financing reforms that have been implemented over the past 10-15 years. Compared with many other countries in a similar economic situation it is evident that the health system is performing relatively well, particularly at the primary health care level, although regional variations exist. Historically, the emphasis on the grass-roots level with a wide organizational reach has 485 Good Practice in Health Financing, p. 486 always been strong and it is likely that it has contributed to the good health outcomes of the country. The government continues to envision a strong development of the sector with an emphasis on medical technology development both in terms of products and services. Moreover, the relatively limited financial inputs compared with other countries along with the favorable health outcomes would suggest that Viet Nam's health sector is relatively efficient. Nonetheless, many difficulties persist in the performance of the health service delivery system. One problem that has been observed is the relatively high reliance on hospital care. When looking at other countries with similar national incomes, Viet Nam ranks among the highest in terms of hospital beds per population only behind some of the countries of the former Soviet Union with its strong emphasis on hospital services (WHS 2006). Related to this is the inadequacy of the referral system which leads to relatively low levels of utilization at the lower levels and high rates of use at upper levels of care even for relatively simple conditions. Moreover, the rational use of care is hampered by some weaknesses in the monitoring system where health information data are not systematically analyzed for planning and management purposes. A third issue is the observed differences in access to and utilization of health services in Viet Nam. It is likely that these differences have increased in the recent past and may even continue to do so for some time to come. As is discussed elsewhere in this report, however, the government has introduced measures to address these inequalities in health, the most important of which are Health Care Funds for the Poor and the policy of free health care to children under 6 years of age. Finally, the problem of irrational use of drugs and antibiotic resistance is an area where the government will have to step up counteracting measures. The important factors of training of medical staff, effective supervision, and the purposeful enforcement of existing regulations are important in this respect. 486 Good Practice in Health Financing, p. 487 Health Coverage Reforms: Enabling factors and lessons from Viet Nam This report has described the main health coverage reforms in Viet Nam over the past two decades. During this period, the health sector has undergone substantial transformations in many regards, including the introduction of co-financing for health care, the private provision of services, and the gradual introduction of various types of pre-payment mechanisms, including social health insurance and free health care for certain target groups. The discussion has also shown that coverage rates have increased steadily both in terms of services and of health insurance. For example, today around 90 percent of all children receive vaccinations compared with only around 55 percent in the mid-eighties. Over the same period, almost 60 percent of the total population has been covered by some type of health insurance. Notwithstanding these successes, many problems remain as illustrated by wide spread child malnutrition, substantial inequalities in access to and utilization of health services, and the existence of catastrophic out-of- pocket payments for care. Enabling factors The relative success of the Vietnamese government in extending health coverage to an increasingly larger share of the population, both in terms of services and with regard to health insurance, can be attributed to many different factors both inside and outside of the health sector itself. With respect to the most important such enabling factors outside of the health sector, at least three are of particular importance. The first of these external factors is the rapid and sustained economic growth over the past 15 to 20 years that has enabled the government to embark upon a series of ambitious health financing reforms and lately, to adopt the goal of universal health insurance by 2010. While health spending as a share of GDP has remained largely constant over this period, per capita spending on health has increased from less than USD 15 to well over USD 25. Moreover, the growth in national income has enabled the government to significantly increase its share of health spending, which has meant that substantial additional resources are now being channeled to the new entitlements of free health care for the poor and for children. The importance of economic growth for sustainable financial support to government provision of health insurance is all the more evident from the fact that Viet Nam has received relatively limited amounts of development assistance in health during this period; on average less than 2 percent of total health spending has come from foreign assistance, compared with the much larger contributions seen in other low-income countries. Without economic growth rates of 6-7 percent per year, these reforms would not have been able to be implemented. Moreover, economic growth in Viet Nam has been truly inclusive in that widespread poverty has been broadly reduced. Given the well established importance of poverty reduction for improved health (and vice versa), this is a critical component of the Vietnamese experience. One final result of the sustained economic growth rates and the limited dependence on development assistance for health is the fact that Viet Nam has been able to develop and 487 Good Practice in Health Financing, p. 488 pursue policies that have been genuinely domestic in their scope and aims. This relative level of independence in policy making is not unique for the health sector but can be seen in the country's overall development approach, and, among other things, this has most likely meant that ownership of programs has been high and the incentive and ability to correct faulty implementation has been present. The second enabling factor external to the health sector is the relatively well structured and organized Vietnamese society. The political structure of the country, with comparatively effective - at least nominally - channels of communication from the central level down to the community level has enabled the government to implement policies at the national level in a comparatively uniform manner and in a timely fashion. Section 4 above, described the organization of the health delivery system, which, among other things, encompasses a clear structure from the village level up to the tertiary hospitals at the central level. In combination with the political stability of the country in the post- American war period, this fact has most likely contributed to the ability of the government to implement many of its policy programs. Moreover, the relatively structured organization of the Vietnamese society will probably be an important favorable asset when the country has managed to computerize its public management system, including the health information system, and enable the government to strengthen its monitoring and evaluation system. A more effective information system will enable the authorities to implement even more effectively the special anti-poverty programs that target certain groups in difficult socioeconomic circumstances. Although recent reports have emphasized the relatively limited amounts of leakage and under coverage in these programs, the scope for further improvements should be evident from the eventual computerization of the health and other information systems. One final enabling factor that can be seen as external to the health sector as such is the high esteem with which health has been held by the Vietnamese people and the government. For example, a significant share of the population (around one third according to the national health survey of 2002) does some kind of regular exercise to keep in good health. And the leadership has always emphasized the need to focus on public health as an important instrument to reduce widespread poverty in the country. These socio-cultural factors in combination with nutritional and healthy food consumption habits most likely also contribute to the relatively good health outcomes in Viet Nam compared with most other countries with which it can be compared, and, of additional importance, it can be reasonably assumed that these factors have contributed to facilitating the uptake of the health coverage reforms as a general understanding of the importance of health for wellbeing has been present among the population. In addition to these external factors, there are at least three elements within the health sector that can help in explaining the success of Viet Nam to provide its population with health service and financial coverage over the past 20 years. The first of these is the historic inheritance of a wide network of primary health care services. The grassroots level of the health care system has long been seen as the most important part of an effective and equitable service delivery system. In the course of the recent reform period, 488 Good Practice in Health Financing, p. 489 the government has built strategically on these past investments to further develop their efficiency in providing the population with needed services and interventions. Moreover, a clear policy goal of the government has long been to provide the necessary prevention and disease eradication interventions. The strategic approach of providing effective preventive care for specific infectious diseases has led to a decline in the burden of disease from those conditions and consequently a relative increase in the disease burden towards conditions which cannot be easily prevented by traditional means. The second important factor within the health sector to explain the relative success of the country's health coverage reform program is the timing, scope, and sequencing of the identified measures. In terms of service provision, the government identified a set of priority health problems and diseases that it wished to focus on in its attempts to diminish the burden of disease by eradicating or reducing these specific diseases. Although other problems may have existed, prioritizing was seen as a necessary means by which scarce resources could be put to the best use. Furthermore, the priority disease interventions have been allowed to adapt to the gradually changing epidemiological pattern of the country, as can be seen, for example, by the inclusion of HIVIA IDS prevention in the national disease priority program and the issuing of policies on tobacco control and accident and injury prevention. Finally, given the fact that providing effective and equitable health care requires, besides skilled health professionals and clinicians, extensive planning and management capacities, the Vietnamese health sector is relatively blessed with such skills. Within all levels of the system, a formal structure delineates the management responsibilities of the particular type of provider, one result of which is the production of clear operational plans for the coming activity period. The weakness of the system can be seen in the implementation and overview of these plans as monitoring is only patchily carried out. Related to the health planning and management skills of the health system is the capacity to develop useful health policies. A strategic effort on behalf of the central health authorities is to develop skills and capacities in health policy making, including analysis of the efficiency and distributional impact of the policies and programs. Lessons from health coverage reforms in Viet Nam While many of the health coverage reforms that Viet Nam has introduced can be implemented also in other countries, some of the enabling factors may not be so easy to replicate. For example, the introduction of special funds to finance the health care needs of certain groups can theoretically be made also in other contexts. The long-term sustainability of such programs, however, presumably requires relatively high rates of economic growth for extended periods of time and expected rapid declines in the size of beneficiary groups. With regard to political economy factors, an important dimension of the Vietnamese experience has probably been the high level economic and political stability. The one- party structure of the Vietnamese governance system has enabled the country to develop policies that have been allowed to be implemented over a long period of time without due 489 Good Practice in Health Financing, p. 490 interference from varying and sometimes competing political fractions. The policy lesson from this factor is the need to reach agreements on health policy reform measures that span the political divide in multi-party democracies. There is otherwise a risk that health reform measures are not allowed to run their course over a sufficiently long period of time to take effect. In terms of good value for money, the importance that Viet Nam has put on the primary health care level probably serves as an important lesson to other low- and middle-income countries that are currently introducing health coverage reforms. As has been noted in this report, the grass-roots level has served as a backbone of the whole health care system and enabled the health sector to gradually introduce ever more advanced organizational and medical procedures, such as provider payment reforms and liberalizations of the health care production markets. Low-income countries need to develop a national capacity to develop, implement, and evaluate health coverage programs and other policies. Moreover, countries should attempt using such capacities most effectively by pooling analytical resources and thereby creating a critical mass for effective health policy analysis. Lessons from Viet Nam in this respect also include the strategic use of development assistance not to become too dominating. Generally, the country has used its development assistance in health to develop health policy and other core competences and to implement the national disease priority programs. In terms of health insurance coverage, the experiences of Viet Nam may be of interest to other countries by looking at the design of the programs and the sequencing of their implementation. While universal coverage may be the long-term goal of the health system, countries should not attempt at reaching this goal too quickly or by means of only one or two alternatives. Viet Nam initiated its health financing coverage attempts by introducing social health insurance in the early 1990s. The eligible groups were well defined and the benefit package in terms of services was identified, although in the case of Viet Nam it may be too broad. Gradually, along with the introduction of other programs, including voluntary health insurance and free health cards for the poor, more people were covered by pre-payment. Importantly, the benefit package of the insurance programs that have been implemented most recently was defined to be in accordance with the initial program, thereby creating a certain level of transparency in the health insurance system This pattern is yet again repeated by the introduction of the Health Care Funds for the Poor as of2003. Creating a uniform benefit package in this way limits the scope for overlap within the health insurance system, which in tum reduces transaction costs. Should the government chose to liberalize also the market for health insurance by allowing private for-profit agents to enter the market, this last point may be even more important as private actors would then compete with existing insurance programs and from the perspective of the client, a relatively uniform and generous benefit package would facilitate the choice between alternative programs and prevent private firms to skim the market. 490 Good Practice in Health Financing. p. 491 Given that countries will attempt to achieve universal health coverage by means of a whole range of alternative health insurance options due to resource limitations, one important lesson from Viet Nam is the ability to target scarce public resource to those most in need. Recent analysis has shown that compared with many other countries, Viet Nam has been able to develop relatively effective target mechanisms with limited rates of leakage or undercoverage (World Bank, 2003a). While it has also been shown that these achievements are the result of a pragmatic approach of local authorities, the lesson from Viet Nam would be that the successful implementation of target programs requires a certain level of decentralization in the operation of these programs. For administrative cost-effectiveness it was also important that the health sector could latch on to the targeting being done by the Hunger Eradication and Poverty reduction program and didn't have to fund the heavy administrative work of doing its own targeting. Finally, the case of Viet Nam also points at the importance of introducing health coverage reforms in a gradual and incremental manner. While the most difficult challenges of increasing coverage rates for the rural poor remain, the fact that Viet Nam introduced social health insurance for the formally employed first, followed by a concerted effort to cover the poor and socially excluded, most likely provided the authorities with important opportunities of learning by doing when undertaking health coverage reforms. The importance of having good information and knowledge about the health care seeking behaviors of different target groups, such as children, the rural poor, and women, cannot be sufficiently emphasized. 491 Good Practice in Health Financing, p. 492 References Adams, S.. 2005. Viet Nam's Health Care System: A Macroeconomic Perspective, Paper Prepared for the International Symposium on Health Care Systems in Asia at Hitotsubashi University, Tokyo, January 21-22, IMF, Hanoi. Aka1, A. .2004. A comprehensive review of health insurance and financing regulations: FAQs, WHO, Hanoi .September. Baulch, B. et al. .2002. Ethnic Minority Development in Viet Nam: A socioeconomic perspective, WPS 2836, The World Bank, Washington, D.C .. May. Chang F.R. and Trivedi P.K., .2003. Economics of Self-medication: Theory and Evidence in Health Economics, 12.9.:721-39 Chien T.T.T .. ed . .2003., Viet Nam Health Report 2002, Hanoi: Medical Publishing House Chien T.T.T .. ed.. 2005., Health care for the poor in Viet Nam, Hanoi: Medical Publishing House. Die1eman, M, Cuong PV, Anh LV and Martineau T .2003., Identifying factors for job motivation of rural health workers in North Viet Nam, in Human Resources for Health. 2003, pp. 1: 10 .http://www.human-resources-health.comlcontent/1/1/10. Dong, P.T., et al. .eds . .2002. User fee, health insurance and utilization of health services, Hanoi: MOH- Central Commission for Science and Education and Viet Nam- Sweden Health Cooperation Programme G1ewwe, P., Nisha Agrawal, and David Dollar, eds .. 2004. Economic Growth, Poverty, and Household Welfare in Viet Nam, World Bank, Washington, D.C. Government of Viet Nam and World Bank .2005. Viet Nam-Management of Public finances for growth and poverty reduction .. Pub1ic Expenditure Review .. Vol. I and II. Hanoi: Finance Publishing House. Government of Viet Nam .2003. The Comprehensive Poverty Reduction and Growth Strategy .CPRGS./Poverty Reduction Strategy Paper .PRSP., GoV/IMFlWorld Bank, Hanoi. GSO .Genera1 Statistical Office .. various years. Viet Nam Household Living Standards Survey, Hanoi. Gwatkin, D, Adam Wagstaff, and Abdo Yazbeck .2005. Reaching the Poor with Health, Nutrition, and Population Services: What Works, What Doesn't, and Why, World Bank, Washington, D.C. 492 Good Practice in Health Financing, p. 493 Health Strategy and Policy Institute and World Health Organization .2004. Assessment of Decision 139 implementation. Huong, P.B. et al. .2006. Alternative approaches to ensure health care for the rural poor: lessons for Viet Nam from the Chinese experience, Hanoi. ILO .2006. Enhancing social protection in Viet Nam-Developing the new social security law, Hanoi. IMF .2005. Viet Nam: 2004 Article IV Consultation-Staff Report, IMF, Washington, D.C .. May. Jowett M, A. Deolalikar, and Peter Martinsson .2004. Health insurance and treatment seeking behaviour: Evidence from a low-income country, Health Economics 13:.845-857. Knowles, James c., Ha, N. T. H. et al. .2003. Making Health Care More Affordable for the Poor: Health Financing in Viet Nam, Volume 1: Main Report, MoH/ADB TA- 3877-VIE, Hanoi. Knowles, James C., Ha, N. T. H. et al. .2003. Making Health Care More Affordable for the Poor: Health Financing in Viet Nam, Volume 2: Annexes, MoH/ADB TA-3877-VIE, Hanoi. Knowles, James c., Ha, N. T. H. et al. .2005. Making Health Care More Affordable for the Poor: Health Financing in Viet Nam, Medical Publishing House, Hanoi. Lieu, D.H., Nguyen H. Long, and Sarah Bales.2005. Study on treatment costs for selected disease groups in provincial general hospitals. Hanoi: Medical Publishing House, forthcoming. Lieu, D.H., N.Q. An, and Nguyen H. Long .2005. Study on health care financing from a household and user perspective. Hanoi: Medical Publishing House forthcoming. Minh, H.V., N. Khang, and N.V. Dung .2004. Health Care Financing for the Poor in Viet Nam, Working Paper No. 2004-8, World Bank. Ministry of Finance .2005. Medium Term Expenditure Framework, Hanoi. MOH .2001. 55 years of developing Viet Nam's revolutionary health system [55 nam phat trien su nghiep y te cach mang Viet Nam] Hanoi: Medical Publishing House. MOH .2003. VNHS 2001102: Main report, MoH, Hanoi. MOH .2004a. Assessment of 1-year implementation of Decision 139, Hanoi. 493 Good Practice in Health Financing, p. 494 MOH .2004b. NHA- the process of implementation and development in Viet Nam. Hanoi: Statistical Publishing House. MOH .2005. Study on treatment costs for selected disease groups at provincial general hospitals, VSHCIHPC, Hanoi .July. MOH .2006. NHA 1998-2003. Hanoi: Statistical Publishing House. Ministry of Planning and Investment .2005. The five-year socio-economic development plan 2006-2010, MoPI, Hanoi .September. Nguyen, V.c. .2003. Assessing the coverage and impact of Viet Nam's programs for targeted transfers to the poor using the Viet Nam Household Living Standard survey 2002, the World Bank, Hanoi .June. Sepehri A, S. Sarma, and W. Simpson .2006. Does non-profit health insurance reduce financial burden?- Evidence from the Viet Nam Living Standards Survey Panel, Health Economics 15:.603-616. Sepehri A, R. Chernomas, and H. Akram-Lodhi .2005. Penalizing patients and rewarding providers: user charges and health care utilization in Viet Nam, Health Policy and Planning 20.2.: 90-99. UNDP .2005. Preliminary Assessment of free health care for under six-year-old children policy at Central Pediatric Hospital, Report to UNDP workshop on Financial Policy Analysis .Hanoi. Valdelin, Jan et al. .2005. Health cooperation at the crossroads: More of the same or making a difference?, Evaluation of the VSHC 2001-2005, Hanoi .November. Van de Walle, D .2004. Testing Viet Nam's public safety net, Journal of Comparative Economics, 32, 661-679. Van de Walle, D and D. Gunewardena, .2000. Sources of Ethnic Inequality in Viet Nam, WPS 2297, The World Bank, Washington, D.C. .March. Wagstaff, Adam .2005. The Economic Consequences of Health Shocks. World Bank Policy Research Paper No. 3655, June. Wagstaff, Adam, and Eddy van Doorslaer .2003. Catastrophe and impoverishment in paying for health care: with applications to Viet Nam 1993-1998, Health Economics 12: 921-934. Wagstaff, Adam, and M Pradhan, .2005. Health Insurance Impacts on Health and Nonmedical Consumption in a Developing Country, World Bank, Washington, D.C. 494 Good Practice in Health Financing, p. 495 World Bank: .2001. Growing Healthy: A Review of Viet Nam's Health Sector, World Bank:, Hanoi. World Bank: .2003a. Viet Nam Development Report 2004: Poverty, World Bank:, Hanoi. World Bank: .2003b. Ninh Thuan PPA, Poverty Task Force, the World Bank:, Hanoi. World Bank: .2005. Viet Nam: Managing public expenditure for poverty reduction and growth - PER and IF A, the World Bank:, Hanoi .March. World Bank: .2004. Viet Nam Development Report 2005: Governance, World Bank:, Hanoi World Bank: .2006. "Good Practice" in Expanding Health Care Coverage: Lessons from Reforms in Low- and Middle-Income Countries: Concept Note, HNP Unit, HDN, The World Bank:, Washington, D.C. WHO (World Health Organization) .2003. Health and ethnic minorities in Viet Nam, WHO, Hanoi .June. WHO (World Health Organization) .2006. World Health Report 2006: Working together for health, World Health Organization, Geneva. 495