VOLUME 7 DISEASE CONTROL PRIORITIES • THIRD EDITION Injury Prevention and Environmental Health DISEASE CONTROL PRIORITIES • THIRD EDITION Series Editors Dean T. Jamison Rachel Nugent Hellen Gelband Susan Horton Prabhat Jha Ramanan Laxminarayan Charles N. Mock Volumes in the Series Essential Surgery Reproductive, Maternal, Newborn, and Child Health Cancer Mental, Neurological, and Substance Use Disorders Cardiovascular, Respiratory, and Related Disorders Major Infectious Diseases Injury Prevention and Environmental Health Child and Adolescent Health and Development Disease Control Priorities: Improving Health and Reducing Poverty DISEASE CONTROL PRIORITIES Budgets constrain choices. Policy analysis helps decision makers achieve the greatest value from limited available resources. In 1993, the World Bank published Disease Control Priorities in Developing Countries (DCP1), an attempt to systematically assess the cost-effectiveness (value for money) of interventions that would address the major sources of disease burden in low- and middle-income countries. The World Bank’s 1993 World Development Report on health drew heavily on DCP1’s findings to conclude that specific interventions against noncommunicable diseases were cost-effective, even in environments in which substantial burdens of infection and undernutrition persisted. DCP2, published in 2006, updated and extended DCP1 in several aspects, including explicit consideration of the implications for health systems of expanded intervention coverage. One way that health systems expand intervention coverage is through selected platforms that deliver interventions that require similar logistics but deliver interventions from different packages of conceptually related interventions, for example, against car- diovascular disease. Platforms often provide a more natural unit for investment than do individual interventions. Analysis of the costs of packages and platforms—and of the health improvements they can generate in given epidemiological environments—can help to guide health system investments and development. DCP3 differs importantly from DCP1 and DCP2 by extending and consolidating the con- cepts of platforms and packages and by offering explicit consideration of the financial risk protection objective of health systems. In populations lacking access to health insurance or prepaid care, medical expenses that are high relative to income can be impoverishing. Where incomes are low, seemingly inexpensive medical procedures can have catastrophic financial effects. DCP3 offers an approach to explicitly include financial protection as well as the distribution across income groups of financial and health outcomes resulting from policies (for example, public finance) to increase intervention uptake. The task in all of the DCP volumes has been to combine the available science about interventions implemented in very specific locales and under very specific conditions with informed judgment to reach reasonable conclusions about the impact of intervention mixes in diverse environments. DCP3 ’s broad aim is to delineate essential intervention packages and their related delivery platforms to assist decision makers in allocating often tightly constrained budgets so that health system objectives are maximally achieved. DCP3 ’s nine volumes are being published in 2015, 2016, 2017, and 2018 in an environment in which serious discussion continues about quantifying the sustainable development goal (SDG) for health. DCP3 ’s analyses are well-placed to assist in choosing the means to attain the health SDG and assessing the related costs. Only when these volumes, and the analytic efforts on which they are based, are completed will we be able to explore SDG-related and other broad policy conclusions and generalizations. The final DCP3 volume will report those conclusions. Each individual volume will provide valuable, specific policy analyses on the full range of interventions, packages, and policies relevant to its health topic. More than 500 individuals and multiple institutions have contributed to DCP3. We convey our acknowledgments elsewhere in this volume. Here we express our particular gratitude to the Bill & Melinda Gates Foundation for its sustained financial support, to the InterAcademy Medical Panel (and its U.S. affiliate, the Institute of Medicine of the National Academy of Sciences), and to World Bank Publications. Each played a critical role in this effort. Dean T. Jamison Rachel Nugent Hellen Gelband Susan Horton Prabhat Jha Ramanan Laxminarayan Charles N. Mock VOLUME 7 DISEASE CONTROL PRIORITIES • THIRD EDITION Injury Prevention and Environmental Health EDITORS Charles N. Mock Rachel Nugent Olive Kobusingye Kirk R. Smith © 2017 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 20 19 18 17 This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://creativecommons.org /licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Mock, C. N., R. Nugent, O. Kobusingye, and K. R. Smith, editors. 2017. Injury Prevention and Environmental Health. Disease Control Priorities (third edition), Volume 7. Washington, DC: World Bank. doi:10.1596/978-1-4648-0522-6. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. ISBNs and DOIs: Softcover: Hardcover: ISBN: 978-1-4648-0522-6 ISBN: 978-1-4648-0521-9 ISBN (electronic): 978-1-4648- 0523-3 DOI:10.1596/978-1-4648-0522-6 DOI:10.1596/978-1-4648-0521-9 Cover photo: © Romana Manpreet | Global Alliance for Clean Cookstoves. Used with the permission of Romana Manpreet | Global Alliance for Clean Cookstoves. Further permission required for reuse. Cover and interior design: Debra Naylor, Naylor Design, Washington, DC Chapter opener photos: chapter 1: © United Nations Photo. Used via a Creative Commons license (https://creativecommons.org /licenses/by-nc-sa/2.0/); chapter 2: © United Nations Photo. Used via a Creative Commons license (https://creativecommons .org/licenses/by-nc-sa/2.0/); chapter 3: © World Bank. Further permission required for reuse; chapter 4: © Harsha K R. Used via a Creative Commons license (https://creativecommons.org/licenses/by-nc-sa/2.0/); chapter 5: © Trocaire. Used via a Creative Commons license (https://creativecommons.org/licenses/by-nc-sa/2.0/); chapter 6: © UN/Logan Abassi. Used with the permission of the UN. Further permission required for reuse; chapter 7: © 2003 Nigel Bruce, Courtesy of Photoshare. Used with permission. Further permission required for reuse; chapter 8: © Mohri UN-CECAR. Used via a Creative Commons license (https://creativecommons.org/licenses/by-nc-sa/2.0/); chapter 9: © DFID. Used via a Creative Commons license (https:// creativecommons.org/licenses/by-nc-sa/2.0/); chapter 10: © Carsten ten Brink. Used via a Creative Commons license (https:// creativecommons.org/licenses/by-nc-sa/2.0/); chapter 11: © 2009 Edson E. Whitney, Courtesy of Photoshare. Used with permission. Further permission required for reuse; chapter 12: © World Bank. Further permission required for reuse; chapter 13: © Land Rover Our Planet. Used via a Creative Commons license (https://creativecommons.org/licenses/by-nc-sa/2.0/). Library of Congress Cataloging-in-Publication Data has been requested. Contents Foreword xi Preface xiii Abbreviations xv 1. Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 1 Charles N. Mock, Kirk R. Smith, Olive Kobusingye, Rachel Nugent, Safa Abdalla, Rajeev B. Ahuja, Spenser S. Apramian, Abdulgafoor M. Bachani, Mark A. Bellis, Alexander Butchart, Linda F. Cantley, Claire Chase, Mark R. Cullen, Nazila Dabestani, Kristie L. Ebi, Xiagming Fang, G. Gururaj, Sarath Guttikunda, Jeremy J. Hess, Connie Hoe, Guy Hutton, Adnan A. Hyder, Rebecca Ivers, Dean T. Jamison, Puja Jawahar, Lisa Keay, Carol Levin, Jiawen Liao, David Mackie, Kabir Malik, David Meddings, Nam Phuong Nguyen, Robyn Norton, Zachary Olson, Ian Partridge, Margie Peden, Ajay Pillarisetti, Fazlur Rahman, Mark L. Rosenberg, John A. Staples, Stéphane Verguet, Catherine L. Ward, and David A. Watkins 2. Trends in Morbidity and Mortality Attributable to Injuries and Selected Environmental Hazards 25 David A. Watkins, Nazila Dabestani, Charles N. Mock, Mark R. Cullen, Kirk R. Smith, and Rachel Nugent 3. Road Traffic Injuries 35 Abdulgafoor M. Bachani, Margie Peden, G. Gururaj, Robyn Norton, and Adnan A. Hyder 4. Nontransport Unintentional Injuries 55 Robyn Norton, Rajeev B. Ahuja, Connie Hoe, Adnan A. Hyder, Rebecca Ivers, Lisa Keay, David Mackie, David Meddings, and Fazlur Rahman 5. Interpersonal Violence: Global Impact and Paths to Prevention 71 James A. Mercy, Susan D. Hillis, Alexander Butchart, Mark A. Bellis, Catherine L. Ward, Xiangming Fang, and Mark L. Rosenberg ix 6. Occupation and Risk for Injuries 97 Safa Abdalla, Spenser S. Apramian, Linda F. Cantley, and Mark R. Cullen 7. Household Air Pollution from Solid Cookfuels and Its Effects on Health 133 Kirk R. Smith and Ajay Pillarisetti 8. Health Risks and Costs of Climate Variability and Change 153 Kristie L. Ebi, Jeremy J. Hess, and Paul Watkiss 9. Water Supply, Sanitation, and Hygiene 171 Guy Hutton and Claire Chase 10. Interventions to Prevent Injuries and Reduce Environmental and Occupational Hazards: A Review of Economic Evaluations from Low- and Middle-Income Countries 199 David A. Watkins, Nazila Dabestani, Rachel Nugent, and Carol Levin 11. Helmet Regulation in Vietnam: Impact on Health, Equity, and Medical Impoverishment 213 Zachary Olson, John A. Staples, Charles N. Mock, Nam Phuong Nguyen, Abdulgafoor M. Bachani, Rachel Nugent, and Stéphane Verguet 12. Household Energy Interventions and Health and Finances in Haryana, India: An Extended Cost-Effectiveness Analysis 223 Ajay Pillarisetti, Dean T. Jamison, and Kirk R. Smith 13. Costs and Benefits of Installing Flue-Gas Desulfurization Units at Coal-Fired Power Plants in India 239 Maureen L. Cropper, Sarath Guttikunda, Puja Jawahar, Kabir Malik, and Ian Partridge DCP3 Series Acknowledgments 249 Volume and Series Editors 251 Contributors 253 Advisory Committee to the Editors 255 Reviewers 257 Policy Forum Participants 259 x Contents Foreword The world continues to suffer from an enormous burden health measures is disappointingly low. Countries, partic- of morbidity, disability, and premature mortality from ularly LMICs, need to do more. Policy makers should injuries and environmental health conditions. Much of seriously consider the recommendations of this volume this burden is unnecessary and can be prevented by when they develop their own essential package of inter- evidence-based, high-impact interventions that can be ventions, which is one of the three key pathways to achieve implemented in all countries, irrespective of income. universal health coverage. International organizations and Injuries are leading causes of death, responsible for development agencies should increase their support to an estimated 5 million deaths and around 9 percent low-income countries to make this possible. of global mortality. Most of the deaths are in low- This volume focuses on another key challenge to public and middle-income countries (LMICs). More than health across the globe. Environmental causes lead to 1.2 million people die too young each year because of more than 8 million deaths per year; outdoor and indoor road traffic injuries. According to the World Health air pollution accounts for more than 5 million of these Organization’s (WHO) Global Status Report 2015, death deaths per year. Climate change, which results from unsus- rates in low-income countries are more than double tainable policies in many sectors, exacerbates air pollution those in high-income countries. The African region has threats and causes additional morbidity and mortality. the highest death rates. Unsafe sanitation and lack of safe water and hygiene cause The current situation presents a major challenge to an estimated 1.4 million deaths, almost all in LMICs. socioeconomic development and has rightly been the Although the enormous burden and serious health focus of attention globally and within all countries. In challenges caused by environmental risk factors are 2010, a decade for action on road safety was established evident, surveillance and systematic monitoring of by the United Nations General Assembly; more recently, the magnitude of the risks and their health impacts in 2015, as part of the Sustainable Development Goals, are severely limited in many countries. The data gaps countries of the world made an ambitious commitment are particularly serious for air quality and air pollution to halving the number of global deaths and injuries due levels. Increasing awareness among high-level govern- to road traffic crashes by 2020. ment officials and policy makers on the seriousness of There is, therefore, a pressing need for action. This environmental risks and the pressing need to take effec- volume of Disease Control Priorities, third edition pro- tive multisectoral action should be a key priority for all vides an excellent evidence-based guide to policy makers countries. This volume makes a strong case for advocacy on the approaches and rational choice of interventions to and for strengthening commitment, and it provides clear address this challenge. Many of the interventions included policy advice on strategic directions to consider. in the volume are among the most cost-effective interven- Despite the fact that there are limited economic anal- tions in public health and can make a substantial impact ysis studies on some environmental health conditions, on reducing the health and socioeconomic burden due to there is clear evidence for a range of interventions recom- injuries, particularly in LMICs. Yet, current progress is too mended in the volume that are cost-effective or supported slow. As highlighted in this volume and documented in by favorable cost-benefit ratios. Implementing them will the Global Status Report, implementation of the key public have a considerable impact on reducing environmental xi health risks and preventing a broad range of com- Finally, making a difference in addressing injuries and mon communicable and noncommunicable conditions environmental health risks will require solid commit- responsible for a major proportion of global disease bur- ments from all parts of governments, particularly sectors den and premature mortality. In addition to their desir- such as transport, energy, industry, agriculture, housing, able health effects, many of these interventions will also and waste management. The health sector will have to have important non-health outcomes that may be part of demonstrate leadership in evidence-based advocacy, the priorities of the non-health sectors involved. governance, technical support, and surveillance. The An important challenge to health and development need to act is highlighted by a range of goals and targets is climate change. Although the importance of climate included in the sustainable development agenda. change challenges is increasingly recognized, and tack- Countries that initiate effective action now will benefit ling them is becoming a global priority, addressing the from improved health and realize considerable health health consequences is not receiving adequate attention. care savings. Policy makers at the highest levels of government and in different sectors need to be made fully aware of the Ala Alwan, MD, FRCP, FFPH seriousness of the health dimensions and the effective Regional Director Emeritus, approaches to mitigate them. WHO Eastern Mediterranean Region xii Foreword Preface The fields of injury prevention and environmental This volume looks at several types of policy approaches health address diverse health problems that arise from to reduce the burden of ill health from environmental and exposure to outside forces, such as chemicals and other occupational risks and injuries. Unlike the other DCP3 toxins, infectious agents, kinetic energy, and thermal volumes, most of the actions proposed in this volume energy. The health problems addressed by these fields speak directly to non-health sectors, where a substantial include the following: portion of disease and injury prevention policies and programs needs to occur. These actions include fiscal and • Unintentional injuries, such as road traffic accidents, intersectoral policies, such as taxes and subsidies; regu- falls, burns, and drowning lations; and policies that affect infrastructure, the built • Intentional injuries, such as interpersonal violence environment, and product design. Also included are infor- • Diseases, such as those caused or aggravated by expo- mation, education, and communication initiatives to pro- sure to airborne and waterborne pollutants mote behavioral changes; these initiatives leverage a range • Occupational hazards, such as injuries and diseases of vehicles, from mass media campaigns to one-on-one caused or aggravated by workplace toxins, counseling. A second major difference from other DCP3 • Effects of climate change due to human greenhouse volumes is that the economic evidence supporting the emissions, such as enhancement of waterborne infec- actions described in this volume is primarily benefit-cost tious diseases. analyses—the benefits and costs may occur outside of the health sector and must be accounted for in common mon- The conditions and risks encompassed in these fields etized units. We include two extended cost-effectiveness account for more than 12 million deaths per year— analyses (ECEAs) that indicate policies that provide strong 21 percent—of the annual global total of 56 million deaths. financial risk protection for individuals and households. Most of these conditions and risks have not been Most of the policies and interventions discussed effectively addressed globally; in recent years, however, in this volume have not been fully implemented in some have received increased attention. This Injury high-income countries (HICs); their implementation in Prevention and Environmental Health volume of Disease low- and middle-income countries (LMICs) is substan- Control Priorities, third edition (DCP3), contributes to tially more incomplete. More complete implementation the understanding of how to address these health prob- would help to reduce the disproportionately high rates lems in the following ways: of death and disability from these conditions in LMICs. Doing so could avert over seven million premature • Elucidating the health burden of these conditions deaths annually from environmental and occupational • Documenting trends in the health burden at different exposures and injuries in LMICs. phases of national development The goal of the editors and authors of Injury • Identifying the most cost-effective and cost-beneficial Prevention and Environmental Health is to provide the interventions requisite evidence-based rationale and guidance to • Describing policies and platforms that can widely and increase implementation of effective strategies to prevent effectively deliver these interventions. injuries and lower environmental risks in countries at all xiii economic levels. We hope to stimulate increased imple- We thank the following individuals who provided mentation of proven effective strategies that have not yet valuable comments and assistance to this effort: Elizabeth been applied widely, let alone universally. We also seek Brouwer, Kristen Danforth, Mary Fisk, Rumit Pancholi, to focus attention on the need to identify new strategies Jinyuan Qi, Shamelle Richards, and Carlos Rossel. The that would be particularly effective in LMICs. Finally, we authors also thank the reviewers organized by the wish to highlight the potentially substantial health haz- National Academy of Medicine and the InterAcademy ards of climate change. This particular environmental Medical Panel listed separately in this volume. We issue will likely become increasingly preeminent in the especially thank Brianne Adderley for her hard work in 21st century. The resultant health problems, including keeping this large endeavor well organized. food and water insecurity, may rival those of other major risk factors. The toll could be especially tragic among the Charles N. Mock world’s poorest people. Enhancing the understanding of Rachel Nugent climate change and identifying effective interventions are Olive Kobusingye likely to become major challenges in the next generation. Kirk R. Smith xiv Preface Abbreviations ALRI acute lower respiratory infection BCA benefit-cost analysis BCR benefit-cost ratio BPL below poverty line CAMx Comprehensive Air Quality Model with Extensions (Eulerian photochemical dispersion model) CATS Community Approach to Total Sanitation CCT conditional cash transfer CDD community-driven development CEA cost-effectiveness analysis CERC Central Electricity Regulatory Authority CI confidence interval CIRCLE Climate Impact Research and Response Coordination for a Larger Europe CLTS community-led total sanitation CO carbon monoxide COPD chronic obstructive pulmonary disease CPLS cost per life saved CRA comparative risk assessment CTC Communities that Care CVD cardiovascular disease DALY disability-adjusted life year DHS Demographic and Health Survey ECEA extended cost-effectiveness analysis ESI Economics of Sanitation Initiative EU European Union FGD flue-gas desulfurization GACC Global Alliance for Clean Cookstoves GBD Global Burden of Disease GDP gross domestic product GIU Give It Up GNI gross national income GNP gross national product xv GPOBA Global Program for Output-Based-Aid GRP gross regional product GW gigawatt HAP household air pollution HAPIT Household Air Pollution Intervention Tool HICs high-income countries HLY healthy life-year HRTWS Human Right to Safe Drinking Water and Sanitation HWTS household water treatment and storage IAQG Indoor Air Quality Guidelines ICER incremental cost-effectiveness ratio IER integrated-exposure response IHD ischemic heart disease IHDS Indian Human Development Survey IHME Institute for Health Metrics and Evaluation IMAGE Intervention with Microfinance for AIDS and Gender Equity IOM Institute of Medicine IPV intimate partner violence ISBI International Society for Burn Injuries JMP Joint Monitoring Programme kWh kilowatt-hour LC lung cancer LICs low-income countries LMICs low- and middle-income countries LPG liquefied petroleum gas MDGs Millennium Development Goals mg/m3 milligrams per cubic meter MHM menstrual hygiene management MICS Multiple Indicator Cluster Survey MICs middle-income countries MW megawatt NCAP New Car Assessment Program NCDs noncommunicable diseases NGO nongovernmental organization NISP National Improved Stove Program NOx oxides of nitrogen OBA output-based aid OECD Organisation for Economic Co-operation and Development OOP out of pocket OR odds ratio OSH occupational safety and health PFD personal flotation device PIC products of incomplete combustion PM particulate matter PPPHW Global Public-Private Partnerships for Handwashing PRB Powder River Basin QALY quality-adjusted life year xvi Abbreviations RCT randomized controlled trial RHS Reproductive Health Survey RTI road traffic injury SDGs Sustainable Development Goals SERC State Electricity Regulatory Commission SFU solid fuel use SLTS School-Led Total Sanitation SO2 sulfur dioxide SPA Service Provision Assessment STHs soil-transmitted helminths STIs sexually transmitted infections SV Smokeless Village swFGD seawater flue-gas desulfurization TSSM Total Sanitation and Sanitation Marketing UMICs upper-middle-income countries UN United Nations UNICEF United Nations Children’s Fund UNRSC United Nations Road Safety Collaboration VSL value per statistical life WASH water, sanitation, and hygiene WASH-BAT Water, Sanitation, and Hygiene-Bottleneck Analysis Tool wFGD wet limestone flue-gas desulfurization WHO World Health Organization WTP willingness to pay YLDs years lived with disability YLLs years of life lost Abbreviations xvii Chapter 1 Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition Charles N. Mock, Kirk R. Smith, Olive Kobusingye, Rachel Nugent, Safa Abdalla, Rajeev B. Ahuja, Spenser S. Apramian, Abdulgafoor M. Bachani, Mark A. Bellis, Alexander Butchart, Linda Cantley, Claire Chase, Mark Cullen, Nazila Dabestani, Kristie L. Ebi, Xiagming Fang, G. Gururaj, Sarath Guttikunda, Jeremy J. Hess, Connie Hoe, Guy Hutton, Adnan A. Hyder, Rebecca Ivers, Dean T. Jamison, Puja Jawahar, Lisa Keay, Carol Levin, Jiawen Liao, David Mackie, Kabir Malik, David Meddings, Nam Phuong Nguyen, Robyn Norton, Zachary Olson, Ian Partridge, Margie Peden, Ajay Pillarisetti, Fazlur Rahman, Mark L. Rosenberg, John A. Staples, Stéphane Verguet, Catherine L. Ward, and David A. Watkins VOLUME SUMMARY water, sanitation, and hygiene (1.4 million); and air Injury Prevention and Environmental Health identifies pollution (5.5 million). The vast majority of these essential prevention strategies and related policies that deaths are in low- and middle-income countries. address substantial population health needs, are cost- • Risk factors for deaths from these diseases vary with effective, and are feasible to implement. This chapter stages of national development in ways that can be summarizes and critically assesses the volume’s four key understood and used in designing prevention strategies. findings. • A range of interventions could effectively address these problems; many of these interventions are • There is a large burden of death and disability among the most cost-effective and cost-beneficial of from injuries and environmental health conditions. all interventions used to prevent disease. Worldwide, injuries result in more than 5 million • This chapter synthesizes the volume’s prevention premature deaths per year out of a global total of strategies to identify an effective essential package of 56 million deaths (based on widely used estimates). interventions and policies, most of which have been There are also large numbers of deaths attributable inadequately applied on a global scale. Better imple- to risk factors related to noninjury occupational mentation of these interventions and policies would exposures (560,000); inadequate access to clean help bring down the high rates of death and disability Corresponding author: Charles N. Mock, Departments of Surgery and Global Health, University of Washington, Seattle, Washington, United States; cmock@uw.edu. 1 from injury and environmental and occupational risks prevent these conditions. Treatment for health condi- in low- and middle-income countries (LMICs) toward tions resulting from injury and environmental risk the lower rates in high-income countries. Doing so factors is covered in other volumes of the third edition could avert more than 7 million deaths annually of Disease Control Priorities (DCP3), as are immuniza- from environmental and occupational exposures and tions and prevention of suicide (Black, Laxminarayan, injuries. and others 2016; Black, Levin, and others 2016; Bundy and others 2017; Debas and others 2015; Mock and others 2015; Patel and others 2015; Patel and others INTRODUCTION 2016; Prabhakaran and others 2017). Injury Prevention and Environmental Health identifies In this review, we identify several key messages. essential prevention strategies and related policies that First, there is a large health burden from injury, occu- address substantial population health needs and pational risk factors, air pollution, unclean water, and that are cost-effective and feasible to implement. This poor sanitation. These conditions are major global volume addresses diverse conditions that arise from health problems to which inadequate attention has exposure to outside forces, such as chemicals and been directed. Second, these disorders and the risk toxins, kinetic energy, or thermal energy. These condi- factors that cause them have predictable patterns tions require similar policy approaches to reducing risk across stages of national development. Understanding and mandate involvement of multiple sectors. Included these patterns can assist with the planning of preven- in this group of conditions are injuries attribut- tion efforts. Third, cost-effective and cost-beneficial able to unintentional mechanisms (road traffic crashes, interventions that can address these conditions already falls, burns, and drowning); injuries attributable to exist and are in established use in most high-income intentional mechanisms (interpersonal violence); dis- countries (HICs). In most low- and middle-income orders caused by or aggravated by exposure to airborne countries (LMICs), these interventions have been toxins (air pollution); occupational issues (injuries implemented only to a modest extent or not at all. On and disorders caused by or aggravated by toxins in the basis of these interventions’ cost-effectiveness and the workplace); and waterborne infectious diseases. their potential to lower the disease burden, we pro- This volume focuses exclusively on interventions to pose a package of policy interventions (box 1.1). Box 1.1 From the Series Editors of Disease Control Priorities, Third Edition Budgets constrain choices. Policy analysis helps Disease Control Priorities in Developing Countries, decision makers achieve the greatest value from lim- second edition (DCP2) published in 2006, updated ited resources. In 1993, the World Bank published and extended DCP1 in several respects, giving Disease Control Priorities in Developing Countries explicit consideration to the implications for health (DCP1), which sought to assess systematically the systems of expanded intervention coverage (Jamison cost-effectiveness (value for money) of interventions and others 2006). One way to expand coverage of addressing the major sources of disease burden in health interventions is through platforms for inter- low- and middle-income countries (Jamison and ventions that require similar logistics but that others 1993). The World Bank’s World Development address heterogeneous health problems. Platforms Report 1993 drew heavily on DCP1’s findings to often provide a more natural unit for investment conclude that specific interventions to combat non- than do individual interventions, but conventional communicable diseases were cost-effective, even in health economics has offered little understanding of environments with substantial burdens of infection how to make choices across platforms. Analysis of and undernutrition (World Bank 1993). the costs of packages and platforms—and of the box continues next page 2 Injury Prevention and Environmental Health Box 1.1 (continued) health improvements they can generate in given volume—and their related delivery platforms. This epidemiological environments—can help guide information is intended to assist decision makers in health system investments and development. allocating often tightly constrained budgets and in achieving health system objectives. DCP3 introduces the notion of packages of interven- tions. Whereas platforms contain logistically related Four of DCP3’s nine volumes were published in 2015 sets of interventions, packages contain conceptually and 2016, and the remaining five will appear in 2017 related ones. The 21 packages developed in the nine and 2018. The volumes appear in an environment in volumes of DCP3 include those for surgery and car- which serious discussion continues about quantifying diovascular disease, for example. In addition, DCP3 and achieving the Sustainable Development Goal explicitly considers health systems’ objective of (SDG) for health (United Nations 2015). DCP3’s anal- financial risk protection. In populations lacking yses are well placed to assist in choosing the means access to health insurance or prepaid care, medical with which to attain the health SDG and assessing the expenses that are high relative to income can be related costs. These volumes, and the analytic efforts impoverishing. Where incomes are low, seemingly on which they are based, will enable researchers to inexpensive medical procedures can have cata- explore SDG-related and other broad policy conclu- strophic financial effects. DCP3 considers financial sions and generalizations. The final volume will report protection and the distribution across income groups those conclusions. Each individual volume provides as outcomes resulting from policies (for example, specific policy analyses on the full range of interven- public finance) to increase intervention uptake and tions, packages, and policies relevant to its health topic. improve delivery quality. All of the volumes seek to combine the available science about interventions Dean T. Jamison implemented in specific locales and conditions with Rachel Nugent informed judgment to reach reasonable conclusions Hellen Gelband about the impact of intervention mixes in diverse Susan Horton environments. DCP3’s broad aim is to delineate Prabhat Jha essential intervention packages—such as those for Ramanan Laxminarayan injury prevention and environmental health in this Charles N. Mock KEY MESSAGES The vast majority (85 percent) of these deaths were in LMICs. The annual incidence of mortality from injury is Disease Burden Addressable by Injury Prevention and considerably higher in LMICs (76 per 100,000) com- Environmental Health pared with HICs (58 per 100,000) (WHO 2016). In most The different topics examined take advantage of one or LMICs, half or more of road traffic crash deaths happen more widely used data sources, such as the World Health to vulnerable road users, such as motorcyclists, bicyclists, Organization (WHO) Global Health Estimates or the and especially pedestrians. Injuries to vehicle occupants Global Burden of Disease (GBD) study. predominate in most HICs. Other leading causes of Other global datasets may show slightly different unintentional injury are falls, drowning, and burns. The relationships, but the patterns would be similar.1 leading cause of intentional injury deaths is suicide. Homicide is the next leading cause, followed at a distant Injury third by deaths directly due to war and other forms of Injuries include those arising from unintentional causes collective violence (Watkins, Dabestani, Mock, and (such as road traffic crashes, falls, and burns) and inten- others 2017; WHO 2016). Interpersonal violence is also tional causes (such as suicide and violence). In 2012, an important yet under-recognized risk factor for injuries altogether caused more than 5 million premature high-risk behaviors, such as unsafe sex, smoking, and deaths globally (table 1.1).2 substance abuse, and, through these behaviors, for some Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 3 Table 1.1 Injuries: Deaths by Cause, All Ages, Both Sexes, 2012 Low- and Middle-Income Countries 2012 High-Income Countries 2012 Total deaths (thousands) Percent of all deaths Total deaths (thousands) Percent of all deaths All causes 44,200 100 11,700 100 Injuries (unintentional and intentional) 4,400 10 750 6 Unintentional injuries 3,220 7 510 4 Road traffic injuries 1,140 3 120 1 Other unintentional injuries 750 2 180 2 Falls 580 1 120 1 Drowning 340 1 40 0 Fire, heat, and hot substances 250 1 20 0 Poisoning 160 0 30 0 Exposure to forces of nature 2 0 0 0 Intentional injuries 1,190 3 240 2 Self-harm 610 1 200 2 Interpersonal violence 460 1 40 0 Collective violence and legal intervention 120 0 0 0 Source: WHO Global Health Estimates 2012 (WHO 2016). Note: Not all totals are exact due to rounding. communicable and noncommunicable diseases, as well many countries. Part of the problem is lack of reporting as for mental health conditions, including anxiety disor- on occupational issues, which is aggravated by the ders, depression, and suicidal ideation. fact that most people in LMICs work in the informal sector, for which accurate, or sometimes any, statistics Occupational Risks are not kept. Occupational health problems encompass Occupational and environmental (water and air) risks lead some that are long-standing, such as agricultural inju- to a substantial health burden. In the usual estimates of ries. Others arise or are aggravated by changes in manu- global disease burden, this burden is reflected in disease- facturing and supply chain practices globally as more specific estimates; for example, unsafe water leads to deaths dangerous jobs are transferred to LMICs, especially to from diarrhea, which are reported in the main global dis- locations with limited environmental and safety safe- ease burden estimates (Watkins, Dabestani, Mock, and guards, and are performed by people with lower levels of others 2017; WHO 2016). Additional analyses discussed training and who usually have limited or no access to later show the burden from the risk factors themselves. protective equipment (Watkins, Dabestani, Mock, and Occupationally related deaths and disabilities include others 2017). on-the-job injuries and exposure to chemicals (such as pesticides, solvents, and heavy metals); heat; and noise; Water, Sanitation, and Hygiene among other risk factors. An estimated 720,000 deaths Inadequate access to safe water, sanitation, and hygiene occur annually from occupational exposures globally, 79 (WASH) was estimated to result in about 1.4 million percent of which are in LMICs. The largest contributors to deaths globally in 2013, virtually all (more than this burden are injuries and exposure to particulate matter, 99 percent) in LMICs (table 1.3). WASH-related deaths gases, and fumes (which contribute to respiratory and car- account for a large proportion of diarrheal disease and diovascular disease and cancers) (table 1.2). Occupational intestinal infectious diseases, almost all among children. ergonomic factors and exposure to noise do not cause The major attributal factors are unsafe water sources mortality, but they contribute significantly to disability. (1,240,000 deaths globally), unsafe sanitation (820,000 Notwithstanding the global estimates in table 1.2, deaths), and lack of hygiene (especially availability of estimates and sources of overall burden of occupa- handwashing with soap: 520,000 deaths), with an uncer- tional deaths and disabilities are not well known for tain degree of overlap in attributable deaths among 4 Injury Prevention and Environmental Health Table 1.2 Occupational Risks: Attributable Deaths by Cause, All Ages, Both Sexes, 2013 Deaths (Thousands) Low- and middle-income countries 2013 High-income countries 2013 Total attributable deaths 23,800 7,000 Total environmental and occupational risks 7,420 760 Occupational risks 570 140 Occupational asthmagens 50 0 Occupational carcinogens 190 110 Occupational ergonomic factors 0 0 Occupational injuries 140 20 Occupational noise 0 0 Occupational particulate matter, gases, 200 10 and fumes Source: Global Burden of Disease (GBD) 2013 Study (IHME 2016). Note: Each of the six major occupational hazards is listed as a subcategory of “occupational risks,” which are a subset of “total environmental and occupational risks,” which are a subset of “total attributable deaths.” Data from GBD 2013 were used because similar data were unavailable from the WHO Global Health Estimates. GBD 2010 and GBD 2015 estimates are somewhat different from GBD 2013. Not all totals are exact due to rounding. Table 1.3 Environmental Risks: Attributable Deaths by Cause, All Ages, Both Sexes, 2013 Deaths (Thousands) Low- and middle-income countries 2013 High-income countries 2013 Total attributable deaths 23,800 7,000 Total environmental and occupational risks 7,420 760 Unsafe water, sanitation, and handwashing 1,390 10 No handwashing with soap 510 10 Unsafe sanitation 820 0 Unsafe water source 1,240 10 Air pollution 4,990 540 Ambient ozone pollution 180 40 Ambient particulate matter pollution 2,430 500 Household air pollution from solid fuels 2,880 10 Source: Global Burden of Disease (GBD) 2013 Study (IHME 2016). Note: Each of the major environmental hazards is listed as a subcategory of the bolded categories. Data from GBD 2013 were used because similar data were unavailable from the WHO Global Health Estimates. There is an unknown degree of overlap between the impacts across the air pollution and unsafe water categories, which is not addressed here. these causes. Water and sanitation were the topics of 2013 (Watkins, Dabestani, Mock, and others 2017). Millennium Development Goal 7 and have received con- Despite these improvements, inadequate access to WASH siderable attention over the past several decades. As a remains a major health problem, accounting for approx- result, there have been significant advances in access to imately 43 percent of under-five mortality in South and clean water and improved sanitation, with related South-East Asia and Sub-Saharan Africa (Humphrey decreases in burden. In addition, better nutrition and 2009; Petri and Miller 2008). rehydration therapy have reduced case fatality substan- tially. The total number of deaths estimated as attribut- Air Pollution able to inadequate WASH has declined by 49 percent, Exposure to airborne pollutants in ambient and house- from 2.7 million deaths in 1990 to 1.4 million deaths in hold settings was estimated to result in more than Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 5 5 million deaths globally in 2013 (table 1.3). In disease This portrayal gives the false impression, however, that burden estimates, air pollution contributes a significant the impact of noncommunicable disease increases with proportion of deaths attributable to respiratory infec- development, which is not the case at large scale. For tions; chronic obstructive pulmonary disease; cere- comparisons of the health status of populations, the brovascular disease; ischemic heart disease; and cancers correct calculation is the age-standardized version. of the trachea, bronchus, and lung. The forms of air The age-standardized version is the true epidemiological pollution evaluated were ambient particulate matter transition, which takes account of the younger age struc- pollution (approximately 2.9 million deaths globally) ture in poor countries, as shown in figure 1.2 (Smith and and household air pollution from solid fuels (approxi- Ezzati 2005). Age-standardized data provide a more accu- mately 2.9 million deaths globally) in the form of rate illustration of the comparative health of someone particle and ozone pollution, although there are other going through the life course in each region, what most categories that have not yet been assessed globally. people consider the important comparison of health sta- Overall, 90 percent of air pollution deaths are in tus across populations. In contrast to what is shown in the LMICs. However, because use of solid cooking fuels mortality transition (figure 1.1), in figure 1.2 all general in households is confined almost entirely to LMICs, disease categories—communicable (category I), non- essentially all impacts occur there. Ambient particulate communicable (category II), and injuries matter air pollution occurs in rural and urban areas (category III)—actually decline across income groups and is related to a variety of emissions sources, includ- after age standardization, substantially so in categories ing motorized transport, power plants, industry, road I and III (communicable and injuries), but definitively for and construction dust, brick kilns, and garbage burning. noncommunicable as well. Thus, as is uncomfortably true Household air pollution occurs primarily in less for many of life’s conditions, it is generally better to live in urbanized areas and is related to use of solid fuels for a richer rather than a poorer society. cooking and heating. It also is a major source of ambi- Many factors other than income affect health, and ent pollution, causing at least a quarter of ambient many of these are amenable to policy. Policy, in turn, is pollution exposures in India and China, for example affected by factors other than income, although income (Chafe and others 2014; Lelieveld and others 2015). remains one primary determinant. All analyses in this Thus, perhaps 16 to 31 percent of the burden attributed section use age-standardized deaths per capita to nor- to ambient pollution actually started in households, malize across the four World Bank income regions and although this burden is not yet well characterized. aggregate large categories of disease and risk that tend to Ambient air pollution is estimated to account for obfuscate individual differences. It should be noted that a larger proportion of cardiovascular and cerebrovas- higher resolution by more subregions, specific diseases, cular diseases, while household air pollution accounts or even by country might show subtleties not revealed for a larger proportion of chronic and acute respiratory by comparison across only four income regions. disease, the latter affecting children (Watkins, Dabestani, Mortality trends are not reflective of the entire picture of Mock, and others 2017). health because nonfatal injury and illness also affect Taken together, the conditions and risks covered in health status. The aggregated patterns shown in this sec- this review comprise more than 12 million deaths per tion, however, show similar trends when disability- year, not accounting for possible overlaps among adjusted life years (DALYs) are used. different categories of attributable causes. Climate As shown in figure 1.3, the overall health impacts change contributes a small portion of the current from environmental and occupational exposures and burden of climate-sensitive health outcomes but, from injuries tend to decline across country income given its trajectory, will become increasingly impor- groups after age standardization. Examined in more tant in future decades. detail, however, the trends for environmental risks can be divided into three categories in what has been termed the environmental risk transition (Smith 1990). Environmental and Injury Risk Transitions All comparisons in this section rely on the widely used Traditional Environmental Health Risks GBD 2015 dataset—other global datasets may show Traditional environmental health risks (poor food, air, slightly different absolute levels and relationships, but water, and sanitation at the household level) tend the patterns will be similar.3,4 A classic portrayal of mor- to decline with economic development, but they do so at tality trends during the national development process is varying rates depending on policy and the degree of the “mortality transition” that documents shifts over income and education equity in societies. This link to time in causes of death (figure 1.1) (Omran 1971). income is observed in figure 1.4, which shows the 6 Injury Prevention and Environmental Health Figure 1.1 Crude Death Rates across Income Categories for All Category I, II, and III Diseases, All Ages, 2015 a. Category I (communicable or infectious, b. Category II (noncommunicable) maternal, neonatal, and nutritional) 800 400 Deaths per 100,000 population Deaths per 100,000 population 600 300 400 200 200 100 0 0 Low income Lower-middle Upper-middle High income Low income Lower-middle Upper-middle High income income income income income c. Category III (injuries, intentional and unintentional) 75 Deaths per 100,000 population 50 25 0 Low income Lower-middle Upper-middle High income income income burden from household air pollution and from poor reach and the point at which they turn downward are water, sanitation, and hygiene steadily declining across strongly determined by preventive policy. Figure 1.5 income groups. Although much diminished in rich illustrates how the burdens from ambient particle pollu- countries, these risks still dominate global environmen- tion, environmental tobacco smoke, and ambient ozone tal health burdens today. air pollution rise and then fall with development. Modern Environmental Health Risks Global Environmental Health Risks Modern environmental health risks from industrializa- The imposition of a set of global environmental risks— tion, urbanization, vehicularization, and agricul- exemplified by release of greenhouse pollutants and tural modernization tend to rise at first during the including other global environmental stressors, such as development process, then peak and fall at higher levels biodiversity loss—has risen with development. The of income and education. Again, the height these risks notable exception is reductions of stratospheric Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 7 Figure 1.2 Age-Standardized Trends in Mortality Risk for Category I, II, and III Diseases by Income, 2015 a. Category I (communicable or infectious, b. Category II (noncommunicable) maternal, neonatal, and nutritional) 750 Deaths per 100,000 population Deaths per 100,000 population 400 500 200 250 0 0 Low income Lower-middle Upper-middle High income Low income Lower-middle Upper-middle High income income income income income c. Category III (injuries, intentional and unintentional) 125 100 Deaths per 100,000 population 75 50 25 0 Low income Lower-middle Upper-middle High income income income ozone–depleting pollutants under the Montreal Protocol, In summary, as shown in figure 1.7, all environmental which is one of the major examples of successful inter- risk factors taken together declined over the develop- national policy. Such global hazards do not dominate ment spectrum because of the strong decline in tradi- current environmental health burdens, but as these tional risks. In general, traditional risks are faced mostly threats continue to rise, they may dominate health bur- at the household level in lower-income countries, where dens later in the century unless strong actions are imple- required behavioral changes and low access to resources mented. The trends for risks from greenhouse gas are barriers to interventions. Modern risks are com- emissions are illustrated in figure 1.6 for the two most monly seen at the community level because they derive important gases, carbon dioxide and methane (Smith, from larger-scale social organization, including industri- Desai, and others 2013). alization and urban design. Global risks arise at larger 8 Injury Prevention and Environmental Health geographic and organizational scales, with most health examination of individual injury categories reveals differ- impacts generally occurring in populations that have ent patterns, recognizing that reporting bias is present contributed little to concentrations of greenhouse gases for many types of injury. Mortality from, in declin- in the atmosphere. ing number, road traffic injuries, falls, drowning, fires, Is there an injury risk transition? Panel c of figure 1.2 occupational injuries, and snakebites (surprisingly preva- illustrates that the impact of all forms of injuries declines lent in poor areas) appears generally to follow the classic with development. A question, however, is whether traditional risk form, declining steadily with development (figure 1.8). Mortality from interpersonal violence and poisoning may also follow the traditional form, but trends Figure 1.3 Age-Standardized Mortality, 2015, from All are not clear at this resolution (four income groups only; Occupational and Environmental Risk Factors Examined figure 1.9). Thus, there is no clear transition from one to in the Global Burden of Disease Study 2015 another type of injury with development, but rather a steady decline across essentially all categories examined here as protective policies and infrastructure are put in place and daily work and living environments evolve. Deaths per 100,000 population 300 Transition frameworks are common in development discussion (for example, demographic, nutrition, and inequality transitions) but should primarily be consid- 200 ered tools for parsing observed patterns rather than generating normative predictions of what will happen. 100 They provide a structure for categorizing changes that occur during development and for designing policies that avoid the worst trends and enhance the best ones. They are not destiny but analytic tools. 0 Low income Lower-middle Upper-middle High income It is important to be aware that the relationships in income income this chapter are cross-sectional and thus cannot take into account the different world situation in place when Note: This figure is based on summed impacts from estimates of the impacts of currently developed regions were developing as com- separate risk factors. It thus includes contributions from communicable diseases (category I), noncommunicable diseases (category II), and injuries (category III). This pared with poor countries today. Nevertheless, they figure contains no contribution from global risks, but as shown in the vertical axis of provide instructive ways to understand and organize figure 1.6, global risks are relatively small at present. current risk patterns. Figure 1.4 Age-Standardized Trends in Mortality Risk for Household Air Pollution and for Poor Water, Sanitation, and Hygiene, 2015 a. Household air pollution from solid fuels b. Poor water, sanitation, and hygiene 200 90 Deaths per 100,000 population Deaths per 100,000 population 150 60 100 50 30 0 0 Low income Lower-middle Upper-middle High income Low income Lower-middle Upper-middle High income income income income income Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 9 Figure 1.5 Age-Standardized Trends in Mortality Risk for Ambient Particle Pollution, Environmental (Secondhand) Tobacco Smoke, and Ambient Ozone Pollution, 2015 a. Ambient particulate pollution b. Secondhand tobacco smoke 125 100 Deaths per 100,000 population Deaths per 100,000 population 20 75 50 10 25 0 0 Low income Lower-middle Upper-middle High income Low income Lower-middle Upper-middle High income income income income income c. Ambient ozone pollution Deaths per 100,000 population 10 5 0 Low income Lower-middle Upper-middle High income income income ECONOMIC EVALUATION OF INJURY because many of the interventions are population-based PREVENTION AND ENVIRONMENTAL policies and regulations that use multisectoral approaches, which are inherently less straightforward to study using INTERVENTIONS economic methods that are more readily applied to Economic evaluation aims to inform decision making by individual-level health interventions. In addition, several quantifying tradeoffs between resource inputs required of the environmental interventions have notable non- for alternative strategies and resulting outcomes. Four health outcomes that are often difficult to cost, such as main approaches are discussed in box 1.2. time savings, reduction in black carbon emissions, and Economic evaluation of the interventions that address lower pressure on forests from shifts in household fuels. the conditions in this review has not been conducted Nevertheless, there is an accumulating body of to the same extent as for many other health problems evidence that many of the interventions addressing (Watkins, Dabestani, Nugent, and Levin 2017), in part injury and environmental health are very cost-effective 10 Injury Prevention and Environmental Health Figure 1.6 Trends of Global Environmental Health Risk by Income Using WHO Regions 10 AFR-E AMR-A EUR-C SEAR-D EMR-D WPR-A AFR-D EMR-B EUR-A EUR-B 1 WPR-B Per capita DALYs (DALYs/1,000 people) AMR-B AFR-D AFR-E AMR-D EMR-D SEAR-B SEAR-D AMR-D EMR-B Ratio of imposed 0.1 to experienced SEAR-B WPR-B AMR-B Region Ratio AMR-A 298 EUR-A 264 EUR-B WPR-A 198 EUR-C 178 EUR-B 23.1 EMR-B 9.08 EUR-C 0.01 AMR-B 5.81 WPR-A AMR-A WPR-B 5.21 SEAR-B 4.29 EUR-A AMR-D 1.51 AFR-D 0.227 EMR-D 0.168 SEAR-D 0.154 AFR-E 0.130 0.001 1,500 3,000 6,000 12,000 24,000 48,000 Per capita GDP, PPP-adjusted (US$/person) Experienced Experienced (exponential) Imposed Imposed (exponential) Source: Smith, Desai, and others 2013. Note: DALY = disability-adjusted life year; GDP = gross domestic product; PPP = purchasing power parity. In key, AFR = African Region; AMR = American Region; EMR = Eastern Mediterranean Region; EUR = European Region; SEAR = South-East Asia Region; WPR = Western Pacific Region. A–E refer to specific groupings of countries by mortality strata within each region. The trend for “experiencing” the risk is inverse to the trend for “imposing the risk.” The latter is based on parsing the total estimated global burden from climate change according to each region’s contribution to emissions of carbon dioxide and methane over time—its natural debt. Figure 1.7 Deaths from All Environmental Risk Factors, in LMICs. For example, studies in LMICs have shown Age Standardized, 2015 that speed bumps at high-risk junctions cost US$12 per DALY averted (in 2012 US$), improved enforce- ment of traffic laws costs US$84 per DALY averted, and 300 enforcing motorcycle helmet use costs US$615 per Deaths per 100,000 population DALY averted (Bishai and Hyder 2006; Ditsuwan and others 2013; Watkins, Dabestani, Nugent, and Levin 200 2017). Swimming lessons and improved supervision of children to prevent drowning cost US$27 and US$256 per DALY averted, respectively (Rahman and others 100 2012; Watkins, Dabestani, Nugent, and Levin 2017). In general, an intervention with a cost-effectiveness ratio of one to three times the per capita gross domestic 0 product of a country is considered cost-effective (Newall, Low income Lower-middle Upper-middle High income Jit, and Hutubessy 2014; Watkins, Dabestani, Nugent, and income income Levin 2017). Thus, for almost all countries, the examples Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 11 Figure 1.8 Age-Standardized Trends in Mortality Risk for Types of Unintentional Injury, 2015 a. Road traffic injuries b. Falls 30 15 Deaths per 100,000 population Deaths per 100,000 population 20 10 10 5 0 0 Low income Lower-middle Upper-middle High income Low income Lower-middle Upper-middle High income income income income income c. Drowning d. Fire, heat, and hot substances 8 8 Deaths per 100,000 population Deaths per 100,000 population 6 6 4 4 2 2 0 0 Low income Lower-middle Upper-middle High income Low income Lower-middle Upper-middle High income income income income income e. Occupational injuries f. Venomous animal contact 4 4 Deaths per 100,000 population Deaths per 100,000 population 3 3 2 2 1 1 0 0 Low income Lower-middle Upper-middle High income Low income Lower-middle Upper-middle High income income income income income 12 Injury Prevention and Environmental Health Figure 1.9 Age-Standardized Trends in Mortality Risk for Interpersonal Violence and Poisoning, 2015 a. Interpersonal violence b. Poisoning 7.5 3 Deaths per 100,000 population Deaths per 100,000 population 5.0 2 2.5 1 0 0 Low income Lower-middle Upper-middle High income Low income Lower-middle Upper-middle High income income income income income Box 1.2 Economic Evaluation of Investments in Injury Prevention and Environmental Health Economic evaluation aims to inform decision mak- (for example, public sector finance of a given ing by quantifying tradeoffs between resource inputs intervention, such as regulation of helmets for required for alternative investments and resulting motorcyclists). This approach, extended cost- outcomes. Four main approaches are relevant to this effectiveness analysis (ECEA), enables assessment chapter: not only of efficiency in improving the health of a population but also of efficiency in achieving the other major goal of a health system—protecting • Assessing how much of a specific health outcome— the population from financial risk of medical for example, serious injuries averted—can be impoverishment—along with the distributional attained for a given level of resource input. consequences of the given policy, such as equity. • Assessing how much of an aggregate measure • Assessing the economic benefits, measured in of health—such as deaths or disabilities or monetary terms, from investment in a health inter- disability-adjusted life years (DALYs) averted— vention and weighing that benefit against its cost can be attained from a given level of resource (benefit-cost analysis, or BCA). BCA enables com- inputs applied to alternative interventions. This parison of the attractiveness of interventions in the approach (cost-effectiveness analysis, or CEA) same sector and across different sectors. Benefit- allows comparisons of the attractiveness of inter- cost ratios greater than 1 identify interventions that ventions addressing different health outcomes represent net positive returns on investment. (for example, motorcycle helmet use versus cesar- ean section) to be made. • Assessing how much health and financial risk CEAs predominate among economic evaluations protection and its distribution across population in injury prevention. Three recent overviews of subgroups can be attained for a given policy CEA findings for injury prevention in low- and box continues next page Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 13 Box 1.2 (continued) middle-income countries (one in this volume) have many greater than 10 (Hutton and Chase 2017; especially focused on road safety and drowning Watkins, Dabestani, Nugent, and Levin 2017). prevention. These studies underpin this chapter’s ECEAs are still a relatively new evaluation approach. conclusion that many injury prevention modalities This volume presents two new ECEAs. One is on the are highly cost-effective even in resource-constrained impact of motorcycle helmet regulation on health, environments (Ditsuwan and others 2013; Rahman equity, and medical impoverishment in Vietnam and others 2012; Watkins, Dabestani, Nugent, and (Olson and others 2017). The other found that a Levin 2017). public-private subsidy for poor Indian households to BCAs predominate among economic evaluations in receive clean fuels could avert 44,000 deaths for environmental health, especially for air pollution US$825 each and about 1.5 million DALYs for US$25 and for water, sanitation, and hygiene. BCAs are each. This result was far cheaper than cookstove especially suitable for these topics because they alternatives, and the subsidy for clean fuels provided are able to consider the benefits of nonhealth out- greater health benefits to all income groups. The comes, such as time savings in procuring water or greatest health benefit is achieved when the clean fuel fuels. These BCAs have consistently identified inter- subsidy is targeted to the poor (Pillarisetti, Jamison, ventions with benefit-cost ratios greater than 1, and and Smith 2017). cited earlier for injury prevention would be considered For household air pollution control, a limited but cost-effective. Likewise, the cost-effectiveness of the inter- growing literature evaluates cost-effectiveness and BCRs ventions is similar to that of many widely implemented associated with transitions to cleaner cooking. Hutton health interventions, for example, treatment of severe and others performed global cost-benefit analyses of malaria (US$5–US$220 per DALY averted), micronutri- scenarios in which households made the transition ent supplementation (US$20–US$100 per DALY averted), away from solid fuels to either clean fuels or clean bio- oral rehydration solution (US$150 per DALY averted), mass stoves and found the transition to clean fuel and treatment of pneumonia (US$300–US$500 per and the transition to improved stoves had BCRs of DALY averted) (Black, Laxminarayan, and others 2016). 4.3 and approximately 60, respectively (Hutton and The area of WASH has undergone extensive economic others 2006; Hutton, Rehfuess, and Tediosi 2007). analysis, primarily using benefit-cost analysis. A benefit- Benefit-cost analysis has been applied in other specific cost ratio (BCR) greater than 1 is generally considered a geographies, including in Nepal (Malla and others 2011; good investment. Favorable BCRs (1.9–5.1) have been Pant 2011), China (Aunan and others 2013), the Western identified for a range of interventions: filters, piped water, Pacific Region (Arcenas and others 2010), and in Kenya boreholes, and private latrines. Combinations of interven- and Sudan (Malla and others 2011). tions have shown even higher BCRs (2–45) for improved Similarly, the few occupational safety and health water, sanitation, and universal basic access (Hutton 2013; interventions that have been studied in LMICs do Hutton and Chase 2016; Hutton and Chase 2017; appear cost-effective or cost-beneficial. Simulation Watkins, Dabestani, Nugent, and Levin 2017). studies using the WHO-CHOICE methodology found Air pollution control has been subjected to limited engineering controls that decrease the release of silica economic analysis in LMICs. Two studies on ambient air into the air at the workplace to be a cost-effective pollution in Mexico found that retrofitting vehicles to method for preventing silicosis in several industries in reduce emissions produced net benefits of US$100– LMICs; these were found to be more cost-effective US$11,000 per vehicle, corresponding to BCRs of 1.1–7.0. than the use of masks and respirators (but all with Measures to decrease pollution from brick kilns, including cost-effectiveness ratios in the range of several hundred filtration systems, switching to natural gas, and relocating U.S. dollars per DALY averted) (Lahiri and others 2005). kilns to less densely populated areas, produced net Similar methodology identified training programs to benefits corresponding to BCRs of 38 and higher prevent back injury to be a cost-effective method for (Blackman and others 2000; Stevens, Wilson, and Hammitt preventing back pain in LMICs globally; these training 2005; Watkins, Dabestani, Nugent, and Levin 2017). programs were found to be more cost-effective than 14 Injury Prevention and Environmental Health engineering controls (but all with cost-effectiveness exposure standards and enhancing surveillance for ratios of less than US$1,000) (Lahiri, Markkanen, and water- and vector-borne infections. Economic analyses Levenstein 2005). Ergonomic changes in footwear of such measures are in their infancy but have nonethe- manufacturing in Brazil had a BCR of 7.2 (Guimarães, less suggested that not addressing climate change will be Ribeiro, and Renner 2012; Watkins, Dabestani, Nugent, very costly to health systems in less than two decades and Levin 2017). (Ebi, Hess, and Watkiss 2017). In summary, although the literature on economic evaluation of injury prevention and environmental health in LMICs is small, consistent evidence is emerging Essential Interventions to Address Injury and that a range of interventions are cost-effective, cost- Environmental Health beneficial, or both. One particular environmental issue On the basis of their cost-effectiveness or cost-benefit, will likely become increasingly preeminent in the twenty- feasibility, and potential to lower the burden of these first century: climate change. The economic conse- conditions, a package of policy interventions can be quences of the resulting health problems and food and recommended (tables 1.4 and 1.5). These interventions water insecurity will potentially rival those of other include policies in the health sector and in other sectors, major risk factors. In addition to lowering greenhouse including taxes and regulations that affect infrastructure gas emissions, a range of countermeasures have been and the built environment—especially interven- considered, such as establishing occupational heat tions that have proven cost-effectiveness in LMICs. Table 1.4 Essential Injury and Occupational Health Policies Fiscal and Intersectoral Policy Infrastructure, built environment, Information, education, Domain of action Taxes and subsidies and product design Regulation and communication Road safety Overall Subsidized public Mass transport infrastructure and Adoption and enforcement transportation land use (bus rapid transit, rail) of harmonized motor vehicle safety standards Pedestrian safety Increased visibility, areas for Increased supervision of children pedestrians separate from fast walking to school motorized traffic Motorcycle safety Exclusive motorcycle lanes Mandatory use of daytime running lights for motorcycles Mandatory motorcycle helmet laws Bicycle safety Increased visibility, lanes for cyclists Social marketing to promote separate from fast motorized traffic helmet use by child bicyclists Child passenger safety Legislation for and enforcement of child restraints (including seats) Speed control Traffic-calming infrastructure (for Setting and enforcement example, speed bumps), especially of speed limits appropriate at dangerous road segments to function of roads Driving under the Setting and enforcement influence of alcohol of blood alcohol concentration limits Seatbelt use Mandatory seatbelt use Social marketing to promote laws for all occupants seatbelt use table continues next page Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 15 Table 1.4 Essential Injury and Occupational Health Policies (continued) Fiscal and Intersectoral Policy Infrastructure, built environment, Information, education, Domain of action Taxes and subsidies and product design Regulation and communication Other unintentional injury Drowning Legislation and Parental or other adult enforcement of use of supervision (for example, use of personal flotation devices crèches) in high-risk areas for recreational and other Swimming lessons for children high-risk boaters Burns Safer stove design Poisoning Child-resistant containers Information, education, and communication for safe storage of hazardous substances Violence Child maltreatment Corporal punishment ban Parent training, including nurse home visitation, for high-risk families Youth violence Social development programs that teach social skills and incorporate training for parents Information sharing between police and hospital emergency departments Gender-based violence Microfinance combined School-based programs to and intimate partner with gender equity address gender norms and violence training attitudes Interventions for problem drinkers (who are also abusive partners) Advocacy support programs (for example, to increase availability and use of shelters for at-risk women) Cross cutting for Reducing availability and Dispensing alcohol in plastic rather Stricter licensing laws multiple types of injury harmful use of alcohol than glass that could be used as a and reduced availability through increased weapon of firearms taxation and decreased availability of outlets Occupational safety and health Engineering controls to decrease Enforcement of safety Training in hazard recognition release of silica and other toxins standards and control relevant to the Safe injection devices, such as Formalization of large work performed (for example, blunt-tip suture needles informal sectors in task-based training for low- and middle-income hazardous tasks) countries Effective use of available personal protective equipment Occupational health workforce development Note: Interventions for treatment—for example, trauma care for injured people—are covered in other DCP3 volumes and are not addressed here. 16 Injury Prevention and Environmental Health Table 1.5 Essential Environmental Policies Fiscal and Intersectoral Policy Domain of Infrastructure and built Information, education, action Taxes and subsidies environment Regulation and communication Water and Targeted subsidies to Quality WASH facilities in Defined WASH standards per setting National awareness sanitation poor and vulnerable schools, workplaces, public (household, outside household) campaigns (for example, groups spaces, and health care on handwashing) Incentives for private facilities WASH behavior-change sector to become more interventions, such as involved with WASH for community-led total supply chain and service sanitation provision Outdoor air Fuel taxes Relocation of industrial Diesel retrofits Updated health pollution Fines for residential trash sources, such as brick kilns Coal to natural gas transition information systems to burning Municipal trash collection include vulnerability, Brick kiln retrofits for emissions control adaptation, and capacity Fines for not controlling Diesel to CNG transition for PM, SO2, and NO2 emissions control assessment construction dust fleets Acceleration of Euro standards for vehicles Tax polluters Movement toward banning solid fuels in cities National regulation to reduce household Cap and trade policies emissions to outdoors for specific pollutants Regular street cleaning to (for example, SO2) control dust Construction and road dust controls No more subsidies Adoption of European Union fuel standards for coal Household air Advanced biomass stove Improved ventilation as part of Lower barriers and expanded licensure Ventilation pollution subsidies building codes and norms requirements for clean fuel distribution HAP health effects Targeted and expanded Enhanced clean fuel Kerosene ban education LPG and other clean fuel distribution networks National regulation on clean household fuels Promotion of kitchen subsidies to the poor Electrification as a health to match UN SE4ALL goals retrofits to encourage Subsidies for clean measure Smoke-free communities HAP-reducing alternatives to kerosene Application of modern digital interventions and Campaigns for middle technology to enhance access behaviors class to give up subsidies to household clean fuel intended for poor Chemical Regulations on hazardous Arsenic: monitoring of groundwater supplies Notification of contamination waste disposal covering land, and provision of alternatives if needed public of locations of air, and water Asbestos: banning of import, export, mining, contaminated sites manufacture, and sale Mercury: monitoring and reduction or elimination of use in artisanal mining, large-scale smelting, and cosmetics Established and enforced toxic element emissions limits for air and water Restricted access to contaminated sites Strict control and movement to selective bans of highly hazardous pesticides table continues next page Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 17 Table 1.5 Essential Environmental Policies (continued) Fiscal and Intersectoral Policy Domain of Infrastructure and built Information, education, action Taxes and subsidies environment Regulation and communication Lead exposure Concessionary financing Minimization of occupational Ban on lead paint and leaded fuels Lead poisoning training for remediation of worst and environmental exposures Ban on lead in water pipes, cookware, drugs, for health care providers conditions in maintaining, renovating, food supplements, and cosmetics and demolishing buildings and other structures with lead paint Reduction in corrosiveness of drinking water National take-back requirements for collecting used lead batteries Regulations governing land-based waste disposal Risk-based limits for lead in air, water, soil, and dust Global climate Carbon tax or cap Mitigation policies and Energy efficiency and fuel efficient vehicles Early warning and change and trade (mitigation) incentives, including land- (mitigation) emergency response Subsidies to use plans, building design, Mainstreaming of climate change into public systems renewable energy transportation, to reduce GHGs health planning and programs, and into Resilient design in buildings health system policies and plans and infrastructure (adaptation) Methane control regulations Consideration of climate change in public health infrastructure (mitigation and adaptation) Note: CNG = compressed natural gas; GHG = greenhouse gas; HAP = household air pollution; LPG = liquefied petroleum gas; NO2 = nitrogen dioxide; PM = particulate matter; SO2 = sulfur dioxide; UN SE4ALL = United Nations Sustainable Energy for All program; WASH = water, sanitation, and hygiene. Interventions for treatment (for example, oral rehydration solution for diarrhea) or other individual-level medical services (for example, deworming, growth monitoring) are covered in other DCP3 volumes and are not addressed here. Interventions in this table include those that have been shown to be cost-effective or cost-beneficial in low- and middle-income countries or for which such cost-effectiveness or cost-benefit can be logically concluded from high-income or other data. For water and sanitation, many of the policy-level interventions mentioned do not have such evidence; however, the individual items promoted by these policies (for example, filters, piped water, boreholes, private latrines) do have a strong evidence base. Unlike interventions with only health benefits, however, many if not most interventions in environmental health bring a range of other benefits lying outside the health sector, for example, time savings, property values, IQ enhancement, and so on. Cost-effectiveness measured solely in health terms, therefore, can be misleading with regard to total social benefit-cost relationships. The package also includes interventions with proven considerably advanced by government policies (includ- cost-effectiveness in HICs with high likelihood of trans- ing taxes and subsidies) that promote alternative safer ferability to LMICs. Finally, the package also includes and more energy-efficient forms of transportation, such interventions that are logical and feasible, but for which as mass transport, especially rail, as well as by promoting there is currently little empirical evidence on cost-effec- and ensuring the safety of walking and cycling. For road tiveness. Details of these policies, including the evi- traffic crashes themselves, promoting safer infrastruc- dence for them, are addressed in the chapters of this ture is a key intervention. For example, traffic-calming volume. infrastructure such as speed bumps, especially at danger- We acknowledge that the list is not exhaustive. ous intersections, is a very cost-effective method for Other policies might be considered essential. For many protecting pedestrians. In similar fashion, safety-related countries, tables 1.4 and 1.5 provide a reasonable starting product design, such as child-resistant containers for point for an essential policy package to comprehensively poisons and medicines, has played a major role in injury address injury prevention and environmental health, prevention. Safety-related product design encompasses although there will be country-specific variations. engineering (as do infrastructure and the built environ- Examples from injury prevention include promoting ment) as well as regulation because safer products are safer forms of transportation. In general, the individual often best promoted by mandating them in legislation. automobile (especially two- and three-wheeled motor- Other key injury prevention regulations include man- ized vehicles) is one of the least safe modes of transpor- dating the use of restraints for automobile occupants tation. The overall field of transport safety would be and helmets for motorcycle riders. 18 Injury Prevention and Environmental Health Within the public health sector, information and of the health impacts of household air pollution is rela- communication strategies can be successfully delivered tively recent, and understanding of the true scale of the through mass media, as with strategies to promote safe impact of other issues, such as lead, has recently been driving behaviors such as seatbelt and helmet use. Such greatly enhanced. Thus, actions in the health sector have strategies usually do not work well in isolation but are lagged the knowledge of potential benefits. best combined with legislation and effective enforcement. Some of the interventions, although listed for one Information and communication strategies can also be condition, have beneficial effects for other conditions. For delivered in smaller group settings and individually, as example, promoting alternatives to private automobiles with many of the violence prevention strategies. For decreases both injury rates and pollutant emissions. example, home visiting programs using skills training to Improved stoves and fuels decrease air pollution and rates promote better parenting skills, especially to high-risk of household burns. Violence prevention strategies (such groups such as young first-time parents from lower as home visiting and life and social skills training) reduce socioeconomic status, have been found to be very effec- substance abuse, mental health problems, and subsequent tive in preventing child maltreatment in HICs. crime and violence, and increase positive outcomes, Occupational safety and health overlap with injury including academic attainment and employment. prevention. However, interventions in this field primar- A set of policies for a specific subset of pollution, ily target the worksite and thus are distinct from those climate change, is presented in table 1.5. Many of these described earlier that target the general population. Key policies have been widely considered and are straightfor- strategies in promoting occupational safety and health ward and logical (such as promoting active transport include regulations such as setting appropriate limits on and early warning and emergency response systems). As work hours. Given higher risks faced by those in the noted above, economic analysis of their impact has just informal work sector, formalizing this sector, including recently begun. encompassing it within appropriate and context-specific Implementation of many of the interventions requires regulatory and organized labor systems, is a key measure intersectoral collaboration. For example, road safety that needs to be promoted globally. On an individual involves law enforcement, ministries of transport, gov- basis, better application of known safe practices and ernment agencies that regulate manufacturing, and known effective personal protective equipment, such as public health agencies. Likewise, surveillance plays a key masks and respirators to prevent inhalation of silica and role. Surveillance includes not only monitoring of trends other airborne toxins, is needed. for disease burden, but also surveillance for risk factors. Many of the individual WASH interventions, such as For example, a key element for managing air pollution filters, piped water, boreholes, and private latrines, have is monitoring of air quality. Such monitoring, which is been documented to be very cost-effective and cost- especially important for lead control, includes such beneficial (table 1.5). However, access to these inter- activities as examining sample surveys of blood in chil- ventions can be difficult for the poor, especially in rural dren and monitoring of levels from hot spots such as areas. Policies to ensure that these interventions reach lead battery manufacturing and recycling sites. Similarly, everyone include financing strategies (such as targeted a key component of improving WASH is a strengthened subsidies to poor and vulnerable groups), strengthen- monitoring and rapid feedback system for the coverage ing supply chains for water and sanitation products and quality of water and sanitation services. and services, and developing national standards on universal access. Pollution-related interventions include those address- CONCLUSIONS ing air pollution (household and ambient, both of which are primarily related to combustion-derived particulate Injury and occupational and environmental risks result matter) as well as a number of chemical contaminants, in a large health burden. Some of this burden tends to such as lead, asbestos, arsenic, and pesticides. The range of decrease with economic development (for example, risks policy levers can be used for these issues: taxes and subsi- from unsafe water and sanitation), whereas some tends dies (such as targeting clean fuel subsidies to the poor); to initially increase with economic development before infrastructure and built environment (such as relocating declining at high-income levels (for example, ambient industrial sources such as brick kilns); regulation and air pollution and transport injuries). A range of inter- international agreements (such as banning the import, ventions can speed the decrease in burden for the former export, mining, manufacture, and sale of asbestos); and or mitigate the rises for the latter. Many of these inter- actions within the health sector (such as establishing ventions have been shown to be among the most cost- environmental lead surveillance). Among these, awareness effective or cost-beneficial of all interventions used to Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 19 prevent or treat disease. The interventions summarized LMICs. This assessment was straightforward for injury in this chapter include these as well as other similar deaths. However, for deaths from occupational and interventions that are reasonable but have not yet been environmental exposures, we considered “attributable subjected to sufficient economic analysis. Given their deaths” (tables 1.2 and 1.3). These are not mutually exclu- potential to lower this significant health burden, these sive, with overlap of some of the categories. For example, interventions are high priorities for future population, deaths from unsafe water and lack of handwashing partly policy, and implementation research. overlap. Hence, differences in mortality rates were con- Implementation of most of the interventions that sidered for the overall categories of WASH, air pollution, address the conditions in this volume (tables 1.4 and 1.5) and occupational health, not by subcategory. There is has been far less than optimal, especially in LMICs. For likely minimal overlap between air pollution and unsafe example, the WHO’s Global Status Report on Road Safety WASH. For simplicity and lack of systematic analysis of (WHO 2015) assessed 180 countries for the status of key these overlaps, they are ignored in this analysis. road safety interventions. Although the majority of Within these caveats, it can be estimated that more countries (105) implemented best practice standards for widespread implementation of the package of interven- seatbelt laws (such as mandatory seatbelts for all occu- tions and policies covered in this review could avert about pants), far fewer had best practice standards for laws 2 million deaths from injury (not including suicide, which on speed control (47), mandatory motorcycle helmet is not addressed in this volume); 200,000 deaths from use (44), and drunk driving (34) (WHO 2015). Formal occupational risk factors (not including injury); 1.4 mil- health-based intervention programs for household air lion deaths from unsafe water and sanitation; and about pollution have not shown major worldwide implemen- 4 million deaths from air pollution (the larger component tation success to date, although local progress during of which is attributable to household fuels). A total of the relatively short period that they have been imple- more than 7 million deaths could be averted (table 1.6). mented is occasionally seen. Nevertheless, clean fuels, Several factors might cause the real number of poten- through nonhealth actions and economic growth, have tially avertable deaths to be lower or higher. For example, brought major health benefits to hundreds of the differences in death rates between countries at differ- millions of people. Finding ways to expand the rate ent economic levels is in part attributable to better of these improvements to cover populations that would prevention, but also to better medical treatment, which not benefit otherwise is clearly a high priority. is not addressed in the policy package considered in this Many of the interventions considered herein need to volume. Therefore, the estimates of deaths averted by be better applied in HICs, but most have been imple- improved prevention alone might be overstated. mented to a lesser extent in LMICs, which has contrib- However, these estimates do not take into account the uted to the higher health burden from injury and from lives that could be saved by addressing some of the other occupational and environmental risk factors in LMICs. nonoccupational toxins, such as lead and arsenic. Finally, To assess the potential gains from more widespread the interventions considered here have not been fully implementation of these interventions, we estimated applied in many HICs, and many deaths could be averted the deaths that could be averted if the age-adjusted mor- there as well. Even within these caveats and limitations, it tality rates for these conditions in HICs pertained in is apparent that a large number of deaths could be Table 1.6 Disease Burden Avertable by Improved Injury Prevention, Occupational Safety and Health, and Environmental Policy in Low- and Middle-Income Countries Total Deaths (Thousands) Current scenario Hypothetical scenario Avertable Injury (excluding suicide) 3,790 1,730 2,060 Occupational risks (excluding injury) 430 220 210 Unsafe water, sanitation, and handwashing 1,390 20 1,370 Air pollution 4,990 950 4,040 Total 7,680 Source: Global Burden of Disease (GBD) 2013 Study (IHME 2016); WHO Global Health Estimates 2012 (WHO 2016). Note: Hypothetical scenario is the disease burden that would occur if age-specific rates for these conditions in high-income countries applied in low- and middle-income countries. Avertable burden is the difference between current and hypothetical scenarios. Three levels of significance are kept to reduce rounding errors, but true uncertainty is possibly higher. Even so, totals may not add due to rounding. 20 Injury Prevention and Environmental Health averted by better implementation of the low-cost and counterfactual based on what is considered feasible to feasible interventions considered in this volume. obtain. The estimates shown for all the other injury categories, however, here assume that 100 percent of the impact can be avoided. ACKNOWLEDGMENTS The Bill & Melinda Gates Foundation provides financial REFERENCES support for the Disease Control Priorities Network project, of which this volume is a part. Brianne Adderley, Arcenas, A., J. Bojö, B. Larsen, and F. Ruiz Ñunez. 2010. “The Economic Costs of Indoor Air Pollution: New Results for Kristen Danforth, and Shamelle Richards provided valu- Indonesia, the Philippines, and Timor-Leste.” Journal of able comments and assistance on this chapter. Natural Resources Policy Research 2 (1): 75–93. Members of the DCP3 Injury Prevention and Aunan, K., L. W. H. Alnes, J. Berger, Z. Dong, L. Ma, and others. Environmental Health Author Group wrote the chapters 2013. “Upgrading to Cleaner Household Stoves and Reducing on which this chapter draws. The Group includes those Chronic Obstructive Pulmonary Disease among Women in listed as coauthors of this chapter, as well as Maureen Rural China—A Cost-Benefit Analysis.” Energy for Sustainable Cropper, Susan D. Hillis, James A. Mercy, and Paul Development 17 (5): 489–96. Watkiss. Bishai, D. M., and A. A. Hyder. 2006. “Modeling the Cost Effectiveness of Injury Interventions in Lower and Middle Income Countries: Opportunities and Challenges.” Cost NOTES Effectiveness and Resource Allocation 4: 2. Black, R. E., R. Laxminarayan, M. Temmerman, and N. Walker, World Bank Income Classifications as of July 2014 are as editors. 2016. Disease Control Priorities (third edition): follows, based on estimates of gross national income (GNI) Volume 2, Reproductive, Maternal, Newborn, and Child per capita for 2013: Health. Washington, DC: World Bank. Black, R. E., C. Levin, N. Walker, D. Chou, L. Liu, and • Low-income countries (LICs) = US$1,045 or less others. 2016. “Reproductive, Maternal, Newborn, and Child • Middle-income countries (MICs) are subdivided: Health: Key Messages from Disease Control Priorities, 3rd a) lower-middle-income = US$1,046 to US$4,125 Edition.” The Lancet 388 (10061): 2713–836. b) upper-middle-income (UMICs) = US$4,126 to US$12,745 Blackman, A., S. Newbold, J. Shih, and J. Cook. 2000. “The • High-income countries (HICs) = US$12,746 or more. Benefits and Costs of Informal Sector Pollution Control: Mexican Brick Kilns.” Discussion Paper 00-46, Resources 1. This chapter cites the source of burden estimates at each for the Future, Washington, DC. use, but these estimates change regularly as new data Bundy, D. A. P., N. de Silva, S. Horton, D. T. Jamison, become available and modeling tools improve. There and G. C. Patton, editors. 2017. Disease Control Priorities are some discrepancies between the estimates done by (third edition): Volume 8, Child and Adolescent Health and different organizations, namely, the Institute for Health Development. Washington, DC: World Bank. Metrics and Evaluation and the WHO, because of differ- Chafe, Z. A., M. Brauer, Z. Klimont, R. Van Dingenen, S. Mehta, ent assumptions and methods. The precision is generally and others. 2014. “Household Cooking with Solid Fuels kept at three places of significance to avoid rounding Contributes to Ambient PM2.5 Air Pollution and the Burden errors, but in reality true uncertainties are much larger. of Disease.” Environmental Health Perspectives 122 (12): 2. Note on terminology: Some definitions of premature 1314–20. deaths involve those deaths below a certain age, for exam- Debas, H. T., P. Donkor, A. Gawande, D. T. Jamison, M. E. Kruk, ple, younger than age 70 years. This table and the other and C. N. Mock, editors. 2015. Disease Control Priorities tables in this chapter consider all of the deaths to be pre- (third edition): Volume 1, Essential Surgery. Washington, mature but not relative to a specific threshold for age. DC: World Bank. 3. Except for figure 1.1, all analyses in this section are pre- Ditsuwan, V., J. L. Veerman, M. Bertram, and T. Vos. 2013. sented with age-standardized deaths per capita to normal- “Cost-Effectiveness of Interventions for Reducing Road ize across the four World Bank income regions using GBD Traffic Injuries Related to Driving under the Influence of 2015 data. Results are similar if using age-standardized Alcohol.” Value in Health 16 (1): 23–30. DALYs, however. Only environmental risks examined in Ebi, K., J. Hess, and P. Watkiss. 2017. “Health Risks and Costs of the GBD 2015 were included. It should be noted that con- Climate Variability and Change.” In Disease Control Priorities ducting the analysis using more subregions or by country (third edition): Volume 7, Injury Prevention and Environmental might show subtleties not revealed by comparison across Health, edited by C. N. Mock, R. Nugent, O. Kobusingye, and only four income regions. K. R. Smith. Washington, DC: World Bank. 4. Like the environmental risk factors, the occupa- Guimarães, L. D., J. L. Ribeiro, and J. S. Renner. 2012. “Cost-Benefit tional injury category was examined in a comparative Analysis of a Socio-Technical Intervention in a Brazilian risk assessment framework, that is, with a nonzero Footwear Company.” Applied Ergonomics 43 (5): 948–57. Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 21 Humphrey, J. 2009. “Child Undernutrition, Tropical Enteropathy, on Health, Equity and Medical Impoverishment.” In Toilets, and Handwashing.” The Lancet 3754 (9694): 1032–35. Disease Control Priorities (third edition): Volume 7, Injury Hutton, G. 2013. “Global Costs and Benefits of Reaching Prevention and Environmental Health, edited by C. N. Mock, Universal Coverage of Sanitation and Drinking-Water R. Nugent, O. Kobusingye, and K. R. Smith. Washington, Supply.” Journal of Water and Health 11 (1): 1–12. DC: World Bank. Hutton, G., and C. Chase. 2016. “The Knowledge Base for Omran, A. R. 1971. “The Epidemiologic Transition: A Theory Achieving the Sustainable Development Goal Targets of the Epidemiology of Population Change.” The Milbank on Water Supply, Sanitation and Hygiene.” International Quarterly 83 (4): 731–57. Journal of Environmental Research and Public Health Pant, K. P. 2011. “Cheaper Fuel and Higher Health Costs 13 (6): E536. among the Poor in Rural Nepal.” AMBIO 41 (3): 271–83. ———. 2017. “Water Supply, Sanitation, and Hygiene.” In Patel, V., D. Chisholm, T. Dua, R. Laxminarayan, and M. E. Disease Control Priorities (third edition): Volume 7, Injury Medina-Mora, editors. 2015. Disease Control Priorities (third Prevention and Environmental Health, edited by C. N. Mock, edition): Volume 4, Mental, Neurological, and Substance Use R. Nugent, O. Kobusingye, and K. R. Smith. Washington, Disorders. Washington, DC: World Bank. DC: World Bank. Patel, V., D. Chisholm, R. Parikh, F. J. Charlson, L. Degenhardt, Hutton, G., E. Rehfuess, and F. Tediosi. 2007. “Evaluation of the and others. 2015. “Addressing the Burden of Mental, Costs and Benefits of Interventions to Reduce Indoor Air Neurological, and Substance Use Disorders: Key Messages Pollution.” Energy for Sustainable Development 11 (4): 34–43. from Disease Control Priorities, 3rd Edition.” The Lancet Hutton, G., E. Rehfuess, F. Tediosi, and S. Weiss. 2006. 387 (10028): 1672–85. Evaluation of the Costs and Benefits of Household Energy and Petri, W., and M. Miller. 2008. “Enteric Infections, Diarrhea, Health Interventions at Global and Regional Levels. Geneva: and Their Impact on Function and Development.” Journal World Health Organization. of Clinical Investigation 118 (4): 1277–90. Jamison, D. T., J. G. Breman, A. R. Measham, G. Alleyne, Pillarisetti, A., D. T. Jamison, and K. Smith. 2017. “Household M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove, Energy Interventions and Health and Finances in editors. 2006. Disease Control Priorities in Developing Haryana, India: An Extended Cost-Effectiveness Analysis.” Countries (second edition). Washington, DC: Oxford In Disease Control Priorities (third edition): Volume 7, Injury University Press and World Bank. Prevention and Environmental Health, edited by C. N. Mock, Jamison, D. T., W. Mosley, A. R. Measham, and J. Bobadilla, R. Nugent, O. Kobusingye, and K. R. Smith. Washington, editors. 1993. Disease Control Priorities in Developing DC: World Bank. Countries (first edition). New York: Oxford University Press. Prabhakaran, D., T. Gaziano, J.-C. Mbanya, Y. Wu, and Rachel Lahiri, S., C. Levenstein, D. I. Nelson, and B. J. Rosenberg. Nugent, editors. 2017. Disease Control Priorities (third 2005. “The Cost Effectiveness of Occupational Health edition): Volume 5, Cardiovascular, Respiratory, and Related Interventions: Prevention of Silicosis.” American Journal of Disorders. Washington, DC: World Bank. Industrial Medicine 48 (6): 503–14. Rahman, F., S. Bose, M. Linnan, A. Rahman, S. Mashreky, Lahiri, S., P. Markkanen, and C. Levenstein. 2005. “The Cost and others. 2012. “Cost-Effectiveness of an Injury and Effectiveness of Occupational Health Interventions: Drowning Prevention Program in Bangladesh.” Pediatrics Preventing Occupational Back Pain.” American Journal of 130 (6): e1621–28. Industrial Medicine 48 (6): 515–29. Smith, K. 1990. “Indoor Air Quality and the Pollution Lelieveld, J., J. S. Evans, M. Fnais, D. Giannadaki, and A. Pozzer. Transition.” In Indoor Air Quality, edited by H. Kasuga. 2015. “The Contribution of Outdoor Air Pollution Sources Berlin, Heidelberg: Springer-Verlag. to Premature Mortality on a Global Scale.” Nature 525: Smith, K., M. A Desai, J. V. Rogers, and R. A. Houghton. 2013. 367–71. “Joint CO2 and CH4 Accountability for Global Warming.” Malla, M. B., N. Bruce, E. Bates, and E. Rehfuess. 2011. Proceedings of the National Academy of Sciences 110 (31): “Applying Global Cost-Benefit Analysis Methods to Indoor E2865–74. Air Pollution Mitigation Interventions in Nepal, Kenya Smith, K., and M. Ezzati. 2005. “How Environmental Health and Sudan: Insights and Challenges.” Energy Policy 39 (12): Risks Change with Development: The Epidemiologic and 7518–29. Environmental Risk Transitions Revisited.” Annual Review Mock, C. N., P. Donkor, A. Gawande, D. T. Jamison, M. E. of Environment and Resources 30: 291–333. Kruk, and others. 2015. “Essential Surgery: Key Messages Stevens, G., A. Wilson, and J. K. Hammitt. 2005. “A Benefit- from Disease Control Priorities, 3rd Edition.” The Lancet Cost Analysis of Retrofitting Diesel Vehicles with Particulate 385 (9983): 2209–19. Filters in the Mexico City Metropolitan Area.” Risk Analysis Newall, A. T., M. Jit, and R. Hutubessy. 2014. “Are Current Cost- 25 (4): 883–99. Effectiveness Thresholds for Low- and Middle-Income United Nations. 2015. Transforming Our World: The 2030 Agenda Countries Useful? Examples from the World of Vaccines.” for Sustainable Development. New York: United Nations. Pharmacoeconomics 32 (6): 525–31. Watkins, D., N. Dabestani, C. Mock, M. Cullen, K. Smith, Olson, Z., J. Staples, C. N. Mock, N. P. Nguyen, A. Bachani, and others. 2017. “Trends in Morbidity and Mortality and others. 2017. “Helmet Regulation in Vietnam: Impact Attributable to Injuries and Selected Environmental 22 Injury Prevention and Environmental Health Hazards.” In Disease Control Priorities (third edition): Priorities (third edition): Volume 7, Injury Prevention and Volume 7, Injury Prevention and Environmental Health, Environmental Health, edited by C. N. Mock, R. Nugent, edited by C. N. Mock, R. Nugent, O. Kobusingye, and O. Kobusingye, and K. R. Smith. Washington, DC: World Bank. K. R. Smith. Washington, DC: World Bank. WHO (World Health Organization). 2015. Global Status Watkins, D., N. Dabestani, R. Nugent, and C. Levin. 2017. Report on Road Safety 2015. Geneva: WHO. “Interventions to Prevent Injuries and Reduce Environmental ———. 2016. “Global Health Estimates (2012).” WHO, Geneva. and Occupational Hazards: A Review of Economic Evaluations World Bank. 1993. World Development Report 1993: Investing in from Low- and Middle-Income Countries.” In Disease Control Health. New York: Oxford University Press. Injury Prevention and Environmental Health: Key Messages from Disease Control Priorities, Third Edition 23 Chapter 2 Trends in Morbidity and Mortality Attributable to Injuries and Selected Environmental Hazards David A. Watkins, Nazila Dabestani, Charles N. Mock, Mark R. Cullen, Kirk R. Smith, and Rachel Nugent INTRODUCTION Self-harm is not covered in detail in this chapter because it is covered in volume 4, chapter 9, of Disease The effects of globalization on low- and middle-income Control Priorities (third edition) (Vijayakumar and others countries (LMICs) have led to major changes in the dis- 2015). Although the conditions presented above are seem- ease burden attributable to injuries and environmental ingly very different, a common feature links them: they risks. On the one hand, rapidly developing regions face a can all be addressed through multisectoral interventions, rising number of road traffic injuries (RTIs) and fatalities, including legal and regulatory frameworks and public as well as health effects from increasingly polluted air. On works investments. These interventions are assessed fur- the other hand, economic development has led to greater ther in the subsequent chapters of this volume. availability of water, sanitation, and hygiene (WASH) ser- This chapter presents two types of burden estimates. vices and a reduced burden of diarrheal and helminthic For injuries, we present deaths and disability-adjusted illness in many settings. These trends are heterogeneous, life years (DALYs). For occupational and environmental however, and very poor countries, and regions within hazards, we present attributable deaths and DALYs. The populous countries such as India, exhibit slower progress. distinction between these two types of estimates is that This chapter presents an overview of trends in the the former are related to specific causes of death, such burden of injuries and environmental health issues in as RTIs, whereas the latter are related to risk factors, LMICs. We focus on five major groups of conditions, such as unimproved water, for specific causes of death, presented as they appear in this volume and not in order such as diarrheal disease. Attributable deaths and of importance: DALYs are estimated using the comparative risk assess- ment (CRA) methodology rather than mortality analy- • Unintentional injuries, which include RTIs and those sis. They often total greater or less than 100 percent resulting from other causes owing to multiple risk factors (or no known risk fac- • Interpersonal and collective violence tors) for various causes of death. Hence, estimates of • Occupational hazards the burden of environmental and occupational risk • WASH-related illnesses factors cannot be directly compared to estimates of the • Health effects of air pollution. burden of injuries. Corresponding author: David A. Watkins, Department of Medicine, University of Washington, Seattle, Washington, United States; davidaw@uw.edu. 25 This chapter presents estimates from two sources of single-highest number of injuries worldwide in any data. Mortality and morbidity data on injuries are taken category, and their number continues to increase. from the World Health Organization’s (WHO) Global RTIs caused 1.1 million deaths and 70 million DALYs Health Estimates database, most recently updated in 2014 in 2012, accounting for 3 percent of total deaths and (WHO 2016). Details on the methods for estimating DALYs in LMICs. RTIs rank among the top 10 causes cause-specific mortality and DALYs—including calcula- of death globally, and LMICs constitute a higher pro- tion of years of life lost and disability weights—are portion of deaths. Most deaths occur among vulnera- available in the relevant documentation from the WHO. ble road users, such as pedestrians, motorcyclists, and Attributable mortality and morbidity data on occupa- cyclists. Poor design, traffic congestion, and lack of tional and environmental risks are taken from the Global road maintenance and safety systems in many LMICs Burden of Disease 2013 Study (GBD 2013) because make prevention more challenging (Peden and similar data were not available from the WHO. Details on others 2004). the methods for estimating attributable mortality and The United Nations Road Safety Collaboration DALYs using CRA are provided in Forouzanfar and others has highlighted the unequal effect of RTIs by socio- (2015). economic status, age, and gender. In general, adoles- We compare trends in total deaths with trends in age- cent and younger working-age adult males are most standardized mortality rates. For injuries, we calculated affected by RTIs, and the increase in deaths between age-standardized rates based on the global population 2000 and 2012 has been largest in this group structure in 2012 and compared mortality in 2000 to (table 2.1). Yet many countries do not regularly col- mortality in 2012. For occupational and environmental lect data on RTIs, thereby limiting awareness of the risks, we used age-standardized attributable rates from problem and potential interventions, such as speed GBD 2013, which based calculations on the global pop- limits, seat belt enforcement, and impaired-driving ulation structure in 2013 and compared attributable laws (WHO 2013). mortality in 1990 to attributable mortality in 2013. The number of RTI-related deaths and DALYs in LMICs increased by 36 percent and 23 percent, respec- tively, between 2000 and 2012 (table 2.1). This increasing UNINTENTIONAL INJURIES burden of RTIs is independent of demographic changes and is related to increasing age-specific mortality rates: Road Traffic Injuries the age-standardized mortality rate from RTIs increased Among unintentional injuries, RTIs remain the most 17 percent from 18 per 100,000 persons to 21 per common cause of deaths and DALYs. RTIs cause the 100,000 persons in LMICs (table 2.2). Notably, no Table 2.1 Road Traffic Injuries: Deaths and DALYs in LMICs by Age and Gender, 2000–12 Deaths DALYs Change Change Gender Age 2000 2012 (%) 2000 2012 (%) Both genders Total 836,500 1,134,800 36 57,241,000 70,385,200 23 (percentage of total deaths or DALYs) (2.0) (2.6) (2.3) (3.0) Male 0–27 days 1,300 1,300 0 120,000 117,000 −3 1–59 months 30,800 27,100 −12 2,820,000 2,485,000 −12 5–14 years 48,600 50,200 3 4,453,000 4,532,000 2 15–29 years 199,700 232,800 17 15,664,000 18,038,200 15 30–49 years 197,200 260,200 32 13,289,000 16,760,000 26 50–59 years 53,900 102,900 91 2,808,000 4,948,000 76 60–69 years 42,400 73,400 73 1,655,000 2,626,000 59 70+ years 41,800 78,500 88 952,000 1,600,000 68 Total 615,700 826,400 34 41,761,000 51,106,200 22 table continues next page 26 Injury Prevention and Environmental Health Table 2.1 Road Traffic Injuries: Deaths and DALYs in LMICs by Age and Gender, 2000–12 (continued) Deaths DALYs Change Change Gender Age 2000 2012 (%) 2000 2012 (%) Female 0–27 days 1,400 1,300 −7 128,000 120,000 −6 1–59 months 23,400 20,200 −14 2,137,000 1,847,000 −14 5–14 years 29,500 30,500 3 2,680,000 2,751,000 3 15–29 years 49,900 61,000 22 4,351,000 5,264,000 21 30–49 years 50,300 70,900 41 3,820,000 5,160,000 35 50–59 years 20,000 37,900 90 1,066,000 1,905,000 79 60–69 years 20,700 36,900 78 768,000 1,289,000 68 70+ years 25,600 49,700 94 530,000 943,000 78 Total 220,800 308,400 40 15,480,000 19,279,000 25 Source: WHO 2016. Note: DALYs = disability-adjusted life years; LMICs = low- and middle-income countries. Estimates and percentage changes may vary slightly because of rounding. Table 2.2 Injuries: Deaths and Mortality Rates by Cause in LMICs, All Ages and Both Genders, 2000–12 Change 2000 2012 (%) Total deaths ASMR per Total deaths ASMR per Cause (thousands) 100,000 persons (thousands) 100,000 persons Deaths ASMR All causes 41,800 1,050 44,200 880 6 −16 Injuries (unintentional and intentional) 4,130 93 4,410 81 7 −13 Unintentional injuries 2,920 66 3,220 60 10 −9 Road traffic injuries 800 18 1,140 21 43 17 Poisonings 220 5 160 3 −27 −40 Falls 440 12 580 12 32 0 Burns (fire, heat, and hot substances) 240 5 250 4 4 −20 Drownings 390 8 340 6 −13 −25 Exposures to forces of nature 0 0 0 0 0 100 Other unintentional injuries 830 18 750 14 −10 −22 Intentional injuries 1,210 27 1,190 21 −2 −22 Self-harm 680 16 610 11 −10 −30 Interpersonal violence 420 9 460 8 10 −11 Collective violence and legal intervention 110 2 120 2 9 0 Source: WHO 2016. Note: ASMR = age-standardized mortality rate; LMICs = low- and middle-income countries. Percentages may vary slightly because of rounding. Trends in Morbidity and Mortality Attributable to Injuries and Selected Environmental Hazards 27 apparent age-specific peak occurred in female deaths rates for nontransport unintentional injuries are all declin- from RTIs. Finally, although RTIs frequently lead to pre- ing substantially with the exception of falls. mature mortality, nonfatal outcomes contribute a sub- stantial proportion of total DALYs (Peden and others Poisonings 2004). The majority of RTI-related DALYs occurs among The burden of unintentional poisoning is declining, males ages 15–49. with age-standardized mortality rates declining 40 percent from 2000 to 2012. In 2012, LMICs experienced an estimated 163,000 deaths and 9.3 million DALYs. Other Unintentional Injuries Most poisoning cases continue to occur among children Although RTIs are the single leading cause of death by who have unintentionally gained access to toxic chemi- injury, nontransport unintentional injuries collectively— cals (Balan and Lingam 2012). including poisonings, falls, burns, and drownings—account for twice the number of deaths and DALYs as do RTIs Falls (tables 2.2 and 2.3). The proportion of deaths and DALYs After RTIs, falls are the most frequent cause of death and as a result of unintentional injuries is higher in LMICs than DALYs owing to unintentional injury, resulting in in high-income countries. Nontransport unintentional 577,000 deaths and 33.5 million DALYs in LMICs in injuries account for more than 6,700 deaths per day and 2012. Age-standardized mortality rates from falls have 2.4 million deaths annually—almost twice the number stagnated since 2000 in contrast to other unintentional of deaths from transport injuries and twice the number of injuries. A high incidence of falls has been observed in deaths from intentional injuries. Nontransport uninten- South-East Asia. Most falls occur among the elderly, even tional injuries also account for over 128 million DALYs in LMICs (Kalula and others 2011; Ranaweera and oth- annually—almost twice the number from transport inju- ers 2013). The burden of falls is also being driven by ries and from intentional injuries. Trends across individual population aging and is exacerbated by lack of treatment causes vary: deaths from falls and burns are increasing, for cognitive problems and by unsafe living environ- while deaths from poisonings and drownings are decreasing ments (Lau and others 2001). Work-related falls and (table 2.2). At the same time, age-standardized mortality other injuries are also a major problem in LMICs. Table 2.3 Injuries: DALYs by Cause in LMICs, All Ages and Both Genders, 2000–12 2000 2012 Total DALYs Percentage of Total DALYs Percentage of Change Cause (thousands) all DALYs (thousands) all DALYs (%) All causes 2,486,000 100 2,355,000 100 −5 Injuries (unintentional and intentional) 266,000 11 264,000 11 −1 Unintentional injuries 197,000 8 197,000 8 0 Road traffic injuries 57,000 2 70,000 3 23 Poisonings 13,000 1 9,000 0 −31 Falls 28,000 1 33,000 1 18 Burns (fire, heat, and hot substances) 17,000 1 17,000 1 0 Drownings 28,000 1 22,000 1 −21 Exposures to forces of nature 0 0 0 0 0 Other unintentional injuries 54,000 2 46,000 2 −15 Intentional injuries 69,000 3 67,000 3 −3 Self-harm 36,000 1 31,000 1 −14 Interpersonal violence 26,000 1 29,000 1 12 Collective violence and legal intervention 7,000 0 7,000 0 0 Source: WHO 2016. Note: DALYs = disability-adjusted life years; LMICs = low- and middle-income countries. Percentages may vary slightly because of rounding. 28 Injury Prevention and Environmental Health Burns average (table 2.2). At the same time, age-standardized In 2012, an estimated 245,000 deaths and 16.8 million mortality rates from interpersonal violence in LMICs DALYs in LMICs were attributable to burns. Deaths remained stable, suggesting that the increase in from burns have remained stable since 2000, especially violence-related deaths was due to demographic changes. in Sub-Saharan Africa, South-East Asia, Europe, and the A significant proportion of cases in LMICs are gender Eastern Mediterranean. Age-standardized mortality based (Jan and others 2010). Female infant deaths among rates from burns declined 20 percent, suggesting that infanticide victims are far more common. The mortality increases in numbers of deaths are being driven by pop- rates among males ages 15–29 are approximately five ulation growth or increases in death rates in specific times those among females in the same age group. groups only. Among injuries, burns are uniquely more Collective violence, which tends to be episodic, accounts prevalent among females. This pattern is due to differen- for a smaller fraction of total deaths and DALYs (less than tial exposure to unsafe cooking appliances and other 1 percent) compared with interpersonal violence. hazards in households in LMICs (Ahuja and Bhattacharya 2002; Ahuja, Bhattacharya, and Rai 2009; Ahuja, Dash, and Shrivastava 2011; Hyder and others 2009; Mabrouk, El Badawy, and Sherif 2000). Another important etiol- OCCUPATIONAL HEALTH RISKS ogy of burns among women is acid attacks, which have Occupational health encompasses numerous issues, received particular media attention in the past few years including chemical, biological, physical, and psychosocial and seem to be more frequent in South Asia (Acid hazards. Relatively more is known about occupational Survivors Foundation 2014). injuries because most countries track these in the aggre- gate, although some LMICs do not separate serious events Drownings and fatalities in the workplace from those of other origin. Drowning-related deaths (337,000) and DALYs (21.8 Substantially less information is available from LMICs million) in LMICs are decreasing. Death rates declined regarding occupational health for many of the same rea- by 13 percent and age-standardized mortality rates by sons that control remains problematic: knowledge is lim- 25 percent. Drowning continues to be most common ited; regulations are either nonexistent or unenforceable among children, adolescents, and young adult males. It because of lack of trained personnel; and the research often occurs as a result of risky behaviors and the exac- community has focused on more salient health issues, erbating effects of harmful alcohol use (Peden and such as infectious diseases or the emerging epidemic of McGee 2003). Drowning among children under age five noncommunicable chronic diseases. Noncommunicable occurs most frequently in settings where swimming edu- diseases may well have a workplace contribution, but this cation and child supervision are inadequate, such as is unstudied. Infectious diseases, such as tuberculosis, rural areas (Rahman and others 2012). have been linked to workplace exposures or living condi- tions, including migrant labor practices. Other Causes Existing estimates of occupational risks in LMICs Exposures to forces of nature and other unintentional suggest that they are generally increasing in parallel with injuries (not elsewhere classified) account for a substan- global trends (table 2.4). From 1990 to 2013, attributable tial additional fraction of total deaths and DALYs. deaths and DALYs from occupational risks increased by Because interventions for these conditions are not spe- 27 percent. They constituted 2.4 percent of total attrib- cifically addressed in Disease Control Priorities (third utable deaths and 5.7 percent of total attributable edition), they are not discussed further here. DALYs in LMICs in 2013. The single-largest causes of attributable deaths were occupational particulate mat- ter, gases, and fumes (196,000), occupational carcino- INTERPERSONAL AND COLLECTIVE gens (189,000), and occupational injuries (135,000). Health loss from carcinogens in particular increased VIOLENCE dramatically from 1990 to 2013. However, age- In 2012, LMICs experienced 461,000 deaths and standardized mortality rates from occupational risks 29 million DALYs related to interpersonal violence. decreased for all causes except occupational carcinogens, Although deaths from interpersonal violence, which is which increased 33 percent. This suggests that most of the major cause of intentional injury covered in chapter the increase in the number of attributable deaths is due 5 of this volume (Mercy and others 2017), are increasing to demographic changes, but that a real increase in worldwide, the rate of increase was 9 percent in LMICs exposure to carcinogens occurs independent of demo- from 2000 to 2012, which was faster than the global graphic patterns. Although occupational ergonomic Trends in Morbidity and Mortality Attributable to Injuries and Selected Environmental Hazards 29 Table 2.4 Occupational Risks: Estimated Attributable Deaths, Mortality Rates, and DALYs in LMICs by Cause, All Ages and Both Genders, 1990–2013 Deaths ASMR DALYs (thousands) per 100,000 persons (thousands) Change Change Change Cause 1990 2013 (%) 1990 2013 (%) 1990 2013 (%) Total 18,700 23,800 27 729 597 −18 880,000 837,000 −5 Total environmental and occupational risks 7,600 7,400 −3 280 188 −33 378,000 271,000 −28 Occupational risks 460 580 26 17 13 −24 37,000 47,000 27 Occupational asthmagens 60 50 −17 3 1 −67 2,600 2,500 −4 Occupational carcinogens 80 190 138 3 4 33 1,800 4,000 122 Occupational ergonomic factors — — — — — — 13,000 18,000 38 Occupational injuries 130 140 8 3 2 −33 8,000 8,100 1 Occupational noise — — — — — — 4,000 6,500 63 Occupational particulate matter, gases, and fumes 190 200 5 8 5 −38 6,600 8,200 24 Source: IHME 2015. Note: ASMR = age-standardized mortality rate; DALYs = disability-adjusted life years; LMICs = low- and middle-income countries; — = not available. Each of the six major occupational hazards is listed as a subcategory of “occupational risks,” which are a subset of “total environmental and occupational risks,” which are a subset of “total” attributable deaths and DALYs. Data from the Global Burden of Disease Study 2013 (IHME 2015) were used because similar data were unavailable from the World Health Organization’s Global Health Estimates database. Percentages may vary slightly because of rounding. factors and occupational noise were not attributable to and regulations. Decreasing gender gaps in many profes- any deaths, they were substantial contributors to DALYs sions are leading to more women being injured in the and appear to be a growing problem. workplace (Kelsh and Sahl 1996; Nordander 2008; In congruence with these estimates, the handful of Turgoose, Carter, and Stride 2006). Children are also empirical studies on occupational health hazards in at much greater risk in many LMICs: between 3 and LMICs that appear annually suggest an acceleration in 75 percent of children ages 7–14 years are informally risks as more dangerous trades move in the global mar- employed, depending on the country and occupation ketplace to regions of lower training and regulation. In (World Bank 2014). Child workers are at especially high particular, the emerging global supply chain in electron- risk because of growing bodies that are more susceptible ics, toys, and textiles is replete with chemical and physical to toxic and carcinogenic substances. Most child workers hazards, including heavy metals, solvents, plastics, noise, are employed in agriculture, where they are exposed to and heat (see chapter 6 in this volume, Abdalla and oth- strenuous labor and pesticides (ILO 2011). ers 2017). Further, a major factor driving occupational Finally, the number of people working in the informal injury trends in LMICs is the export of hazardous indus- sector worldwide is increasing, and such work is fraught tries to these countries because of lower wages. Workers with poor regulations, inadequate standards, and insuffi- in these hazardous positions number 1.52 billion—an cient availability of protective equipment (Charmes 2012; increase of 23 million since 2009 (ILO 2012). ILO 2002). Although research on the informal sector is The gender and age distributions of occupational challenging, it clearly shows that informal workers fre- injuries are noteworthy. Women are at higher risk for quently live in poverty, routinely face adverse working injuries and experience more severe injuries than men in conditions (Muntaner and others 2010), and generally many high-income workplaces, after accounting for job have limited access to health care (Noe and others 2004). tasks. Women also appear to be more severely affected Strategies for prevention and control of disease are gener- by occupational risks in the Middle East and North ally very limited for this population of globally staggering Africa, although the data are less satisfactory (Abdalla proportion. Occupational health and safety are generally and others 2017). However, many occupational injuries viewed as an area within the broad province of primary appear to implicate high-hazard sectors that tend to health care, rather than the focus of specialized, separate employ males, such as construction, fishing, and mining. strategies, as in high-income countries. Expanded training These industries often lack safety equipment, training, of the health care workforce in this domain is essential. 30 Injury Prevention and Environmental Health WATER, SANITATION, AND HYGIENE– and Miller 2008). Nonfatal enteric infections can have RELATED ILLNESSES long-term health consequences as well: pathogenic bacte- ria can cause inflammation in children’s intestines, reduc- Health loss owing to unsafe water, poor sanitation, and ing proper absorption of nutrition. This process generates poor handwashing practices is an important indicator of a cycle of malnutrition and enteropathy, which then con- overall population health and poverty in LMICs. The tributes further to under-five mortality and to chronic Millennium Development Goal target for improved nutritional deficiency that can extend into adulthood drinking water was met in 2010. However, many LMICs (Black and others 2008; Black and Victora 2013). still lack WASH intervention coverage because countries The social and psychological disadvantages of having have set different standards for improved water supply poor access to WASH services are also noteworthy. and sanitation (Roaf and Khalfan 2005). Disparities Significant stress can be traced to lack of improved water within countries persist, particularly in rural areas. sources and sanitation among the poor (Hutton and In rural Sub-Saharan Africa, for example, many individ- others 2014). Women generally bear the burden of col- uals must walk long distances to collect adequate drinking lecting water. Children who have difficulty accessing water, while piped water is readily available in urban water or practice open defecation can readily spread areas. waterborne diseases to others. Infected children are less Although the rate of open defecation has decreased likely to attend school. Further, low WASH intervention globally, it is still common practice in many rural areas. coverage has a large environmental effect: elimination in Poor hygiene practices, such as infrequent use of soap for bodies of water negatively affects ecosystems and dis- handwashing, exacerbate the spread of pathogens rupts natural resources (Corcoran and Nellemann 2011; (Strickland 2000). Helminths are commonly transmitted Rabalais and Turner 2013). through feces and drinking water sources and are particu- Table 2.5 provides estimates of morbidity and mor- larly problematic for agricultural workers and children in tality attributable to lack of WASH services. These esti- rural areas (Lozano and Naghavi 2010). As a consequence, mates represent a lower bound, because many other diarrheal illnesses are responsible for approximately health effects that are not easily measured but are none- 43 percent of under-five mortality in South and South- theless linked to WASH likely exist. The estimates East Asia and Sub-Saharan Africa (Humphrey 2009; Petri demonstrate the effect of having met the Millennium Table 2.5 Environmental Risks: Estimated Attributable Deaths, Mortality Rates, and DALYs in LMICs by Cause, All Ages and Both Genders, 1990–2013 Deaths ASMR DALYs (thousands) per 100,000 persons (thousands) Change Change Change Cause 1990 2013 (%) 1990 2013 (%) 1990 2013 (%) Total 18,700 23,800 27 729 597 −18 880,000 837,000 −5 Total environmental and occupational risks 7,600 7,400 −3 280 188 −33 378,000 271,000 −28 Air pollution 4,100 5,000 22 179 132 −26 145,000 133,000 −8 Ambient ozone pollution 100 180 80 6 6 0 2,000 4,000 100 Ambient particulate matter air pollution 1,600 2,400 50 68 62 −9 56,000 62,000 11 Household air pollution from solid fuels 2,800 2,900 4 122 79 −35 101,000 81,000 −20 Unsafe water, sanitation, and handwashing 2,700 1,400 −48 69 30 −57 190,000 83,000 −56 No handwashing with soap 1,000 500 −50 26 11 −58 70,000 31,000 −56 Unsafe sanitation 1,700 800 −53 45 17 −62 124,000 49,000 −60 Unsafe water source 2,400 1,200 −50 62 27 −56 169,000 75,000 −56 Source: IHME 2015. Note: ASMR = age-standardized mortality rate; DALYs = disability-adjusted life years; LMICs = low- and middle-income countries. Each of the major environmental hazards is listed as a subcategory. Data from the Global Burden of Disease Study 2013 (IHME 2015) were used because similar data were unavailable from the World Health Organization’s Global Health Estimates database. Percentages may vary slightly because of rounding. Trends in Morbidity and Mortality Attributable to Injuries and Selected Environmental Hazards 31 Development Goal target; attributable deaths and for a relatively larger proportion of chronic and acute DALYs have decreased by 48 percent and 56 percent, respiratory disease, the latter affecting children. respectively, and age-standardized mortality rates have Of all LMICs, India experiences the most significant decreased by 57 percent. Nevertheless, the health effect effect from air pollution because of population size, remains significant: 6 percent of deaths and 10 percent weak regulation, and rapid industrialization; two-thirds of DALYs in LMICs in 2013 were attributable to WASH- of the population still uses solid fuels (Smith 2015). At related illnesses. the same time, the major source of air pollution in less urbanized areas is household air pollution, which results from the use of solid fuels for cooking and heat in homes HEALTH EFFECTS OF AIR POLLUTION (Smith 2015). In contrast to outdoor air pollution, mor- Air pollution continues to have significant health effects bidity from household air pollution appears to be in LMICs, where it accounted for nearly 21 percent of decreasing in LMICs, as is suggested by a 20 percent attributable deaths and 16 percent of attributable reduction in DALYs from 1990 to 2013. However, most DALYs in 2013. Total attributable deaths from air pollu- of this reduction is due to the decline in background tion have increased 22 percent since 1990, while DALYs child pneumonia rates related to improvements in have decreased 8 percent. Further, age-standardized health care and nutrition. attributable mortality rates have decreased 26 percent Surveillance of air pollution is challenging for several (IHME 2015). reasons. Measurement devices are typically located in Table 2.5 shows recent estimates that about 2.5 million urban areas, so less is known about air pollution in rural deaths in LMICs in 2013 were attributable to airborne areas. Satellite observations combined with modeling, pollutants in public settings. In the disease burden esti- however, are becoming a major source of information on mates, air pollution contributes a significant proportion ambient air quality in rural and other unmonitored of deaths as a result of respiratory infections; chronic areas (Brauer and others 2016). Linking production, obstructive pulmonary disease; cerebrovascular disease; exposure, and health effects is problematic, because dis- ischemic heart disease; and cancers of the trachea, bron- persion of particulate matter can be widespread. Health chus, and lung. The forms of air pollution that have been effects per unit of pollution emitted are thought to be evaluated are (1) ambient particulate matter air pollu- greatest among people living close to household sources tion (approximately 2.6 million attributable deaths in and roads (Smith 2015). LMICs) in the form of particle and ozone pollution and Integrated exposure-response relationships are now (2) household air pollution from solid fuels (approxi- used to determine health burdens and suggest policies. mately 2.9 million deaths in LMICs), although other These relationships link results from epidemiological categories have not yet been assessed globally. (Because studies across a wide range of exposures to combustion of a degree of overlap, totals are less than the sum of particles—from ambient air pollution, secondhand individual components.) tobacco smoke, household air pollution, and active Overall, 90 percent of air pollution deaths are in smoking (Burnett and others 2014). These are linked, in LMICs. However, because household solid cookfuel use turn, to global models of population exposure based on is essentially confined entirely to LMICs, essentially all a wide range of data sources. The analyses suggest that effects occur in these settings. Ambient particulate mat- outdoor air pollution accounts for a relatively larger ter air pollution occurs in both rural and urban areas proportion of cardiovascular disease and cancer, while and is related to a variety of emission sources, including household air pollution accounts for acute (children motorized transport, power plants, industry, road and only) and chronic respiratory disease (IHME 2015). construction dust, brick kilns, garbage burning, and the like. Household air pollution occurs primarily in less urbanized areas and is related to the use of solid fuels CONCLUSIONS for cooking and heat in homes. It is also a major source of ambient pollution, causing at least one-fourth of Recent decades have seen dramatic shifts in the patterns ambient pollution exposures in India and China, for of health loss from injuries and occupational and envi- example. Thus, perhaps 15 percent of the burden ronmental hazards in LMICs: accounted to ambient pollution actually began in households in LMICs. • The burden of RTIs and falls, in particular, is increas- Ambient particulate matter air pollution accounts for ing substantially, resulting in a net increase in the a relatively larger proportion of cardiovascular and cere- overall burden of injuries despite a decline in the brovascular diseases. Household air pollution accounts burden of drownings and poisonings. 32 Injury Prevention and Environmental Health • In keeping with trends in economic development REFERENCES in LMICs, health loss from occupational hazards is Abdalla, S., S. S. Apramian, L. F. Cantley, and M. R. Cullen. increasing. The rapid growth in the effect of occu- 2017. “Occupation and Risk for Injuries.” In Disease Control pational carcinogens in these settings is of particular Priorities (third edition): Volume 7, Injury Prevention and concern. Environmental Health, edited by C. N. Mock, R. Nugent, • As a group, environmental risks are declining, in O. Kobusingye, and K. R. Smith. Washington, DC: World Bank. particular, risks related to unsafe water and poor Acid Survivors Foundation. 2014. Annual Report 2014. sanitation. Dhaka, Bangladesh. http://www.acidsurvivors.org • Household air pollution also appears to be declining, /images/frontImages/Annual_Report-2014.pdf. but it is being replaced by ambient particulate matter Ahuja, R., and S. Bhattacharya. 2002. “An Analysis of 11,196 air pollution from vehicles and industrial sources, Burn Admissions and Evaluation of Conservative particularly in urban areas in populous countries Management Techniques.” Burns 28 (6): 555–61. such as India and China. Ahuja, R., S. Bhattacharya, and A. Rai. 2009. “Changing Trends of an Endemic Trauma.” Burns 35 (5): 650–56. Ahuja, R., K. Dash, and P. Shrivastava. 2011. “A Comparative Designing interventions and policies to address inju- Analysis of Liquefied Petroleum Gas (Lpg) and Kerosene ries and occupational and environmental hazards Related Burns.” Burns 37 (8): 1403–10. requires up-to-date information on the relative magni- Balan, B., and L. Lingam. 2012. “Unintentional Injuries among tude of these conditions and their trends over time. Children in Resource Poor Settings: Where Do the Fingers Surveillance for many of these conditions is politically Point?” Archives of Disease in Childhood 97 (1): 35–38. and technically challenging, so the estimates presented in Black, R., L. Allen, Z. Bhutta, L. Caulfield, M. de Onis, this chapter likely reflect a lower bound on their total and others. 2008. “Maternal and Child Undernutrition: burden in LMICs. The need to set priorities around each Global and Regional Exposures and Health Consequences.” of these conditions should be explicitly linked to efforts The Lancet 371 (9608): 243–60. Black, R., and C. Victora. 2013. “Maternal and Child to improve information systems, both within the health Undernutrition and Overweight in Low-Income and sector and elsewhere, for example, in the labor and envi- Middle-Income Countries.” The Lancet 382: 427–51. ronment sectors. Brauer, M., G. Freeman, J. Frostad, A. von Donkelaar, The conditions presented in this volume are impor- R. V. Martin, and others. 2016. “Ambient Air Pollution tant contributors to the overall burden of disease in Exposure Estimation for the Global Burden of Disease LMICs. Injuries are responsible for about 9 percent of 2013.” Environmental Science and Technology 50 (1): 79–88. deaths and 7 percent of DALYs, whereas occupational Burnett, R. T., C. A. Pope III, M. Ezzati, C. Olives, S. S. Lim, and and environmental risks are responsible for about others. 2014. “An Integrated Risk Function for Estimating 29 percent of attributable deaths and 31 percent of the Global Burden of Disease Attributable to Ambient attributable DALYs. With some notable exceptions, their Fine Particulate Matter Exposure.” Environmental Health importance continues to increase in parallel with eco- Perspectives 122 (4). doi:10.1289/ehp.1307049. Charmes, J. 2012. “The Informal Economy Worldwide: nomic development, urbanization, and the epidemio- Trends and Characteristics.” Margin: the Journal of Applied logical transition. Hence, policies that focus on these Economic Research 6 (2): 103–32. conditions must also account for and attempt to address Corcoran, E., and C. Nellemann. 2011. Sick Water? The Central the complex social, demographic, and economic factors Role of Wastewater Management in Sustainable Development: that are driving health trends. A Rapid Response Assessment. Arendal, Norway: United Nations Environment Programme, United Nations Human Settlements Programme, GRID-Arendal. NOTE Forouzanfar, M., L. Alexander, H. R. Anderson, V. F. Bachman, S. Biryukov, and others. 2015. “Global, Regional, and World Bank Income Classifications as of July 2014 are as National Comparative Risk Assessment of 79 Behavioural, follows, based on estimates of gross national income Environmental and Occupational, and Metabolic Risks or (GNI) per capita for 2013: Clusters of Risks in 188 Countries, 1990–2013: A Systematic Analysis for the Global Burden of Disease Study 2013.” • Low-income countries (LICs) = US$1,045 or less The Lancet 386 (10010): 2287–323. • Middle-income countries (MICs) are subdivided: Humphrey, J. 2009.“Child Undernutrition, Tropical Enteropathy, a) Lower-middle-income countries = US$1,046 to Toilets, and Handwashing.” The Lancet 3754: 1032–35. US$4,125 Hutton, G., U. Rodriguez, A. Winara, N. Anh, K. Phyrum, b) Upper-middle-income countries (UMICs) = US$4,126 and others. 2014. “Economic Efficiency of Sanitation to US$12,745 Interventions in Southeast Asia.” Journal of Water, Sanitation • High-income countries (HICs) = US$12,746 or more. and Hygiene for Development 4 (1): 23–36. Trends in Morbidity and Mortality Attributable to Injuries and Selected Environmental Hazards 33 Hyder, A., D. Sugerman, P. Puvanachandra, J. Razzak, and Emergency Department Based Injury Surveillance System H. El Sayed. 2009. “Global Childhood Unintentional Injury in Nicaragua.” Injury Prevention 10 (4): 227–32. Surveillance in Four Cities in Developing Countries: A Pilot Nordander, C. 2008. “Gender Differences in Workers with Study.” Bulletin of the World Health Organization 87 (5): Identical Repetitive Industrial Tasks: Exposure and 345–52. Musculoskeletal Disorders.” International Archives of IHME (Institute for Health Metrics and Evaluation). 2015. Occupational and Environmental Health 81 (8): 939–47. Global Burden of Disease 2013 Study. Seattle, WA: IHME. Peden, M., and K. McGee. 2003. “The Epidemiology of ILO (International Labour Organization). 2002. Women Drowning Worldwide.” Injury Control and Safety Promotion and Men in the Informal Economy: A Statistical Picture. 10 (4): 195–99. Geneva: ILO. Peden, M., R. Scurfield, D. Sleet, D. Mohan, A. Hyder, and ———. 2011. Children in Hazardous Work: What We Know, others. 2004. World Report on Road Traffic Injury Prevention. What We Need to Do. International Programme on the Geneva: World Health Organization. Elimination of Child Labour. Geneva: ILO. Petri, W., and M. Miller. 2008. “Enteric Infections, Diarrhea, ———. 2012. “Global Employment Trends 2012: World Faces and Their Impact on Function and Development.” Journal a 600 Million Jobs Challenge, Warns ILO.” Press release, of Clinical Investigation 118 (4): 1277–90. ILO, Geneva, January 24. Rabalais, N., and E. Turner. 2013. Coastal Hypoxia: Jan, S., G. Ferrari, C. Watts, J. Hargreaves, J. Kim, and others. Consequences for Living Resources and Ecosystems. Coastal 2010. “Economic Evaluation of a Combined Microfinance and Estuarine Studies 58. Washington, DC: American and Gender Training Intervention for the Prevention of Geophysical Union. Intimate Partner Violence in Rural South Africa.” Health Rahman, F., S. Bose, M. Linnqn, A. Rahman, S. Mashreky, and Policy and Planning 26: 366–72. others. 2012. “Cost-Effectiveness of an Injury and Drowning Kalula, S., V. Scott, A. Dowd, and K. Brodrick. 2011. “Falls and Prevention Program in Bangladesh.” Pediatrics 130 (6): 1–10. Fall Prevention Programmes in Developing Countries: Ranaweera, A., P. Fonseka, A. Pattiya, and S. Siribaddana. Environmental Scan for the Adaptation of the Canadian 2013. “Incidence and Risk Factors of Falls among the Falls Prevention Curriculum for Developing Countries.” Elderly in the District of Colombo.” Ceylon Medical Journal of Safety Research 42 (6): 461–72. Journal 58: 100–6. Kelsh, M., and J. Sahl. 1996. “Sex Differences in Work-Related Roaf, V., and A. Khalfan. 2005. “Monitoring Implementation of Injury Rates among Electric Utility Workers.” American the Right to Water: A Framework for Developing Indicators.” Journal of Epidemiology 143 (10): 1050–58. Global Issue Papers 14, Heinrich Boll Foundation, Berlin. Lau, E., P. Suriwongpaisal, J. Lee, S. Das De, and M. Festin. Smith, K. R. 2015. “Report of the Steering Committee on Air 2001. “Risk Factors for Hip Fracture in Asian Men and Pollution and Health Related Issues.” Ministry of Health Women: The Asian Osteoporosis Study.” Journal of Bone and Family Welfare, Government of India, New Delhi. and Mineral Research 16 (3): 572–80. Strickland, G. 2000. Hunter’s Tropical Medicine and Emerging Lozano, R., and M. Naghavi. 2010. “Global and Regional Infectious Diseases, sixth edition. Philadelphia, PA: WB Mortality from 235 Causes of Death for 20 Age Groups in Saunders. 1990 and 2010: A Systematic Analysis for the Global Burden Turgoose, L., A. Carter, and C. Stride. 2006. “Encouraging an of Disease Study 2010.” The Lancet 380: 2095–128. Increase in the Employment of Women Returners in Areas Mabrouk, A., A. El Badawy, and M. Sherif. 2000. “Kerosene of Skill Shortage in Traditionally Male Industries.” DoTaI Stove as a Cause of Burns Admitted to the Ain Shams Burn University of Sheffield and Institute of Work Psychology, Unit.” Burns 26 (5): 474–77. Sheffield, U.K. Mercy, J., S. D. Hillis, A. Butchart, M. A. Bellis, C. L. Ward, and Vijayakumar, L., M. R. Phillips, M. M. Silverman, D. Gunnell, and others. 2017. “Interpersonal Violence: Global Impact and V. Carli. 2015. “Suicide.” In Disease Control Priorities (third Paths to Prevention.” In Disease Control Priorities (third edition): Volume 4, Mental, Neurological, and Substance edition): Volume 7, Injury Prevention and Environmental Use Disorders, edited by V. Patel, D. Chisholm, T. Dua, Health, edited by C. N. Mock, R. Nugent, O. Kobusingye, and R. Laxminarayan, and M. E. Medina-Mora. Washington, DC: K. R. Smith. Washington, DC: World Bank. World Bank. http://www.dcp-3.org/mentalhealth. Muntaner, C., O. Solar, C. Vanroelen, J. Martinez, M. Vergara, WHO (World Health Organization). 2013. “Road Safety: Basic and others. 2010. “Unemployment, Informal Work, Facts.” Fact Sheet #1, WHO, Geneva. Precarious Employment, Child Labor, Slavery, and Health ———. 2016. Global Health Estimates 2012 (database). Inequalities: Pathways and Mechanisms.” International WHO, Geneva. http://www.who.int/healthinfo/global Journal of Health Services 40 (2): 281–95. _burden_disease/en/. Noe, R., J. Rocha, C. Clavel-Arcas, C. Aleman, M. Gonzales, World Bank. 2014. World Development Indicators: Children at and others. 2004. “Occupational Injuries Identified by an Work. Washington, DC: World Bank. 34 Injury Prevention and Environmental Health Chapter 3 Road Traffic Injuries Abdulgafoor M. Bachani, Margie Peden, G. Gururaj, Robyn Norton, and Adnan A. Hyder INTRODUCTION further inform the global discourse on reducing RTIs worldwide, with a special focus on LMICs, where Road traffic injuries (RTIs) are the leading cause of 90 percent of fatal RTIs occurred yet only 54 percent of unintentional injuries, accounting for the greatest pro- global vehicles were registered (WHO 2015a). portion of deaths from unintentional injuries. They are the leading cause of injury-related disability-adjusted life years (DALYs), and they pose a significant economic and HEALTH BURDEN OF ROAD TRAFFIC societal burden. Despite this burden, RTIs remain a largely neglected public health problem, especially in INJURIES low- and middle-income countries (LMICs), where Each day, more than 3,400 people die on the world’s urbanization and motorization are rapidly increasing. roads (1.25 million people each year), making RTIs Unfortunately, reliable data on the burden of RTIs and the ninth leading cause of death globally (WHO 2014). cost-effective interventions in LMICs are sorely lacking. The global rate of mortality resulting from RTIs has In 2010, global efforts to reduce the burden of road increased 46 percent since 1990 (Lozano and others safety injuries received a major boost when the United 2012). Latest estimates from the Global Health Estimates Nations (UN) General Assembly launched the Decade of (WHO 2014) show that road traffic crashes were respon- Action for Road Safety 2011–2020, with a goal of saving sible for 24 percent of all injury-related deaths globally 5 million lives worldwide by 2020 (United Nations Road (figure 3.1)—and a total of 78.7 million DALYs lost Safety Collaboration 2010). Since then, awareness of in 2012, up from 69.1 million in 2000 (WHO 2014). road safety and its close relationships to economic and Current trends suggest that RTIs will become the seventh social development has grown significantly, and activi- leading cause of death by 2030 unless action is taken ties that promote road safety at international and (WHO 2015a). national levels have gained new momentum. Across World Health Organization (WHO) regions, This chapter uses the latest global and regional esti- the highest road traffic mortality rate was in Africa (26.6 mates to characterize the burden of RTIs, including their per 100,000 population); the lowest was Europe (9.3 per mortality; morbidity; and economic and social impacts 100,000) (WHO 2015a). Over the past two decades, in on individuals, families, and society. It summarizes eco- the absence of effective road safety programs, mortality nomic evidence on proven and promising interventions resulting from RTIs has increased steadily in East Asia, that address the burden. The goal of this chapter is to South Asia, and Eastern and Western Sub-Saharan Africa Corresponding author: Abdulgafoor M. Bachani, Department of International Health, Johns Hopkins International Injury Research Unit, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, United States; abachani@jhu.edu. 35 (Odero, Khayesi, and Heda 2003; WHO 2014). This trend Barberia 2013; WHO 2013a, 2014).Importantly, within contrasts with that in high-income countries (HICs), the same region, considerable disparity exists in death where road traffic fatalities are on a downward trajectory rates across countries of different income status. In following the implementation of safety programs over Europe, for example, low-income countries (LICs) the past decade (table 3.1) (Garcia-Altes, Suelves, and had RTI mortality rates more than twice those for HICs (18.8 per 100,000 versus 8.3 per 100,000, respectively) Figure 3.1 Global Mortality from All Injuries, 2012 (WHO 2015a).1 LMICs overall bear a disproportionally high bur- den of RTIs (Hyder, Labinjo, and Muzaffar 2006; Hyder, Muzaffar, and Bachani 2008; Hyder and others 2013; Hyder and Peden 2003; WHO 2013a). They have Intentional Road injuries injuries (self-harm, 24.4% a little more than 50 percent of the world’s vehicles assault, violence) but more than 90 percent of the road traffic deaths 27.8% (WHO 2015a). More than twice as many individuals Poisonings per 100,000 population die from RTIs in LMICs com- 3.8% pared to HICs (WHO 2014, 2015a) (table 3.1). Even Other within HICs, individuals from lower socioeconomic Falls unintentional backgrounds are more likely to be involved in road 13.5% injuries 18.1% traffic crashes than their more affluent counterparts Burns (WHO 2015b). 5.2% All types of road users are at risk of RTIs, but marked differences exist in the fatality rates. In particular, vul- Drownings nerable road users (such as pedestrians and users of 7.2% two-wheelers) are at greater risk compared to motor- Source: WHO 2014. vehicle occupants, and they usually bear the greatest Table 3.1 Death Rates and Rates of DALY Losses Resulting from Road Traffic Injuries, by Gender and Income, 2012 and 2000 2012 Global LMICs HICs Total Men Women Total Men Women Total Men Women Deaths (per 100,000 population) All unintentional injuries 52.5 67.0 37.8 55.6 70.6 40.1 39.0 50.3 28.0 RTIs 17.7 25.6 9.7 19.6 28.2 10.8 9.2 13.9 4.8 DALY losses (per 100,000 population) All unintentional injuries 3,211.4 4,216.2 2,190.2 3,434.0 4,477.8 2,361.6 2,216.5 3,011.7 1,446.4 RTIs 1,112.6 1,603.8 613.5 1,217.3 1,744.2 676.0 644.8 957.3 342.1 2000 Deaths (per 100,000 population) All unintentional injuries 57.7 74.6 40.4 60.1 76.4 43.3 47.5 66.9 29.0 RTIs 16.7 24.3 8.9 17.0 24.7 9.1 15.4 22.9 8.3 DALY losses (per 100,000 population) All unintentional injuries 3,772.6 4,942.0 2,585.8 4,008.3 5,152.6 2,830.0 2,807.4 4,047.0 1,621.7 RTIs 1,129.0 1,636.1 614.3 1,163.0 1,672.4 638.4 989.9 1,481.8 519.4 Source: WHO 2014. Note: DALY = disability-adjusted life year; HICs = high-income countries; LMICs = low- and middle-income countries; RTIs = road traffic injuries. 36 Injury Prevention and Environmental Health burden of injury (Peden and others 2004). For example, and HICs in 2012, and the gender disparity has persisted almost 50 percent of the global road traffic deaths over the past decade (table 3.1) (WHO 2014). occur among vulnerable road users—motorcyclists (23 percent), pedestrians (22 percent), and cyclists (4 percent) (WHO 2015a). In many LMICs, where ECONOMIC AND SOCIETAL BURDEN OF the proportion of vulnerable road users is as high as ROAD TRAFFIC INJURIES 57 percent, few, if any, interventions are in place to pro- tect these road users; pedestrian deaths account for Economic Burden almost 40 percent of all road injury fatalities in LICs In addition to the health burden, RTIs account for and about 20 percent in middle-income countries profound economic costs to individuals, families, and (Bachani, Koradia, and others 2012; Bachani, Zhang, societies. In resource-constrained settings, assessing RTI- and others 2014; Hyder and Bishai 2012; Hyder, Ghaffar, related costs would help policy makers and health plan- and Masood 2000; WHO 2013a). ners to prioritize and choose the most appropriate Definite data on the number of people who survive interventions to control and prevent RTIs (Bishai and RTIs but live with disabilities are almost nonexistent. Bachani 2012). However, accurately quantifying these However, estimates suggest that for every one RTI- costs is not easy. The tangible costs—direct costs, such as related death, an additional 20–50 more individuals medical costs, and indirect costs, including lost produc- suffer some disability (Peden and others 2004). The tivity and economic opportunity—can be estimated in WHO estimates that RTIs accounted for a total of almost economic terms; the intangible costs associated with suf- 14 million life years lost annually due to disability fering and pain, however, often are more difficult to assess. in 2012 globally; RTIs represented 30 percent of the Three approaches have been developed to estimate costs injury-related disability burden (WHO 2014). of injury: the human capital, willingness-to-pay, and gen- Empirical evidence in LMICs (although limited eral equilibrium frameworks (Bishai and Bachani 2012). and with varied quality) supports these estimates. For example, a study in Arkhangelsk, the Russian • The human capital approach estimates the aggre- Federation, that investigated trends in traffic crashes gated injury costs at societal, national, and regional between 2005 and 2010 found 217 fatalities and 5,964 levels as the sum of the costs at the individual level, non-fatal injuries. The study used police data, which including direct medical costs, indirect lost produc- was considered the most reliable existing data source tivity costs, and intangible psychological costs of pain for this purpose (Kudryavtsev and others 2013). and suffering. The strategies for measuring pain and Another study in China (using a national disability suffering in this model are not fully developed, and survey) estimated the prevalence of RTI-related dis- most studies using this approach exclude this compo- ability to be 1.12 per 1,000 population in 2006 (Lin nent. Because of its structured nature and the ability and others 2013). Given the high burden of disability to compartmentalize costs into different categories, associated with RTIs, better measurement of this dis- the human capital framework remains the most ability is necessary not only to highlight but also to common approach to value RTI-related injury and develop appropriate strategies for addressing this death, especially in LMICs (Bishai and Bachani 2012). burden. Recent applications of approaches to obtain- • The willingness-to-pay approach estimates the value ing empirical population-level data on the prevalence of pain and suffering by asking what people would be and impact of disability in LMICs is a step in the right willing to pay to live in a world with a lower risk of direction (Bachani, Galiwango, and others 2014, 2015; injuries. By placing monetary values on injuries that Madans and Loeb 2013; Madans, Loeb, and Altman are grounded in the consumers’ own preferences, this 2011). approach provides an option for including estimates The significant burden of RTIs in terms of both pre- of the value of pain and suffering to determine the mature mortality and disability is attributable to the fact cost of injuries. that young adults (ages 15–44 years) are among the • The general equilibrium approach provides strategies most affected age group. More than 460,000 young peo- for actually measuring the costs from a broader mac- ple under age 30 years die in road traffic crashes each roeconomic perspective using simulation-modeling year—about 1,262 a day (WHO 2007, 2013b, 2014). techniques. The estimates using this approach are a Among them, more than 75 percent of the deaths occur dynamic assessment of the present value of forgone among young men (WHO 2015b). The rates of both consumption opportunities resulting from injuries. injury-related death and DALY losses were about three However, this approach has not been applied to esti- times higher among men than women in both LMICs mating costs of injuries. Road Traffic Injuries 37 Comparisons across these approaches are not appro- ranged between US$0.02 million to US$0.10 million priate because of the different methodologies and differ- (Mofadal, Kanitpong, and Jiwattanakulpaisarn 2015). ent level of data (micro versus macro) used in the three Although these studies clearly demonstrate the adverse measures. impact of RTIs on economic and social development, Because of the demand for epidemiologic data on more studies and improved health information systems the number and nature of RTIs, as well as the challenges in LMICs are needed to document and understand the of measuring intangible costs, few studies have full extent and nature of this burden. attempted to estimate RTI-related costs, but this has been changing over the past decade. One large 21-country study estimated that the global cost of RTIs Societal Burden was US$518 billion; the costs of RTIs at the national Despite the progress made in understanding the epi- level in most cases exceeded 1 percent of the gross demiology and economic burden of RTIs, understand- national product (GNP) (Jacobs, Aeron-Thomas, and ing of the long-term societal impact of RTIs remains Astrop 2000). Another study that used the human capi- inadequate. Evidence of the significant societal impact of tal approach in 11 HICs gave an average cost equivalent RTIs is limited and mostly available only for HICs. For to 1.3 percent of the GNP in the 1990s—ranging from example, the European Commission estimates that more 0.5 percent for the United Kingdom to 2.8 percent for than 30,000 people were killed and more than 120,000 Italy (Elvik 2000). More recent studies in Australia were permanently disabled by RTIs in 2011; as a result, (Connelly and Supangan 2006), the Republic of Korea nearly 150,000 families struggled with the consequent (Lim, Chung, and Cho 2011), New Zealand (O’Dea and devastation (European Commission 2014). A similar Wren 2010), and the United States (Blincoe and others study in the United Kingdom estimated that about 2015) have also highlighted the significant burden that 1.1 percent of the total population (more than 130,000 RTIs impose on a nation’s economy. A WHO analysis individuals) in the whole of England and Wales had lost reveals similar economic burden of RTIs across a close family member in a fatal RTI since 1971, subject- countries—ranging from 0.2 percent of the gross ing many of them to mental health and other conse- domestic product (GDP) in Chile and Jamaica to quences (Sullivan and others 2009). 7.8 percent in South Africa (WHO 2015a). In LMICs, because of the scarcity of good medical Cost studies on RTIs in LMICs often are scant care, rehabilitation services, and financial protection because of the poor capacity of health information sys- mechanisms, individuals often rely heavily on their tems in these settings (WHO 2013a). Studies show that social networks for support. In these settings, injuries RTIs cost approximately US$89.6 billion a year (in 2012 often have far-reaching implications that need to be US$) in LMICs, or 1–2 percent of their GNPs (Jacobs, understood to better address the burden. Studies exam- Aeron-Thomas, and Astrop 2000). The high RTI-related ining the social impact of RTIs in LMICs are almost costs as a share of GNP have also been shown in a nonexistent (Peden and others 2004). However, those few country-specific studies, including Bangladesh that do exist show that road traffic crashes and resultant (Mashreky and others 2010), Belize (Perez-Nunez and deaths or disabilities can take a heavy toll on families and others 2010), China (Zhou and others 2003), Uganda friends of injured persons, many of whom experience (Benmaamar, Dunkerley, and Ellis 2002), and Vietnam adverse financial, physical, social, and psychological (Nguyen and others 2013; Nguyen and others 2015). stresses. For example, families and friends of injured Using the human capital approach, researchers in persons reallocate work or change work patterns to Vietnam estimated that each RTI cost about 6 months of provide care. Often, debts are incurred because of the average salary during hospitalization (US$420 [in 2012 expensive rehabilitation services and reduced income US$]), and the average costs during recovery (12 months (Mock, Arreola-Risa, and Quansah 2003). Children in after hospital discharge) were equivalent to an entire these households can be pressured to leave school or can year of income (US$919 [in 2012 US$]) (Nguyen and suffer from decreased supervision (Mock, Arreola-Risa, others 2013, Nguyen and others 2015). Similarly, the and Quansah 2003). total economic costs of injury including direct and indi- rect costs in Belize represented 0.9 percent of the GDP in 2007 (Perez-Nunez and others 2010). In addition, RISK FACTORS FOR ROAD TRAFFIC INJURIES researchers using the willingness-to-pay approach esti- The Haddon matrix revolutionized the understanding of mated that each motorist fatality cost $0.55 million (in the multifactorial nature of the causes and risk factors of 2012 US$) in Malaysia (Mohd Faudzi, Mohamad, and RTIs, and it has made a substantial contribution to the Ghani 2011), and the value of a Sudanese pedestrian reduction of RTIs (Haddon 1968, 1973). The matrix 38 Injury Prevention and Environmental Health Table 3.2 Risk Factors of Road Traffic Injuries: The Haddon Matrix Environment Agent (vehicles Phase Host (human) and equipment) Physical Socioeconomic Precrash Speeding Insufficient lighting Flaws in road design (for example, Lack of comprehensive lack of lane separation) traffic safety law Driving while impaired (for Compromised braking Flaws in road layout (for example, Inadequate licensing system example, alcohol-impaired driving) lack of separation of vehicles and for drivers vulnerable road users) Inexperienced and young drivers Inadequate Improper speed limits Economic pressure (for maintenance example, social deprivation) Distracted driving Lack of pedestrian facilities Poor road user eyesight Substance use Lack of alternative modes of traveling Insufficient visibility Crash Failure to use restraints (for Lack of occupant Non-forgiving roadside (for example, seatbelt, child seat) restraints example, lack of crash barriers) Failure to wear a helmet Compromised braking Poorly designed and maintained Insufficient crash- roads protective design Postcrash No first-aid skills Fire risk Inadequate rescue facilities Inadequately trained EMSa and rehabilitation personnel Lack of access to medical Leakage of hazardous Congestion Inadequate prehospital care personnel materials a. Emergency Medical Services. provided a framework to integrate the traditional epi- Crash Risk Factors demiological triangle of host, vector, and environment Risk factors at the crash phase mainly affect the out- with a temporal perspective in terms of precrash, crash, comes in terms of injury severity and fatality. Risk and postcrash phases (table 3.2) (Haddon 1973). This factors at the individual level include failure to use seat- approach facilitates the analysis of potential interven- belts, helmets, and child restraints. Vehicles without tions covering the spectrum from primary prevention to occupant restraints and crash-protective design or with rehabilitation. The matrix has been broadly applied in compromised braking lead to a higher risk of injury both HICs and LMICs to assist with a systematic under- death and more severe disability. At the environmental standing of the epidemiology and risk factors, and to level, poorly designed and maintained roads, low visibil- facilitate the ability to prioritize the most appropriate ity, and lack of crash-protective roadside objects also put preventive and curative measures (Brice and others 2011; road users in danger. Although failure to use seatbelts, Chorba 1991; Short, Mushquash, and Bedard 2013). helmets, or child restraints significantly increases risk of RTIs and deaths among vehicle occupants, many LMICs have no mandatory requirements; even if they do, com- Precrash Risk Factors pliance and law enforcement often are limited (Peden Risk factors at the precrash phase include those that and others 2004). predispose individuals to be involved in a crash. At the individual level, these include speeding, driving while impaired, driving while distracted, being inexperi- Postcrash Risk Factors enced or young, and using substances; at the vehicle While preventing road traffic crashes is always desirable, level, these include compromised braking and inade- a comprehensive road safety strategy is incomplete with- quate lighting and maintenance; and at the environ- out a focus on improving postcrash care for injured mental level, these include both physical and persons to reduce fatalities and improve outcomes. socioeconomic factors (Herbert and others 2011). Many LMICs lack appropriate and adequate postcrash Road Traffic Injuries 39 care, contributing to the high burden of deaths and dis- • Improving postcrash outcomes, from appropriate ability resulting from RTIs (Khorasani-Zavareh and and life-saving measures at the scene of the crash others 2009; Miranda and others 2013; Paravar and oth- through rehabilitation services. ers 2013; Solagberu and others 2014). In 2007, global efforts to improve postcrash care, In addition to these fundamental principles, political including trauma and emergency care services, gained will and commitment are essential to reducing the bur- major momentum when a World Health Assembly den of RTIs. adopted a resolution that called on governments and the The Decade of Action for Road Safety 2011–2020 WHO to increase their efforts to improve care for vic- adopts a systems approach to addressing the burden of tims of injuries and other medical emergencies (WHO RTIs, and proposes five pillars: road safety management, 2011). It also called on the WHO to raise awareness safer roads and mobility, safer vehicles, safer road users, about affordable ways in which trauma and emergency and postcrash care (United Nations Road Safety care services can be strengthened, especially through Collaboration 2010). universally applicable means, such as improvements in organization and planning (WHO 2011). Other studies from LMICs have highlighted a similar need and oppor- Safer Road Users tunities to improve care for injured patients (Hyder and Effective legislation that establishes safety codes and Razzak 2013). Documented case studies have shown that punishes unsafe behavior is the first and foremost inter- improvements can be made even in the poorest and vention needed to reduce RTIs. Currently, 91 out of most difficult settings (Mock and others 2010). For 180 countries have national laws that address the key risk example, the simple administration of tranexamic acid factors, including speeding; driving under the influence; to actively bleeding patients in the acute care phase could and failing to use motorcycle helmets, seat belts, and prevent thousands of premature deaths (Ker and others child restraints. Since 2011, 17 countries have amended 2012). Therefore, implementing interventions based on their laws on one or more key risk factors for RTIs to the assessment of risk factors, together with good post- bring them in line with best practice (WHO 2015a). crash care practices, has the potential to save and However, little progress has been made globally in improve the lives of RTI victims and move closer to the extending the coverage of national laws to include all key goal of the Decade of Action for road safety (United risk factors (WHO 2015a). Nations Road Safety Collaboration 2010). Encouraging a culture of safe road behavior guided by legislation requires not only a high level of enforce- ment but also a high public perception of enforcement INTERVENTIONS (WHO 2013a). A large body of research (although few Most road traffic deaths and serious injuries are prevent- studies were conducted in LMICs) shows that: able, because crash risk is largely predictable; therefore, many proven or promising countermeasures can be imple- • Establishing and enforcing speed limits according to mented. RTIs respond well to targeted interventions that designated functions of the roads can reduce RTIs by prevent the occurrence of the injury, minimize the severity up to 34 percent (WHO 2013a). of the injury sustained, and mitigate the sequelae. • Setting legislative limits on blood alcohol concen- Although no blueprint for road safety exists, a broad trations at 0.05 grams per deciliter (g/dl) and con- consensus exists on several principles for interventions: ducting random breath tests can significantly reduce alcohol-related RTIs (Elvik and others 2009; Shults • Reducing risk exposure by stabilizing motorization and others 2001). Despite global progress in strength- levels, providing alternative modes of travel, and ening legislation that penalizes alcohol-impaired improving land-use planning practices driving, LMICs are less likely than HICs to adopt the • Reducing risk factors directly related to crash practices (WHO 2013a). causation, such as speeding, drinking and driving, • Introducing and enforcing the use of motorcycle using unsafe vehicles on unsafe roads (with inade- helmets can reduce the risk of death by 40 percent quate safety features for the traffic mix), and failing and the risk of serious head injuries by more than to enforce road safety laws effectively 70 percent, yet LMICs are less likely to adopt the • Reducing severity of injuries by mandating and practices (Liu and others 2008). enforcing the use of seat belts, child restraints, and • Introducing and enforcing the use of seatbelts can helmets, as well as by improving road infrastructure reduce the risk of fatal injuries by up to 50 percent for and vehicle design to protect all road users front seat occupants and up to 75 percent for rear seat 40 Injury Prevention and Environmental Health occupants (Zhu and others 2007). Although most coun- Safer vehicles in LMICs are scarce, however, because tries have mandatory seatbelt laws, the legislation often of costs and inadequate government safety regulations does not extend to rear seat occupants (WHO 2013a). on the automotive industry (IIHS 2013). For example, • Mandating the use of child restraints can reduce the Latin New Car Assessment Program (NCAP) evalu- the likelihood of a fatal crash for children by up to ated car models in the Latin America market and found 80 percent (Zaza and others 2001). However, such that those earning the lowest rating in safety equipment laws do not always exist in LMICs. For example, only (one out of five stars) were among the top selling cars 1 out of 10 South-East Asian countries has a law (IIHS 2013). Furthermore, while frontal airbags for the requiring child restraints (WHO 2015a). driver and front passenger have been standard equip- ment on vehicles in the United States since 1999, they Effective enforcement of traffic laws in low-resource typically were optional equipment on car models in settings could provide economic benefits. Research LMICs (IIHS 2013). shows that observance of traffic codes (Bishai and In addition to four-wheeled vehicles, the surge of Hyder 2006) and the use of motorcycle helmets (Bishai motorized two-wheelers (motorcycles and electric bikes, and Hyder 2006) and seatbelts (Chisholm and others or e-bikes) in LMICs is even more concerning, especially 2012) can be very cost-effective in preventing RTIs in in South-East Asia and Africa. For example, in Malaysia LMICs. While a paucity of good evidence in LMICs of and Thailand, these vehicles were adopted at a ratio the effectiveness of education exists (as indicated by a of three persons per vehicle and four persons per vehicle, systematic review of 15 randomized controlled trials on respectively, in 2011 (Sekine 2014). Both countries had the effectiveness of safety education programs), some significantly higher fatality rates in motorcycles crashes: have testified to the synergistic effects of approaches 62 percent in Malaysia and 73 percent in Thailand that combine education with legislation and enforce- (WHO 2015a). ment (Duperrex, Bunn, and Roberts 2002; Sedlák, Looking to address the safety of vehicles in LMICs, Grivna, and Cihalova 2006). the Global NCAP offers a stakeholder movement (as part of the UN’s Decade of Action for Road Safety) to encourage adoption and enforcement of harmonized Safer Vehicles motor vehicle standards in LMICs to promote safer vehi- More than 1.8 billion vehicles are registered globally, and cles (NCAP 2011). more than half of them are in LMICs (WHO 2015a). The increasing demand for mobility has led to rapid motorization (especially in LMICs), creating challenges Safer Infrastructure for safer transport. Strategies focusing on safer vehicles Poorly designed road networks that lack sufficient safety have expanded, from protecting those inside of vehi- measures significantly increase RTIs. Results of the cles to protecting those outside of vehicles. As auto- International Road Assessment Program in LMICs show makers have refined advanced technology designed to that about half of the roads assessed in these countries prevent or mitigate crashes, they have introduced it into are rated in the highest risk category, largely because passenger vehicle models. While limited data on the 84 percent of the roads assessed where pedestrians are effectiveness of safety technologies exists, some (such as present have no footpaths (WHO 2013a). This contrib- crash avoidance systems) showed the potential to miti- utes in part to the high proportion (60 percent) of all gate RTIs (Jermakian 2011; WHO 2013a). A study in road traffic deaths in these countries among vulnerable France shows that while public safety measures (such as road users (WHO 2013a). speed cameras) contributed to a greater than 75 percent A growing number of countries have amended their reduction in road crash fatalities, enhanced vehicle safety national transport policies to encourage alternative modes technologies directly saved 27,365 car occupants and of transport, such as walking and cycling, or to increase 1,083 pedestrian from fatal crashes between 2000 and investment in public transport systems to deal with 2010 (Page, Hermitte, and Cuny 2011). Furthermore, a increased motorization and RTIs (WHO 2013a). However, literature review on road safety interventions shows that these approaches often have lacked the appropriate strat- electronic stability control systems were associated with egies for heterogeneous traffic environments or the a 2–41 percent reduction in RTIs (Novoa, Pérez, and required resources to ensure the safety of vulnerable road Borrell 2009). The study also noted that the most suc- users; these deficits have the potential to counteract the cessful interventions are those that reduce or eliminate intended effect of the interventions (WHO 2013a). For the hazard of RTIs and do not rely on changes in road example, separating vulnerable road users (pedestrians, users’ behavior (Novoa, Pérez, and Borrell 2009). motorcyclists, and cyclists) from larger and faster vehicles Road Traffic Injuries 41 while promoting programs such as city cycling has been Proven and Promising Interventions shown to reduce injuries and fatalities (Herrstedt 1998; The World Report on Road Traffic Injury Prevention Radin, Mackay, and Hills 2000; Vieira Gomes and Cardoso remains the seminal document discussing proven and 2012; Wittink 2001). However, only 91 countries have promising interventions for road traffic injury preven- policies that physically separate vulnerable road users tion (Peden and others 2004). Randomized controlled from other road users (WHO 2015a). Other safety fea- trials (RCTs) are the gold standard for assessing effec- tures with proven effectiveness include adequate lighting tiveness of interventions; however, given the resources (Radin, Mackay, and Hills 1996, 2000); adequate lane that such trials require and the ethical issues of random- markings or signage (Ward and others 1989); appropriate izing life-saving interventions, RCTs are rarely used to pedestrian crossings (Dalby 1981); and roadside barriers evaluate road safety interventions. (Bambach, Mitchell, and Grzebieta 2013), among others Consequently, proven interventions rely heavily on (Duduta and others 2011; Fuentes and Hernandez 2013; case-control or before-and-after studies, but even these Mock, Arreola-Risa, and Quansah 2003). are largely concentrated in HICs. Road safety approaches Traffic calming measures (such as the use of speed in LMICs in recent years have focused on adapting strat- bumps or rumble strips) are effective in reducing RTIs egies that worked in HICs and achieved good results. As (Changchen and others 2010; Lines and Machata table 3.3 shows, some interventions focus on reducing or 2000; Novoa, Pérez, and Borrell 2009). Those and other eliminating exposure to risk factors among vulnerable measures that limit vehicle speed in areas with high con- road users, such as promoting alternative modes of centrations of vulnerable road users were found to transport (Duduta and others 2011), constructing exclu- reduce the risk of vehicle crashes with pedestrians by sive lanes for motorcyclists (Radin, Mackay, and Hills 67 percent (WHO 2013a). However, only 47 countries 2000), increasing visibility of pedestrians and cyclists representing 950 million people have set effective urban (Radin, Mackay, and Hills 1996, 2000), and supervising speed limits; of those, only 27 countries rate their children walking to school (Muchaka and Behrens 2012; enforcement of the speed laws as good (WHO 2015a). Muda and Ali 2006). Table 3.3 Examples of Proven and Promising Road Safety Interventions Implemented in LMICs Implementation and evaluation in LMICs Interventions proven in HICs Country Study design Results Providing and encouraging Guadalajara, Before-and-after study of the impact of 46 percent reduction in crashes after Macrobus was use of alternative forms of Mexico Macrobus on crashes implemented (Duduta and others 2011) mass transportation Increasing the visibility of Seremban and Shah Time series study of the use of daytime 29 percent reduction in visibility-related motorcycle pedestrians and cyclists Alam, Malaysia running lights for motorcycles crashes (Radin, Mackay, and Hills 1996, 2000) Supervising children Kuala Teregganu, Case–control study assessing the risk of Risk of injury was reduced by 57 percent among walking to school Malaysia injury to children walking or cycling to supervised children (Muda and Ali 2006) school who were supervised by parents Separating different types Selagor, Malaysia Video observational study of crashes and 39 percent reduction in motorcycle crashes, and 600 of road users outcomes after introduction of an exclusive percent decrease in fatalities (Radin, Mackay, and motorcycle lane Hills 2000) Reducing average speeds China Before-and-after study of simple Average speed dropped by 9 percent in three of through traffic calming engineering measures (such as speed four intervention sites; overall number of casualties measures humps, raised intersections, and dropped by 60 percent (Changchen and others 2010) crosswalks) on speed and casualties Setting and enforcing Londrina, Brazil Time series study on enforcement of speed Reduction in mortality to 27.2 per 100,000 population speed limits appropriate to control, seat belt use, new traffic code, after one year of implementing a new traffic code the function of roads and improved prehospital care (De Andrade and others 2008) table continues next page 42 Injury Prevention and Environmental Health Table 3.3 Examples of Proven and Promising Road Safety Interventions Implemented in LMICs (continued) Implementation and evaluation in LMICs Interventions proven in HICs Country Study design Results Setting and enforcing Kampala, Uganda Time series study on enforcement of 17 percent reduction in traffic fatalities after blood alcohol alcohol-impaired driving and speed laws intervention (Bishai and others 2008) concentration limits Villa Clara, Cuba Time series study on enforcement of 29.9 percent reduction in traffic crashes, 70.8 percent alcohol-impaired driving during weekends reduction in deaths, and 58.7 percent reduction in injuries, compared with previous year (2002) (Garcell and others 2008) Setting and enforcing the Iran, Islamic Rep. Before-and-after study of seat belt and Death rates reduced from 38.2 per 100,000 population use of seat belts for all helmet enforcement and social marketing in 2004 to 31.8 in 2007 (p < 0.001); death rate per motor vehicle occupants 10,000 vehicles reduced from 24.2 to 13.4. (Soori and others 2009) Guangzhou, China Before-and-after study of enhanced 12 percent increase in prevalence of seat belt use enforcement and social marketing on seat- (p = 0.001) (Stevenson and others 2008) belt wearing Setting and enforcing Cali, Colombia Time series analysis of fatalities following 52 percent reduction in motorcyclist deaths (Espitia- motorcycle helmet use implementation of mandatory helmet Hardeman and others 2008) law, reflective vests, restrictions on when motorcycles can be used, and compulsory driving training Thailand Before-and-after survey using trauma Helmet use increased 5-fold, injuries decreased by registry data following implementation of 41 percent, and deaths decreased by 20.8 percent helmet law (Ichikawa, Chadbunchachai, and Marui 2003) Vietnam Time series observational study in three 16 percent reduction in injuries, and 18 percent provinces following introduction of manda- reduction in deaths (Passmore, Tu, and others 2010) tory motorcycle helmet law Malaysia Time series study of motorcycle-related 25 percent reduction in motorcycle-related crashes, crashes, injuries, and fatalities following 27 percent reduction in motorcycle-related casualties, implementation of a Motorcycle Safety and 35 percent reduction in motorcycle fatalities (Law Program using annual police statistics and others 2005) Encouraging helmet use Czech Republic Case–control study of helmet enforcement, 100 percent increase in helmet use, and 75 percent among child bicycle riders education, and reward campaign at schools reduction in head injury admission rates (Sedlák, Grivna, and Cihalova 2006) Note: LMICs = low- and middle-income countries. Other interventions focus on addressing the five uated in LMICs (or the results of studies yet to be pub- major behavioral risk factors of RTIs by setting blood lished in the peer-reviewed literature) are as follows: alcohol concentration limits (Bishai and others 2008; Garcell and others 2008), setting or reducing speed • Setting and enforcing lower blood alcohol limits for limits (Changchen and others 2010; De Andrade and novice drivers others 2008), and enforcing the use of seatbelts for • Setting and enforcing the usage of appropriate child drivers and passengers and helmets for motorcyclists restraints and bicyclists (Espitia-Hardeman and others 2008; • Reducing speed limits around areas with high pedes- Ichikawa, Chadbunchachai, and Marui 2003; Law and trian densities, such as schools and hospitals others 2005; Passmore, Nguyen, and others 2010; • Implementing graduated driver licensing systems for Passmore, Tu, and others 2010; Sedlák, Grivna, and new drivers. Cihalova 2006; Soori and others 2009; Stevenson and others 2008). However, challenges exist when adapting interven- Four additional types of interventions that have tions to the LMIC context. The “effectiveness” realized proven applicability in HICs but that have yet to be eval- often is subject to a variety of factors, including the Road Traffic Injuries 43 law-making process and long-standing values, norms, “non-cost-effective” if it produces a healthy life year for and behaviors. Moreover, when trying to identify and more than three times the GDP per capita (Tan-Torres quantify the interventions in LMICs, research and imple- Edejer and others 2003). The authors found four cost-ef- mentation capacity (as well as access to funding and fectiveness studies, which have been updated to US$ costs) play important roles in the effectiveness element 2012: (Perel and others 2007). • Bishai and Hyder (2006) modeled the cost- effectiveness of four potential interventions to Economic Analysis of Interventions increase enforcement of traffic codes (including Data on the economic benefits of these interventions, media coverage, speed bumps, bicycle helmet especially in LMICs, are limited. Although some data legislation, and motorcycle helmet legislation) in are available in HICs (such as the net economic benefits several LMICs, using previous research findings of these interventions), the starkly different costs asso- on effective interventions in LMICs. The results ciated with property losses, disability, and medical care indicated that the average costs per DALY averted make simply translating the conclusions from HICs to (discounted at 3 percent) are US$12 for install- LMICs difficult. A recent systematic review of studies ing speed bumps at high-risk junctions where on costs, cost-effectiveness, and economic benefits of 25 percent of RTIs occurred, US$84 for providing interventions for RTIs and other types of unintentional traffic enforcement, and US$615 for setting and injuries in LMICs found that, of the 30 economic eval- enforcing motorcycle helmet use, all of which were uations published before February 2013, only two stud- very cost-effective. ies analyzed the costs of road safety interventions or • Chisholm and others (2012) studied the global devices (Wesson and others 2013). The costs reported public health responses to RTIs by estimating the below have been updated to 2012 US$ for easier population costs and effects of five enforcement comparison. strategies—speed cameras, alcohol-impaired driving Bishai and others (2003) estimated that the budge- and breath testing campaigns, seatbelt use, helmets tary expenditure on road safety at all levels of govern- for motorcyclists, and helmets for bicyclists—on ment in Uganda and Pakistan is US$0.12 and US$0.11 reducing the RTI burden in South-East Asia and per capita, respectively. Hendrie and others (2004) Sub-Saharan Africa. In addition to confirming the examined availability, urban price, and affordability previous studies, the results suggested that simulta- of child and family safety devices across 18 economi- neous enforcement of multiple road safety laws could cally diverse countries and found that child safety lead to the most health gains at the least expense. seats and bicycle helmets were more expensive in • Ditsuwan and others (2013) focused on RTIs related lower-income countries than higher-income coun- to alcohol-impaired driving in Thailand and associ- tries. For example, a bicycle helmet cost 10 hours of ated interventions. From a health sector perspective, factory work in lower-income countries, while the they found that, when compared with doing nothing cost in higher-income countries was equivalent to less and considering only intervention costs (average than one hour of work. The study also noted that costs per DALY averted), selective breath testing booster seats were usually not available in lower- (US$555), random breath testing (US$611), mass income countries, and the average price of one was media campaign (US$440), selective breath testing US$277 based on limited data from eight LMICs in with mass media campaign (US$542), and random the study sample. breath testing with mass media campaign (US$576) The systematic review (Wesson and others 2013) also were all very cost-effective. They also estimated that includes six cost-effectiveness analyses exploring costs implementing all the interventions together would associated with RTI interventions. When comparing potentially reduce the burden of alcohol-related RTIs across interventions that report costs in terms of years of by 24 percent in Thailand. life saved (YLSs) or DALYs averted, the authors applied • Bishai and others (2008) modeled the costs and the WHO standards of the Choosing Interventions that potential effectiveness of enhanced traffic safety Are Cost-Effective (WHO-CHOICE) project. WHO- patrols in the capital of Uganda from the perspective CHOICE considers an intervention “very cost-effective” of the police department. The evaluation concluded if it generates a healthy life year for less than the GDP that traffic enforcement could be very cost-effective per capita; “cost-effective” if it produces a healthy life (US$32 per YLS) in low-income countries, even from year for less than three times the GDP per capita; and a government perspective. 44 Injury Prevention and Environmental Health Although limited, these studies demonstrate that • Developing epidemiologically sound and robust road safety interventions are among the most cost- information systems on road crashes, injuries, and effective interventions. In environments of limited fatalities resources and competing priorities, such studies have • Promoting intersectoral approaches resonated with policy makers. More economic evalua- • Developing a suitable choice of evidence-based sci- tions of road safety interventions need be conducted in entific interventions in conjunction with integrated LMICs to advance this important agenda. monitoring and evaluation (Schopper, Lormand, and Waxweiler 2006). IMPLEMENTATION OF PREVENTION AND The safe systems approach builds on the unique CONTROL PROGRAMS strength of each sector—ministries, other governmental agencies, private organizations, and NGOs—to inte- Safe Systems Approach grate road safety into different policies systematically, Road traffic crashes and their outcomes depend on com- both vertically within each sector and horizontally plex interactions, which makes a systems approach to across sectors. The European Commission, for example, addressing road safety desirable. The safe systems advocates that road safety policies need to utilize other approach recognizes that multiple sectors need to work related policy avenues to identify areas of integration, in harmony to minimize the occurrence of these crashes thereby creating opportunities for useful synergies that and their impacts (SafetyNet 2009). This approach has are in line with the safe systems approach (Elvik and taken center stage and is being adapted in many set- others 2009; International Transport Forum 2008). The tings globally (Elvik and others 2009; Gururaj 2011; United Nations Road Safety Collaboration (UNRSC) is WHO 2009). Among the key principles of this approach a great example of bringing together different sectors are recognizing human error in transport systems; and stakeholders at the global level to advocate for appreciating human physical vulnerability and fallibil- comprehensive multisectoral approaches to addressing ity; promoting accountability of systems and shared the burden of RTIs (United Nations Road Safety responsibilities; integrating interventions; developing Collaboration 2010). Another great example is the intersectoral approaches; highlighting ethical values; Bloomberg Philanthropies Global Road Safety Program and promoting societal values for economic develop- (box 3.1), a large-scale initiative that brings together a ment, human health, and individual choices (WHO multisectoral consortium at the global and national 2013a). Some well-known and successful examples of levels to implement promising interventions to reduce such an approach include the Swedish Vision Zero the burden of RTIs in LMICs (Hyder and others 2012; (Swedish Road Safety 2013), the Sustainable Safety Peden 2010). Model of the Netherlands (SWOV 2006), and the Safe The health sector is well-positioned to play a lead- Systems approach of Australia (Australian Transport ing role in developing and integrating road safety Council 2011). into its mainstream agenda. Reducing occurrence of RTIs not only improves population health but also likely has far-reaching health benefits by address- Road Safety Policies and Integrated Approaches ing the key risk factors for road safety (Schopper, To work effectively, the safe systems concept needs to be Lormand, and Waxweiler 2006). For example, limiting part of an integrated policy framework and a national alcohol-impaired driving will help control non- road safety plan that define goals and objectives based on communicable diseases, as well as improve the social burden of RTIs at population level. Some components of welfare of the population (Global Road Safety the integrated strategic approach for road safety include Partnership 2007; Gururaj and others 2011). Similarly, the following: health professionals can use their close involvement in the delivery of trauma care and rehabilitation ser- • Developing a sound road safety management system vices to advocate road safety practices, such as use of • Building institutional capabilities and mechanisms motorcycle helmets, seat belts, and child restraints. In for interaction short, the health sector needs to expand its tradi- • Developing sustainable policies tional caregiving role and be involved in areas that • Strengthening human and financial resources and are relevant to promoting road safety, such as data capabilities collection, advocacy, policy development, and capac- • Providing advocacy approaches ity building (WHO 2013a). Road Traffic Injuries 45 Box 3.1 Case Study: Improving Seatbelt Use in the Russian Federation Background: The Russian Federation is an Key Stakeholders and Setting: The program is upper-middle-income country with one of the high- administered by the Department of Road Safety est road traffic injuries (RTIs) mortality rates (18.9 within the Russian Ministry of Interior and jointly per 100,000 population in 2013 in the European implemented by other governmental departments at Region (WHO 2015a). Every year, nearly 30,000 the national and regional levels in two intervention people are killed from RTIs, and an additional sites: Ivanovskaya and Lipetskaya Oblast. 260,000 are injured or permanently disabled on Results: Prevalence of seatbelt and child restraint Russian roads (Department of the Federal Road use was monitored using observation studies. Results Safety Inspectorate of the Russian Ministry of from these studies show a steady increase in seatbelt Interior [http://www.gibdd.ru/stat/]; Institute for use rates in the two sites over time. As figure B3.1.1 Health Metrics and Evaluation 2013). shows, the overall prevalence of seatbelt use increased from 47.5 percent to 88.8 percent among all occu- Intervention: The Bloomberg Philanthropies Global pants in Ivanovskaya Oblast. Similar trends were Road Safety Program (the Global Road Safety observed in Lipetskaya Oblast, where overall seatbelt Program) in Russia aims to support the govern- use increased from 52.4 percent to 73.5 percent over ment’s implementation of its national objectives in the same period. Although lower than seatbelt use, preventing deaths and serious injury on the coun- child restraint use also has increased over this period try’s roads. The program focuses on increasing the in both intervention regions. use of seatbelts and child restraints, as well as speed The preliminary results of observational studies management, through three key activities: legisla- show promising signs that seatbelt use is moving in tion, enhanced police enforcement, and social mar- the right direction in both Oblasts since the imple- keting campaigns. mentation of the measures. Figure B3.1.1 Seatbelt Use in Ivanovskaya, Russia, following Implementation of a Seatbelt Program Ivanovskaya 100 90 % wearing seatbelts or child restraints 80 70 60 50 40 30 20 10 0 Feb-11 Jun-11 Sep-11 Dec-11 Apr-12 Jul-12 Oct-12 Jan-13 May-13 Aug-13 Nov-13 Drivers Front seat passengers All occupants All passengers Rear seat passengers Child restraint use Source: Slyunkina and others 2013. 46 Injury Prevention and Environmental Health OPERATIONALIZING ACTION FOR ROAD Targeting High-Risk Individuals and Groups SAFETY Continuing to document and highlight the significance of the health and economic burdens of RTIs on individ- The information presented throughout this chapter can uals and their families is a major part of a profile-raising be crystallized into actionable items that can be under- strategy. Highlighting the burden in high-risk popula- taken by countries or organizations to enhance road tions (such as adolescents and young people) might safety. As described in previous sections of the chapter, well prove to be a more effective strategy than a countries around the world have diverse landscapes for population-wide approach, given the overwhelming road safety, with different financial and infrastructural burden of RTIs in these age groups. Additionally, high- contexts, policy and legislative environments, as well as lighting the greater impact of these injuries on poor human resource capacities. As such, a one-size fits all list people might provide an impetus for some governments of “must-dos” for road safety may not be practical, but and some funders to take action. the principles of injury prevention and evidence base of road safety must guide all actions. Accordingly, we Promoting Effective Interventions highlight five key areas of focus: resource mobilization; Continuing to identify and promote cost-effective, policy and legislative environment; intervention imple- evidence-based strategies for the prevention of RTIs mentation; data systems; and capacity development. could form an important component of a profile-raising strategy. In particular, promoting the implementation and evaluation of the initiatives that have produced sus- Resource Mobilization tained reductions in RTI-related crashes or those that are Despite the increasing burden of deaths and disabilities proven cost-effective and feasible among low socio- from RTIs, generating financial and political support for economic groups might be particularly effective. road safety has not been without its challenges. The health sector, for example, often pays relatively little Developing a Clinical Research Agenda attention to RTIs as a significant health issue, which has Partnering with clinicians involved in the acute and contributed to the limited support from government postacute care for the victims might form another strand health sectors and health funders generally. This calls for of the approach to fostering intersectoral engagements a multifaceted approach that could involve the follow- on this issue. Our knowledge about the longer-term ing areas: physical, psychological, and economic impacts is still scant, as is our knowledge about the impact of RTIs on • Forming intersectoral partnerships health care systems. Consequently, developing a research • Targeting high-risk individuals and groups agenda in partnership with clinicians to access this • Promoting effective interventions information might provide a useful stimulus to mobilize • Developing a clinical research agenda. resources and action. Forming Intersectoral Partnerships While the health sector primarily deals with treating Policy and Legislative Environment and caring for RTIs, effective solutions to road safety The WHO has published Strengthening Road Safety require a multisectoral approach. In order to contrib- Legislation, a manual that outlines the strategies and ute to the evidence base in this area and ultimately to resources that might be used to facilitate implementation reductions in the incidence and burden of RTIs, health and enforcement of such legislation (WHO 2013c). The professionals ought to work with nonhealth sector col- manual presents some enabling factors for countries to leagues. Given their expertise and experience in dealing adopt and implement legislation, including the following: with RTIs, health professionals, for example, could make significant contributions to the design of preven- • Recent trends in injuries and fatalities tive interventions, or provide input to product manu- • Social norms and values facturers working to improve the effectiveness of safety • Financial, human, and other resources. devices. This would enable them to leverage financial and political support for their activities from non- The manual outlines a framework to support govern- health sector funders. Countries or cities could pro- ments and those working with governments to facilitate mote these linkages by supporting intersectoral the implementation of legislation. The framework working groups, providing seed funding for multidisci- includes conducting an institutional assessment to iden- plinary research, or both. tify local, regional, and national bodies responsible for Road Traffic Injuries 47 making and enforcing legislation; reviewing and assess- instrumental in guiding a country’s health system in ing the gaps in national laws and regulations; and planning for and addressing the burden. In addition to improving their comprehensiveness based on evidence. mortality and morbidity estimates, reliable information The manual also outlines an advocacy process to facili- and data on modifiable risk factors, costs associated with tate the legislative and regulatory changes. RTIs, and age- and gender-specific RTI data at both the In addition to a focus on behavioral risk factors, poli- national and local levels could inform researchers and cies and legislation to prevent RTIs need to focus on issues policy makers about cost-effective interventions, as well such as safe road infrastructure, protection of vulnerable as provide implications of the future health and eco- road users, land use, and safer vehicles. Furthermore, nomic burden—which could be a powerful advocacy tool research examining factors that influence policy change for action. around the prevention of RTIs is much needed. Such Current efforts in HICs such as the EU project JAMIE research, especially in LMICs, ought to include interven- (2011–2014, Joint Action for Injury Monitoring in tion studies to test what approaches have the greatest Europe) have enabled participating member states to success in bringing about legislation, as well as studies that have a relatively limited but useful set of injury data col- show which approaches might be the most cost-effective. lected from emergency departments. This project has Unfortunately, funding for such implementation or policy significantly improved comparable injury surveillance research is woefully inadequate, and a significant chal- systems across EU Member States (Bauer and others lenge remains in undertaking such research and develop- 2014; Rogmans 2012). ing a strong, policy-oriented evidence base. In LMICs, however, the absence or limited availability of strong and robust injury information systems presents a significant challenge to obtaining consistent and qual- Intervention Implementation ity data on injuries. These measurement limitations As evidenced from the findings in the most recent Global render demonstrating the magnitude of the injury prob- Status Report on Road Safety, those countries without lem or even tracking a nation’s progress in addressing it adequate laws were almost exclusively LMICs (WHO difficult. Establishing simple yet robust data systems in 2015a), and the implementation challenge for road LMICs would facilitate the flow of continuous, reliable, safety interventions is greatest in these countries. An and systematic information on key variables to all stake- implementation research agenda may help in overcom- holders (Chandran, Hyder, and Peek-Asa 2010; Hofman ing this challenge. and others 2005; Kruk and others 2010; Lett, Kobusingye, In the case of legislation implementation, undertaking and Sethi 2002; Mock and others 2004; Razzak, Sasser, and research to gain a systematic understanding of why rele- Kellermann 2005). Integrating systems for collecting key vant legislation has not been implemented might provide information on risk factors and outcomes into new and a useful starting point to determining what sort of addi- existing programs to address RTIs in LMICs therefore is tional research might be needed to facilitate change. essential to begin closing this gap (Bachani, Koradia, and Some governments, even when evidence of efficacy is others 2012; Bachani and others 2013; Hyder and others strong, require the evidence of effectiveness within their 2013; Slyunkina and others 2013). specific jurisdictions. However, to provide such evi- dence, legislative action must be implemented first, which usually is difficult. In such cases, the most useful Capacity Development approach would be to undertake small scale efforts or A recurring theme in the preceding sections is the scar- even simulation exercises that could show governments city of appropriately skilled human resources in LMICs the potential reductions in disease burden and the to address the burden of RTIs effectively. This scarcity is potential cost-savings of introducing specific legislations evidenced by the relatively few studies on the burden or interventions. (health, economic, and social) of RTIs and effectiveness The identification of evidence to support the efficacy of interventions for RTIs originating from LMICs in the and effectiveness of non-legislative interventions must peer-reviewed literature (Wesson and others 2013). also be a continuing endeavor. Clearly, the level of investment in research and develop- ment on RTIs in LMICs must increase. This investment will be critical for generating local evidence and for pro- Data Systems moting injury on the global public health agenda. Key Accurately and regularly collecting comprehensive data areas for such capacity include epidemiological research on RTIs is vital to monitoring a country’s progress in to describe the existing burden, causes, and distribution addressing road safety. Such information can be of RTIs, as well as intervention research. 48 Injury Prevention and Environmental Health Table 3.4 Example of Intersectoral Contributions across the Five Domains to Increase the Use of Seatbelts and Child Restraints Health Police Finance/donors NGOs Academia Resource mobilization for Leadership; Stakeholder — Funding Advocacy Generation of increasing seatbelt/child engagement evidence/data restraint use Seatbelt/child restraint Review of laws Implement Leverage networks/ Review of laws Policy analysis policy and legislation law influence Intervention implementation Technical assistance Enforcement Funding Creating awareness; Monitoring implementation Data systems Indicators defined Evidence for — Technical or logistical Evaluation; enforcement support technical support Capacity development Technical training — Funding — Training Note: — = not available; NGO = nongovernmental organization. Any technical assistance delivered to countries for road been renewed. By implementating interventions and safety must include a capacity development component, legislation targeted to behavioral factors, vehicle and with the ultimate goal of improving local capacity to con- equipment factors, and infrastructure, as well as the duct injury research, plan services needed, and reduce the availability of adequate postcrash care, addressing this burden of injuries. The Global Road Safety Partnership, burden is possible, especially in LMICs. However, more an organization that works with LMICs to promote the research is needed to better understand the specific Decade of Action for Road Safety, is a good example needs in LMICs, as well as policy and legislation frame- (United Nations Road Safety Collaboration 2010). works that may be appropriate for such settings. Systems More accessible training and mentoring programs for must be established that will yield the data necessary to road safety also are needed. Although many road safety inform these activities; adequately trained human training programs exist globally, not all are accessible to resources also are needed both to generate new research interested individuals from LMICs, mainly because of the and design and to implement the appropriate policies training programs’ locations or associated costs or both. and programs. A few (such as the Teach-VIP and Mentor-VIP developed by the WHO) make training materials and mentorship ACKNOWLEDGMENTS for LMIC researchers available at no cost (Hyder, Meddings, and Bachani 2009; Meddings 2010, 2015; The authors would like to express their gratitude to Meddings and others 2005). Another online training Xiaoge Julia Zhang and Jeffrey C. Lunnen for the editorial program for prevention and control offered by the Johns support they provided in the preparation of this chapter. Hopkins International Injury Research Unit takes advan- tage of the increasing internet connectivity in LMICs to provide free formal classroom-type instruction on key NOTES topics, ranging from understanding the burden of RTIs WHO Member States are grouped into six geographical to selecting and implementing interventions and evaluat- regions: African, the Americas, South-East Asia, Europe, ing them (JHU-IIRU 2013). The reach and effectiveness Eastern Mediterranean, and Western Pacific. of these new approaches have not yet been determined; World Bank Income Classifications as of July 2014 are as follows, based on estimates of gross national income (GNI) however, they are a step in the right direction, and more per capita for 2013: such efforts are needed to improve road safety globally. An example of an action agenda for increasing seatbelt • Low-income countries (LICs) = US$1,045 or less use using the five elements described is provided in table 3.4. • Middle-income countries (MICs) are subdivided: a) lower-middle-income = US$1,046 to US$4,125 CONCLUSIONS b) upper-middle-income (UMICs) = US$4,126 to US$12,745 • High-income countries (HICs) = US$12,746 or more. RTIs continue to contribute to a significant amount of the health, social, and economic burden to society, and 1. The Global Status Report on Road Safety 2015 by the World global interest in slowing or even halting this trend has Health Organization aims to describe the burden of road Road Traffic Injuries 49 traffic injuries and implement effective interventions in all Reports.” Bundesgesundheitsblatt Gesundheitsforschung Member States using a standardized methodology, and it Gesundheitsschutz 57 (6): 673–80. aims to assess changes since the first and second Global Status Benmaamar, M., C. Dunkerley, and S. D. Ellis. 2002. Reports in 2009 and 2013. The data presented in the report “Urban Transport Services in Sub-Saharan Africa: were collected from 180 countries and areas, covering 6.97 Recommendations for Reforms in Uganda.” Presentation billion people (97.3 percent of the world’s population). Data at the 81st Annual Meeting of the Transportation Research collection in each country was coordinated by a National Board, Washington, DC, January 2002. Transport Research Data Collector and driven by a number of individual Laboratory, Crowthorne, United Kingdom. respondents from different sectors within a country, each Bishai, D., B. Asiimwe, S. Abbas, A. A. Hyder, and W. Bazeyo. of whom completed a self-administered questionnaire with 2008. “Cost-Effectiveness of Traffic Enforcement: Case information on key variables. This group was then required Study from Uganda.” Injury Prevention 14 (4): 223–27. to come to a consensus on the data that best represented their Bishai, D., and A. Bachani. 2012. “Chapter 19: Injury Costing country, which is presented in the report. Response rates by Frameworks.” In Injury Research: Theories, Methods, and region covered were between 95 percent of the population in Approaches, edited by G. Li and S. P. Baker, 371–79. the European region to 99.6 percent in the Western Pacific New York, NY: Springer Science + Business Media, LLC. region. Data collection was carried out in 2014; accordingly, Bishai, D., and A. A. Hyder. 2006. “Modeling the Cost while data on legislation and policies were related to 2014, Effectiveness of Injury Interventions in Lower and Middle data on fatalities were related to 2013 (WHO 2015a). Income Countries: Opportunities and Challenges.” Cost Effectiveness and Resource Allocation 4 (2). Bishai, D., A. A. Hyder, A. Ghaffar, R. H. Morrow, and O. Kobusingye. 2003. “Rates of Public Investment for Road REFERENCES Safety in Developing Countries: Case Studies of Uganda Australian Transport Council. 2011. National Road Safety and Pakistan.” Health Policy and Planning 18: 232–35. Strategy 2011–2020. Canberra, Australia. https://www Blincoe, L., T. R. Miller, E. Zaloshnja, and A. B. Lawrence. 2015. .infrastructure .gov.au/roads/safety/national_road_safety “The Economic and Societal Impact of Motor Vehicle _strategy/files/NRSS_2011_2020_15Aug11.pdf. Crashes, 2010 (Revised).” Technical Report DOT HS-812 Bachani, A. M., E. Galiwango, D. Kadobera, J. A. Bentley, 013. Washington, DC: National Center for Statistics and D. Bishai, and others. 2014. “A New Screening Instrument Analysis, National Highway Traffic Safety Administration. for Disability in Low-Income and Middle-Income Brice, J. H., J. R. Studnek, B. L. Bigham, C. Martin-Gill, C. B. Settings: Application at the Iganga-Mayuge Demographic Custalow, and others. 2011. “EMS Provider and Patient Safety Surveillance System (IM-DSS), Uganda.” BMJ Open 4 (12): during Response and Transport: Proceedings of an Ambulance e005795. doi:10.1136/bmjopen-2014-005795. PubMed Safety Conference.” Prehospital Emergency Care 16 (1): 3–19. PMID: 25526793; PubMed Central PMCID: PMC4275668. Chandran, A., A. A. Hyder, and C. Peek-Asa. 2010. “The Global Bachani, A. M., E. Galiwango, D. Kadobera, J. A. Bentley, Burden of Unintentional Injuries and an Agenda for D. Bishai, and others. 2015. “Characterizing Disability Progress.” Epidemiologic Reviews 32 (1): 110–20. at the Iganga-Mayuge Demographic Surveillance System Changchen L., Z. Gaowuiang, Z. Jianjun, and Z. Hao. 2010. (IM-DSS), Uganda.” Disability and Rehabilitation: 1–9. “First Engineering Practice of Traffic Calming in Zhaitang [Epub ahead of print] PubMed PMID: 26457663. Town in China.” Proceedings of the 2010 International Bachani, A. M., Y. W. Hung, S. Mogere, D. Akungah, J. Nyamari, Conference on Optoelectronics and Image Processing (volume and others. 2013. “Prevalence, Knowledge, Attitude and 1), 565–68. Washington, DC: Institute of Electrical and Practice of Speeding in Two Districts in Kenya: Thika and Electronic Engineering (IEEE). Computer Society. Naivasha.” Injury 44 (Suppl 4): S24–30. doi:10.1016/S0020 Chisholm, D., H. Naci, A. A. Hyder, N. T. Tran, and M. Peden. -1383(13)70209-2. PubMed PMID: 24377774. 2012. “Cost Effectiveness of Strategies to Combat Road Bachani, A. M., P. Koradia, H. K. Herbert, S. Mogere, D. Akungah, Traffic Injuries in Sub-Saharan Africa and South East and others. 2012. “Road Traffic Injuries in Kenya: The Health Asia: Mathematical Modelling Study.” BMJ 344: e612. Burden and Risk Factors in Two Districts.” Traffic Injury doi:10.1136/bmj.e612. Prevention 13 (Suppl 1): 24–30. doi:10.1080/15389588 Chorba, T. L. 1991. “Assessing Technologies for Preventing .2011.633136. PubMed PMID: 22414125. Injuries in Motor Vehicle Crashes.” International Journal of Bachani, A. M., X. J. Zhang, K.A. Allen, and A.A. Hyder. Technology Assessment in Health Care 7 (3): 296–314. 2014. Injuries and Violence in the Eastern Mediterranean Connelly, L. B., and R. Supangan. 2006. “The Economic Costs Region: A Review of the Health, Economic and Social of Road Traffic Crashes: Australia, States and Territories.” Burden.” Eastern Mediterranean Health Journal 20 (10): Accident and Analysis Prevention 38 (6): 1087–93. 643–52. Review. PubMed PMID: 25356696. Dalby, E. 1981. “Applications of Low-Cost Road Accident Bambach, M. R., R. J. Mitchell, and R. H. Grzebieta. 2013. “The Countermeasures According to an Area-Wide Strategy.” Protective Effect of Roadside Barriers for Motorcyclists.” Traffic Engineering and Control 22: 567–74. Traffic Injury Prevention 14 (7): 756–65. De Andrade, S. M., D. A. Soares, T. Matsuo, C. L. Barrancos Bauer, R., M. Steiner, R. Kisser, S. M. Macey, and D. Thayer. 2014. Liberatti, and M. L. Hiromi Iwakura. 2008. “Road Injury- “Accidents and Injuries in the EU. Results of the EuroSafe Related Mortality in a Medium-Sized Brazilian City after 50 Injury Prevention and Environmental Health Some Preventive Interventions.” Traffic Injury Prevention Haddon, W., Jr. 1968. “The Changing Approach to the 9 (5): 450–55. Epidemiology, Prevention, and Amelioration of Trauma: Department of the Federal Road Safety Inspectorate of the The Transition to Approaches Etiologically Rather Than Russian Ministry of Interior. http://www.gibdd.ru/stat/. Descriptively Based.” American Journal of Public Health and Ditsuwan, V., J. Lennert Veerman, M. Bertram, and T. Vos. the Nation’s Health 58 (0002-9572): 8. 2013. “Cost-Effectiveness of Interventions for Reducing ———. 1973. “Energy Damage and the Ten Countermeasure Road Traffic Injuries Related to Driving under the Influence Strategies.” Journal of Trauma 13: 321–31. of Alcohol.” Value in Health 16 (1): 23–30. Hendrie, D., T. R. Miller, M. Orlando, R. S. Spicer, C. Taft, and Duduta, N., C. Adriazola, C. Wass, D. Hidalgo, and L. A. Lindau. others. 2004. “Child and Family Safety Device Affordability 2011. Traffic Safety on Bus Corridors: Pilot Version—Road by Country Income Level: An 18 Country Comparison.” Test. Washington, DC: EMBARQ. Injury Prevention 10: 338–43. Duperrex, O., F. Bunn, and I. Roberts. 2002. “Safety Education Herbert, H. K., A. A. Hyder, A. Butchart, and R. Norton. 2011. of Pedestrians for Injury Prevention: A Systematic Review “Global Health: Injuries and Violence.” Infectious Disease of Randomised Controlled Trials.” British Medical Journal Clinics of North America 25 (3): 653–68. 324 (7346): 1129. Herrstedt, L. 1998. “Planning and Safety of Bicycles in Urban Elvik, R. 2000. “How Much Do Road Accidents Cost the Areas.” In Proceedings of the Traffic Safety on Two Continents National Economy?” Accident Analysis and Prevention Conference, 43–58. Lisbon, September 22–24, 1997. 32: 849–51. Linköping: Swedish National Road and Transport Research Elvik, R., T. Vaa, A. Hoye., and M. Sorensen, eds. 2009. The Institute. Handbook of Road Safety Measures. UK: Emerald Group Hofman, K., A. Primack, G. Keusch, and S. Hrynkow. 2005. Publishing Limited. “Addressing the Growing Burden of Trauma and Injury in Espitia-Hardeman, V., L. Vélez, E. Muñoz, M. I. Gutiérrez- Low- and Middle-Income Countries.” American Journal of Martínez, R. Espinosa-Vallín, and others. 2008. “Impact Public Health 95 (1): 13–17. of Interventions Directed toward Motorcyclist Death Hyder, A. A., K. A. Allen, G. Di Pietro, C. A. Adriazola, R. Sobel, Prevention in Cali, Colombia: 1993–2001.” Salud Pública de and others. 2012. Addressing the Implementation Gap México 50 (Suppl 1): S69–77. in Global Road Safety: Exploring Features of an Effective European Commission. 2014. Statistics—Accidents Data. Response and Introducing a 10-Country Program. American Brussels, Belgium. http://ec.europa.eu/transport/road Journal of Public Health 102 (6): 1061–67. doi:10.2105 _safety/specialist/statistics/index_en.htm. /AJPH.2011.300563. Epub 2012 Apr 19. PubMed PMID: Fuentes, C. M., and V. Hernandez. 2013. “Spatial Environmental 22515864; PubMed Central PMCID: PMC3483956. Risk Factors for Pedestrian Injury Collisions in Ciudad Juarez, Hyder, A. A., K. A. Allen, D. H. Peters, A. Chandran, and Mexico (2008–2009): Implications for Urban Planning.” D. Bishai. 2013. Large-Scale Road Safety Programmes in International Journal of Injury Control and Safety Promotion Low- and Middle-Income Countries: An Opportunity to 20 (2): 169–78. doi:10.1080/17457300.2012.724690. Generate Evidence.” Global Public Health 8 (5): 504–18. Garcell, H. G., T. S. Enríquez, F. G. Garcia, C. M. Quesada, R. P. doi:10.1080/17441692.2013.769613. Epub 2013 Feb 27. Sandoval, and others. 2008. “Impact of a Drink-Driving PubMed PMID: 23445357. Detection Program to Prevent Traffic Accidents [Villa Clara Hyder, A. A., and D. Bishai. 2012. “Road Safety in 10 Countries: Province, Cuba]. Impacto de un programa de detección de A Global Opportunity.” Traffic Injury Prevention 13 (Suppl conductores bajo los efectos del alcohol en la prevención 1): 1–2. doi:10.1080/15389588.2011.650023. PubMed de accidentes de tráfico (provincia de Villa Clara [Cuba]).” PMID: 22414120. 22 (4): 344–47. Hyder, A. A., A. Ghaffar, and T. I. Masood. 2000. “Motor Garcia-Altes, A., J. M. Suelves, and E. Barberia. 2013. “Cost Vehicle Crashes in Pakistan: The Emerging Epidemic.” Savings Associated with 10 Years of Road Safety Policies in Injury Prevention 6 (3): 199–202. Review. PubMed PMID: Catalonia, Spain.” Bulletin of the World Health Organization 11003185; PubMed Central PMCID: PMC1730645. 91 (1): 28–35. Hyder, A. A., M. Labinjo, and S. S. Muzaffar. 2006. A New Global Road Safety Partnership. 2007. Drinking and Driving: Challenge to Child and Adolescent Survival in Urban Africa: A Road Safety Manual for Decision-Makers and Practitioners. An Increasing Burden of Road Traffic Injuries.” Traffic Injury Geneva: Global Road Safety Partnership. Global New Car Prevention 7 (4): 381–88. Review. PubMed PMID: 17114096. Assessment Program, http://www.globalncap.org. Hyder, A. A., D. Meddings, and A. M. Bachani. 2009. “MENTOR- Gururaj, G. 2011. “Road Safety in India: A Framework for VIP: Piloting a Global Mentoring Program for Injury and Action.” Publication 83, National Institute of Mental Health Violence Prevention.” Academic Medicine 84 (6): 793–96. and Neurosciences, Bangalore. doi:10.1097/ACM.0b013e3181a407b8. Review. PubMed Gururaj, G., P. Murthy, G. N. Rao, and V. Benegal. 2011. PMID: 19474562. “Alcohol Related Harm: Implications for Public Health Hyder, A. A., S. S. Muzaffar, and A. M. Bachani. 2008. Road and Policy in India.” Publication 73, National Institute Traffic Injuries in Urban Africa and Asia: A Policy Gap of Mental Health and Neurosciences, Bangalore, India. in Child and Adolescent Health. Public Health 122 (10): http://www.nimhans.kar.nic.in/cam/CAM/Alcohol_report 1104–10. doi:10.1016/j.puhe.2007.12.014. Epub 2008 Jul 1. _NIMHANS.pdf. PubMed PMID: 18597800. Road Traffic Injuries 51 Hyder, A. A., and M. Peden. 2003. “Inequality and Road-Traffic Targeted Motorcycle Safety Programme on Motorcycle- Injuries: Call for Action.” The Lancet 362 (9401): 2034–35. Related Accidents, Injuries and Fatalities in Malaysia. PubMed PMID: 14697797. International Journal of Injury Control and Safety Promotion Hyder, A. A., and J. A. Razzak. 2013. The Challenges of Injuries 12 (1): 9–21. and Trauma in Pakistan: An Opportunity for Concerted Lett, R., O. Kobusingye, and D. Sethi. 2002. “A Unified Action.” Public Health 127 (8): 699–703. doi:10.1016/j Framework for Injury Control: The Public Health Approach .puhe.2012.12.020. Epub 2013 Mar 13. Review. PubMed and Haddon’s Matrix Combined.” Injury Control and Safety PMID: 23489711; PubMed Central PMCID: PMC4313547. Promotion 9 (3): 199–205. Ichikawa, M., W. Chadbunchachai, and E. Marui. 2003. “Effect Lim, S. J., W. J. Chung, and W. H. Cho. 2011. “Economic of the Helmet Act for Motorcyclists in Thailand.” Accident Burden of Injuries in South Korea.” Injury Prevention Analysis and Prevention 35 (2): 183–89. 17 (5): 291–96. Institute for Health Metrics and Evaluation (IHME). 2013. Lin, T., N. Li, W. Du, X. Song, and X. Zheng. 2013. “Road Traffic “Global Burden of Disease Data.” http://www.healthdata Disability in China: Prevalence and Socio-Demographic .org/gbd/data. Disparities.” Journal of Public Health (Oxford) 35 (4): 541–7. Insurance Institute for Highway Safety (IIHS). 2013. “Safety doi:10.1093/pubmed/fdt003. Epub 2013 Feb 5. PubMed Gains Aren’t Global: Some Regions Lag U.S., Europe, Australia PMID: 23386326. in Protecting People in Crashes.” Status Report. 48 (5). http:// Lines, C., and K. Machata. 2000. “Changing Streets, Protecting www.iihs.org/iihs/sr/statusreport/article/48/5/1. People: Making Roads Safer for All.” In Proceedings of Best in International Transport Forum. 2008. “Towards Zero: Ambitious Europe 2000 Road Safety Conference, Brussels, September 12. Road Safety Targets and the Safe System Approach.” Liu, B. C., R. Ivers, R. Norton, S. Boufous, S. Blows, and S. K. Lo. Organisation for Economic Co-operation and Development. 2008. “Helmets for Preventing Injury in Motorcycle Riders.” Paris. http://www.internationaltransportforum.org/Pub Cochrane Database of Systematic Reviews 1: CD004333. /pdf/08TowardsZeroE.pdf. Lozano, R., M. Naghavi, K. Foreman, S. Lim, K. Shibuya, and Jacobs, G., A. Aeron-Thomas, and A. Astrop. 2000. Estimating others. 2012. “Global and Regional Mortality from 235 Global Road Fatalities. Report 445. Crowthorne, United Causes of Death for 20 Age Groups in 1990 and 2010: Kingdom: Transport Research Laboratory. A Systematic Analysis for the Global Burden of Disease Jermakian, J. S. 2011. “Crash Avoidance Potential of Four Study 2010.” The Lancet 380 (9859): 2095–128. Passenger Vehicle Technologies.” Accident Analysis Prevention Madans, J. H., and M. Loeb. 2013. “Methods to Improve 43 (3): 732–40. International Comparability of Census and Survey JHU-IIRU (Johns Hopkins International Injury Research Measures of Disability.” Disability and Rehabilitation 5 (13): Unit). 2013. Courses in Injury Prevention. http://www 1070–73. doi:10.3109/09638288.2012.720353. Epub 2012 .jhsph.edu/research/centers-and-institutes/johns Oct 1. PubMed PMID: 23020151. -hopkins-international-injury-research-unit/training/. Madans, J. H., M. E. Loeb, and B. M. Altman. 2011. Measuring Ker, K., J. Kiriya, P. Perel, P. Edwards, H. Shakur, and others. Disability and Monitoring the UN Convention on the Rights 2012. “Avoidable Mortality from Giving Tranexamic Acid to of Persons with Disabilities: The Work of the Washington Bleeding Trauma Patients: An Estimation Based on WHO Group on Disability Statistics. BioMed Central Public Health Mortality Data, A Systematic Literature Review and Data 11 (Suppl 4): S4. doi:10.1186/1471-2458-11-S4-S4. PubMed from the CRASH-2 Trial.” BMC Emergency Medicine 12: 3 PMID: 21624190; PubMed Central PMCID: PMC3104217. (Published online). Mashreky, S. R., A. Rahman, T. F. Khan, M. Faruque, Khorasani-Zavareh, D., B. J. Haglund, R. Mohammadi, L. Svanstrom, and others. 2010. “Hospital Burden of Road M. Naghavi, and L. Laflamme. 2009. “Traffic Injury Deaths Traffic Injury: Major Concern in Primary and Secondary in West Azarbaijan Province of Iran: A Cross-Sectional Level Hospitals in Bangladesh.” Public Health 124 (4): Interview-Based Study on Victims’ Characteristics and Pre- 185–89. Hospital Care.” International Journal of Injury Control and Meddings, D. R. 2010. “WHO Launches TEACH-VIP Safety Promotion 16 (3): 119–26. E-Learning.” Injury Prevention 16 (2): 143. doi:10.1136 Kruk, M. E., A. Wladis, N. Mbembati, S. K. Ndao-Brumblay, /ip.2010.026468. PubMed PMID: 20363825. R. Y. Hsia, and others. 2010. “Human Resource and Funding ———. 2015. “MENTOR-VIP and Broader Capacity Building Constraints for Essential Surgery in District Hospitals for Injury and Violence Prevention.” Injury Prevention 2: in Africa: A Retrospective Cross-Sectional Survey.” PLoS 142. doi:10.1136/injuryprev-2015-041585. PubMed PMID: Medicine 7 (3): 1–11. 25805772. Kudryavtsev A. V., O. Nilssen, J. Lund, A. M. Grjibovski, and Meddings, D. R., L. M. Knox, M. Maddaleno, A. Concha- B. Ytterstad. 2013. “Road Traffic Crashes with Fatal and Eastman, and J. S. Hoffman. 2005. “World Health Non-Fatal Injuries in Arkhangelsk, Russia in 2005–2010.” Organization’s TEACH-VIP: Contributing to Capacity International Journal of Injury Control and Safety Promotion Building for Youth Violence Prevention.” American Journal 20 (4): 349–57. doi:10.1080/17457300.2012.745576. Epub of Preventive Medicine 5 (Suppl 2): 259–65. PubMed PMID: 2012 Dec 7. PubMed PMID: 23216194. 16376728. Law, T. H., R. S. Umar, S. Zulkaurnain, and S. Kulanthayan. Miranda, J. J., E. Rosales-Mayor, D. A. Quistberg, A. Paca-Palao, 2005. Impact of the Effect of Economic Crisis and the C. Gianella, and others. 2013. “Patient Perspectives on the 52 Injury Prevention and Environmental Health Promptness and Quality of Care of Road Traffic Incident Paravar, M., M. Hosseinpour, S. Salehi, M. Mohammadzadeh, Victims in Peru: A Cross-Sectional, Active Surveillance A. Shojaee, and others 2013. “Pre-Hospital Trauma Care in Study.” F1000Res 2: 167. Road Traffic Accidents in Kashan, Iran.” Archives of Trauma Mock, C. N., C. Arreola-Risa, and R. Quansah. 2003. Research 1 (4): 166–71. “Strengthening Care for Injured Persons in Less Developed Passmore, J., L. H. Nguyen, N. Phuong Nguyen, and J. M. Olivé. Countries: A Case Study of Ghana and Mexico.” Injury 2010. “The Formulation and Implementation of a National Control and Safety Promotion 10 (1–2): 45–51. Helmet Law: A Case Study from Viet Nam.” Bulletin of Mock, C. N., C. Juillard, M. Joshipura, and J. Goosen, eds. World Health Organization 88 (10): 783–87. 2010. Strengthening Care for the Injured: Success Stories and Passmore, J., N. T. H. Tu, M. A. Luong, N. D. Chinh, and N. Lessons Learned from Around the World. Geneva: World P. Nam. 2010. “Impact of Mandatory Motorcycle Helmet Health Organization. Wearing Legislation on Head Injuries in Viet Nam: Results of a Mock, C. N., R. Quansah, R. Krishnan, C. Arreola-Risa, and Preliminary Analysis.” Traffic Injury Prevention 11 (2): 202–06. F. Rivara. 2004. “Strengthening the Prevention and Care of Peden, M. 2010. Road Safety in 10 Countries. Injury Prevention Injuries Worldwide.” The Lancet 363 (9427): 2172–79. 16 (6): 433. doi:10.1136/ip.2010.030155. Epub 2010 Nov 11. Mofadal, A. I., K. Kanitpong, and P. Jiwattanakulpaisarn. 2015. PubMed PMID: 21071768. “Analysis of Pedestrian Accident Costs in Sudan Using Peden, M., R. Scurfield, D. Sleet, D. Mohan, A. A. Hyder, and the Willingness-to-Pay Method.” Accident Analysis and others. 2004. World Report on Road Traffic Injury Prevention. Prevention 78: 201–11. doi:10.1016/j.aap.2015.02.022. Epub Geneva: World Health Organization. 2015 Mar 17. Perel, P., K. Ker, R. Ivers, and K. Blackhall. 2007. “Road Safety in Mohd Faudzi, M. Y., N. A. Mohamad, and N. Ghani. 2011. Low- and Middle-Income Countries: A Neglected Research “Malaysian Value of Fatal and Non Fatal Injury Due to Area.” Injury Prevention 13 (4): 227. Road Accident: The Willingness to Pay Using Conjoint Perez-Nunez, R., M. Hijar-Medina, I. Heredia-Pi, S. Jones, Analysis Study.” Eastern Asia Society for Transportation and E. M. Silveira-Rodrigues. 2010. “Economic Impact of Studies 8. Fatal and Nonfatal Road Traffic Injuries in Belize in 2007.” Muchaka, P., and R. Behrens. 2012. “Evaluation of a ‘Walking Revista Panamericana de Salud Pública 28 (5): 326–36. Bus’ Demonstration Project in Cape Town: Qualitative Radin, U. R. S., G. M. Mackay, and B. L. Hills. 1996. “Modelling Findings, Implications and Recommendations.” Paper pre- of Conspicuity-Related Motorcycle Accidents in Seremban sented to the 31st Southern African Transport Conference, and Shah Alam, Malaysia.” Accident Analysis and Prevention Pretoria, July 9–12. 28 (3): 325–32. Muda, F., and O. Ali. 2006. “Road Traffic Accidents among Radin, U. R. S., G. M. Mackay, and B. L. Hills. 2000. “Multivariate Primary School Children Who Cycle or Walk to School Analysis of Motorcycle Accidents and the Effect of Exclusive in Kuala Terengganu District, 1996.” Journal Kesihatan Motorcycle Lanes in Malaysia.” Journal of Crash Prevention Masyarakat 12 (1): 1. and Injury Control 2: 11–17. NCAP (National Car Assessment Programs). 2011. “Vehicle Razzak, J. A., S. M. Sasser, and A. L. Kellermann. 2005. Safety Is Global.” Global NCAP. “Injury Prevention and Other International Public Health Nguyen, H., R. Q. Ivers, S. Jan, A. L. Martiniuk, L. Segal, Initiatives.” Emergency Medicine Clinics of North America and others. 2015. Cost and Impoverishment 1 Year after 23 (1): 85–98. Hospitalisation Due to Injuries: A Cohort Study in Thai Rogmans, W. H. 2012. “Joint Action on Monitoring Injuries in Binh, Vietnam. Injury Prevention. Europe (JAMIE).” Archives of Public Health 70 (1): 19. Nguyen, H., R. Q. Ivers, S. Jan, A. L. Martiniuk, Q. Li, SafetyNet. 2009. Road Safety Management. European Road and others. 2013. The Economic Burden of Road Traffic Safety Observatory. Brussels, Belgium: European Injuries: Evidence from a Provincial General Hospital in Commission. http://ec.europa.eu/transport/road_safety Vietnam.” Injury Prevention 19 (2): 79–84. /specialist/knowledge/pdf/road_safety_management.pdf. Novoa, A. M., K. Pérez, and C. Borrell. 2009. “Evidence-Based Schopper, D., J. D. Lormand, and R. Waxweiler, eds. 2006. Effectiveness of Road Safety Interventions: A Literature Developing Policies to Prevent Injuries and Violence: Review.” Gaceta Sanitaria 23 (6): 553–e1. Guidelines for Policy-Makers and Planners.” Geneva: World O’Dea, D., and J. Wren. 2010. New Zealand Estimates of the Health Organization. Total Social and Economic Cost of “All Injuries” and the Six Sedlák, M., M. Grivna, and J. Cihalova. 2006. “On Bike Priority Areas Respectively, at June 2008 Prices: Technical in Helmet Only: Results of a Three-Year Community Report Prepared for NZIPS Evaluation. Wellington, New Campaign Promoting Bicycle Helmets for Children.” Zealand: Accident Compensation Corporation. Book of Abstracts, 1st European Conference on Injury Odero, W., M. Khayesi, and P. M. Heda. 2003. Road Traffic Injuries Prevention and Safety Promotion, Vienna, Austria, June in Kenya: Magnitude, Causes, and Status of Intervention. Injury 25–27. Control and Safety Promotion 10 (1–2), 53–61. Sekine, T. 2014. “Utilization of Probe Powered Two-Wheeler Page, Y., T. Hermitte, and S. Cuny. 2011. How Safe is Vehicle Vehicles to Realize a Safe Mobile Society.” IATSS Research Safety? The Contribution of Vehicle Technologies to the 38 (1): 58–70. Reduction in Road Casualties in France from 2000 to 2010. Short, M. M., C. J. Mushquash, and M. Bedard. 2013. “Motor Annals of Advances in Automotive Medicine 55: 101–12. Vehicle Crashes among Canadian Aboriginal People: Road Traffic Injuries 53 A Review of the Literature.” Canadian Journal of Rural in a Portuguese Multilane Road.” Accident Analysis and Medicine 18 (3): 86–98. Prevention 48: 346–52. Shults, R. A., R. W. Elder, D. A. Sleet, J. L. Nichols, M. O. Alao, and Ward, H., J. Norrie, R. E. Allsop, and A. P. Sang. 1989. Urban others. 2001. “Reviews of Evidence Regarding Interventions Safety Project: The Bristol Scheme. Contractor Report 192. to Reduce Alcohol-Impaired Driving.” American Journal of Crowthorne, Berkshire, U.K: Transport and Road Research Preventive Medicine 21 (Suppl 4): 66–88. Laboratory. Slyunkina, S. E., V. E. Kliavin, E. A. Gritsenko, A. B. Petruhin, Wesson, H. K., N. Boikhutso, A. M. Bachani, K. J. Hofman, F. Zambon, and others. 2013. “Activities of the Bloomberg and A. A. Hyder. 2013. “The Cost of Injury and Trauma Philanthropies Global Road Safety Program (formerly Care in Low- and Middle-Income Countries: A Review of RS10) in Russia: Promising Results from a Sub-National Economic Evidence.” Health Policy and Planning 29 (6): Project.” Injury 44 (Suppl 4): S64–69. 795–808. Solagberu, B. A., R. I. Osuoji, N. A. Ibrahim, M. A. Oludara, Wittink, R. 2001. “Promotion of Mobility and Safety of R. A. Balogun, and others. 2014. “Child Pedestrian Injury Vulnerable Road Users: Final Report of the European and Fatality in a Developing Country.” Pediatric Surgery Research Project PROMISING (Promotion of Measures International 30 (6): 625–32. for Vulnerable Road Users).” D-2001-3. Leidschendam, Soori, H., M. Royanian, A. R. Zali, and A. Movahedinejad. 2009. Netherlands: SWOV Institute for Road Safety Research. “Road Traffic Injuries in Iran: The Role of Interventions http://www.swov.nl/rapport/d-2001-03.pdf. Implemented by Traffic Police.” Traffic Injury Prevention WHO (World Health Organization). 2007. “Youth and Road 10 (4): 375–78. Safety.” WHO, Geneva. Stevenson, M., J. Yu, D. Hendrie, L. P. Li, R. Ivers, and oth- ———. 2009. European Status Report on Road Safety Towards ers. 2008. “Reducing the Burden of Road Traffic Injury: Safer Roads and Healthier Transport Choices. Geneva: WHO. Translating High-Income Country Interventions to Middle- ———. 2011. 66th World Health Assembly Adopts Resolution Income and Low-Income Countries.” Injury Prevention Calling for Better Health Care for People with Disabilities. 14 (5): 284–89. WHO news archive. http://www.who.int/disabilities/media Sullivan, R., P. Edwards, A. Sloggett, and C. E. Marshall. /news/2013/28_05/en/index.html. 2009. Families Bereaved by Road Traffic Crashes: Linkage ———. 2013a. Global Status Report on Road Safety 2013: of Mortality Records with 1971–2001 Censuses. Injury Supporting a Decade of Action. Geneva: WHO. Prevention 15 (6): 364–68. ———. 2013b. Fact Sheet on Road Safety. WHO, Geneva. Swedish Road Safety. 2013. “Vision Zero: A Safe Road Traffic ———. 2013c. Strengthening Road Safety Legislation: A Practice Concept.” http://www.swedishroadsafety.se/vision-zero.html. and Resource Manual for Countries. Geneva: WHO. SWOV Institute for Road Safety Research. 2006. Advancing ———. 2014. Global Health Estimates 2014. Geneva, WHO. Sustainable Safety: National Road Safety Outlook for 2005– http://www.who.int/healthinfo/global_burden_disease/en/. 2020, edited by F. Wegman and L. Aarts. The Netherlands: ———. 2015a. Global Status Report on Road Safety 2015. SWOV Institute for Road Safety Research. http://www Geneva: WHO. .sustainablesafety.nl. ———. 2015b. Fact Sheet on Road Safety. WHO, Geneva. Tan-Torres Edejer, T., R. Baltussen, T. Adam, R. Hutubessy, Zaza, S., D. A. Sleet, R. S. Thompson, D. M. Sosin, and J. C. A. Acharya, and others, eds. 2003. “Making Choices in Bolen. 2001. “Reviews of Evidence Regarding Interventions Health: WHO Guide to Cost-Effectiveness Analysis.” to Increase Use of Child Safety Seats.” American College of Geneva, World Health Organization. Preventive Medicine 21 (Suppl 4): 31–47. United Nations Road Safety Collaboration. 2010. Global Plan Zhou, Y., T. D. Baker, K. Rao, and G. Li. 2003. “Productivity for the Decade of Action for Road Safety 2011–2020. http:// Loses from Injury in China.” Injury Prevention 9: 124–27. www.who.int/roadsafety/decade_of_action/plan/en/. Zhu, M., P. Cummings, H. Chu, and L. J. Cook. 2007. “Association Vieira Gomes, S., and J. L. Cardoso. 2012. “Safety Effects of of Rear Seat Safety Belt Use with Death in a Traffic Crash: A Low-Cost Engineering Measures: An Observational Study Matched Cohort Study.” Injury Prevention 13 (3): 183–85. 54 Injury Prevention and Environmental Health Chapter 4 Nontransport Unintentional Injuries Robyn Norton, Rajeev B. Ahuja, Connie Hoe, Adnan A. Hyder, Rebecca Ivers, Lisa Keay, David Mackie, David Meddings, and Fazlur Rahman INTRODUCTION that the numbers of deaths from most nontransport Injuries are most commonly categorized as unintentional unintentional injuries are decreasing globally, with or intentional, based on the injured party’s presumed the exception of deaths from falls and possibly from intent (Norton and Kobusingye 2013). Unintentional burns, which are increasing. injuries comprise both transport and nontransport inju- This chapter places injuries in a global context but ries. This chapter examines in detail the leading causes of documents the burden and known risk factors for non- nontransport unintentional injuries, namely falls, drown- transport unintentional injuries in LMICs. It also provides ing, burns, and poisoning. an overview of the best available evidence about interven- The chapter also briefly discusses the burden of inju- tions and policies that are shown to effectively reduce such ries resulting from the other two main categories of injuries in those countries. The key focus of the chapter is nontransport unintentional injuries, namely exposure preventive strategies, although the importance of acute to forces of nature and all other unintentional injuries care and rehabilitation is clear, as discussed elsewhere in combined. All other unintentional injuries combined this volume. Where data are available, the costs and eco- constitute approximately 38 percent of nontransport nomic benefits of these interventions are outlined. unintentional injuries. However, because the numbers A consistent theme for every category of cause- of deaths for each cause-specific injury within this specific, nontransport unintentional injury is the dearth group are comparatively small, and because the nature of reliable evidence from LMICs on risk factors, interven- of, risk factors for, and interventions for each cause are tions, and cost-effective approaches to prevention. This unique, this chapter does not include a detailed exami- theme reflects the limited availability of human and other nation of risk factors or interventions for this group as resources that would enable researchers to access such a whole, nor for any individual cause-specific injury. information, and it also reflects the low priority key stake- Individuals in low- and middle-income countries holders place on addressing the burden of such injuries. (LMICs) sustain a higher proportion of deaths and The final section makes recommendations about disability-adjusted life years (DALYs) from nontrans- what policy makers need to do to continue the trend of port unintentional injuries compared with those in declines in the burden of death and disability from non- high-income countries (HICs). The mortality rates transport unintentional injuries; to achieve similar for almost all of these injuries are higher in LMICs declines for falls; and to reduce the disparities in injury than in HICs. The best available evidence suggests rates between HICs and LMICs. Corresponding author: Robyn Norton, Principal Director, The George Institute for Global Health; Professor of Global Health, University of Oxford, Oxford, United Kingdom; Professor of Public Health, University of Sydney, Sydney, Australia; rnorton@georgeinstitute.org. 55 This chapter follows the World Health Organization unintentional injuries, deaths from falls have increased (WHO) classification of regions: Africa, the Americas, since 2000, in large part as a consequence of the increas- South-East Asia, Europe, the Eastern Mediterranean, and ing numbers of older people, who are at greatest risk. the Western Pacific. Falls are the leading cause of DALYs; between 2000 and 2012, the numbers of DALYs from falls increased by 19.2 percent (figure 4.2). BURDEN OF NONTRANSPORT Men account for a slightly higher proportion of deaths UNINTENTIONAL INJURIES from falls (54 percent) than women, with approximately 50 percent of all fall-related deaths occurring in individ- Recent estimates of the global burden of death and uals ages 70 years and older. The rates of death from falls disability resulting from nontransport unintentional in that age group (96.6 per 100,000 population) are strik- injuries are available from the Global Burden of Disease ingly higher than in all other age groups (table 4.1). study for 2013 (Haagsma and others 2015) and from Although the death rates in LMICs are comparable to the Global Health Estimates provided through WHO those in HICs, they are highest in the LMICs of South- for 2012 (WHO 2014). East Asia (16 per 100,000 population) (table 4.2). Global Health Estimates data suggest that, collec- tively, nontransport unintentional injuries account for more than 6,700 deaths a day and 2.4 million deaths Drowning annually (WHO 2014)—almost twice the number of Drowning is the second most common cause of death deaths from transport injuries and twice the number of and DALYs from nontransport unintentional injuries. deaths from intentional injuries. The total is comparable Drowning accounts for approximately 15 percent of to the number of deaths from HIV/AIDS and tuberculo- both deaths and DALYs, with approximately 372,000 sis combined. Nontransport unintentional injuries also individuals dying each year as a consequence (figures 4.1 account for more than 148 million DALYs annually— and 4.2). Over the past two decades, deaths and DALYs almost twice the number from transport injuries and from drowning have decreased by approximately 20 from intentional injuries (WHO 2014). percent and 30 percent, respectively, although those figures may be underestimates, given the known data limitations in LMICs (WHO 2012). Falls Almost all drowning deaths occur in LMICs Falls are the leading cause of nontransport unintentional (95 percent), and rates of drowning are substantially injury deaths, accounting for almost 700,000 deaths a higher in almost all LMICs compared with HICs year (figure 4.1). In contrast to most other nontransport (table 4.2). Death rates from drowning are about twice as Figure 4.1 Nontransport Unintentional Injury Deaths, by Cause, 2000 and 2012 1,200 1,000 Number of deaths (thousands) 800 600 400 200 0 Falls Drowning Fire, heat, and hot Poisonings Exposure to forces Other unintentional substances of nature injuries 2000 2012 Source: WHO 2014. Note: The WHO data suggest that negligible numbers of deaths resulted from forces of nature in low- and middle-income countries. 56 Injury Prevention and Environmental Health Figure 4.2 Nontransport Unintentional Injury Disability-Adjusted Life Years, by Cause, 2000 and 2012 70,000 60,000 Number of DALYs (thousands) 50,000 40,000 30,000 20,000 10,000 0 Falls Drowning Fire, heat, and hot Poisonings Exposure to forces Other unintentional substances of nature injuries 2000 2012 Source: WHO 2014. Note: The WHO data suggest that negligible numbers of deaths resulted from forces of nature in low- and middle-income countries. Table 4.1 Nontransport Unintentional Injury Deaths, by Proportion of Males and Age Group, 2012 Deaths per 100,000 population Age group (years) Cause-specific injury category Male (percent) <5 5–14 15–29 30–49 50–69 70+ Falls 54 5.0 2.5 2.5 4.0 13.5 96.6 Drowning 67 10.1 6.1 4.2 3.2 4.9 10.9 Burns 53 9.6 3.4 2.7 2.3 3.1 9.7 Poisoning 61 3.6 1.0 2.1 2.7 3.9 6.4 Forces of nature 60 0.0 0.0 0.0 0.0 0.1 0.1 Other 63 17.8 6.3 8.1 9.3 16.9 61.1 Source: WHO 2014. Table 4.2 Nontransport Unintentional Injury Deaths, by Income and Region, 2012 Deaths per 100,000 population Income and region High-income countries Low- and middle-income countries Cause-specific South-East Eastern Western injury category Africa Americas Asia Europe Mediterranean Pacific Falls 9 10 5 16 5 4 8 Drowning 3 8 3 7 4 5 4 Burns 2 14 1 4 3 4 1 Poisoning 2 4 1 3 2 4 3 Forces of nature 0 0 0 0 0 0 0 Other 14 20 10 14 14 16 8 Source: WHO 2014. Note: The WHO data suggest that negligible numbers of deaths resulted from forces of nature in low- and middle-income countries. Nontransport Unintentional Injuries 57 high for men and boys than for women and girls Poisoning (table 4.3). Death rates are highest for those younger Poisoning constitutes approximately 8 percent of non- than age five years, for both genders, followed by those transport unintentional injury deaths and 7 percent of ages 50 years and older. DALYs, resulting in an estimated 193,000 deaths and almost 11 million DALYs annually (figures 4.1 and 4.2). Of poisoning victims, 61 percent are men and boys; Burns the highest rates are in those under age five years and Burns are the third most common cause of death and over age 50 years (table 4.1). The rates of poisoning DALYs from nontransport unintentional injuries, deaths in most LMICs are higher than those in HICs, accounting for approximately 15 percent of deaths and with the rates being twice as high in the LMIC regions of 14 percent of DALYs. Between 250,000 and 350,000 Africa and the Eastern Mediterranean (table 4.2). individuals die each year as a consequence of burns (figures 4.1 and 4.2). Unlike most other unintentional injuries, rates of burn deaths for women and girls are Forces of Nature comparable to those for men and boys (47 percent ver- Deaths and DALYs caused by forces of nature include sus 53 percent, as shown in table 4.1). Rates of death those arising from exposure to excessive natural heat or from burns are highest in those younger than age five cold, earthquakes, and floods. These can vary signifi- years and those ages 70 years and older. However, as cantly by year. For 2012, WHO data suggest that forces of with drowning, almost all deaths occur in LMICs (97 nature accounted for only around 2,000 deaths. However, percent); rates of death are highest in LMICs in Africa, the numbers of deaths and DALYs have increased in followed by South-East Asia, Europe, and the Eastern recent years. Mediterranean (table 4.2). The rates in Africa, for exam- WHO data also suggest that men are at greater risk ple, are 14 per 100,000 population, compared with 2 per than women and that forces of nature primarily impact 100,000 in HICs. older adults (table 4.1). Although these figures clearly identify a significant burden of death and disability, almost all LMICs lack comprehensive data on burns. The International Society All Other Nontransport Unintentional Injuries for Burn Injuries (ISBI), WHO, the U.S. Centers for All other causes combined account for approximately Disease Control and Prevention (CDC), and the Global 38 percent of deaths from nontransport unintentional Alliance for Clean Cookstoves (GACC) have launched injuries and a similar proportion for DALYs (figures 4.1 initiatives to develop minimal datasets and software and 4.2). WHO Global Health Estimates give little addi- platforms for better surveillance of burn incidence in tional detail on these. However, the Global Burden of domestic settings. Disease report breaks out several subgroups, the largest of which are exposure to mechanical forces, adverse effects of medical treatment, and animal contact (Lozano Table 4.3 Drowning Deaths, by Gender, Age Group, and and others 2012). Country Type, 2012 As with most deaths from unintentional injuries, Deaths per 100,000 population men and boys account for a disproportionate number of Gender Age group (years) Global LMICs all other causes (63 percent) (table 4.1). The rates of death are highest in those ages 70 years and older, and Males 0–4 11.5 12.6 among deaths occurring in LMICs, the rates are highest 5–14 7.9 8.7 in Africa (table 4.2). 15–49 5.5 5.9 50+ 8.1 8.9 RISK FACTORS FOR INJURIES Total 7.0 7.6 Females 0–4 8.6 9.6 Falls 5–14 4.3 4.8 Falls in Older People The high burden of fall-related deaths in older people is 15–49 1.8 2.0 due in part to the physical, sensory, and cognitive 50+ 4.8 5.9 changes associated with aging, in combination with Total 3.5 4.0 environments that are not adapted for this population Source: WHO 2014. (Lord and others 2007). Risk factors in LMICs are largely 58 Injury Prevention and Environmental Health similar to those in HICs: age, female gender, previous five years have the highest drowning mortality rates falls, mobility problems, declining vision, medication worldwide. Deaths in this age group frequently occur as use, unsafe environments, and chronic health problems a result of children’s inherent vulnerability—the inabil- (Kalula and others 2011; Ranaweera and others 2013). ity to keep their airway clear of water—combined with a However, the nature of the environmental risks differs lapse in adult supervision. Individuals with lower educa- in LMICs, with more falls resulting from factors relating tion levels are at increased risk of drowning; across all to street and house design, transport, violence, and rural regions and countries, lower socioeconomic groups are locations (Dandona and others 2010; Jagnoor and others more vulnerable to drowning than higher socioeco- 2014; Jitapunkul, Yuktananandana, and Parkpian 2001; nomic groups (Giashuddin and others 2009). Kalula and others 2011; Ranaweera and others 2013). The absence of, or a lapse in, adult supervision has Often, access to water is limited only to locations outside been shown to be an important, potentially modifiable of the home (Hestekin and others 2013). The risk factors risk factor for drowning incidents in children (Bierens for fall-related injuries, including osteoporotic fractures, 2006; Chalmers, McNoe, and Stephenson 2004; may differ across settings because of variations in diet International Life Saving Federation 2007; WHO 2006; and in load-bearing exercise (Lau and others 2001). Yang and others 2007), and individuals with few swim- ming skills or those who have not received swimming Falls in People of Working Age lessons have been shown to be at increased risk (Yang Research to examine risk factors for falls in people of and others 2007). Alcohol consumption is one of the working age in HICs is scarce, but studies of falls in most frequently reported contributory factors associated the home may have some relevance. Those have with adolescent and adult drowning (WHO 2006). Some highlighted the role of alcohol (Kool and others 2008) as medical conditions such as epilepsy, which are often well as structural or environmental hazards (Kool and poorly controlled in LMICs, also place individuals at others 2010). Falls in people of working age in LMICs are increased risk (Bell and others 2008). reported more commonly for men and are reported as occupational injuries, including those on farms (Dandona Environmental Factors and others 2010; Gururaj, Sateesh, and Rayan 2008). Children who live near open water sources are particu- larly at risk (Peden and McGee 2003). People who work Falls in Children on or near water, travel on water, or use surface water or The risk factors for falls for children in HICs include open wells for household water are all likely to face male gender, younger age, and low socioeconomic status. increased risk of unintentional immersion in a water Fall-related injuries are commonly sustained on play- hazard. Similarly, those who live in settings susceptible to grounds; on bunk beds and equipment, such as baby flash floods, river flooding, storm surges, or tsunamis are change tables or baby walkers (Khambalia and others at increased risk of drowning. 2006); or from windows (Harris, Rochette, and Smith 2011). The risk factors in LMICs are similar, with falls reported from ladders or stairs, or beds or other furni- Burns ture (Hyder and others 2009). More falls occur in boys The etiological factors responsible for the majority of and in rural locations (Jiang and others 2010). burn injuries in LMICs are very different from those in HICs. In the United States, for example, 69 percent of burns happen at home, with factors such as alcohol, Drowning smoking, and high bathing temperatures dominating The International Life Saving Federation World (American Burn Association 2012). Almost 50 percent Drowning Report (International Life Saving Federation of burn deaths have been attributed to the combination 2007) divides drowning risk factors into two groups: of alcohol and smoking. human factors and environmental factors. In LMICs a large proportion of burn injuries is sus- tained in the kitchen or cooking area and is related to the Human Factors nature of the cooking appliances, the source of heat, and Sociodemographic factors, socioeconomic conditions, the heating of liquids (Hyder and others 2009; Mashreky behavioral factors, and medical conditions have all been and others 2010). Several studies have implicated kero- postulated or shown to be risk factors for drowning. sene stoves in a large percentage of burn injuries (Ahuja Higher rates among men and boys purportedly result and Bhattacharya 2002; Ahuja, Bhattacharya, and Rai from their increased exposure to water and riskier 2009; Ahuja, Dash, and Shrivastava 2011; Mabrouk, El behaviors (Peden and McGee 2003). Children under age Badawy, and Sherif 2000). As the source of heating Nontransport Unintentional Injuries 59 moves up the energy ladder from biomass products to INTERVENTIONS, EFFECTIVENESS, kerosene to liquefied petroleum gas (LPG) to electricity, AND COVERAGE the fuel becomes safer, cleaner, and more expensive. Cooking appliances that use LPG appear to be safer and Evidence for the effectiveness of interventions and less polluting than those fueled by kerosene, but they still policies associated with nontransport unintentional pose serious risks if not properly used and maintained injuries, and especially interventions that are effective in (Ahuja, Dash, and Shrivastava 2011). LMICs, is extremely limited. Few randomized con- In India, most domestic burns are sustained by trolled trials have been undertaken; some before-and- women ages 16–35 years; almost 70 percent of these after studies are available, but much information derives injuries are due to the traditional practice of cooking at from observational studies. This section outlines the floor level or over an open fire, compounded by wearing best available evidence and highlights those interven- loose-fitting clothing made from non-flame-retardant tions that show the greatest likelihood of being effective fabric (Sanghavi, Bhalla, and Das 2009). in LMICs (table 4.4). Burns from incidents involving traditional home- made bottle lamps or commercial wick lamps are a cause of major morbidity and mortality in Bangladesh, India, Falls Mozambique, Nepal, and Sri Lanka. In a study in Sri Falls in Older People Lanka, 41 percent of the burns in patients admitted with Substantial progress has been made in the development unintentional flame burns resulted from homemade of effective fall prevention programs for older people in kerosene bottle lamps tipping over (Laloë 2002). A case- HICs. The incidence of falls in older people living in the control study of childhood burn injuries in 2008 in rural community has been reduced by either group and home- Bangladesh revealed that households using traditional based exercise programs, usually containing some balance kerosene lamps (kupi bati) had a greater than threefold and strength-training exercises, or by Tai Chi programs risk of childhood burns relative to households not using (Gillespie and others 2012). Successful multifactorial such lamps (Mashreky and others 2010). interventions include home safety modifications, cataract surgery, withdrawal of psychotropic medication, and insertion of a pacemaker for those with carotid sinus Poisoning hypersensitivity (Gillespie and others 2012). The effec- Although some studies have been conducted in LMICs tiveness of fall prevention programs for older people in to examine victims of all ages admitted to hospitals acute and subacute hospital settings, though promising, for unintentional poisoning (Akbaba and others 2007; especially in high-risk groups, is limited and requires fur- Peiris-John and others 2013; Sawalha and others ther investigation (Cameron and others 2012). 2010), much of the available literature on unintentional Although a substantial body of work is emerging on poisoning in these countries focuses on young children. the burden and risk factors for falls in older people in Findings from these studies show that, consistent with LMICs, little or no evidence exists about the effectiveness overall trends, boys tend to be at higher risk than girls of fall prevention programs in these settings (Kalula and (Balan and Lingam 2012; Lifshitz and Gavrilov 2000; others 2011). Although many of the interventions shown Soori 2001). Paraffin and kerosene, other types of chem- to be effective in HICs might be effective in LMICs, ical products, medicines, and drugs are the most com- implementing them can be difficult. Competing health mon agents in unintentional child poisoning cases care demands that are perceived to be more urgent, com- (Balan and Lingam 2012; Balme and others 2012; Kohli bined with a lack of trained health care profession- and others 2008; Lifshitz and Gavrilov 2000; Ozdemir als, create challenges for implementing or translating and others 2012; Zia and others 2012). evidence-based policy for fall prevention in LMICs. Case-control studies in LMICs have also highlighted Further, the lack of systematic care for older people the importance of risk factors such as unsafe storage of places much of the burden on family members. Without chemicals or medicines (Ahmed, Fatmi, and Siddiqui substantial investment in prevention programs, elder- 2011; Chatsantiprapa, Chokkanapitak, and Pinpradit care facilities, acute-care hospital services, and rehabili- 2001; Ramos and others 2010; Soori 2001); history of tation, the burden on families will increase. previous poisoning (Ahmed, Fatmi, and Siddiqui 2011; Soori 2001); distraction or lack of adult supervision Falls in People of Working Age (Ramos and others 2010; Soori 2001); hyperactive child Fall prevention for those of working age in LMICs behavior; low socioeconomic status; and low maternal requires a systematic approach, with a focus on indus- educational status (Ahmed, Fatmi, and Siddiqui 2011). trial and construction safety standards. Little work has 60 Injury Prevention and Environmental Health Table 4.4 Interventions for Cause-Specific Injuries, with Promising or Good Evidence, in HICs and LMICs Cause-specific injury Age group HICs LMICs Falls Older people Group and home-based exercise programs, containing — balance and strength-training exercises, or Tai Chi (Gillespie and others 2012) Multifactorial interventions, including home safety — modifications (Gillespie and others 2012) Targeted interventions involving cataract surgery, — withdrawal from psychotropic medication, and pacemaker insertion (Gillespie and others 2012) Working age Company-oriented safety campaigns and drug-free — workplace programs (van der Molen and others 2012) Children Home safety interventions providing free, low-cost, or — subsidized safety equipment (Kendrick and others 2012) Drowning Children Legislation and enforcement of swimming pool fencing Parental or other adult supervision and (Stevenson and others 2003; Thompson and Rivara 2000) swimming lessons (Rahman 2010; Rahman and others 2009; Rahman and others 2012) Provision of swimming lessons (Brenner, Saluja, and — Smith 2003; Brenner and others 2009) Legislation and enforcement of PDF use for recreational — boaters (Bugeja and others 2014) Burns All ages Installation and maintenance of smoke detectors (Mock Improvements in stove design (Mock and and others 2011; Norton and others 2006) others 2011) Education, legislation, and enforcement to regulate the — temperature of household taps (Norton and others 2006) Poisoning Children Home safety education, with the provision of safety Community-based educational interventions equipment (Kendrick and others 2013) (Schwebel and others 2009) Child-resistant containers (Krug and others 1994) Note: — = not available; HICs = high-income countries; LMICs = low- and middle-income countries. been done to evaluate the effectiveness of programs to although no evidence suggests that such programs reduce falls at building sites or in industrial settings, increase the possession of window locks, screens, or either in high- or low-income environments. Low- windows with limited openings (Kendrick and others quality evidence suggests that company-oriented safety 2008; Kendrick and others 2012). Interventions that interventions–such as multifaceted safety campaigns provide free, low-cost, or subsidized safety equipment and drug-free workplace programs—can reduce non- appear to be more effective in improving safety prac- fatal injuries among construction workers (van der tices than interventions that do not do so (Kendrick and Molen and others 2012). Further improvements in con- others 2012). However, little research has been con- struction safety standards and regulations are likely to ducted on adapting known effective interventions to reduce fall-related injuries, but these will require the LMICs (Kendrick and others 2008). development and implementation of appropriate poli- Although many important challenges face efforts to cies, as well as education and enforcement. prevent falls for both young and older people in rural settings, the increasing and rapid urbanization of LMICs Falls in Children will present additional challenges. The development of In HICs, home safety interventions for the prevention high-rise apartments is likely to increase the risk of falls of falls in children have been shown to increase the use from windows and stairways, particularly in poorly lit of stair gates and to reduce the use of baby walkers, buildings. Urban slums and squatter camps pose Nontransport Unintentional Injuries 61 particular risks (Rizvi and others 2006). Urban planning swimming in the village pond, which had been converted and architectural design can play a major role in mitigat- into a safe training site. Both program components ing the risks of falls, as can regulation of sidewalks to appeared to reduce the incidence of drowning in the provide a safe walking environment free of roadside stalls. intervention villages (Rahman 2010; Rahman and others 2009; Rahman and others 2012). The World Report on Child Injury Prevention summa- Drowning rized the evidence on key strategies to prevent drown- Evidence for Drowning Prevention ing among children. It suggested that four interventions The scientific literature on drowning prevention studies should be considered as effective: removing or covering published since the late 1990s identifies a number of pos- water hazards, requiring isolation fencing around swim- sible and promising options for drowning prevention. In ming pools, wearing PFDs, and ensuring immediate HICs, much of the evidence relates to the prevention of resuscitation (Peden and others 2008). The report sug- drowning in recreational settings. Evidence from observa- gested that although other strategies are promising, tional studies suggests that legislation and enforcement of including ensuring the presence of lifeguards in swimming swimming pool fencing are likely to significantly reduce areas and raising targeted awareness about drowning, for drowning, especially among children (Stevenson and oth- the remainder, the evidence is insufficient, ineffective, or ers 2003; Thompson and Rivara 2000). Also, a growing potentially harmful. body of evidence shows the contribution of alcohol con- sumption to recreational drowning in young people and Prevention Challenges in LMICs adults, so legislation and enforcement to control alcohol A 2012 report prepared on behalf of the Working Group use, especially in relation to aquatic activities, are likely to on Child Drowning in LMICs has highlighted the have an important effect (Diplock and Jamrozic 2006). challenges in addressing drowning prevention, especially Some evidence indicates that providing swimming among children (Linnan and others 2012): lessons may reduce drowning risks (Brenner, Saluja and Smith 2003; Brenner and others 2009; Rahman and oth- • Most LMICs are predominantly rural. ers 2012). Increased knowledge of water safety, both for • Water and other environmental hazards are ubiquitous children and adults, also may decrease the risk of drown- around the home and throughout the community. ing. However, little evidence shows that water safety • Building codes and zoning ordinances are lacking or knowledge alone leads to improved safety (Kendrick and unenforced. others 2007; Moran 2006; Solomon and others 2013). By • Universal primary education is a goal, not a reality, comparison, increasingly strong evidence supports legis- resulting in high levels of illiteracy across large seg- lation requiring and enforcing the use of personal flota- ments of the population. tion devices (PFDs) by recreational boaters as an effective • Parents often have many children and must rely on intervention strategy (Bugeja and others 2014). older children to supervise younger ones. In LMICs, evidence has shown that both increased • Essential social services are lacking, such as emergency parental or other adult supervision of and swimming medical and rescue services that extend lifesaving ser- lessons for children reduce child drownings (Rahman vices outside hospitals or other safety infrastructure. 2010; Rahman others 2009; Rahman and others 2012). • Sufficient financial resources are lacking. The Prevention of Child Injury through Social • Adequate human resources for drowning prevention Intervention and Education (PRECISE) was imple- are lacking. mented in Bangladesh between 2006 and 2010 and cov- ered more than 750,000 people in rural villages in three In contrast, the report suggests, HICs have built a separate subdistrict intervention areas. The research culture of water safety on these foundations, using the design involved a comparison between very large cohorts wealth of financial and social capital that they possess. of children participating in the interventions with non- Introducing drowning prevention and the creation of a participating children who were matched for age, gender, culture of water safety was a natural progression in the and location of residence. For children ages one to five process of developing strong public health and public years, a village crèche (child care) program called Anchal safety institutions connected to effective civil governance was established to provide a safe haven where mothers and enforcement structures. could drop off their children for four hours a day while Although, in theory, the principles underlying they tended to domestic work. Children ages four years drowning prevention are the same among all popula- and older received training in a program called SwimSafe, tion groups, whether in LMICs or HICs, they require which taught water safety, safe rescue, and survival thoughtful and extensive adaptation, given the different 62 Injury Prevention and Environmental Health societal contexts and norms (Hyder and others 2008). It Platform Cooking may not be possible to adapt the drowning prevention Floor-level cooking has been implicated in increasing strategies for HICs in low-resource settings. The report the incidence of burn injuries, whether on a woodstove, by Linnan and others (2012) highlights, in particular, a kerosene stove, or an open three-stone fire. Cooking on that of the four interventions deemed to have sufficient a platform immediately distances children from fires and evidence for effectiveness in the World Report on Child from toppling cooking vessels. Platform cooking also Injury Prevention, three—fencing around swimming renders irrelevant, to an extent, the type of clothing pools, legislating the use of PFDs, and ensuring imme- worn by women while cooking. Loose-fitting clothes are diate resuscitation—are likely to be unfeasible or unsus- much less likely to get caught in the fire if the stove is on tainable in LMICs. a platform. However, there is no published literature outlining the development and implementation of platform design, nor evaluation of their effectiveness. Burns In HICs, prevention efforts have focused on education and on the installation and maintenance of smoke Poisoning detectors for the prevention of fire-related burns. The traditional three “E” approach to preventing To reduce the incidence of scald-related burns, efforts injuries—education, engineering, and enforcement—can have included legislation and enforcement to regulate be used as a framework to select intervention strategies for the temperature of household taps (Mock and others preventing unintentional poisonings in LMICs. Because 2011; Norton and others 2006). In LMICs, strategies these three Es generally refer to two broad concepts, have primarily focused on the prevention of fire- behavior and environment, the focus is first on strategies related burns. to change the behavior of individuals and communities and then on strategies to alter the environment. Education and Increasing Awareness in Communities Education alone is unlikely to lead to behavioral changes. Behavioral Strategies for the Prevention of Poisoning However, in many LMICs, especially in areas with high Many experts have highlighted the need to target behav- levels of literacy, educating the public on safe practices ioral change to prevent accidental poisoning. Suggested may be an important strategy in improving awareness interventions include safe storage of poisons, that is, levels. This education may lead to increased pressure on where they are stored as well as the types of containers authorities to pass appropriate prevention legislation used (Ahmed, Fatmi, and Siddiqui 2011; Kohli and and to provide the necessary impetus for resources to others 2008; Schwebel and Swart 2009). Ozdemir and address the problem. others (2012), for example, recommended storing poi- sons in high places or locked cupboards after finding Improvements in Stove Design that 70 percent of responsible agents were easily acces- Between 1992 and 1994, a household randomized trial, sible to child victims. RESPIRE (randomized exposure study of pollution Unfortunately, the few studies that have assessed the indoors and respiratory effects), was implemented in effectiveness of behavioral approaches to addressing unin- rural highland Guatemala to determine the effects of hav- tentional poisoning, such as paraffin- and kerosene- ing an improved wood stove with a chimney (plancha) on related injuries in LMICs, have mixed results (Schwebel the health of young children younger than 18 months, and Swart 2009). One strategy, known as community- compared with continued use of open fires (Smith and based educational interventions, has shown promise in others 2011). Prior to the intervention, the burn incidence South Africa. In 2008, Swart and others evaluated the rate among young children was 42.1 per 1,000 per year. effectiveness of a paraprofessional home visitation pro- Six months postintervention the rates were 18.1 and 35.2 gram to prevent child injuries. Intervention households per 1,000 children per year among the intervention and received four visits from home visitors recruited from the control groups, respectively. In addition, the plancha community. During these visits, the trained home visitors group had fewer serious burns (Mock and others 2011). gave caregivers safety information, as well as devices to Another intervention with tremendous potential to improve safety. Although statistically insignificant, the prevent burn injuries is the use of a safer and cleaner findings showed a decline in risks associated with poison- kerosene stove design that is competitively priced. The ing, as well as other injuries (Swart and others 2008). Global Alliance for Clean Cookstoves is investing sig- Complementing this work, Schwebel and others nificant resources in research and design improvements (2009) examined the effectiveness of a trainer-to-trainer for kerosene stoves (http://www.cleancookstoves.org). model, in which experts from the Paraffin Safety Nontransport Unintentional Injuries 63 Association of Southern Africa trained local community unintentional injuries in LMICs, the published data on members to distribute educational materials to an inter- the costs and cost-effectiveness of interventions are lim- vention community in South Africa. The educational ited. This section presents information on the costs materials were based on the theory of health behavioral incurred by such injuries, where data are available, and change. The findings showed that the intervention was on potential cost-effective interventions for LMICs, sup- effective at significantly changing the level of kerosene plemented by the best available information from HICs safety knowledge in the intervention community. The (table 4.5). All costs presented in this section have been researchers found slight behavioral changes related to converted to 2012 U.S. dollars. kerosene safety as well as to perceptions of risk (Schwebel and others 2009). Results from these two studies are consistent with Falls findings from HICs. A meta-analysis conducted in 2013, The costs of falls are well documented in HICs, but few for example, showed that home safety education used in data are available on the costs of falls in LMICs. However, interventions that included the provision of safety the costs are likely to be substantial. Falls and road traffic equipment were effective at increasing safety practices injuries accounted for the largest out-of-pocket health for preventing injury, including poisoning (Kendrick care costs for those hospitalized for injuries in Vietnam. and others 2013). Of road traffic victims, 26 percent experience cata- strophic expenditure as a result of their injuries (Nguyen Environmental Strategies for the Prevention of and others 2013). Poisoning There are no published data showing the cost- Some of the identified risk factors point to the need to effectiveness of fall prevention programs in LMICs. target broader environmental risk factors, such as by However, growing evidence suggests that some enacting and enforcing poisoning prevention legislation. community-based fall prevention programs among older Krug and others (1994), for example, demonstrated in a people in HICs can be cost-effective. Data from Australia controlled before-and-after study in South Africa that the on older adults living in the community show that the incidence rate of paraffin ingestion decreased by 47 most cost-effective intervention is the practice of Tai Chi; percent when child-resistant containers were widely dis- the cost per quality-adjusted life year (QALY) is US$49,119; tributed (Krug and others 1994). However, government the incremental cost per fall avoided is US$3,484 (Church policies mandating the use of such child-resistant con- and others 2011). And although evidence for effectiveness tainers do not exist in many LMICs (Balan and Lingam is still emerging, the data suggest that cataract surgery is 2012). In Turkey, only a limited number of medications potentially extremely cost-effective (the cost per QALY is are sold with child safety caps (Ozdemir and others 2012); US$3,818; the incremental cost per fall avoided is US$275. in Pakistan, a call for child-resistant packaging legislation For those taking psychotropic medications, medication has been made (Ahmed, Fatmi, and Siddiqui 2011). withdrawal is also highly cost-effective (with the cost per Other types of legislation, such as laws that mandate QALY of $22,711, and the incremental cost per fall standards for wick stoves in South Africa, are lacking in avoided of US$1,251 (Church and others 2011). many LMICs (Schwebel and Swart 2009). Suggested The evidence on cost-effective interventions in resi- interventions include ensuring that labels possess all the dential care settings in Australia is still emerging. The necessary safety information and are in languages that data suggest that medication review and the use of hip people can understand. Nonyelum and others (2010), for protectors among medium- and high-risk groups are example, showed that safety warnings on pharmaceutical highly cost-effective strategies. The former is more effec- and consumer products still need improvement in tive and less costly than no intervention; the cost per Nigeria. Their study revealed that only 70 percent of the QALY of the latter is US$2,002, and the incremental cost 600 products examined had adequate warning labels. per fall avoided is US$114 (Church and others 2011). Moreover, despite English being Nigeria’s official lan- Among all individuals living in residential settings, vita- guage, 5 percent of products had only non-English labels. min D supplementation has the potential to be extremely cost-effective at a cost of US$7,970 per QALY, and an incremental cost per fall avoided of US$444. COSTS AND COST-EFFECTIVENESS OF INTERVENTIONS Drowning Not surprisingly, given the dearth of evidence on effec- Cost and cost-effectiveness data for interventions to pre- tive interventions for the prevention of nontransport vent drowning in LMICs are scarce, given the paucity of 64 Injury Prevention and Environmental Health Table 4.5 Promising and Cost-Effective Interventions for Cause-Specific Injuries for HICs and LMICs, US$ 2012 Cause-specific injury HICs LMICs Falls Tai Chi: cost per QALY—$49,119; incremental cost per — fall avoided—$3,484 (Church and others 2011) Cataract surgery: cost per QALY—$3818; incremental — cost per fall avoided—$275 (Church and others 2011) Psychotropic medication withdrawal: cost per — QALY—$22,711; incremental cost per fall avoided—$1,251 (Church and others 2011) In residential settings, medication review: more — effective and less costly than no intervention (Church and others 2011) In residential settings, use of hip protectors: cost — per QALY—$2,002; incremental cost per fall avoided—$114 (Church and others 2011) In residential settings, vitamin D supplementation: — cost per QALY—$7,970; incremental cost per fall avoided—$444 (Church and others 2011) Drowning Fencing of residential swimming pools in homes with Supervision of children: $256 per DALY averted and children younger than age 18 years: cost per QALY— $8,703 per death averted (Rahman and others 2012) $35,212 to $43,663 (Segui-Gomez 2001) Purchase of personal flotation devices for boats: cost Swimming training for children ages four years and per QALY—$5,634 (Segui-Gomez 2001) older: $27 per DALY averted and $949 per death averted (Rahman and others 2012) Poisoning Distribution of child-resistant containers: $127 per DALY averted; $3,329 per death averted (Norton and others 2006) Note: — = not available; DALY = disability-adjusted life year; HICs = high-income countries; LMICs = low- and middle-income countries; QALY = quality-adjusted life year. information on effective interventions in these settings. younger than age 18 years ranged from US$35,212 to However, data have been published on the cost-effectiveness US$43,663 per QALY gained, depending on whether the of the PRECISE study in Bangladesh (Rahman and others fenced pools belonged to homes with children of differ- 2012). Cost-effectiveness was calculated using the WHO- ent age subgroups and whether an incremental installa- CHOICE guidelines (CHOosing Interventions that are tion was considered (Segui-Gomez 2001). Cost Effective), by determining the numbers of DALYs and Modeling of the cost-effectiveness of PFDs resulted deaths averted and the costs associated with both (http:// in figures of US$5,634 per QALY gained. Sensitivity www.who.int/choice/interventions/en/?) analyses were also conducted, suggesting that installing The cost-effectiveness of Anchal—the component of fencing around in-ground pools in homes with children the intervention that involved supervision of children younger than age 18 years and purchasing PFDs for ages one to five years in a community crèche—in reduc- recreational boats resulted in cost-effectiveness figures ing mortality was US$256 per DALY averted and well below those of many interventions implemented in US$8,703 per death averted. The cost-effectiveness of the clinical and public health realms. SwimSafe—the component of the intervention that involved children ages four years and older receiving swimming training—was US$27 per DALY averted and Burns US$949 per death averted. Overall, the cost-effectiveness Many burn injuries lead to prolonged hospital stays. In of PRECISE was US$114 per DALY averted and US$3,970 addition to acute burn care, patients often require a pro- per death averted. tracted period of rehabilitation. Only recently has the By comparison, earlier research focusing on interven- cost of providing reasonable burn care in LMICs been tions in HICs has shown that the cost-effectiveness of reported. Ahuja and Goswami (2013) calculated the cost fencing around residential pools in homes with children per patient (all medications and consumables, dressing Nontransport Unintentional Injuries 65 material, investigations, blood products, dietary costs, proportion of all injuries than women and girls. However, and salaries of all personnel) in a third-level teaching the reliability and validity of data from LMICs remain hospital in northern India to be US$1,102. uncertain, and improved data collection in these coun- Although the cost of burn care is relatively easy to tries needs to be prioritized. calculate and reflects the cost of survival from a major injury, albeit with disability, the cost-effectiveness of prevention programs is not easy to calculate. National or Risk Factors, Interventions, and Cost-Effectiveness regional statistics need to be available to measure the Falls effectiveness of prevention interventions. Interventions Despite evidence of a rising burden of falls in older people need to be combined with educational campaigns to in LMICs worldwide, few evidence-based prevention pro- institute safe behavioral practices, and studies evaluating grams have been implemented in these countries. all costs against all benefits with regard to burn injuries Governments have failed to recognize the costs of this are not available. Also, to establish the cost-effectiveness burden, resulting in inadequate policy development and of any action for preventing burns, one needs to factor in investment in prevention programs or prevention research. the elimination of the high cost of burn care and the Although HICs have an increasingly strong evidence prevention of disability, in addition to the decrease in base for effective and cost-effective programs to prevent burn incidence. falls in older people, policymakers in these countries need to better understand how such programs may be put into practice, both at the community level and in Poisoning residential care settings. Further, significant work needs Limited data exist on the cost of unintentional poison- to be undertaken by health care providers to adapt fall ings in LMICs. A study in Pakistan revealed that the costs prevention programs from HICs to LMICs, where risk of treatment to patients were considerable; approxi- factors may vary. Such programs will likely have to be mately 37 percent had to pay out of pocket. However, substantially modified for LMIC environments. only 9 percent of the patients were able to obtain govern- Consequently, more research is needed to enhance ment support to cover the treatment cost (Zia and others understanding of the likely contextual factors and unique 2012). contributors to falls among older people in LMICs. Such A detailed analysis of the cost-effectiveness of provid- factors include the influence of diet, physical activity, ing child-resistant containers in 2006 showed that, as a environment, and transportation, and the role of health means of preventing paraffin poisoning among children services. Fall prevention programs that target the physi- in South Africa, the intervention had a cost-effectiveness cal environment, inside and outside the home, may sig- ratio of $3,329 per death averted (Norton and others nificantly affect the success of such programs in older 2006). The impact of this intervention was calculated to be people in those countries. 263 DALYs averted, and cost-effectiveness was estimated Falls are also a significant cause of death and injury in to be US$127 per DALY at a 3 percent discount rate. children and working-age adults in HICs and LMICs. Although HIC studies suggest that environmental factors, including urban and street design as well as CONCLUSIONS building design, contribute significantly to falls in these population groups, the evidence base is weak. Burden of Unintentional Injuries Nevertheless, fall prevention programs involving envi- Recent global estimates provide a strong foundation for ronmental modification may have more of an impact in understanding the burden of death and disability associ- LMICs, particularly in countries with rapid urbanization ated with nontransport unintentional injuries. Falls, or areas with high levels of poverty. Similarly, focusing which are the most important cause of death and dis- on building design and safety standards for construction ability, are likely to become even more important as sites and workplaces is likely to reduce falls for people of populations in LMICs continue to age. Drowning and working age. Increasing regulation of consumer prod- burns are important contributors to the burden and ucts by governments, as well as community education on predominantly affect LMICs, especially younger chil- the appropriate use of such products, will be relevant as dren. Poisonings constitute the next leading contributor the use of these products increases in LMICs. to the unintentional injury burden and affect both HICs and LMICs, and particularly adults ages 70 years and Drowning older. With the exception of burn injuries, and to a lesser The need to address the burden of drowning in LMICs extent falls, men and boys account for a much higher is a neglected health issue in many countries, with very 66 Injury Prevention and Environmental Health few researchers focused on identifying effective interven- children, despite the fact that older adults are at highest tions or on examining differences in risk factors between risk. Data on the cost of unintentional poisoning are HICs and LMICs. The development of successful drown- limited, so the true economic burden of this public ing prevention strategies in LMICs faces a number of health problem is unknown. Few researchers have inves- obstacles: tigated the effectiveness of interventions in these set- tings; even fewer have studied their cost-effectiveness. • The absence of and need by researchers to better More studies are urgently needed. Cost studies, as well as identify risk factors for drowning, not only among benefit-cost analysis of successful behavioral programs, young children, but also among other age groups for example, can be vitally important. Findings could • The absence of and need by researchers to identify encourage donors and governments to invest in preven- effective intervention strategies, especially for older tive measures in LMICs. age groups • The need by governments and other stakeholders to scale up effective drowning interventions into Summary Conclusions national, regional, and global programs Nontransport unintentional injuries are comparable to • Capacity building for implementing drowning pre- transport injuries in terms of the burden of death and vention at all levels of program development disability, but they have not received the same attention • The need to stimulate and sustain investment in from government agencies or researchers. Recognition drowning prevention interventions and activities of the need to prevent falls and fall-related injuries • Incorporation of research into program design and among older people is likely to grow in LMICs as gov- implementation. ernments in these countries begin to address the grow- ing numbers of older people and the potential Burns cost-effectiveness of prevention strategies. However, Major causes of burn injuries in LMICs include poverty given the observed declines in the burden of other unin- and hazardous work environments, including in the tentional injuries, it seems less likely that government home. Therefore, progress on burn prevention can be initiatives will drive a strengthened evidence base or expected with countries’ socioeconomic growth and facilitate prevention initiatives for these injuries. government enforcement of regulations. In HICs, the Consequently, global support from the United Nations, introduction of smoke detectors and flame-retardant WHO, academia, nongovernmental organizations, and sleepwear, along with enforced safety practices in the commercial enterprises, in tandem with the injury con- workplace, have led to significant reductions in fire inju- trol community and health practitioners, will be impor- ries. These efforts will be less effective in LMICs until the tant to move the unintentional injury agenda forward. infrastructure improves. In the meantime, policymakers and health providers need to develop a better understanding of (1) risk factors NOTE for burn injuries in LMICs, (2) the economic impact of World Bank Income Classifications as of July 2014 are as burn injuries on survivors, and (3) the effectiveness and follows, based on estimates of gross national income (GNI) cost-effectiveness of burn prevention programs. per capita for 2013: Sufficient data are available from HICs to support the claim that burn injuries can be successfully prevented • Low-income countries (LICs) = US$1,045 or less using education, engineering changes, enforcement of • Middle-income countries (MICs) are subdivided: legislative protections, and environmental modifications a) lower-middle-income = US$1,046 to US$4,125 b) upper-middle-income (UMICs) = US$4,126 to (Peck, Molnar, and Swart 2009). US$12,745 In addition to focusing on prevention strategies, • High-income countries (HICs) = US$12,746 or more. health care organizations should encourage providers to be involved in specialized burn treatment at a local level. Moreover, encouraging participation with global initia- REFERENCES tives, such as the Global Alliance for Clean Cookstoves, Ahmed, B., Z. Fatmi, and A. R. Siddiqui. 2011. “Population can further the success of local initiatives. Attributable Risk of Unintentional Childhood Poisoning in Karachi Pakistan.” PLoS One 6 (10): e26881. Poisoning Ahuja, R. B., and S. Bhattacharya. 2002. “An Analysis of Most of the available studies on risk factors associated 11,196 Burn Admissions and Evaluation of Conservative with unintentional poisoning in LMICs focus on young Management Techniques.” Burns 28 (6): 555–61. Nontransport Unintentional Injuries 67 Ahuja, R. B., S. Bhattacharya, and A. Rai. 2009. “Changing Dandona, R., G. A. Kumar, R. Ivers, R. Joshi, B. Neal, and Trends of an Endemic Trauma.” Burns 35 (5): 650–56. others. 2010. “Characteristics of Non-Fatal Fall Injuries in Ahuja R. B., J. K. Dash, and P. Shrivastava. 2011. “A Comparative Rural India.” Injury Prevention 16 (3): 166–71. Analysis of Liquefied Petroleum Gas (LPG) and Kerosene Diplock, S., and K. Jamrozic. 2006. “Legislative and Regulatory Related Burns.” Burns 37 (8): 1403–10. Measures for Preventing Alcohol-Related Drownings and Ahuja, R. B., and P. Goswami. 2013. “Cost of Providing Near-Drownings.” Australian and New Zealand Journal of Inpatient Burn Care in a Tertiary Teaching Hospital of Public Health 30 (4): 314–17. North India.” Burns 39 (4): 558–64. Giashuddin, S. M., A. Rahman, F. Rahman, S. R. Mashreky, S. M. Akbaba, M., E. Nazlican, H. Demirhindi, Z. Sütoluk, and Y. Chowdhury, and others. 2009. “Socioeconomic Inequality Gökel. 2007. “Etiological and Demographical Characteristics in Child Injury in Bangladesh—Implication for Developing of Acute Adult Poisoning in Adana, Turkey.” Human and Countries.” International Journal for Equity in Health 8: 7. Experimental Toxicology 26 (5): 401–6. Gillespie, L. D., M. C. Robertson, W. J. Gillespie, C. Sherrington, American Burn Association. 2012. National Burn Repository. S. Gates, and others. 2012. “Interventions for Preventing Chicago, IL: American Burn Association. Falls in Older People Living in the Community.” Cochrane Balan, B., and L. Lingam. 2012. “Unintentional Injuries Database of Systematic Reviews 9: CD007146. among Children in Resource Poor Settings: Where Do Gururaj, G., V. Sateesh, and A. Rayan. 2008. Bengaluru Injury/ the Fingers Point?” Archives of Disease in Childhood 97 Road Traffic Injury Surveillance Programme: A Feasibility (1): 35–38. Study. Bengaluru: National Institute of Mental Health and Balme, K. H., J. C. Roberts, M. Glasstone, L. Curling, and Neuro Sciences. M. D. Mann. 2012. “The Changing Trends of Childhood Haagsma, J. A., N. Graetz, I. Bolliger, M. Naghavi, H. Higashi, Poisoning at a Tertiary Children’s Hospital in South Africa.” and others. 2015. “The Global Burden of Injury: Incidence, South African Medical Journal 102 (3): 142–46. Mortality, Disability-Adjusted Life Years and Time Bell, G. S., A. Gaitatzis, C. L. Bell, A. L. Johnson, and J. W. Sander. Trends from the Global Burden of Disease Study 2013.” 2008. “Drowning in People with Epilepsy: How Great Is the Injury Prevention. 0: 1–15. E-published 3 December 2015. Risk?” Neurology 71: 578–82. doi:10.1136/injuryprev-2015-041616. Bierens, J. J. L. M. 2006. Handbook on Drowning: Prevention, Harris, V. A., L. M. Rochette, and G. A. Smith. 2011. “Pediatric Rescue, Treatment. Berlin: Springer-Verlag. Injuries Attributable to Falls from Windows in the United Brenner, R. A., G. Saluja, and G. S. Smith. 2003. “Swimming States in 1990–2008.” Pediatrics 128 (3): 455–62. Lessons, Swimming Ability, and the Risk of Drowning.” Hestekin, H., T. O’Driscoll, J. S. Williams, P. Kowal, K. Peltzer, Injury Control and Safety Promotion 10 (4): 211–16. and others. 2013. “Measuring Prevalence and Risk Factors Brenner, R. A., G. S. Taneja, D. L. Haynie, A. C. Trumble, C. Qian, for Fall-Related Injury in Older Adults in Low- and Middle- and others. 2009. “Association between Swimming Lessons Income Countries: Results from the WHO Study on Global and Drowning in Childhood: A Case-Control Study.” Archives AGEing and Adult Health (SAGE).” SAGE Working Paper of Pediatrics and Adolescent Medicine 163 (3): 203–10. 6, World Health Organization, Geneva. http://cdrwww.who Bugeja, L., E. Cassell, L. R. Brodie, and S. J. Walter. 2014. .int/healthinfo/sage/SAGEWorkingPaper6_Wave1Falls.pdf. “Effectiveness of the 2005 Compulsory Personal Hyder, A. A., N. N. Borse, L. Blum, R. Khan, S. El Arifeen, and Flotation Device (PFD) Wearing Regulation in Reducing others. 2008. “Childhood Drowning in Low- and Middle- Drowning Deaths among Recreational Boaters in Income Countries: Urgent Need for Intervention Trials.” Australia.” Injury Prevention 20 (6): 387–92. doi:10.1136 Journal of Paediatrics and Child Health 44 (4): 221–27. /injuryprevention-2014-041169. Hyder, A. A., D. E. Sugerman, P. Puvanachandra, J. Razzak, Cameron, I. D., L. D. Gillespie, M. C. Robertson, G. R. H. El-Sayed, and others. 2009. “Global Childhood Murray, K. D. Hill, and others. 2012. “Interventions for Unintentional Injury Surveillance in Four Cities in Preventing Falls in Older People in Care Facilities and Developing Countries: A Pilot Study.” Bulletin of the World Hospitals.” Cochrane Database of Systematic Reviews 12: Health Organization 87 (5): 345–52. CD005465. International Life Saving Federation. 2007. World Drowning Chalmers, D., B. McNoe, and S. Stephenson. 2004. Drowning, Report. Gemeenteplein, Belgium: International Life Saving Near-Drowning and Other Water-Related Injury: Literature Association. Review and Analysis of National Injury Data. A Report to Jagnoor, J., L. Keay, N. Jaswal, M. Kaur, and R. Ivers. 2014. the Accident Compensation Corporation. Dunedin: Injury “A Qualitative Study on the Perceptions of Preventing Prevention Research Centre. Falls as a Health Priority among Older People in Northern Chatsantiprapa, K., J. Chokkanapitak, and N. Pinpradit. 2001. India.” Injury Prevention 20 (1): 29–34. “Host and Environment Factors for Exposure to Poisons: Jiang, X., Y. Zhang, Y. Wang, B. Wang, Y. Xu, and others. 2010. A Case-Control Study of Preschool Children in Thailand.” “An Analysis of 6215 Hospitalized Unintentional Injuries Injury Prevention 7 (3): 214–17. among Children Aged 0–14 in Northwest China.” Accident Church, J., S. Goodall, R. Norman, and M. Haas. 2011. An Analysis and Prevention 42 (1): 320–26. Economic Evaluation of Community and Residential Aged Jitapunkul, S., P. Yuktananandana, and V. Parkpian. 2001. “Risk Care Falls Prevention Strategies in NSW. Sydney: NSW Factors of Hip Fracture among Thai Female Patients.” Journal Ministry of Health. of the Medical Association of Thailand 84 (11): 1576–81. 68 Injury Prevention and Environmental Health Kalula, S. Z., V. Scott, A. Dowd, and K. Brodrick. 2011. “Falls Lozano, R., M. Naghavi, K. Foreman, S. Lim. K. Shibuya, and Fall Prevention Programmes in Developing Countries: V. Aboyans, and others. 2012. “Global and Regional Environmental Scan for the Adaptation of the Canadian Mortality from 235 Causes of Death for 20 Age Groups Falls Prevention Curriculum for Developing Countries.” in 1990 and 2010: A Systematic Analysis for the Global Journal of Safety Research 42 (6): 461–72. Burden of Disease Study 2010.” The Lancet 380 (9859): Kendrick, D., C. Coupland, C. Mulvaney, J. Simpson, S. J. Smith, 2095–128. and others. 2007. “Home Safety Education and Provision Mabrouk, A., A. El Badawy, and M. Sherif. 2000. “Kerosene of Safety Equipment for Injury Prevention.” Cochrane Stove as a Cause of Burns Admitted to the Ain Shams Burn Database of Systematic Reviews 1: CD005014. Unit.” Burns 26 (5): 474–77. Kendrick, D., M. C. Watson, C. A. Mulvaney, S. J. Smith, A. Mashreky, S. R., A. Rahman, T. F. Khan, L. Svanström, and J. Sutton, and others. 2008. “Preventing Childhood Falls F. Rahman. 2010. “Determinants of Childhood Burns in at Home: Meta-Analysis and Meta-Regression.” American Rural Bangladesh: A Nested Case-Control Study.” Health Journal of Preventive Medicine 35 (4): 370–79. Policy 96 (3): 226–30. Kendrick, D., B. Young, A. J. Mason-Jones, N. Ilyas, F. A. Achana, Mock, C., M. Peck, C. Juillard, D. Meddings, A. Gielen, and and others. 2012. “Home Safety Education and Provision others. 2011. Burn Prevention: Success Stories and Lessons of Safety Equipment for Injury Prevention.” Cochrane Learned. Geneva: World Health Organization. Database Systematic Reviews 9: CD005014. Moran, K. 2006. “Water Safety Knowledge, Attitudes and ———. 2013. “Home Safety Education and Provision of Safety Behaviours of Asian Youth in New Zealand.” In Proceedings Equipment for Injury Prevention (Review).” Evidence- of the Second International Asian Health and Wellbeing Based Child Health 8 (3): 761–939. Conference, November 13–14, edited by S. Tse, M. Hoque, Khambalia, A., P. Joshi, M. Brussoni, P. Raina, B. Morrongiello, K. Rasanathan, M. Chatterji, R. Wee, and others. Auckland, and others. 2006. “Risk Factors for Unintentional Injuries New Zealand: University of New Zealand. Due to Falls in Children Aged 0–6 Years: A Systematic Nguyen, H., R. Ivers, S. Jan, A. Martiniuk, and C. Pham. 2013. Review.” Injury Prevention 12 (6): 378–81. “Catastrophic Household Costs Due to Injury in Vietnam.” Kohli, U., V. S. Kuttiat, R. Lodha, and S. K. Kabra. 2008. “Profile Injury 44 (5): 684–90. of Childhood Poisoning at a Tertiary Care Centre in North Nonyelum, S. C., N. Nkem, C. N. Ofeyinwa, and O. E. Orisakwe. India.” Indian Journal of Pediatrics 75: 791–94. 2010. “Safety Warnings and First Aid Instructions on Kool, B., S. Ameratunga, M. Lee, E. Robinson, S. Crengle, and Consumer and Pharmaceutical Products in Nigeria: Has others. 2010. “Prevalence of Risk and Protective Factors There Been an Improvement?” Journal of the Pakistan for Falls in the Home Environment in a Population-Based Medical Association 60 (10): 801–4. Survey of Young and Middle-Aged Adult New Zealanders.” Norton, R., A. A. Hyder, D. Bishai, and M. Peden. 2006. Australian and New Zealand Journal of Public Health 34 (1): “Unintentional Injuries.” In Disease Control Priorities in 63–66. Developing Countries, second edition, edited by D. T. Jamison, Kool, B., S. Ameratunga, E. Robinson, S. Crengle, and R. Jackson. J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, 2008. “The Contribution of Alcohol to Falls at Home among D. B. Evans, P. Jha, A. Mills, and P. Musgrove. Washington, Working-Aged Adults.” Alcohol 42 (5): 383–88. DC: Oxford University Press and World Bank. Krug, A., J. B. Ellis, I. T. Hay, N. F. Mokgabudi, and J. Robertson. Norton, R., and O. Kobusingye. 2013. “Injuries.” New England 1994. “The Impact of Child-Resistant Containers on the Journal of Medicine 368 (18): 1723–30. Incidence of Paraffin (Kerosene) Ingestion in Children.” Ozdemir, R., B. Bayrakci, Ö. Teks ¸ am, B. Yalçın, and G. Kale. South African Medical Journal 84 (11): 730–34. 2012. “Thirty-Three-Year Experience on Childhood Laloë, V. 2002. “Epidemiology and Mortality of Burns in a Poisoning.” Turkish Journal of Pediatrics 54 (3): 251–59. General Hospital of Eastern Sri Lanka.” Burns 28 (8): 778–81. Peck, M., J. Molnar, and D. Swart. 2009. “A Global Plan for Lau, E. M., P. Suriwongpaisal, J. K. Lee, S. Das De, M. R. Festin, Burn Prevention and Care.” Bulletin of the World Health and others. 2001. “Risk Factors for Hip Fracture in Asian Organization 87 (10): 802–03. Men and Women: The Asian Osteoporosis Study.” Journal Peden, M., K. Oyegbite, J. Ozanne-Smith, A. A. Hyder, of Bone and Mineral Research 16 (3): 572–80. C. Branche, and others, eds. 2008. World Report on Child Lifshitz, M., and V. Gavrilov. 2000. “Acute Poisoning in Children.” Injury Prevention. Geneva: World Health Organization. The Israel Medical Association Journal 2 (7): 504–6. Peden, M. M, and K. McGee. 2003. “The Epidemiology of Linnan, M., A. Rahman, J. Scarr, T. Reinten-Reynolds, Drowning Worldwide.” Injury Control and Safety Promotion H. Linnan, and others. 2012. “Child Drowning: Evidence 10 (4): 195–99. for a Newly Recognized Cause of Child Mortality in Low Peiris-John, R., B. Kafoa, I. Wainiqolo, R. K. Reddy, E. McCaig, and Middle Income Countries in Asia.” Working Paper and others. 2013. “Population-Based Characteristics of 2012-07, Special Series on Child Injury 2, UNICEF Office Fatal and Hospital Admissions for Poisoning in Fiji: TRIP of Research, Florence. Project-11.” Injury Prevention 19 (5): 355–57. Lord, S. R., C. Sherrington, H. B. Menz, and J. C. T. Close. Rahman, A. 2010. “A Community Based Child Drowning 2007. Falls in Older People: Risk Factors and Strategies Prevention Programme in Bangladesh: A Model for Low for Prevention. 2nd ed. Cambridge, U.K.: Cambridge Income Countries.” Doctoral thesis, Public Health Science University Press. Department, Karolinska Institutet, Solna, Sweden. Nontransport Unintentional Injuries 69 Rahman, A., S. R. Mashreky, S. M. Chowdhury, M. S. Giashuddin, Household Air Pollution on Childhood Pneumonia in I. J. Uhaa, and others. 2009. “Analysis of the Childhood Fatal Guatemala (RESPIRE): A Randomised Controlled Trial.” Drowning Situation in Bangladesh: Exploring Prevention The Lancet 378 (9804): 1717–26. Measures for Low-Income Countries.” Injury Prevention Solomon, R., M. J. Giganti, A. Weiner, and M. Akpinar-Elci. 15: 75–79. 2013. “Water Safety Education among Primary School Rahman, F., S. Bose, M. Linnan, A. Rahman, S. Mashreky, Children in Grenada.” International Journal of Injury and others. 2012. “Cost-Effectiveness of an Injury and Control and Safety Promotion 20 (3): 266–70. Drowning Prevention Program in Bangladesh.” Pediatrics Soori, H. 2001. “Developmental Risk Factors for Unintentional 130 (6): e1621–28. Childhood Poisoning.” Saudi Medical Journal 22 (3): Ramos, C. L., H. M. Barros, A. T. Stein, and J. S. Costa. 2010. 227–30. “Risk Factors Contributing to Childhood Poisoning.” Jornal Stevenson, M. R., M. Rimajova, D. Edgecombe, and K. Vickery. de Pediatria 86 (5): 435–40. 2003. “Childhood Drowning: Barriers Surrounding Private Ranaweera, A. D., P. Fonseka, A. PattiyaArachchi, and Swimming Pools.” Pediatrics 3 (2): e115–19. S. H. Siribaddana. 2013. “Incidence and Risk Factors of Swart, L., A. van Niekerk, M. Seedat, and E. Jordaan. 2008. Falls among the Elderly in the District of Colombo.” Ceylon “Paraprofessional Home Visitation Program to Prevent Medical Journal 58: 100–106. Childhood Unintentional Injuries in Low-Income Rizvi, N., S. Luby, S. I. Azam, and F. Rabbani. 2006. “Distribution Communities: A Cluster Randomized Controlled Trial.” and Circumstances of Injuries in Squatter Settlements of Injury Prevention 14 (3): 164–69. Karachi, Pakistan.” Accident Analysis and Prevention 38 (3): Thompson, D. C., and F. P. Rivara. 2000. “Pool Fencing for 526–31. Preventing Drowning in Children.” Cochrane Database of Sanghavi, P., K. Bhalla, and V. Das. 2009. “Fire-Related Deaths Systematic Reviews 2: CD001047. in India in 2001: A Retrospective Analysis of Data.” The van der Molen, H. F., M. M. Lehtola, J. Lappalainen, Lancet 373 (9671): 1282–88. P. L. Hoonakker, H. Hsiao, and others. 2012. “Interventions Sawalha, A. F., W. M. Sweileh, M. T. Tufaha, and D. Y. Al-Jabi. to Prevent Injuries in Construction Workers.” Cochrane 2010. “Analysis of the Pattern of Acute Poisoning in Patients Database of Systematic Reviews 12: CD006251. Admitted to a Governmental Hospital in Palestine.” Basic WHO (World Health Organization). 2006. Guidelines for Safe and Clinical Pharmacology and Toxicology 107 (5): 914–18. Recreational Water Environments. Vol. 2: Swimming Pools Schwebel, D. C., and D. Swart. 2009. “Preventing Paraffin-Related and Similar Environments. Geneva: WHO. Injury.” Journal of Injury and Violence Research 1 (1): 3–5. ———. 2012. “Drowning.” Fact Sheet, WHO, Geneva. http:// Schwebel, D. C., D. Swart, J. Simpson, P. Hobe, and S. K. Hui. 2009. www.who.int/mediacentre/factsheets/fs347/en/index.html. “An Intervention to Reduce Kerosene-Related Burns and ———. 2014. “Global Health Estimates.” http://www.who.int Poisonings in Low-Income South African Communities.” /healthinfo/global_burden_disease/en/. Health Psychology 28 (4): 493–500. Yang, L., Q. Q. Nong, C. L. Li, Q. M. Feng, and S. K. Lo. 2007. Segui-Gomez, M. 2001. “Cost Effectiveness of Interventions “Risk Factors for Childhood Drowning in Rural Regions to Prevent Drowning and Near-Drowning.” Abstract at of a Developing Country: A Case-Control Study.” Injury 129th Annual Meeting of the American Public Health Prevention 13 (3): 178–82. Association, October 23. https://apha.confex.com/apha Zia, N., U. R. Khan, J. A. Razzak, P. Puvanachandra, and /129am/techprogram/paper_23731.htm. A. A. Hyder. 2012. “Understanding Unintentional Smith, K. R., J. P. McCracken, M. W. Weber, A. Hubbard, Childhood Home Injuries: Pilot Surveillance Data from A. Jenny, and others. 2011. “Effect of Reduction in Karachi, Pakistan.” BMC Research Notes 5: 37. 70 Injury Prevention and Environmental Health Chapter 5 Interpersonal Violence: Global Impact and Paths to Prevention James A. Mercy, Susan D. Hillis, Alexander Butchart, Mark A. Bellis, Catherine L. Ward, Xiangming Fang, and Mark L. Rosenberg INTRODUCTION NATURE AND BURDEN OF INTERPERSONAL Interpersonal violence is a pervasive public health, VIOLENCE human rights, and development challenge (Rosenberg Globally, the three primary forms of violence are and others 2006). Its effects reverberate through interpersonal violence; self-directed violence, includ- families, communities, and nations and across ing suicide; and collective violence, including war, generations. It is a leading cause of death among terrorism, and state-perpetrated violence in the form adolescents and young adults in most parts of the of genocide or torture (Dahlberg and Krug 2002). This world. Exposure to interpersonal violence increases chapter focuses on interpersonal violence. individuals’ lifelong vulnerability to a broad range of emotional, behavioral, and physical health problems. Interpersonal violence directly affects health care Definitions of Interpersonal Violence expenditures worldwide; indirectly, it affects national The World Health Organization (WHO) defines vio- and local economies—stunting development, lence as follows: “The intentional use of physical force increasing inequality, and eroding human capital or power, threatened or actual, against oneself, another (WHO 2008). person, or against a group or community that either Attention to interpersonal violence as a global issue results in or has a high likelihood of resulting in injury, has expanded dramatically since the World Health death, psychological harm, mal-development, or Assembly identified violence as a public health prior- deprivation” (Dahlberg and Krug 2002, 5). This defini- ity in 1996. Reports by the United Nations (UN) have tion encompasses interpersonal, self-directed, and contributed greatly to increased awareness (Krug and collective violence. others 2002; Pinheiro 2006; UN 2006). These and Interpersonal violence involves the intentional use other efforts culminated in specific targets for of physical force or power against other persons by an eliminating interpersonal violence in the UN’s post- individual or small group of individuals. Interpersonal 2015 Action Agenda for Sustainable Development violence may be physical, sexual, or psychological (UN 2015). (also called emotional violence), and it may involve Corresponding author: James A. Mercy, U.S. Centers for Disease Control and Prevention, Atlanta, Georgia, United States; jam2@cdc.gov. 71 deprivation and neglect. Acts of interpersonal violence the estimated rate of deaths due to interpersonal can be further divided into family or partner violence violence or homicide in LMICs was 8.0 per 100,000 and community violence. people, compared with 3.3 per 100,000 in high-income countries (HICs). • Family or partner violence refers to violence within Rates and patterns of violent death varied by region the family or between intimate partners. It includes (figure 5.1). Homicide rates were highest in LMICs in child maltreatment, dating and intimate partner vio- Latin America and the Caribbean and in Sub-Saharan lence (IPV), and elder maltreatment. Africa, and lowest in East Asia and the Pacific and in • Community violence occurs among individuals who some countries in northern Africa. In 2010, homicide are not related by family ties but who may know each was the leading cause of years of life lost in tropical and other. It includes youth violence, bullying, assault, central Latin America, the fourth leading cause in rape or sexual assault by acquaintances or strangers, southern Sub-Saharan Africa, and the eighth leading and violence that occurs in institutional settings such cause in the Caribbean and Eastern Europe (Lozano and as schools, workplaces, and prisons. others 2012). Poorer countries, especially those with large gaps between the rich and the poor, tend to have higher rates of homicide than wealthier countries The Burden of Interpersonal Violence (Butchart and Engstrom 2002). Information on the magnitude, nature, and conse- Homicide rates differed markedly by age and gen- quences of interpersonal violence is critical for program der (table 5.2). For infants and young children ages and policy development. 0–4 years, the rates for male and female homicide victims are 2.9 and 3.2 per 100,000, respectively; this Deaths Resulting from Interpersonal Violence is the only age range in which the female rate exceeds WHO’s Global Health Estimates (GHE) indicate that the male rate. For the 15- to 29-year-old age group, approximately 1.4 million people died in 2011 as a result rates for males were nearly five times those for of all three major forms of violence (table 5.1). Of those females; for the remaining older age groups, rates for deaths, 35.3 percent, or 504,587, were due to interper- males were around two to four times those for females. sonal violence. GBD estimates find that 83 percent of Homicide rates for females doubled between the ages all violence-related deaths occur in low- and middle- of 5–14 and 15–29 years and then decreased; however, income countries (LMICs), and 91.4 percent of deaths the rates increased again in women ages 70 years and due to interpersonal violence occur in LMICs. In 2011, above. Rates for males increased almost tenfold for Table 5.1 Estimated Violence-Related Deaths, by Type and Income Level, 2011 Category Number Rate per 100,000 peoplea Proportion of total (percent) Suicide 803,900 11.4 56.3 LMICs 606,698 10.5 75.5 HICs 197,201 15.2 24.5 Interpersonal 504,587 7.1 35.3 LMICs 461,429 8.0 91.4 HICs 43,158 3.3 8.6 Conflict related 119,463 1.7 8.4 LMICs 117,131 2.0 98.0 HICs 2,332 0.2 2.0 All types of violence 1,427,949 20.2 100.0 LMICs 1,185,259 20.5 83.0 HICs 242,691 18.8 17.0 Source: WHO 2014. Note: HIC = high-income country; LMIC = low- and middle-income country. a. Age standardized. 72 Injury Prevention and Environmental Health Figure 5.1 Homicide, Suicide, and Conflict-Related Fatality Rates, by Region, 2011 25 20 15 Percent 10 5 0 AFRO AMRO EMRO EURO SEARO WPRO Suicide Homicide Conflict related Source: WHO 2014. Note: AFRO = African Regional Office; AMRO = Americas Regional Office; EMRO = Eastern Mediterranean Regional Office; EURO = European Regional Office; SEARO = South-East Asia Regional Office; WPRO = Western Pacific Regional Office. Table 5.2 Estimated Global Homicide and Suicide Rates, by Age and Gender, 2012 Homicidesa Suicidesa Age Males Females Males Females 0–4 years 2.9 3.2 0.0 0.0 5–14 years 2.0 1.6 1.2 1.2 15–29 years 19.4 3.7 15.6 11.3 30–49 years 15.4 2.8 17.6 7.8 50–59 years 9.6 2.0 21.7 9.3 60–69 years 6.8 2.2 24.6 13.1 70+ years 6.4 3.5 43.2 20.8 Total 11.4 2.8 14.5 8.2 Source: WHO 2014. a. Per 100,000 people. males 15 to 29, declined slightly for those 30 to 49, and three homicides of females, are perpetrated by an then decreased with age. Overall, homicides resulted intimate partner (Stöckl and others 2013). in the deaths of slightly more than 4 males for every Firearms are associated with a substantial number 1 female. Recent estimates indicate that, globally, of homicides around the world. In 2010, an estimated about one in seven homicides, and more than one in 196,200 firearm homicides were committed in Interpersonal Violence: Global Impact and Paths to Prevention 73 nonconflict situations (Lozano and others 2012). In youth ages 10 to 24 years, interpersonal violence Firearm suicides are also an important problem in was the fifth leading cause of disability-adjusted life many countries, such as in the United States, where years (DALYs) in 2004, accounting for 3.5 percent of all more than 60 percent of all firearm deaths are suicides. DALYs in this age group (Gore and others 2011). The number of suicides committed with firearms Unfortunately, few studies in LMICs examine the nonfa- globally is unknown. tal consequences of youth violence. This research gap urgently requires filling. Nonfatal Interpersonal Violence Violence against Women In recent years, multiple reports using household survey The prevalence of violence against women has been doc- data have characterized the prevalence of interpersonal umented by Demographic and Health Surveys con- violence. ducted in Sub-Saharan African countries (Cameroon, Kenya, Malawi, Rwanda, Uganda, Zambia, and Violence against Children and Youth Zimbabwe) and by Reproductive Health Surveys (RHS) Results from the United Nations Children’s Fund conducted in Central and South American countries (UNICEF) Multiple Indicator Cluster Surveys and (Ecuador, El Salvador, Guatemala, Jamaica, Nicaragua, Demographic and Health Surveys (DHS) in and Paraguay). In Sub-Saharan Africa, the DHS findings 33 countries found that an average of 76 percent of show that the prevalence of physical violence against children ages 2 to 14 years had experienced some form women ranged from 30 percent in Malawi to 60 percent of violent physical or psychological discipline during in Uganda, with most perpetrators being intimate the previous month (UNICEF 2010). Surveys of vio- partners; the prevalence of sexual and emotional vio- lence against children in five countries (Haiti, Kenya, lence by intimate partners was also high (Borwankar, Swaziland, Tanzania, and Zimbabwe) found that the Diallo, and Sommerfelt 2008). For Central and South prevalence of sexual violence against girls was America, RHS findings show prevalence of physical vio- 26 percent to 38 percent, and against boys it was lence against women ranging from 17 percent in Jamaica 9 percent to 21 percent (CDC, INURED, and the to 31 percent in Ecuador (Bott and others 2012). Global Comité de Coordination, 2014; Reza and others 2009; and regional estimates of violence against women UNICEF, CDC, and KNBS 2012; UNICEF, CDC, and demonstrate that 35 percent of women worldwide have Muhimbili University 2012; ZIMSTAT, UNICEF, and experienced physical or sexual violence (or both), and CCORE 2012). The prevalence of physical violence most of that violence was perpetrated by intimate against girls was 61 percent to 74 percent, and against partners (WHO 2013). boys it was 57 percent to 76 percent. The prevalence of emotional violence against girls was 24 percent to Violence against Elderly People 35 percent, and against boys it was 27 percent to Elder maltreatment has been examined using 39 percent. Regardless of the type of violence, perpe- population-based surveys and records from adult trators are largely known to the victim, and violence protective services. In surveys, 6.0 percent of older tends to occur in homes. people reported significant abuse in the past month, and In meta-analyses of studies worldwide, 11 percent 5.6 percent of couples reported physical violence in their to 22 percent of girls and 4 percent to 19 percent of relationship in the past year (Cooper, Selwood, and boys have experienced child sexual abuse, 14 percent to Livingston 2008). In studies involving vulnerable elders 55 percent have experienced child physical abuse, in nursing or care homes, nearly 25 percent reported 12 percent to 22 percent have experienced physical significant levels of psychological abuse. Rates of abuse neglect, and 13 percent to 25 percent have experienced reported to adult protective services are generally very emotional neglect (Stoltenborgh and others 2011; low (1 percent to 2 percent). Stoltenborgh and others 2013a; Stoltenborgh, Bakermans-Kranenburg, and van Ijzendoorn 2013b). The hidden nature of child sexual and physical abuse Consequences of Interpersonal Violence is poignant. When compared with official reports, The consequences of experiencing interpersonal vio- self-reported prevalence of child sexual abuse was lence are pervasive and enduring. Evidence confirms more than 30 times the official rate (Stoltenborgh and that exposure to violence increases the risks of injuries, others 2011); self-reported prevalence of physical infectious diseases, mental health problems, reproduc- abuse was more than 75 times the official rate tive health problems, and noncommunicable diseases (Stoltenborgh and others 2013a). (NCDs). 74 Injury Prevention and Environmental Health Cause of Physical or Psychological Injury Increased Risk of Mental Health Problems Although injury historically has been defined as an indi- Globally, studies from high-, middle-, and low-income vidual’s experience of physical damage, the definition has countries document that violent experiences lead to var- been expanded to include damage that is psychological, ious mental health consequences. The WHO World with the potential to lead to maldevelopment or depriva- Mental Health Survey findings from 21 countries tion (Norton and Kobusingye 2013). Whether they are demonstrate that violence during childhood is associ- physical or psychological, violence-associated injuries ated with mood, anxiety, behavior, and substance commonly go unrecognized and range from self-limiting disorders, as well as suicidal behavior, during adulthood to severe. Physical injuries include lacerations, bruises, (Kessler and others 2010). Furthermore, studies involv- wounds, fractures, broken teeth, ocular damage, burns, ing 21,000 women from Asia, Sub-Saharan Africa, and internal injuries, and head injuries. Such injuries, espe- Latin America and the Caribbean confirm strong associ- cially those associated with highly lethal means such as ations between various forms of violence—including firearms, may lead to disability, including brain damage, experiencing and witnessing IPV, nonpartner physical amputations, or paralysis (Buchanan 2013). violence, and childhood sexual abuse—and suicides (Devries and others 2011). Link to Infectious Diseases The association between sexual and physical violence and Increased Risk of Future Violence infectious diseases, particularly sexually transmitted infec- Exposure to violence during childhood increases the tions (STIs) and human immunodeficiency virus/acquired risk of experiencing or perpetrating violence later in immune deficiency syndrome (HIV/AIDS) is well sup- life. Experiencing child maltreatment and witnessing ported. Evidence across multiple studies demonstrates that partner abuse have consistently been shown to increase these associations are strong, largely consistent, graded, and the risk of becoming either a perpetrator or a victim of biologically plausible (Andersson, Cockcroft, and Shea sexual violence and IPV as an adult (Capaldi and 2008; Machtinger, Wilson, and others 2012). Emerging evi- others 2012; Tharp and others 2012). This intergener- dence suggests that violence may be associated with the ational effect of childhood violence increases the risk transmission and progression of infections, increases in that men will become perpetrators and that women antiretroviral failure, high-risk behaviors, and an indepen- will become victims. An assessment of Reproductive dently elevated risk of HIV/AIDS-associated death Health Surveys in six countries in the Americas found (Machtinger, Haberer, and others 2012). The importance of that the proportion of women reporting IPV was more gender-based violence as a driver of HIV/AIDS in women is than twice as high for those who experienced sexual or so prominent that multilateral donors such as UN Women physical abuse in childhood as for those who did not view elimination of violence against women and children as (Bott and others 2012). a key strategy for advancing prevention (IOM 2013). Increased Risk to Special Populations Increased Risk of Reproductive Problems Although most reports addressing interpersonal violence Multiple studies document the reproductive conse- focus on the general population, some recent studies quences of exposure to child maltreatment and IPV. have addressed infectious, reproductive, and mental These forms of violence are associated with unintended health consequences of violence for children outside of pregnancy and teen pregnancy, and they influence vic- family care, including street children, trafficked children, tims’ associated risk behaviors, such as multiple partners those affected by crises and armed conflict, and those and early initiation of sexual activity (Hillis and others living in institutions such as orphanages. For street 2004). The intergenerational effects of exposure to child- children, studies from LMICs report HIV/AIDS sero- hood violence may be extreme. For example, violence prevalences of 40 percent and higher among those who against girls increases the future risk of adverse preg- experienced childhood violence, in contrast to general nancy outcomes, such as fetal death (Hillis and others population prevalences of 1 percent (Kissin and others 2004). In addition, mortality for young children is sig- 2007; Robbins and others 2010). Associations between nificantly higher when their mothers are victims of IPV violence and elevated risks of HIV/AIDS and STIs, preg- (Silverman and others 2011). A review of studies from nancy, psychiatric pathology, depression, anxiety, post- 17 LMICs shows that IPV leads to an increased prevalence traumatic stress disorder, and suicide have been reported of pregnancy-associated mental health disorders, such as among victims of sex trafficking and armed conflict, postpartum depression, which impair a mother’s ability as well as among those in orphanages (Reed and others to provide a safe, stable, and nurturing environment for 2012; Silverman and others 2009; Zapata and others her children (Fisher and others 2012). 2011, 2013). Interpersonal Violence: Global Impact and Paths to Prevention 75 Increased Risk of NCDs Framework for Estimating the Costs of Violence Violence during childhood is also associated with non- Challenges in creating such a framework include communicable diseases (NCDs) that often only “weaknesses in the knowledge base both in economic become evident decades later. Exposure to childhood costing and in violence prevention, difficulty in creating violence leads to consistent and graded increases in the a universal algorithm for diverse settings, and disagree- four NCDs that accounted for nearly 60 percent of the ments in types of costs to include” (IOM and 53 million deaths globally in 2010—cardiovascular NRC 2012, 7). Although no methodology exists to enu- disease, cancer, chronic lung disease, and diabetes merate the full impact of violence, costs that are com- (Lozano and others 2012; Norman and others 2012). monly considered include direct costs, which arise In both HICs and LMICs, childhood violence has been proximal to the violent event, and indirect costs, which associated with major risk factors for these diseases, result from consequences, externalities, or lost opportu- including alcohol abuse, tobacco use, physical inac- nities (IOM and NRC 2012). In general, direct costs tivity, and obesity (Anda and others 2010). Beyond typically include those associated with medical care, health effects, serious psychosocial effects of child- psychological care, property damage, policing, incarcer- hood violence that are observed decades later include ation, and residential treatment; indirect costs are those severe problems with finances, family, jobs, anger, and commonly associated with lost wages and decreased stress (Hillis and others 2004). productivity. Current approaches that largely confine estimates of health-related costs to proximal conse- Basic Science Evidence quences lead to marked underestimates because they fail The biological underpinnings of the empirical associ- to incorporate costs of HIV/AIDS, chronic diseases, and ations between exposure to violence and subsequent other conditions attributable to violence. major causes of mortality in adulthood have been A comprehensive approach to estimating costs will established through basic science. Recent evidence strengthen global efforts to elevate the urgency of vio- demonstrates that traumatic stress, such as that associ- lence prevention (WHO 2004). Comparisons across ated with violence in childhood, impairs brain archi- countries of the costs of interpersonal violence are com- tecture (both structure and function), immune status, plicated by variations in definitions, types of costs, dis- metabolic systems, and cellular inflammatory count rates, comparable data, and methodology. Reports responses (Anda and others 2010). It is clear that early vary greatly in the types of costs they include, whether exposure to toxic stress in childhood confers lasting such costs are disaggregated, and whether they include damage at the most basic levels of the nervous, costs associated with both victimization and perpetra- endocrine, and immune systems, and that such expo- tion. Although the absolute costs appear to be higher in sures can alter the physical structure of DNA (epige- HICs than LMICs, the relative costs of violence as a netic effects) (Danese and McEwen 2012). Important proportion of government spending are often high in research summarizing the effects of early childhood both types of economies (table 5.3). experiences suggests that those multifaceted Reports of direct and indirect societal costs of inter- gene-environment interactions that cause negative personal violence in general vary widely, ranging from health consequences after exposure to chronic stress US$75.2 million (2013 dollars) for homicide in New also appear to confer positive health consequences Zealand in 1992 to US$579.4 billion (2013 dollars) for after exposure to early environments that are engaging homicide, child abuse, sexual and other assault, and rob- and nurturing (Heim and Binder 2012). Epidemiologic bery in the United States in 1993 (Fanslow and others research complements these findings, demonstrating 1997; Miller, Cohen, and Wiersema 1996). Estimates of that early nurturing in the home leads to sustained the costs of both interpersonal and collective violence in positive economic and psychosocial consequences up the Americas show that direct and indirect economic to five decades later (Hillis and others 2010). costs ranged from 5.1 percent of 1997 gross domestic product (GDP) in Peru to 24.9 percent of 1997 GDP in ECONOMIC BURDEN OF INTERPERSONAL El Salvador (Buvinic, Morrison, and Shifter 1999). Other estimates found that interpersonal violence accounted VIOLENCE for 4.0 percent of GDP in Jamaica in 2006, 1.2 percent in Given the high prevalence of interpersonal violence and Brazil in 2004, and 0.4 percent in Thailand in 2005 its extensive consequences, the associated economic (Butchart and others 2008; Ward and others 2009). impact is substantial. However, no comprehensive Reports that estimate the national costs of child mal- framework for estimating the total economic burden of treatment are largely from HICs, whereas those that violence exists. address IPV also include LMICs. In HICs, for example, 76 Injury Prevention and Environmental Health Table 5.3 Summary of Costs of Interpersonal Violence, by Types of Violence Violence-associated Violence- Costs converted costs as percentage associated costs Discount to 2013 U.S. of national health as percentage of Study Country Costs included (time frame) rate Total costs (per year) dollars* expenditures GDP Interpersonal violence in general Miller, Cohen, and United States Victim costs: medical and mental health n.a. US$358.0 billion for US$577.2 billion n.a. n.a. Wiersema 1996 care, victim services, productivity and homicide, child abuse, sexual quality of life losses (1993) (lifetime) and other assault, and robbery (1993) Bellis and England and For homicide, wounding, assault, and n.a. £29.9 billion (2008–09) US$53.8 billion n.a. n.a. others 2012 Wales sexual assault: criminal justice system, health and victim services, foregone output, and physical and emotional costs (lifetime) Mayhew 2003 Australia For homicide, assault, and sexual assault: n.a. $A2.6 billion (2001) US$1.8 billion n.a. n.a. medical costs, lost wages, intangible costs such as pain, suffering, and reduced quality of life (annual) Fanslow and New Zealand Lost earnings, legal fees, incarceration, and n.a. $NZ83.0 million, averaging US$74.9 million, n.a. n.a. others 1997 policing costs associated with homicide $NZ1 million per homicide averaging (annual) (1992) US$900 thousand per homicide Interpersonal Violence: Global Impact and Paths to Prevention Buvinic, Morrison, Peru For both interpersonal violence and n.a. n.a. n.a. n.a. 5.1 percentage of and Shifter 1999 collective violence: health impacts; private 1997 GDP and public expenditures on police and security services; citizens’ willingness to pay to live without violence; and value of goods lost, ransoms, and bribes (annual) Buvinic, Morrison, Brazil For both interpersonal violence and n.a. n.a. n.a. n.a. 10.5 percentage of and Shifter 1999 collective violence: health impacts; private 1997 GDP and public expenditures on police and security services; citizens’ willingness to pay to live without violence; and value of goods lost, ransoms, and bribes (annual) Buvinic, Morrison, Venezuela, For both interpersonal violence and n.a. n.a. n.a. n.a. 11.8 percentage of and Shifter 1999 RB collective violence: health impacts; private 1997 GDP and public expenditures on police and security services; citizens’ willingness to pay to live without violence; and value of goods lost, ransoms, and bribes (annual) table continues next page 77 78 Table 5.3 Summary of Costs of Interpersonal Violence, by Types of Violence (continued) Violence-associated Violence- Costs converted costs as percentage associated costs Injury Prevention and Environmental Health Discount to 2013 U.S. of national health as percentage of Study Country Costs included (time frame) rate Total costs (per year) dollars* expenditures GDP Buvinic, Morrison, Mexico For both interpersonal violence and n.a. n.a. n.a. n.a. 12.3 percentage of and Shifter 1999 collective violence: health impacts; private 1997 GDP and public expenditures on police and security services; citizens’ willingness to pay to live without violence; and value of goods lost, ransoms, and bribes (annual) Buvinic, Morrison, Colombia For both interpersonal violence and n.a. n.a. n.a. n.a. 24.7 percentage of and Shifter 1999 collective violence: health impacts; private 1997 GDP and public expenditures on police and security services; citizens’ willingness to pay to live without violence; and value of goods lost, ransoms, and bribes (annual) Buvinic, Morrison, El Salvador For both interpersonal violence and n.a. n.a. n.a. n.a. 24.9 percentage of and Shifter 1999 collective violence: health impacts; private 1997 GDP and public expenditures on police and security services; citizens’ willingness to pay to live without violence; and value of goods lost, ransoms, and bribes (annual) Butchart and Brazil Lifetime medical and productivity losses 3 percent R$16.1 billion (2004) US$6.4 billion 0.4 percent 1.2 percentage of others 2008 associated with self-directed and 2004 GDP interpersonal violence (lifetime) Butchart and Jamaica Lifetime medical and productivity losses 3 percent J$29.6 billion (2006) US$530.0 million 12.0 percent 4.0 percentage of others 2008 associated with self-directed and 2006 GDP interpersonal violence (lifetime) Butchart and Thailand Lifetime medical and productivity losses 3 percent B15.7 billion (2005) US$453.3 million 4.0 percent 0.4 percentage of others 2008 associated with self-directed and 2005 GDP interpersonal violence (lifetime) Child maltreatment Fang and others United States Childhood health care, adult medical, lost 3 percent • US$124.0 billion lifetime US$132.5 billion n.a. n.a. 2012 productivity, child welfare, criminal justice, costs from new cases of and special education costs (lifetime) fatal and nonfatal child maltreatment (2010) table continues next page Table 5.3 Summary of Costs of Interpersonal Violence, by Types of Violence (continued) Violence-associated Violence- Costs converted costs as percentage associated costs Discount to 2013 U.S. of national health as percentage of Study Country Costs included (time frame) rate Total costs (per year) dollars* expenditures GDP • US$210,012 average lifetime costs per victim of nonfatal child maltreatment • US$1,272,900 average lifetime costs per death (2010 dollars) Habetha and Germany Costs associated with health care, social, n.a. a11.1 billion (2008) US$9.7 billion n.a. n.a. others 2012 and educational services; foster care; productivity losses (lifetime) Mendonca, Alves, Recife, Brazil Hospital costs (annual) n.a. n.a. n.a. Violence against n.a. and Cabral Filho children and 2002 adolescents accounted for 65.1 percent of hospital admissions and 77.9 percent of all hospital costs in the state of Pernambuco Interpersonal Violence: Global Impact and Paths to Prevention (1999) Intimate partner violence CDC 2003 United States Medical and mental health care costs, lost n.a. US$5.8 billion (1995) US$8.9 billion n.a. n.a. productivity (annual) Day 1995 Canada Costs of health care, policing, legal n.a. Can$1.2 billion (1992) US$1.7 billion n.a. n.a. fees, incarceration, lost earnings, and psychological trauma (lifetime) Morrison and Chile Lost productive capacity of abused women n.a. Ch$1.6 billion (1996) US$5.8 million n.a. 2.0 percentage of Orlando 1999 (annual) 1996 GDP Morrison and Nicaragua Lost productive capacity of abused women n.a. C$29.5 million (1996) US$5.2 million n.a. 1.6 percentage of Orlando 1999 (annual) 1996 GDP table continues next page 79 80 Injury Prevention and Environmental Health Table 5.3 Summary of Costs of Interpersonal Violence, by Types of Violence (continued) Violence-associated Violence- Costs converted costs as percentage associated costs Discount to 2013 U.S. of national health as percentage of Study Country Costs included (time frame) rate Total costs (per year) dollars* expenditures GDP International Center Bangladesh Costs of health care, justice, and lost n.a. n.a. n.a. n.a. 4.5 percentage of for Research on productivity (annual) 2007 per capita Women 2009 gross national income International Center Morocco Costs of health care, justice, and lost n.a. n.a. n.a. n.a. 22.0 percentage for Research on productivity (annual) of 2007 per capita Women 2009 gross national income International Center Uganda Costs of health care, justice, and lost n.a. n.a. n.a. n.a. 4.0 percentage of for Research on productivity (annual) 2007 per capita Women 2009 gross national income Roldós and Corso Ecuador Costs of medical and legal services, and n.a. US$109.0 million (2012) US$110.6 million n.a. n.a. 2013 productivity losses (annual) Youth violence Miller, Fisher, and Pennsylvania, Total victim costs, including quality of life 2.5 percent US$5.4 billion (1993) US$8.7 billion n.a. n.a. Cohen 2001 United States and productivity losses of juvenile violence (lifetime) Note: * Conversions to 2013 U.S. dollars using Department of Labor Consumer Price Index Inflation Calculator; GDP = gross domestic product; n.a. = not applicable. Fang and others (2012) estimated the U.S. total lifetime Orlando 1999). Although adolescents and young adults economic burden resulting from new cases of child mal- commit a disproportionate share of all violence and, treatment in 2008 to be US$135 billion (2013 dollars); therefore, account for a high proportion of its cost, an analysis using similar methods showed total costs of youth violence has been the subject of few economic cost US$19.0 billion (2013 dollars) in Germany (Habetha studies (WHO 2004). and others 2012). HIC estimates of the annual direct and indirect costs of IPV against women exceeded US$8.9 billion in the United States and US$1.7 billion (all 2013 Risk and Protective Factors for Violence dollars) in Canada (CDC 2003; Day 1995). For LMICs, Violence results from the interplay of risk factors and costs of the lost productive capacity of abused women as protective factors. In table 5.4 these risk factors are a percentage of GDP were equivalent to 2.0 percent in organized by levels of the ecological model, which exam- Chile, 1.6 percent in Nicaragua, and 22.0 percent ines the relationship between individual and contextual in Morocco (table 5.3) (ICRW 2009; Morrison and factors and considers violence as the product of these Table 5.4 Risk for Perpetrating Violence Level of the ecological model Risk factors Individual • Early exposure to violence and adverse events, including child maltreatment and intimate partner violence • Male gender • Youth • Neuropsychological deficits, including attention-deficit hyperactivity disorder and learning disabilities • Personality disorders • Alcohol and substance misuse • History of violence Household • Intimate partner violence • Household members with criminal records • Harsh, cold, or inconsistent parenting • Low socioeconomic status Peer group • Association with others who use and endorse the use of violence Community • High residential mobility • High unemployment • High population density • Poverty • Drug trade • Inadequate victim care services Societal • Rapid social change • Economic inequality • Gender inequality • Policies that sustain or increase inequalities • Patriarchal norms that prioritize men’s power over women and adults’ power over children • Societal norms that support violence • Poor rule of law • Weak criminal justice system • Availability of lethal means, for example, firearms Source: Krug and others 2002. Interpersonal Violence: Global Impact and Paths to Prevention 81 multiple levels of influence on behavior (Dahlberg and risk of homicides, especially among women (Miller, Krug 2001). Identifying these factors is important, Azrael, and Hemenway 2013). The same authors because increasing protection and decreasing risk under- report that in cross-national comparisons of HICs, lie effective prevention. Several cross-cutting risk factors higher homicide rates have been associated with for perpetrating violence are described in table 5.4. greater access to firearms. Many of these factors also increase an individual’s likeli- hood of being a victim; for example, young men ages 15 to 44 are most likely to be both victims and perpetra- INTERVENTIONS AND THEIR APPLICABILITY tors in any country. Additional factors, such as having a TO LOW- AND MIDDLE-INCOME COUNTRIES disability (Hughes and others 2012), increase the risk of becoming a victim but not a perpetrator. Public health interventions aim to prevent violence from The recognition that different types of interpersonal occurring. Prevention efforts addressing common violence share common risk factors, often occur in underlying risk factors have the potential to decrease combination, and may be causal factors for one other is several different forms of violence simultaneously. Such important (Reza, Mercy, and Krug 2001). For example, efforts include two broad groups of interventions: child maltreatment is a risk factor for youth violence and IPV. Three cross-cutting risk factors bear particular • The first group targets documented risk and pro- mention because they represent factors that, if success- tective factors (for example, enhancing support fully addressed by prevention initiatives, could have for parents, reducing the availability and abuse of substantial impact: parenting, substance abuse (particu- alcohol, and reducing access to lethal means) in larly alcohol), and the availability of lethal means. well-defined target groups, such as adolescents. This Harsh, cold, and inconsistent parenting has been group includes specific violence prevention programs linked to youth violence (van der Merwe, Dawes, and implemented at the community, state and provincial, Ward 2012), IPV (Ireland and Smith 2009), and the and national levels. increased risk of abusing one’s own children (Thornberry, • The second group consists of policies and programs Knight, and Lovegrove 2012). In contrast, children who that address the social determinants of violence, receive warm, consistent parenting tend to have better including efforts to improve the conditions of daily outcomes (Eisenberg and others 2005; Smith, Landry, life and to promote more equitable distribution of and Swank 2010). power, money, and resources. Substance abuse, and particularly alcohol abuse, is implicated in a number of ways in both victimization Policy makers understand that a single intervention and perpetration of violence (Monteiro 2007; WHO or policy will not solve the whole problem, nor will one 2006). Maternal use of alcohol during pregnancy can sector solve it alone; as with automobile safety, the solu- result in fetal alcohol spectrum disorders, with their tions will be incremental and will require multisectoral attendant executive functioning disorders (Mattson and collaboration among policy makers in criminal justice, others 2013), which increase the risk of aggression in public health, education, and other areas. affected children. Alcohol use also reduces self-control The design, targeting, monitoring, and evaluation of and the ability to process information (Giancola 2000), both groups of interventions are enabled by the availabil- making it more likely both that drinkers will use violence ity of timely and reliable surveillance information about in response to perceived threats and that they will be outcomes of interest, including homicides, nonfatal inju- vulnerable to victimization (Klosterman and Fals- ries treated in emergency departments, and self-reported Stewart 2006). Alcohol misuse has been implicated violence recorded through surveys. Indeed, one program across all forms of violence, including perpetration of that has significantly reduced violence-related injuries in child maltreatment (Gilbert and others 2009) and elder Cardiff, Wales, is based on the systematic sharing of abuse (Lachs and Pillemer 2004), and in both victimiza- anonymous data from hospital emergency rooms and tion and perpetration of youth violence (van der Merwe, the police to better identify high-risk locations for vio- Dawes, and Ward 2012) and IPV (Jewkes 2002). lence (Florence and others 2013). Such locations become Access to lethal means of perpetrating interpersonal the focus of situation-specific interventions to reduce violence, such as firearms and sharp objects, contrib- risks, for example, by increasing the presence of police utes substantially to the likelihood that such violence patrols at high-risk times where alcohol is served, alter- will result in death or serious injury (Beaman and ing practices around the serving of alcohol (such as the others 2000). In the United States, the presence of a mandatory use of plastic barware), and instituting crowd firearm in the home is associated with an increased control at public transportation stops. 82 Injury Prevention and Environmental Health Specific Violence Prevention Programs Project in Canada (Wolfe and others 2009) are Most of the scientific evidence for specific prevention evidence-based approaches that could be adapted to programs to date is from HICs. Although conditions LMICs. Life skills and social development training differ in LMICs, table 5.5 shows seven categories of vio- programs are popular in LMICs. Some include evalu- lence prevention programs in HICs that are scientifically ations of effectiveness (for example, PREPARE in credible, along with the types of violence they prevent South Africa). However, because programs are typi- and considerations for their applicability in LMICs. cally delivered in schools, they depend on the readi- ness of the educational system to implement the program and reinforce its effects. Oversight and man- Developing Safe, Stable, and Nurturing Relationships agement structures must be in place before such pro- between Parents or Caregivers and Children grams are implemented. Interventions that support the development of safe, sta- ble, and nurturing relationships between parents or Reducing the Availability and Harmful Use of Alcohol caregivers and children in their early years can prevent Alcohol availability can be regulated by restricting the child maltreatment and reduce childhood aggression hours of sale and reducing the number of alcohol (Bilukha and others 2005; Kaminski and others 2008). retail outlets (Cohen 2007; Duailibi and others 2007; Emerging evidence suggests that such relationships can Nemtsov 1998). Reduced hours of sales have been also reduce violence in adolescence and early adulthood; associated with reduced violence, and higher outlet theoretical grounds exist for assuming they decrease IPV densities have been associated with higher levels of and self-directed violence in later life (Caldera and violence. Empirical evidence has shown that higher others 2007; Olds and others 1998; Walker and others prices for alcohol can decrease consumption and 2011). In addition, these relationships offer the potential reduce mortality attributed to alcohol (Zhao and oth- to prevent problem behaviors, such as substance misuse, ers 2013). Moreover, economic modeling suggests that eating disorders, and unsafe sex, which are important price increases can reduce violence (Markowitz and risk factors for NCDs; STIs, including HIV/AIDS; and Grossman 1998, 2000). Brief interventions and unintentional injuries. longer-term treatment for problem drinkers have Although most evidence for the effectiveness of par- been shown to reduce child maltreatment and IPV enting programs comes from HICs (Knerr, Gardner, and (Dinh-Zarr and others 2004). Interventions in and Cluver 2013; Mikton and Butchart 2009), several initia- around drinking establishments that target crowding, tives to evaluate such programs in LMICs have recently management practice, physical design, staff training, been established, for example, the Children and Violence and access to late night transportation also show Evaluation Challenge Fund. In addition, parenting pro- promise in reducing violence (Bellis and Hughes grams are quite widely implemented in LMICs to sup- 2008; Graham and Homel 2008). port early child development, raising the possibility that Although most evidence for the effectiveness of violence prevention components could be integrated interventions in preventing violence comes from into those programs. HICs, several success stories come from LMICs, including the reduced trading hours in Brazil (Duailibi Developing Life Skills in Children and Adolescents and others 2007). Given that LMICs show some of the Social development programs to build social, emotional, greatest increases in alcohol consumption, more out- and behavioral competencies can prevent violence come evaluations of strategies to address alcohol-related (Hahn and others 2007; Hawkins and others 1999; violence in these settings are urgently needed. Two Klevens and others 2009). Preschool enrichment pro- areas that should be explored are the effectiveness of grams that provide children with academic and social (1) minimum drinking-age laws and (2) efforts to skills at an early age appear promising (Baker- regulate the marketing of alcohol. However, in many Henningham and others 2012; Nelson, Westhues, and LMICs, a large proportion of alcohol consumed is MacLeod 2003). However, outcomes vary greatly across produced at home. In such settings, establishing poli- programs, and relatively few programs have been evalu- cies to regulate alcohol production and sale is an ated for their effects on violence (Durlak, Weissberg, and important prerequisite for effective prevention (WHO Pachan 2010). and Liverpool John Moores University 2006). As School-based programs can address gender norms WHO Member States, all LMICs are committed to and attitudes with the aim of preventing dating vio- implementing the global plan of action on alcohol lence. The Safe Dates program in the United States and health (WHO 2010). That plan includes the (Foshee and others 2005) and the Youth Relationship interventions described. Interpersonal Violence: Global Impact and Paths to Prevention 83 84 Injury Prevention and Environmental Health Table 5.5 Overview of Violence Prevention Strategies Showing Evidence for Effectiveness and Applicability in Low- and Middle-Income Countries Type of violence Child Intimate partner Sexual Youth Elder Intervention maltreatment violence violence violence abuse LMIC applicability Developing safe, stable, and nurturing relationships between children and their parents and caregivers Parent training, including home visitation by nurses • ° These programs are likely to be highly applicable in LMICs. Few such programs in LMICs have been evaluated for violence prevention outcomes; several such studies are underway. Parent-child programs ° ° Programs are resource-intensive and need to be adapted to the requirements of and assets available in LMICs. Developing life skills in children and adolescents Preschool enrichment programs ° Most programs are delivered in schools and depend on the school system to deliver the program and reinforce its effects. Social development programs School-based programs to address gender • Oversight and management structures must be in place before norms and attitudes • ° such programs are implemented. Reducing the availability and harmful use of alcohol Regulation of alcohol sales ° Alcohol is an established risk factor for all types of violence. Some LMICs have seen rapid increases in alcohol consumption. Increases in alcohol prices ° Strategies to address alcohol may eventually be relevant to Interventions for problem drinkers Well-managed and well-designed drinking • ° LMICs where alcohol is currently not available. environments Reducing access to lethal means Restrictive firearm licensing and purchase ° There is limited evidence of the effectiveness of programs and policies policies in reducing access to lethal means of perpetrating violence in LMICs. Emerging evidence from LMICs that have Enforced bans on carrying firearms in public ° changed policies will shed light on the effectiveness of strategies for firearm injury prevention. Promoting gender equality to prevent violence against women Microfinance combined with gender equity ° Strong evidence for the effectiveness of such programs is training limited to outcome evaluation studies in low-resource, rural communities in South Africa. Several outcome evaluation Life skills interventions ° studies of similar programs are underway in other LMICs. table continues next page Table 5.5 Overview of Violence Prevention Strategies Showing Evidence for Effectiveness and Applicability in Low- and Middle-Income Countries (continued) Type of violence Child Intimate partner Sexual Youth Elder Intervention maltreatment violence violence violence abuse LMIC applicability Changing cultural and social norms that support violence Social marketing to modify social norms ° ° Programs that aim to change social norms supportive of violence through standalone mass-media campaigns are popular in LMICs. However, there is no evidence that such standalone programs are effective. They should be delivered in combination with other programs that address risk and protective factors more directly. Instituting victim identification, care, and support programs Screening and referral ° Stark differences exist in access to services (for example, Advocacy support programs • between high- and low-income groups, and between urban and rural settings). The shortage of highly trained, well-supervised Psychosocial interventions ° staff has been a barrier to the implementation of services in Protection orders ° LMICs. However, in the psychosocial arena, new approaches using community health workers are promising. Source: WHO 2009. • Note: LMICs = low- and middle-income countries; = well supported by evidence (multiple randomized controlled trials with different populations); ° = emerging evidence. Interpersonal Violence: Global Impact and Paths to Prevention 85 Reducing Access to Lethal Means • Reducing the lethality of guns (for example, designing Evidence from North America is the primary basis of firearms to make them safer and addressing magazine two systematic reviews and one meta-analysis that size). summarize the effects of various strategies to prevent • Evaluating strategies to reduce the use of military firearm-related violence. One systematic review (Hahn firearms in the aftermath of war or conflict, includ- and others 2005) concluded that the evidence is insuf- ing strategies to disarm former combatants, disband ficient to determine whether firearm laws have any armed groups, and reintegrate former combatants effect on violence. Such laws include bans on specified into civilian society. firearms or ammunition, restrictions on the acquisi- tion of firearms, waiting periods for acquisition, fire- Sound data and evidence on firearm injuries are arms registration, licensing of owners, “shall issue” needed to determine what programs and policies actu- carry laws that allow people who pass background ally work in preventing these injuries while preserving checks to carry concealed weapons, child access pre- the rights of legitimate gun owners. vention laws, and zero tolerance laws for firearms in schools. Another systematic review (Koper and Mayo- Promoting Gender Equality to Prevent Violence Wilson 2009) found that directed police patrols focus- against Women ing on illegal gun carrying can prevent gun crimes Several outcome evaluation studies demonstrate the (including murders, shootings, robberies in which effectiveness of multisector interventions to prevent vio- guns are used, and armed assaults). One meta-analysis lence against women by promoting gender equality. The (Makarios and Pratt 2012) suggests that bans on the Intervention with Microfinance and Gender Equity in sale of firearms had small effects, and law enforcement South Africa, which combines microloans and gender strategies had moderate effects in reducing gun equity training, reduced rates of self-reported violence violence. by more than 50 percent (Pronyk and others 2006). The More recent evidence suggests that the use of street Stepping Stones program implemented in Asia and Sub- outreach workers to mediate conflicts and provide Saharan Africa is a life skills training program that social support, such as job referrals and access to social addresses gender-based violence, relationship skills, services in the U.S. context, may be effective in reduc- assertiveness training, and communication about HIV/ ing youth homicides and firearm offenses (Webster AIDS and has shown promising results (Jewkes and oth- and others 2012). In addition, studies from Brazil and ers 2008; Paine and others 2002). The popularity of South Africa have found that stricter licensing and microfinance and conditional cash transfer programs in reduced circulation of firearms accounted for signifi- LMICs, into which violence prevention objectives could cant decreases in firearm-related injuries (Marinho de be integrated, further underscores their applicability. Souza and others 2007; Matzopoulos, Thompson, and Myers 2014). These reports therefore suggest, from a Changing Cultural and Societal Norms That limited evidence base, that some strategies addressing Support Violence access to firearms show promise, but additional Interventions that challenge cultural and social norms research is needed. supporting violence are widely used, and their rela- Public health can make a critical contribution to pre- tively low cost makes them a popular option. Such venting firearm injuries and deaths by collecting data interventions are often restricted to standalone mass- and evidence. A range of strategies exists for reducing media campaigns that are intended to raise awareness firearm-related violence, but further research and evi- about the harmful effects of violence, thereby reducing dence are needed to assess their effectiveness (IOM and the likelihood of future acts of violence. No evidence NRC 2013). Strategies identified by the Institute of shows that such campaigns are effective; however, Medicine (IOM) and others as being in particular need some evidence suggests that programs combining of additional research include the following: awareness-raising efforts with other mechanisms to change norms (for example, social development and • Increasing efforts to control access to firearms by life skills training and legislation) are effective. In individuals at risk of harming themselves or others South Africa, the Soul City initiative used television (for example, the safe storage of guns, waiting peri- (through a soap opera series), radio, and nationally ods, and background checks) distributed information booklets to raise awareness of • Changing how firearms are used (for example, where new IPV laws. The intervention increased the propor- firearms may be carried and provision of safety tion of people who saw such violence as unacceptable education) (Usdin and others 2005). 86 Injury Prevention and Environmental Health Implementing Victim Identification, Care, and effective, low-cost interventions to reduce violence is an Support Programs urgent priority. However, the same challenges that com- Interventions to identify victims of interpersonal plicate measuring the costs of violence also complicate violence and to provide effective care and support are measuring the benefits associated with its prevention critical for protecting health and breaking cycles of vio- (Barnett 1993; WHO 2004). Despite widely varying lence from one generation to the next. Evidence of effec- methodologies, most studies show that behavioral, legal, tiveness is emerging in several areas: and regulatory interventions are cost-effective (WHO 2004). Evidence addressing specific types of violence, • Screening tools to identify victims of IPV and refer largely from HICs, has identified a variety of cost-effective them to appropriate services (Ramsay and others interventions to prevent child maltreatment, IPV, and 2002) youth violence (table 5.6). Despite the disproportionate • Psychosocial interventions, such as trauma-focused effects of violence in LMICs, economic evaluations of cognitive behavioral therapy, to reduce mental health interventions are rare; therefore, systematic research to problems associated with violence (Bass and others measure the economic benefits of violence prevention 2013; Kornør and others 2008) efforts in LMICs would fill a critical gap. • Protection orders, which prohibit perpetrators of IPV from contacting victims (Holt and others 2003), to reduce repeat victimization. IMPLEMENTATION OF PREVENTION STRATEGIES Several trials have shown that advocacy support The gap between the science and the practice of violence programs—which offer services such as counseling, prevention is growing. Although numerous effective safety planning, and referral—increase victims’ safety programs, policies, and innovations have been identified, behaviors and reduce the risk of further harm (McFarlane they are unlikely to have a substantial public health and others 2006). impact unless they are widely disseminated, implemented with quality and scale, and sustained over the long term Policies and Programs to Address the Social (Rhoades, Bumbarger, and Moore 2012). Moreover, the Determinants of Violence benefits of their implementation must also be monitored. The infrastructure needed to support the dissemination, Violence is strongly associated with social determinants, scaling up, and sustenance of effective programs and such as employment, income equity, rapid social change, policies is slowly emerging. and access to education. The expectation that policies Given the rapid expansion of and increasing demand and programs can prevent violence by addressing social for evidence-based violence prevention innovations, determinants derives from ecological studies that use especially in LMICs, building an infrastructure that can cross-sectional and time-series methods to document more effectively move innovations from research to associations between social determinants and violence. action is increasingly important. That infrastructure Comprehensive violence prevention strategies should do requires attention to three interrelated sets of functions more than just address the risk factors targeted by the and activities that should be coordinated across global, specific programs; such strategies should be integrated country, and local levels: prevention synthesis and trans- with policies directed at the inequities that fuel violence. lation, prevention support, and prevention delivery This integration is particularly important in LMICs, (Wandersman and others 2008). where daily living conditions can undermine the opportunities for positive early child development. For example, the context can include economic and social Prevention Synthesis and Translation policies that exacerbate gaps between rich and poor and between men and women. The greater the extent to which innovations for vio- lence prevention are accessible (both from informa- tional and financial perspectives), user-friendly, and clearly communicated, the more likely it is that effective ECONOMIC EVALUATION OF INTERVENTIONS approaches will be successfully disseminated and imple- Economic evaluation provides a way to compare gains mented (Clancy and Cronin 2005). The seven-part resulting from an intervention, which has its own costs series on Violence Prevention: The Evidence is an exam- and risks. Given the high prevalence of interpersonal ple of an effort to synthesize and translate the scientific violence and its direct and indirect costs, identifying evidence into easily understandable and accessible Interpersonal Violence: Global Impact and Paths to Prevention 87 88 Injury Prevention and Environmental Health Table 5.6 Summary of Economic Evaluations of Interventions to Prevent Interpersonal Violence, by Type of Violence Type of violence Intervention type Intervention details Cost-effectiveness Child maltreatment Home visiting Nurse-family partnerships provide home visiting for low-income mothers to improve Net benefit (in 2011 dollars) for each program prenatal health-related behaviors, provide more responsible and competent care of infants participant was US$13,181 (US$13,617 in 2013 and toddlers, and improve parents’ economic self-sufficiency (Lee and others 2012). dollars); benefit-to-cost ratio was US$2.37 for Location: United States every US$1.00 spent (US$2.46 in 2013 dollars). Parent-child interaction therapy Empirical treatment for conduct disorders is based on behavioral interventions to improve Net benefit (in 2011 dollars) for each program parent-child interaction (Lee and others 2012). participant was US$5,617 (US$5,820 in 2013 Location: United States dollars); benefit-to-cost ratio was US$4.62 for every $1.00 spent (US$4.79 in 2013 dollars). Educational and family support The Child-Parent Center Program is a program for economically disadvantaged children Range of benefit-to-cost ratio (in 2002 dollars) and parents that provides a stable early learning environment and educational and support was US$5.98 (US$7.76 in 2013 dollars) to services for parents (Temple and Reynolds 2007). US$10.15 (US$13.18 in 2013 dollars) for every Location: Chicago, Illinois US$1.00 spent. Intimate partner Microfinance with gender and The Intervention with Microfinance for AIDS and Gender Equity (IMAGE) provides a Cost per disability-adjusted life year averted violence (IPV) human immunodeficiency virus/ combination of microfinance with gender and HIV/AIDS training for women to improve for the initial scale-up was US$2,307 (in 2004 acquired immune deficiency health, income, behavioral skills, communication, and norms (Jan and others 2011). dollars; US$2,852 in 2013 dollars). syndrome (HIV/AIDS) training Location: Rural South Africa Education for primary care Training allows clinicians to increase their identification and referral of survivors of IPV Incremental cost-effectiveness ratio was providers (Norman and others 2010). £2,450 per quality-adjusted life year (in 2005 Location: United Kingdom pounds) ($4,228 in 2013 dollars). Shelters for victims Shelters provide a safe haven for women and child victims of IPV, including support and Net social benefit was US$3.4 million dollars safety planning (Chanley, Chanley, and Campbell 2001). (in 1999 dollars); minimum benefit-to-cost ratio Location: Arizona was US$4.60 for every US$1.00 spent (US$6.45 in 2013 dollars). Youth violence Anonymized information sharing The Cardiff model provides information to direct targeted prevention measures. Youth The cumulative social benefit-to-cost ratio was between police and hospital were among those most likely to benefit (Florence and others 2013). £82 (in 2003 pounds) for every £1.00 spent emergency department Location: Cardiff, United Kingdom (US$163 in 2013 dollars). Multicomponent, long-term Fast Track, a program for at-risk children in grades 1–10, includes tutoring, parent support, The intervention was not cost-effective at a school and family-based child social-skills training, and home visits (Foster, Jones, and the Conduct Problems threshold of US$50,000 willingness to pay (in program Prevention Research Group 2006). 2004 dollars; US$61,817 in 2013 dollars) for an Location: Durham, North Carolina; Nashville, Tennessee; Seattle, Washington; and rural act of interpersonal violence. central Pennsylvania table continues next page Table 5.6 Summary of Economic Evaluations of Interventions to Prevent Interpersonal Violence, by Type of Violence (continued) Type of violence Intervention type Intervention details Cost-effectiveness Mobilization of community Communities That Care (CTC) mobilizes stakeholders to collaborate on preventing Very cost beneficial in the United States; net stakeholders to implement adolescent substance use, delinquency, and interpersonal violence (Kuklinski and others present benefit of CTC was US$5,250 per youth evidence-based systems 2012). (in 2004 dollars; US$6,491 in 2013 dollars), Location: 24 communities in seven states with a benefit-to-cost ratio of US$5.30 for every US$1.00 spent (US$6.55 in 2013 dollars). Educational incentives Incentives included four years of cash and other incentives to induce disadvantaged high For every US$1.0 million (in 1993 dollars; school students to graduate (Greenwood and others 1996), $1.62 million in 2013 dollars), 258 serious Location: California crimes were prevented. Parent training and family The intervention included training for parents and therapy for families with young school- For every US$1.0 million (in 1993 dollars; therapy age children who have shown aggressive behavior (Greenwood and others 1996). US$1.62 million in 2013 dollars), 157 serious Location: California crimes were prevented. Supervision The intervention included monitoring and supervising high school–age youth who have For every US$1.0 million (in 1993 dollars; exhibited delinquent behavior (Greenwood and others 1996). US$1.62 million in 2013 dollars), 72 serious Location: California crimes were prevented. Home visiting and day care The intervention included home visits by child care professionals from birth through the For every US$1.0 million (in 1993 dollars; first two years of childhood, followed by four years of day care (Greenwood and others 1996). US$1.62 million in 2013 dollars), 11 serious Location: California crimes were prevented. Interpersonal Violence: Global Impact and Paths to Prevention 89 briefing documents that demonstrate the effectiveness CONCLUSIONS of interventions to prevent interpersonal and self- directed violence (Liverpool John Moores University The primary rationale for addressing interpersonal 2013; WHO 2009). violence as a public health problem has been its role in causing physical injury and homicide. Evidence has shown that interpersonal violence also plays an impor- tant role in the etiology of mental illness, chronic dis- Prevention Support ease, and even infectious diseases such as HIV/AIDS. Synthesizing and translating information about vio- Unfortunately, such wide-ranging effects remain lence prevention innovations, although important, largely invisible to public health leaders, policy makers, are likely to be insufficient to change prevention and the public. Violence is often hidden, victims rarely practices. Countries, districts, and communities seek- come into contact with official or service agencies, and ing to apply violence prevention innovations need many of the health and social consequences are not the capacity to be successful in scaling up effective evident until years after exposure. Greater awareness programs with fidelity (Wandersman and others of these impacts is now leading to actions that can 2008). A growing body of research suggests that pro- reduce the enormous health and social burden of viding support in the form of specialized training, violence. monitoring of fidelity, technical assistance, and Many LMICs face daunting challenges, including the coaching, along with improving the skills and moti- HIV/AIDS epidemic, ongoing wars and conflicts, cardio- vation of implementing organizations, increases the vascular and other chronic diseases, suicide, and traffic use and successful implementation of innovations injuries. Given the effects of violence on these outcomes, (Fixsen and other 2005; Mihalic and Irwin 2003; preventing interpersonal violence can become a power- Wandersman and others 2008). ful lever that, if successfully engaged, will allow LMICs to An example of a well-functioning prevention support more effectively address a broad range of challenges. system in the United States is the state of Pennsylvania’s The study of violence crosses many domains, and Evidence-Based Prevention and Intervention Support collaboration across different government sectors and Center (EPISCenter) at the University of Pennsylvania across different disciplines and professions is critical, (Rhoades, Bumbarger, and Moore 2012). The EPISCenter both to fully understand the problem and to effectively uses flexible, targeted, and research-based technical prevent it. Violence affects almost every government assistance to develop the capacity of communities to sector, including justice and law enforcement, social support the implementation of evidence-based violence services, protection of women and children, education, prevention programs. transportation, finance, health care and public health, labor, tourism, foreign affairs, interior affairs, commerce, and tourism. The disciplines that have Prevention Delivery important contributions to make include law, psychol- The successful implementation of evidence-based ogy, sociology, social work, medicine and almost every innovations requires that they be carried out and sus- medical specialty, anthropology, engineering, business, tained in organizational settings (Wandersman and architecture and design, and urban planning. Given this others 2008). Organizations’ capacities to deliver the influence, the involvement of foundations, multilateral violence prevention innovations include maintain- agencies, and corporations in programs to prevent vio- ing a well-functioning organization; recruiting and lence is also expanding. maintaining well-trained staff members; developing Progress in preventing interpersonal violence is community support; working with other organiza- advancing rapidly, and clearly the global public health tions; and improving skills in selecting, implementing, community’s increased understanding and capacity to and sustaining an innovation over time (Mihalic and prevent interpersonal violence will make a difference. Irwin 2003; Wandersman and others 2008). The Parent The lessons learned during their brief experience with Centre in South Africa and Raising Voices in Uganda violence prevention efforts are consistent with the les- are two examples of organizations in LMICs that are sons from the community’s much longer experience seeking to sustain the implementation of interventions with the prevention of infectious and chronic diseases. to prevent violence against women and children Violence can be prevented if citizens, their governments, (Butchart and Hendricks 2000; The Parent Centre and the global community start now, act wisely, and 2013; Raising Voices 2013). work together. 90 Injury Prevention and Environmental Health NOTE Borwankar, R., R. Diallo, and A. E. Sommerfelt. 2008. Gender-Based Violence in Sub-Saharan Africa: A Review of Disclaimer: The findings and conclusions in this chapter are Demographic and Health Survey Findings and Their Use in those of the authors and do not necessarily represent the offi- National Planning. Washington, DC: United States Agency cial position of the Centers for Disease Control and Prevention for International Development. or the World Health Organization (WHO). Bott, S., A. Guedes, M. Goodwin, and J. A. Mendoza. 2012. World Bank Income Classifications as of July 2014 are as Violence against Women in Latin America and the Caribbean: follows, based on estimates of gross national income per capita A Comparative Analysis of Population-Based Data from for 2013: 12 Countries. Washington, DC: Pan American Health Organization. • Low-income countries (LICs) = US$1,045 or less Buchanan, C., ed. 2013. Gun Violence, Disability and Recovery. • Middle-income countries (MICs) are subdivided: Sydney: Surviving Gun Violence Project. a) Lower-middle-income = US$1,046 to US$4,125 Butchart, A., D. Brown, A. Khanh-Huynh, P. Corso, N. Florquin, b) Upper-middle-income (UMICs) = US$4,126 to and others. 2008. Manual for Estimating the Economic Costs US$12,745 of Injuries Due to Interpersonal and Self-Directed Violence. • High-income countries (HICs) = US$12,746 or more. Geneva: World Health Organization. Butchart, A., and K. Engstrom. 2002. “Sex- and Age-Specific Effects of Economic Development and Inequality on REFERENCES Homicide Rates in 0 to 24 Year Olds: A Cross-Sectional Anda, R. F., A. Butchart, V. J. Felitti, and D. W. Brown. Analysis.” Bulletin of the World Health Organization 2010. “Building a Framework for Global Surveillance of 80 (October): 797–805. the Public Health Implications of Adverse Childhood Butchart, A., and G. Hendricks. 2000. “The Parent Experiences.” American Journal of Preventive Medicine Centre.” In Behind the Mask: Getting to Grips with Crime and 39 (1): 93–98. Violence in South Africa, edited by T. Emmett and A. Butchart, Andersson, N., A. Cockcroft, and B. Shea. 2008. “Gender- 147–76. Pretoria: Human Sciences Research Council. Based Violence and HIV: Relevance for HIV Prevention Buvinic, M., A. R. Morrison, and M. Shifter. 1999. “Violence in Hyperendemic Countries of Southern Africa.” AIDS in the Americas: A Framework for Action.” In Too Close 22 (S4): S73–86. to Home: Domestic Violence in the Americas, edited by Baker-Henningham, H., S. Scott, K. Jones, and S. Walker. 2012. A. R. Morrison and M. L. Biehl, 3–34. New York: Inter- “Reducing Child Conduct Problems and Promoting Social American Development Bank. Skills in a Middle-Income Country: Cluster Randomised Caldera, D., L. Burrell, K. Rodriguez, S. S. Crowne, C. Rohde, Controlled Trial.” British Journal of Psychiatry 201 (2): 101–18. and others. 2007. “Impact of a Statewide Home Visiting Barnett, W. S. 1993. “Economic Evaluation of Home Visiting Program on Parenting and on Child Health and Programs.” The Future of Children 3 (3): 93–112. Development.” Child Abuse and Neglect 31 (8): 829–52. Bass, J. K., J. Annan, S. McIvor Murray, D. Kaysen, S. Griffiths, Capaldi, D. M., N. B. Knoble, J. W. Shortt, and H. K. Kim. 2012. and others. 2013. “Controlled Trial of Psychotherapy for “A Systematic Review of Risk Factors for Intimate Partner Congolese Survivors of Sexual Violence.” New England Violence.” Partner Abuse 3 (2): 231–80. Journal of Medicine 368 (23): 2182–91. CDC (Centers for Disease Control and Prevention). 2003. Costs Beaman, V., J. L. Annest, J. A. Mercy, M. Kresnow, and of Intimate Partner Violence against Women in the United D. A. Pollock. 2000. “Lethality of Firearm-Related Injuries States. Atlanta: National Center for Injury Prevention and in the United States Population.” Annals of Emergency Control. Medicine 35 (3): 258–66. CDC, INURED (Interuniversity Institute for Research and Bellis, M. A., and K. Hughes. 2008. “Comprehensive Strategies Development), and the Comité de Coordination. 2014. to Prevent Alcohol-Related Violence.” In Institute for Violence against Children in Haiti: Findings from a National the Prevention of Crime Review Volume 2: Towards More Survey, 2012. Port-au-Prince, Haiti: Centers for Disease Comprehensive Approaches to Prevention and Safety, edited Control and Prevention. by R. Hastings and M. Bania, 137–68. Ottawa: ON: Institute Chanley, S. A., J. J. Chanley, and H. E. Campbell. 2001. for Prevention of Crime. “Providing Refuge: The Value of Domestic Violence Shelter Bellis, M. A., K. Hughes, C. Perkins, and A. Bennett. 2012. Services.” American Review of Public Administration 31 (4): Protecting People, Promoting Health: A Public Health 393–413. Approach to Violence Prevention for England. Liverpool, Clancy, C. M., and K. Cronin. 2005. “Evidence-Based Decision U.K.: Department of Health and National Health Service. Making: Global Evidence, Local Decisions.” Health Affairs Bilukha, O., R. A. Hahn, A. Crosby, M. T. Fullilove, A. Liberman, 24 (1): 151–62. and others. 2005. “The Effectiveness of Early Childhood Home Cohen, A. B. 2007. “Sobering Up: The Impact of the 1985–1988 Visitation in Preventing Violence: A Systematic Review.” Russian Anti-Alcohol Campaign on Child Health.” American Journal of Preventive Medicine 28 (2): S11–39. Unpublished paper, Tufts University, Boston, MA. Interpersonal Violence: Global Impact and Paths to Prevention 91 Cooper, C., A. Selwood, and G. Livingston. 2008. “The Florence, C., J. Shepherd, I. Brennan, and T. R. Simon. 2013. Prevalence of Elder Abuse and Neglect: A Systematic “An Economic Evaluation of Anonymised Information Review.” Age and Aging 37 (2): 151–60. Sharing in a Partnership between Health Services, Police Dahlberg, L. L., and E. G. Krug. 2002. “Violence: A Global and Local Government for Preventing Violence-Related Public Health Problem.” In World Report on Violence and Injury.” Injury Prevention 20 (2): 108–14. doi:10.1136 Health, edited by E. G. Krug, L. L. Dahlberg, J. A. Mercy, /injuryprev-2012-040622. A. B. Zwi, and R. Lozano, 1–21. Geneva: World Health Foshee, V. A., K. E. Bauman, S. T. Ennett, C. Suchindran, Organization. T. Benefield, and others. 2005. “Assessing the Effects of the Danese, A., and B. McEwen. 2012. “Adverse Childhood Dating Violence Prevention Program ‘Safe Dates’ Using Experiences, Allostasis, Allostatic Load, and Age-Related Random Coefficient Regression Modeling.” Prevention Disease.” Physiology and Behavior 106 (1): 29–39. Science 6 (3): 245–57. Day, T. 1995. “The Health Related Costs of Violence against Foster, E. M., D. Jones, and the Conduct Problems Prevention Women: The Tip of the Iceberg.” Centre for Research on Research Group. 2006. “Can a Costly Intervention Be Cost Violence against Women and Children Publication Series, Effective? An Analysis of Violence Prevention.” Archives of University of Western Ontario, London, Ontario. General Psychiatry 63 (11): 1284–91. Devries, K., C. Watts, M. Yoshihama, L. Kiss, L. B. Schraiber, Giancola, P. 2000. “Executive Functioning: A Conceptual and others. 2011. “Violence against Women Is Strongly Framework for Alcohol-Related Aggression.” Experimental Associated with Suicide Attempts: Evidence from the WHO and Clinical Psychopharmacology 8 (4): 576–97. Multi-Country Study on Women’s Health and Domestic Gilbert, R., C. S. Widom, K. Browne, D. Fergusson, E. Webb, Violence against Women.” Social Science and Medicine and S. Janson. 2009. “Burden and Consequences of Child 73 (1): 79–86. Maltreatment in High-Income Countries.” The Lancet 373 Dinh-Zarr, T. B., C. Goss, E. Heitman, E. Roberts, and (9657): 68–81. C. DiGuiseppi. 2004. “Interventions for Preventing Injuries Gore, F. M., P. J. N. Bloem, G. C. Patton, J. Ferguson, V. Joseph, in Problem Drinkers.” Cochrane Database of Systematic and others. 2011. “Global Burden of Disease in Young Reviews 3 (3): CD001857. People Aged 10–24 Years: A Systematic Analysis.” The Lancet Duailibi, S., W. Ponicki, J. Grube, I. Pinsky, R. Laranjeira, and 377 (9783): 2093–102. M. Raw. 2007. “The Effect of Restricting Opening Hours Graham, K., and R. Homel. 2008. Raising the Bar: Preventing on Alcohol-Related Violence.” American Journal of Public Aggression in and around Bars, Pubs and Clubs. Devon, U.K.: Health 97 (12): 2276–80. Willan Publishing. Durlak, J. A., R. P. Weissberg, and M. Pachan. 2010. Greenwood, P. W., K. E. Model, C. P. Rydell, and J. Chiesa. 1996. “A Meta-Analysis of After-School Programs That Seek Diverting Children from a Life of Crime: Measuring Costs to Promote Personal and Social Skills in Children and and Benefits. Santa Monica, CA: Rand. Adolescents.” American Journal of Community Psychology Habetha, S., S. Bleich, J. Weidenhammer, and J. M. Fegert. 45 (3–4): 294–309. 2012. “A Prevalence-Based Approach to Societal Costs Eisenberg, N., Q. Zhou, T. L. Spinrad, C. Valiente, R. A. Fabes, Occurring in Consequence of Child Abuse and Neglect.” and others. 2005. “Relations among Positive Parenting, Child and Adolescent Psychiatry and Mental Health 6: 35. Children’s Effortful Control, and Externalizing Problems: doi:10.1186/1753-2000-6-35. A Three-Wave Longitudinal Study.” Child Development Hahn, R. A., O. Bilukha, A. Crosby, M. T. Fullilove, A. Liberman, 76 (5): 1055–71. and others. 2005. “Firearms Laws and the Reduction Fang, X., D. S. Brown, C. S. Florence, and J. A. Mercy. 2012. of Violence: A Systematic Review.” American Journal of “The Economic Burden of Child Maltreatment in the Preventive Medicine 28 (2S1): 40–71. United States and Implications for Prevention.” Child Abuse Hahn, R. A., D. Fuqua-Whitley, H. Wethington, J. Lowy, and Neglect 36 (2): 156–65. A. Crosby, and others. 2007. “Effectiveness of Universal Fanslow, J., C. Coggan, B. Miller, and R. Norton. 1997. “The School-Based Programs to Prevent Violent and Aggressive Economic Cost of Homicide in New Zealand.” Social Behavior: A Systematic Review.” American Journal of Science and Medicine 45 (7): 973–97. Preventive Medicine 33 (2S): S114–29. Fisher, J., M. Cabral de Mello, V. Patel, A. Rahman, T. Tran, Hawkins, J. D., R. F. Catalano, R. Kosterman, R. Abbott, and others. 2012. “Prevalence and Determinants of and K. G. Hill. 1999. “Preventing Adolescent Health-Risk Common Perinatal Mental Disorders in Women in Low- Behaviors by Strengthening Protection during Childhood.” and Lower-Middle-Income Countries: A Systematic Archive of Pediatrics and Adolescent Medicine 153 (3): Review.” Bulletin of the World Health Organization 226–34. 90 (2): 139–49. Heim, C., and E. B. Binder. 2012. “Current Research Trends in Fixsen, D. L., S. F. Naoom, K. A. Blase, R. M. Friedman, and Early Life Stress and Depression: Review of Human Studies F. Wallace. 2005. Implementation Research: A Synthesis of the on Sensitive Periods, Gene-Environment Interactions, and Literature. Tampa, FL: University of South Florida, Louis Epigenetics.” Experimental Neurology 233 (1): 102–11. de la Parte Florida Mental Health Institute, The National Hillis, S. D., R. F. Anda, S. R. Dube, V. J. Felitti, P. A. Marchbanks, Implementation Research Network. and J. S. Marks. 2004. “The Association between Adverse 92 Injury Prevention and Environmental Health Childhood Experiences and Adolescent Pregnancy, Long- Klevens, J., J. W. Martinez, B. Le, C. Rojas, A. Duque, and Term Psychosocial Outcomes, and Fetal Death.” Pediatrics others. 2009. “Evaluation of Two Interventions to Reduce 113 (2): 320–27. Aggressive and Antisocial Behavior in First and Second Hillis, S. D., R. F. Anda, S. R. Dube, V. J. Felitti, P. A. Marchbanks, Graders in a Resource-Poor Setting.” International Journal and others. 2010. “The Protective Effect of Family Strengths of Educational Research 48 (5): 307–19. in Childhood against Adolescent Pregnancy and Its Long- Klosterman, K. C., and W. Fals-Stewart. 2006. “Intimate Partner Term Psychosocial Consequences.” Permanente Journal 14 Violence and Alcohol Use: Exploring the Role of Drinking (3): 18–27. in Partner Violence and Its Implications for Intervention.” Holt, V. L., M. A. Kernic, M. E. Wolf, and F. P. Rivara. 2003. “Do Aggression and Violent Behavior 11 (6): 587–97. Protection Orders Affect the Likelihood of Future Partner Knerr, W., F. Gardner, and L. Cluver. 2013. “Improving Positive Violence and Injury?” American Journal of Preventive Parenting Skills and Reducing Harsh and Abusive Parenting Medicine 24 (1): 16–21. in Low- and Middle-Income Countries: A Systematic Hughes, K., M. A. Bellis, L. Jones, S. Wood, G. Bates, and others. Review.” Prevention Science 14 (4): 352–63. 2012. “Prevalence and Risk of Violence against Adults with Koper, C. S., and E. Mayo-Wilson. 2009. “Police Strategies Disabilities: A Systematic Review and Meta-Analysis of to Reduce Illegal Possession and Carrying of Firearms: Observational Studies.” The Lancet 80 (9845): 899–907. Effects on Gun Crime.” Campbell Systematic Reviews 8 (11). doi:10.1016/S0410-6736(11)61851-5. doi:10.4073/csr.2012.11. ICRW (International Center for Research on Women). Kornør, H., D. Winje, Ø. Ekeberg, L. Weisæth, I. Kirkehei, 2009. “Intimate Partner Violence: High Costs to and others. 2008. “Early Trauma-Focused Cognitive- Households and Communities.” Policy Brief, ICRW, Behavioural Therapy to Prevent Chronic Post-Traumatic Washington, DC. Stress Disorder and Related Symptoms: A Systematic IOM (Institute of Medicine). 2013. Evaluation of PEPFAR. Review and Meta-Analysis.” BMC Psychiatry 8: 81. Washington, DC: National Academies Press. Krug, E. G., L. I. Dahlberg, J. A. Mercy, A. B. Zwi, and R. Lozano, IOM and NRC (National Research Council). 2012. Social and eds. 2002. World Report on Violence and Health. Geneva: Economic Costs of Violence. Washington, DC: National World Health Organization. Academies Press. Kuklinski, M. R., J. S. Briney, J. D. Hawkins, and R. F. Catalano. ––––. 2013. Priorities for Research to Reduce the Threat of 2012. “Cost-Benefit Analysis of Communities That Care Firearm-Related Violence. Washington, DC: National Outcomes at Eighth Grade.” Prevention Science 13 (2): Academies Press. 150–61. Ireland, T., and C. Smith. 2009. “Living in Partner-Violent Lachs, M. S., and K. Pillemer. 2004. “Elder Abuse.” The Lancet Families: Developmental Links to Antisocial Behavior and 364 (9441): 1263–72. Relationship Violence.” Journal of Youth and Adolescence 38 Lee, S., S. Aos, E. Drake, A. Pennucci, M. Miller, and L. Anderson. (3): 323–39. 2012. “Return on Investment: Evidence-Based Options Jan, J., G. Ferrari, C. H. Watts, J. R. Hargreaves, J. C. Kim, to Improve Statewide Outcomes, April 2012.” Document and others. 2011. “Economic Evaluation of a Combined No. 12-04-1201, Washington State Institute for Public Microfinance and Gender Training Intervention for the Policy, Olympia, WA. Prevention of Intimate Partner Violence in Rural South Liverpool Johns Moores University. 2013. “Violence Prevention Africa.” Health Policy and Planning 26 (5): 366–72. Evidence Base.” Database, Centre for Public Health, Jewkes, R. K. 2002. “Intimate Partner Violence: Causes and Liverpool John Moores University, Liverpool, U.K. http:// Prevention.” The Lancet 359 (9315): 1423–29. www.preventviolence.info/EvidenceBase. Jewkes, R. K., M. Nduna, J. Levin, N. Jama, K. Dunkle, and Lozano, R., M. Naghavi, K. Foreman, S. Lim, K. Shibuya, and others. 2008. “Impact of Stepping Stones on Incidence others. 2012. “Global and Regional Mortality from 235 of HIV and HSV-2 and Sexual Behaviour in Rural Causes of Death for 20 Age Groups in 1990 and 2010: South Africa: Cluster Randomised Controlled Trial.” British A Systematic Analysis for the Global Burden of Disease Medical Journal 337: a506. doi:10.1136/bmj.a506. Study 2010.” The Lancet 380 (9859): 2095–128. Kaminski, J. W., L. A. Valle, J. H. Filene, and C. L. Boyle. Machtinger, E. L., J. E. Haberer, T. C. Wilson, and D. S. Weiss. 2008. “A Meta-Analytic Review of Components Associated 2012. “Recent Trauma Is Associated with Antiretroviral with Parent Training Program Effectiveness.” Journal of Failure and HIV Transmission Risk Behavior Abnormal Child Psychology 36 (4): 567–89. among HIV-Positive Women and Female-Identified Kessler, R. C., K. A. McLaughlin, J. G. Green, M. J. Gruber, Transgenders.” AIDS and Behavior 16 (8): 2160–70. N. A. Sampson, and others. 2010. “Childhood Adversities Epub 2012/03/20. and Adult Psychopathology in the WHO World Mental Machtinger, E. L., T. C. Wilson, J. E. Haberer, and D. S. Weiss. Health Surveys.” British Journal of Psychiatry 197 (5): 2012. “Psychological Trauma and PTSD in HIV-Positive 378–85. Women: A Meta-Analysis.” AIDS and Behavior 16 (8): Kissin, D. M., L. Zapata, R. Yorick, E. N. Vinogradova, 2091–100. G. V. Volkova, and others. 2007. “HIV Seroprevalence in Makarios, M. D., and T. C. Pratt. 2012. “The Effectiveness of Street Youth, St. Petersburg, Russia.” AIDS 21 (17): 2333–40. Policies and Programs That Attempt to Reduce Firearm Interpersonal Violence: Global Impact and Paths to Prevention 93 Violence: A Meta-Analysis.” Crime and Delinquency 58 (2): Nelson, G., A. Westhues, and J. MacLeod. 2003. “A Meta-Analysis 222–44. of Longitudinal Research on Preschool Prevention Marinho de Souza, M. de F., J. Macinko, A. P. Alencar, Programs for Children.” Prevention and Treatment 6 (1): D. C. Malta, and O. L. de Morais Neo. 2007. “Reductions in 31A. doi:10.1037/1522-3736.6.1.631a. Firearm-Related Mortality and Hospitalizations in Brazil Nemtsov, A. V. 1998. “Alcohol-Related Harm and Alcohol after Gun Control.” Health Affairs 26 (2): 575–84. Consumption in Moscow before, during and after a Major Markowitz, S., and M. Grossman. 1998. “Alcohol Regulation Anti-Alcohol Campaign.” Addiction 93 (10): 1501–10. and Domestic Violence towards Children.” Contemporary Norman, R., A. Spencer, S. Eldridge, and G. Feder. 2010. Economic Policy 16 (3): 309–20. “Cost Effectiveness of a Programme to Detect and Provide ———. 2000. “The Effects of Beer Taxes on Physical Child Better Care for Female Victims of Intimate Partner Abuse.” Journal of Health Economics 19 (2): 271–82. Violence.” Journal of Health Services and Research Policy Mattson, S. N., S. C. Roesch, L. Glass, B. N. Deweese, 15 (3): 143–49. C. D. Coles, and others. 2013. “Further Development Norman, R. E., M. Byambaa, R. De, A. Butchart, J. Scott, and of a Neurobehavioral Profile of Fetal Alcohol Spectrum others. 2012. “The Long-Term Health Consequences of Disorders.” Alcoholism: Clinical and Experimental Research Child Physical Abuse, Emotional Abuse, and Neglect: A 37 (3): 517–28. Systematic Review and Meta-Analysis.” PLoS Med 9 (11): Matzopoulos, R. G., M. L. Thompson, and J. E. Myers. 2014. e1001349. doi:10.1371/journal.pmed.1001349. “Firearm and Non-Firearm Homicide in Five South African Norton, R., and O. Kobusingye. 2013. “Injuries.” New England Cities: A Retrospective Population-Based Study.” American Journal of Medicine 368 (18): 1723–30. Journal of Public Health 104 (3): 455–60. Olds, D., C. R. Henderson, R. Cole, J. Eckenrode, H. Kitzman, Mayhew, P. 2003. “Counting the Costs of Crime in Australia.” and others. 1998. “Long-Term Effects of Nurse Home Trends and Issues in Crime and Criminal Justice, no. 247. Visitation on Children’s Criminal and Antisocial Behavior: Australian Institute of Criminology, Canberra. 15-year Follow-Up of a Randomized Controlled Trial.” McFarlane, J. M., J. Y. Groff, J. A. O’Brien, and K. Watson. Journal of the American Medical Association 280 (14): 2006. “Secondary Prevention of Intimate Partner Violence: 1238–44. A Randomized Controlled Trial.” Nursing Research 55 (1): Paine, K., G. Hart, M. Jawo, S. Ceesay, M. Jallow, and others. 52–61. 2002. “Before We Were Sleeping, Now We Are Awake: Mendonca, R. N., J. G. Alves, and J. E. Cabral Filho. 2002. Preliminary Evaluation of the Stepping Stones Sexual “Hospital Costs Due to Violence against Children and Health Programme in The Gambia.” African Journal of Adolescents in Pernambuco State, Brazil, during 1999.” Cad AIDS Research 1 (1): 41–52. Saude Publica 18 (6): 1577–81. The Parent Centre. 2013. “The Parent Centre: Helping Children Mihalic, S., and K. Irwin. 2003. “Blueprints for Violence through Parents.” http://www.theparentcentre.org.za/. Prevention: From Research to Real World Settings: Factors Pinheiro, P. S. 2006. World Report on Violence against Children. Influencing the Successful Replication of Model Programs.” Geneva: United Nations. Youth Violence and Juvenile Justice 1: 307–29. Pronyk, P. M., J. R. Hargreaves, J. C. Kim, L. A. Morison, G. Phetla, Mikton, C., and A. Butchart. 2009. “Child Maltreatment and others. 2006. “Effect of a Structural Intervention for the Prevention: A Systematic Review of Reviews.” Bulletin of the Prevention of Intimate-Partner Violence and HIV in Rural World Health Organization 87 (5): 353–61. South Africa: A Cluster Randomised Trial.” The Lancet Miller, M., D. Azrael, and D. Hemenway. 2013. “Firearms and 368 (9551): 1973–83. Violent Death in the United States.” In Reducing Gun Violence Raising Voices. 2013. “Raising Voices: Preventing Violence in America, edited by D. W. Webster and J. S. Vernick, 1–20. against Women and Children.” http://raisingvoices.org Baltimore, MD: Johns Hopkins University Press. /about. Miller, T. R., M. A. Cohen, and B. Wiersema. 1996. Victim Costs Ramsay, J., J. Richardson, Y. H. Carter, L. L. Davidson, and and Consequences: A New Look. National Institute of Justice G. Feder. 2002. “Should Health Professionals Screen Women Research Report, National Institute of Justice, United States for Domestic Violence? Systematic Review.” British Medical Department of Justice, Washington, DC. Journal 325 (7359): 314. Miller, T. R., D. A. Fisher, and M. A. Cohen. 2001. “Costs Reed, R., M. Fazel, L. Jones, C. Panter-Brick, and A. Stein. of Juvenile Violence: Policy Implications.” Pediatrics 2012. “Mental Health of Displaced and Refugee Children 107 (1): e3. Resettled in Low-Income and Middle-Income Countries: Monteiro, M. 2007. Alcohol and Public Health in the Americas: Risk and Protective Factors.” The Lancet 379 (9812): 250–65. A Case for Action. Washington, DC: Pan American Health Reza, A., J. A. Mercy, and E. G. Krug. 2001. “The Epidemiology Organization. of Violent Deaths in the World.” Injury Prevention 7 (2): Morrison, A. R., and M. B. Orlando. 1999. “Social and Economic 104–11. Costs of Domestic Violence: Chile and Nicaragua.” In Too Reza, A., M. J. Breiding, G. Gulaid, J. A. Mercy, C. Blanton, and Close to Home: Domestic Violence in the Americas, edited by others. 2009. “Sexual Violence and Its Health Consequences A. R. Morrison and B. E. Orlando, 51–80. New York: Inter- for Female Children in Swaziland: A Cluster Survey Study.” American Development Bank. The Lancet 373 (9679): 1966–72. 94 Injury Prevention and Environmental Health Rhoades, B. L., B. K. Bumbarger, and J. E. Moore. 2012. “The Child-Parent Centers and Related Programs.” Economics of Role of a State-Level Prevention Support System in Education Review 26 (1): 126–44. Promoting High-Quality Implementation and Sustainability Tharp, A. T., S. Degue, L. A. Valle, K. A. Brookmeyer, of Evidence-Based Programs.” American Journal of G. M. Massetti, and others. 2012. “A Systematic Qualitative Community Psychology 50 (3): 386–401. Review of Risk and Protective Factors for Sexual Violence Robbins, C. L., L. Zapata, D. M. Kissin, N. Shevchenko, Perpetration.” Trauma, Violence, and Abuse 14 (2): 133–67. R. Yorick, and others. 2010. “Multicity HIV Seroprevalence Thornberry, T. P., K. E. Knight, and P. J. Lovegrove. 2012. “Does in Street Youth, Ukraine.” International Journal of STDs and Maltreatment Beget Maltreatment? A Systematic Review AIDS 21 (7): 489–96. of the Intergenerational Literature.” Trauma, Violence and Roldós, M. I., and P. Corso. 2013. “The Economic Burden Abuse 13 (3): 135–52. of Intimate Partner Violence in Ecuador: Setting the UN (United Nations). 2006. Secretary General’s In-Depth Study Agenda for Future Research and Violence Prevention on All Forms of Violence against Women. New York: United Policies.” Western Journal Emergency Medicine 14 (4): Nations. 347–53. ––––. 2015. Transforming Our World: The 2030 Agenda Rosenberg, M. L., A. Butchart, J. Mercy, V. Narasimhan, for Sustainable Development; Resolution Adopted by H. Waters, and others. 2006. “Interpersonal Violence.” In the General Assembly. 70/1. New York: United Nations. Disease Control Priorities in Developing Countries, 2nd http://www.un.org/ga/search/view_doc.asp?symbol=A ed., edited by D. T. Jamison, J. G. Breman, A. R. Measham, /RES/70/1&Lang=E. R. G. Alleyne, M. Claeson, and others, 755–70. Washington, UNICEF (United Nations Children’s Fund). 2010. Child DC: Oxford University Press and World Bank. Disciplinary Practices at Home: Evidence from a Range of Silverman, J. G., M. R. Decker, D. M. Cheng, K. Wirth, Low- and Middle-Income Countries. New York: UNICEF. N. Saggurti, and others. 2011. “Gender-Based Disparities UNICEF, CDC (Centers for Disease Control and Prevention), in Infant and Child Mortality Based on Maternal Exposure and KNBS (Kenya National Bureau of Statistics). 2012. to Spousal Violence: The Heavy Burden Borne by Indian Violence against Children in Kenya: Findings from a Girls.” Archives of Pediatric and Adolescent Medicine 165 (1): 2010 National Survey. Nairobi: UNICEF Kenya Country 22–27. Office. Silverman, J. G., R. Michele, M. R. Decker, L. Heather, UNICEF, CDC, and the Muhimbili University of Health and M. S. McCauley, and others. 2009. “Regional Assessment Allied Science. 2012. Violence against Children in Tanzania: of Sex Trafficking and STI/HIV in Southeast Asia: Findings from a National Survey 2009. Dar es Salaam, Connections between Sexual Exploitation, Violence, and Tanzania: UNICEF Tanzania. Sexual Risk.” United Nations Development Programme Usdin, S., E. Scheepers, S. Goldstein, and G. Japhet. 2005. Regional Center, Colombo. http://www.undp.org/content “Achieving Social Change on Gender-Based Violence: A /dam/undp/library/hivaids/English/SexTrafficking.pdf. Report on the Impact Evaluation of Soul City’s Fourth Smith, K. W., S. H. Landry, and P. R. Swank. 2010. “The Series.” Social Science and Medicine 61 (11): 2434–45. Influence of Early Patterns of Positive Parenting on van der Merwe, A., A. Dawes, and C. L. Ward. 2012. “The Children’s Preschool Outcomes.” Early Education and Development of Youth Violence: An Ecological Development 11 (2): 147–69. Understanding.” In Youth Violence in South Africa: Sources Stöckl, H., K. Devries, A. Rotsein, N. Abrahams, J. Campbell, and Solutions, edited by C. L. Ward, A. van der Merwe, and and others. 2013. “The Global Prevalence of Intimate A. Dawes. Cape Town, South Africa: UCT Press. Partner Homicide: A Systematic Review.” The Lancet Walker, S. P., S. M. Chang, M. Vera-Hernandez, and S. Grantham- 382 (9895): 859–65. McGregor. 2011. “Early Childhood Stimulation Benefits Stoltenborgh, M. A., M. H. van Ijzendoorn, E. M. Euser, Adult Competence and Reduces Violent Behavior.” M. J. Bakermans-Kranenburg. 2011. “Global Perspective on Pediatrics 127 (5): 849–57. Child Sexual Abuse: Meta-Analysis of Prevalence around Wandersman, A., J. Duffy, P. Flaspohler, R. Noonan, K. Lubell, the World.” Child Maltreatment 16 (2): 79–101. and others. 2008. “Bridging the Gap between Prevention Stoltenborgh, M. A., M. J. Bakermans-Kranenburg, M. H. van Research and Practice: The Interactive Systems Framework Ijzendoorn, and L. R. Alink. 2013a. “Cultural-Geographical for Dissemination and Implementation.” American Journal Differences in the Occurrence of Child Physical Abuse? A of Community Psychology 41 (3–4): 171–81. Meta-Analysis of Global Prevalence.” International Journal Ward, E., T. McCartney, D. W. Brown, A. Grant, A. Butchart, of Psychology 48 (2): 81–94. and others. 2009. “Results of an Exercise to Estimate the Stoltenborgh, M. A., M. J. Bakermans-Kranenburg, and Costs of Interpersonal Violence in Jamaica.” West Indian M. H. van Ijzendoorn. 2013b. “The Neglect of Child Medical Journal 58 (5): 446–51. Neglect: A Meta-Analytic Review of the Prevalence of Webster, D. W., J. M. Whitehill, J. S. Vernick, and F. C. Curriero. Neglect.” Social Psychiatry and Psychiatric Epidemiology 2012. “Effects of Baltimore’s Safe Streets Program on 48 (3): 345–55. Gun Violence: A Replication of Chicago’s CeaseFire Temple, J. A., and A. J. Reynolds. 2007. “Benefits and Costs of Program.” Bulletin of the New York Academy of Medicine Investments in Pre-School Education: Evidence from the 90 (1): 27–40. Interpersonal Violence: Global Impact and Paths to Prevention 95 WHO (World Health Organization). 2004. The Economic http://www.who.int/violence_injury_prevention/violence Dimensions of Interpersonal Violence, edited by H. Waters, /world_report/factsheets/pb_violencealcohol.pdf. A. Hyder, Y. Rajkotia, S. Basu, J. A. Rehwinkel, and Wolfe, D. A., C. Crooks, P. Jaffe, D. Chiodo, R. Hughes, and others. Geneva: WHO. http://whqlibdoc.who.int others. 2009. “A School-Based Program to Prevent Adolescent /publications/2004/9241591609.pdf. Dating Violence: A Cluster Randomized Trial.” Archives of ––––. 2006. “Youth Violence and Alcohol Fact Sheet” Pediatrics and Adolescent Medicine 163 (8): 692–99. (accessed June 30, 2013). http://www.who.int/violence Zapata, L. B., D. M. Kissin, C. L. Robbins, E. Finnerty, _injury _prevention/violence/world_report/factsheets /ft H. Skipalska, and others. 2011. “Multi-City Assessment _youth.pdf. of Lifetime Pregnancy Involvement among Street Youth, ––––. 2008. Preventing Violence and Reducing Its Impact: How Ukraine.” Journal of Urban Health 88 (4): 779–92. Development Agencies Can Help. Geneva: WHO. Zapata, L. B., D. M. Kissin, O. Bogoliubova, R. V. Yorick, ––––. 2009. Violence Prevention: The Evidence. Overview. J. M. Kraft, and others. 2013. “Orphaned and Abused Youth Geneva: WHO. Are Vulnerable to Pregnancy and Suicide Risk.” Child Abuse ––––. 2010. Global Strategy to Reduce the Harmful Use of and Neglect 37 (5): 310–19. Alcohol. Geneva: WHO. Zhao, J., T. Stockwell, G. Martin, S. Macdonald, K. Vallance, and ––––. 2013. Global and Regional Estimates of Violence against others. 2013. “The Relationship between Minimum Alcohol Women: Prevalence and Health Effects of Intimate Partner Prices, Outlet Densities and Alcohol-Attributable Deaths in and Non-Partner Sexual Violence. Geneva: WHO. British Columbia, 2002–2009.” Addiction 108 (6): 1059–69. ––––. 2014. “Global Health Estimates Summary Tables: doi:10.1111/add.12139. Deaths by Cause, Age and Sex, by WHO Region.” ZIMSTAT (Zimbabwe National Statistics Agency), UNICEF WHO, Geneva. http://www.who.int/healthinfo /global (United Nations Children’s Fund), and CCORE _burden_disease/en/. (Collaborating Centre for Operational Research and WHO (World Health Organization) and Liverpool John Evaluation). 2012. National Baseline Survey on Life Moores University. 2006. “Interpersonal Violence Experiences of Adolescents in Zimbabwe 2011. Preliminary and Alcohol.” WHO Policy Briefing. Geneva: WHO. Report. ZIMSTAT, Harare, Zimbabwe. 96 Injury Prevention and Environmental Health Chapter 6 Occupation and Risk for Injuries Safa Abdalla, Spenser S. Apramian, Linda F. Cantley, and Mark R. Cullen INTRODUCTION by OSH regulations and enforcement; are difficult to The world of work has changed dramatically. Globalization reach with traditional OSH services; and have greater, but affects the structure of workplaces, the way work is largely hidden, risk for accident and injury. Consequently, performed, and occupational safety and health (OSH). although the true burden of occupational injury in HICs Despite great strides in improving OSH during the past remains uncertain, an estimated 6.9 million worker inju- century, an estimated 317 million nonfatal occupational ries occurred in the European Union (EU) during 2006 injuries and 321,000 occupational fatalities occur globally and 8.5 million occurred in the United States during each year, that is, 151 workers sustain a work-related 2007 (Chau and others 2014; Leigh and Marcin 2012). accident every 15 seconds (ILO 2013a). Poor workplace Occupational injuries and fatalities take an even greater safety and health place a substantial economic burden on toll in LMICs, where a large portion of the population individuals, employers, and society. Estimates from the works in the informal sector or in high-hazard sectors, International Social Security Association (ISSA) suggest including agriculture, construction, fishing, and mining, that costs associated with nonfatal workplace accidents with associated costs as high as 10 percent of GDP. alone equal approximately 4 percent of world gross domes- The great recession of 2007–09 had a negative effect tic product (GDP) each year (ISSA 2014; SafeWork 2012). on OSH in many countries. Corporations downsized, Although virtually every job entails some risk for restructured, and outsourced or transferred work to injury, the magnitude of risk varies widely across jobs, third-party employers, temporary employment agencies, sectors, geographic regions, and individuals. Occupational or independent contractors. As of 2011, 22.3 million injury rates have been rising in low- and middle-income fewer adults and 6.4 million fewer youths participated in countries (LMICs), but declining in high-income the labor force than anticipated according to global countries (HICs), although the effect of globalization has trends before the downturn. From 2007 to 2010, the been mixed. The steady decline in Australia, North ratio of jobs to population declined sharply—from 61.2 America, and Western Europe is due, at least in part, to percent in 2007 to 60.2 percent in 2010 (ILO 2012c)— the export of labor-intensive and often more dangerous and the number of workers in precarious or vulnerable industrial production to regions where salaries are lower, employment reached an estimated 1.52 billion, an workplace regulations are less stringent, and working increase of nearly 23 million since 2009 and 136 million conditions are generally poorer. However, in HICs the since 2000. Latin America and the Caribbean, the Middle number of small firms and informal sector jobs has East, South Asia, South-East Asia and the Pacific, and grown markedly. These firms and jobs are underserved Sub-Saharan Africa experienced the largest increase in Corresponding author: Safa Abdalla, Stanford University, Palo Alto, California, United States; sabdalla@stanford.edu. 97 vulnerable employment. Women were disproportion- covering many sectors and work activities and often offer- ately affected in the Middle East, North Africa, and ing flexible work environments. Given their small size, Sub-Saharan Africa. many SMEs have limited resources and lack formal OSH This chapter discusses the many changes in work and programs and training. Results from a national survey of work-related injuries in seven sections. Following this U.S. firms with fewer than 250 employees found that few introduction, the second section reviews the current state had an employee safety committee, 50 percent had no of occupational injury and safety in HICs, with an formal safety policies, and only 60 percent provided safety emphasis on recent developments and observations, and training to new employees (NFIB 2002). Moreover, a the third focuses on the situation in LMICs, again with an survey of major health insurers in Germany indicated that emphasis on recent developments. The fourth section SMEs in the manufacturing sector had reduced their OSH reviews the effect of global supply chains on global busi- management since the economic recession, a concerning ness practices. The fifth section discusses the economic trend (Kraemer 2010). effects of these changes and interventions used in amelio- Employees of small business enterprises are exposed rating the problems raised, drawing heavily from observa- to higher health and safety risks than are employees of tions in the preceding sections to the extent possible. The larger enterprises (Fabiano, Currò, and Pastorino 2004; sixth section provides a brief synopsis of the contributions Sinclair, Cunningham, and Schulte 2013). They also have that workplace physical, chemical, and biologic exposures much greater difficulty assessing and controlling these may make to the occurrence of acute and chronic medical risks (Eakin, Champoux, and MacEachen 2010; Sørensen, conditions. A final section provides conclusions. Hasle, and Bach 2007). Furthermore, most OSH laws and regulatory agencies are designed for large enterprises in the formal economy OCCUPATIONAL INJURY IN HICs and either exempt or do not cover the informal economy Several developments have the potential to raise occupa- and SMEs. Hence, there is little reporting on these sec- tional injury rates. In HICs, temporary work and other tors or enforcement of laws and regulations even in forms of flexible employment have risen, including HICs. Employees who work in small enterprises far out- contingent work, home-based work, part-time contracts, number those who work in larger enterprises in many unregulated work, and other nontraditional work. Most countries, so addressing these gaps is critical (Hasle and of these arrangements are precarious; they are unstable, Limborg 2006). offer little social protection, and pay low wages. The substantial increase in outsourcing in recent Consistent evidence has shown that workers in precari- decades has shifted some work to workers’ homes and ous or vulnerable work arrangements experience more other informal settings. Whereas many home-based health and safety hazards and poorer health and safety workers are self-employed, others may work under some outcomes than do other workers. Labor statistics often form of outsourcing arrangement and fall under the capture only precarious workers in temporary employ- broad umbrella of precarious employment. ment, underestimating the true burden of precarious Studies examining the health and safety effects of employment on OSH (Benavides and others 2006; outsourcing or subcontracting and home-based work Virtanen, Janlert, and Hammarström 2011). Temporary have reported poorer OSH outcomes, using a range workers have twice the risk for occupational injury as of measures (Quinlan and Bohle 2008). Questions permanent workers, but the reasons for this higher risk regarding the mechanisms by which outsourcing and are poorly defined. They are likely to include less job home-based work negatively affect health remain, but experience, less recognition of workplace hazards, and several factors may contribute (table 6.1). inadequate or ineffective safety training (Virtanen, There is no universally accepted definition of migrant Janlert, and Hammarström 2011). Despite substantially workers. However, evidence from several press investiga- higher rates of occupational injury, temporary workers tions and published reports suggests that regardless of have lower absence rates, perhaps fearing the loss of their their legal status, migrant workers experience various job (Benavides and others 2006; Virtanen, Janlert, and forms of exploitation at work, although conditions are Hammarström 2011). typically worse for undocumented workers (McKay, Economic and employment growth in many regions Craw, and Chopra 2006). Migrant workers are less likely relies to a great extent on small and medium enterprises to receive workplace health and safety training in many (SMEs). More than 90 percent of businesses in the EU and sectors. When such training is provided, language barri- the United States employ fewer than 20 employees, and ers may prevent workers from understanding basic SMEs account for an estimated 82 percent of all occupa- safety procedures or knowing how or where to report tional injuries (Ecorys 2012). SMEs are extremely diverse, safety or health concerns. 98 Injury Prevention and Environmental Health Table 6.1 Negative OSH Consequences Potentially Arising from Home-Based and Outsourced Work Contributing factor Hazard Economic and reward pressure • Work intensification and compromised OSH Disorganization of management systems • Limited worker training or supervision • Poorly designed work settings • Inadequate safety protocols • Obscured mechanisms for workers to raise concerns Dispersed workforce with complex management • Less regulatory oversight and enforcement structures • Workforce logistically difficult for inspectorates to reach • Less worker understanding of employer OSH obligations and worker rights Source: Lippel 2005. Note: OSH = occupational safety and health. Although migrant workers face similar workplace • Awkward postures of the neck, trunk, and extremities hazards as local workers in similar sectors and jobs, their • Whole-body or segmental vibration safety and health may be at greater jeopardy for reasons • Mechanical contact stress from work positions or specific to their situation. Migrant workers are more handling of tools and equipment likely to be employed in hazardous work, work longer • High levels of ambient noise hours with fewer breaks, perform shift work, and be • Extreme temperatures temporary workers or subcontractors and are less likely • Work performed from heights to report accidents (Premji, Lippel, and Messing 2008; • Work performed around operating machinery. Premji and others 2010). Migrant workers are often overqualified for the work they perform in host coun- Although many jobs in HICs continue to require tries, and workplace injury may limit their occupational manual labor, exposure to workplace physical hazards is mobility, perpetuating the education–job mismatch. not limited to manual workers. Results from the 2010 There is a great need to address barriers to suitable European Working Conditions Survey indicate that employment and to improve health and safety strategies 33 percent of European workers handle heavy loads for targeting recent immigrants. at least 25 percent of their working time and 23 percent are exposed to workplace vibration (Eurofound 2010). Additionally, 30 percent of European workers are exposed Workplace Hazards to tiring positions 25–75 percent of their working time, Research conducted in HICs is the predominant source and 16 percent of workers are exposed to tiring positions of information about the contribution of workplace 100 percent of their working time (Eurofound 2010). hazards (physical, psychosocial, and work organization) In the United States, approximately 27 percent of and individual factors (gender, age, and health status) to working adults are exposed continually to repetitive occupational injury risk as well as ways to mitigate risk. motion, 25 percent spend more than half of their time at work either bending or twisting, an estimated 10 percent Physical Exposures are exposed to cramped workspaces that require them to Physical exposures related to job tasks, workplace envi- assume awkward postures every day, and 2.7 percent are ronment, use of tools and materials, machine operation, exposed to whole-body vibration (Tak and Calvert 2011). and machine-paced work affect workers in different Workplace physical hazards clearly persist in HICs, occupations and employment sectors (Chau and others underscoring the importance of mitigating these hazards 2009; Vandergrift and others 2012). The following phys- even as countries move toward becoming largely ical workplace exposures are strongly associated with service-based economies. injury risk: Psychosocial Exposures • Manual handling Significant changes in technology and management • Forceful exertions ideologies, combined with increases in global competi- • Highly repetitive motions with short work cycles tion, are responsible for the trend toward more difficult, Occupation and Risk for Injuries 99 faster, more productive labor with less control over tasks Workers with rotating shifts often experience sleep (Green 2005). Workplace psychosocial hazards arising deficits and fatigue, which decrease their mental agility, from evolving work demands, in conjunction with reduce performance efficiency, and increase error rates. changing economic and social contexts of work, are Evidence suggests that night workers have higher risk for emerging threats to physical and mental health injury than do day workers, with successive night shifts (EU OSHA 2007; NIOSH 2002). Evidence has been further elevating the risk (Folkard and Tucker 2003). amassed suggesting an association between stressors and Increased risk for injury has also been associated the risk for work-related injury (Glasscock and others with working overtime, long hours, and 12-hour shifts 2006; Kim and others 2009; Nakata and others 2006) (Dembe, Delbos, and Erickson 2008; Folkard and and musculoskeletal disorder (MSD) (Bongers and Lombardi 2006). Shorter sleep duration and longer work others 2006; da Costa and Vieira 2010; Ghaffari and hours are independently associated with the risk for others 2008). work-related injury, mainly because fatigue impairs Psychosocial stressors include the following: cognitive functioning and slows response time (Lombardi and others 2010). The understanding of the effect of • Work intensification fatigue on performance is complicated by the existence • Highly monotonous work of individual differences in vulnerability to fatigue. • Time pressure or deadlines These differences may be critical for workers in round- • Significant mental workload the-clock operational settings (Van Dongen, Caldwell, • Ambiguous or conflicting roles and Caldwell 2011). • Lack of decision-making authority As women increasingly participate in the labor force • Machine-paced work or piecework and men assume a progressively larger role in domestic • Isolation duties in most HICs, in combination with substantial • Weak supervisor support changes in family composition and labor force demo- • Demand or reward imbalance graphics, balancing work and family demands has • Job insecurity. become more challenging (Valcour 2007). Work–life conflict has been associated with adverse outcomes, Physical and psychosocial workplace exposures including work-related MSDs (Hämmig and others increase the risk for injury and MSD. In combination, 2011), sleep disorders and fatigue (Wirtz, Nachreiner, they create even greater risk (Lapointe and others 2009; and Rolfes 2011), and reduced labor force participation Magnavita and others 2011), with important implica- and its economic consequences (Jansen and others tions for OSH interventions. 2010). Associations between work–life conflict and adverse outcomes extend beyond the high-risk sectors. Work Organization For example, in the retail sector, Sunday work signifi- The modern 24-hour society has greatly affected cantly increases the risk for accidents (Wirtz, Nachreiner, the timing of work hours. The development of new and Rolfes 2011). Workplace interventions that reduce technologies and global economic competition conflicts between work and private life and address require that goods and services be made available at other risk factors are needed to prevent workplace all hours of the day and night (Costa 2010). The tra- injury and MSDs. ditional schedule of regular, mainly daytime working hours has given way to a variety of work patterns for many workers in HICs. Results from the 2000 Individual Factors European Working Conditions Survey indicate that Individuals have varying susceptibilities to workplace only 25 percent of employed workers and less than injury, and this variability is related to occupational and 10 percent of self-employed workers have traditional individual characteristics (Clarke 2011; Schulte and work schedules. The vast majority work irregu- others 2012). Many reports have found a consistently lar hours, including some combination of com- elevated risk for injury among younger workers (Breslin pressed work hours, variable work hours, shift work and others 2007; Breslin, Smith, and Moore 2011) and or night work, weekend work, part-time work, and workers with lower educational attainment (Breslin on-call work (Costa and others 2004). A large body of 2008; Strong and Zimmerman 2005). evidence suggests that shift work and night work The increased risk for injury among novice workers interfere with circadian rhythms, decrease efficiency, compared to their longer-tenured counterparts remains and strain social and family relationships (Costa and despite adjustment for confounders, including age, Di Milia 2010). sex, and job (Kubo and others 2013; Morassaei and 100 Injury Prevention and Environmental Health others 2013). The reasons include a combination that future attention should focus specifically on under- of unfamiliarity with job tasks or work environment, standing the physical, psychosocial, and training needs failure to recognize workplace hazards, ineffective or of women. inadequate safety training, and differential exposure to more hazardous tasks at the beginning of a job (Breslin Age and Smith 2006; Morassaei and others 2013). Among younger workers, differential distribution by type Workers with lower levels of education appear to of job, workplace environment, and organizational struc- be particularly vulnerable, possibly because of their ture plays an important role in their elevated risk for greater exposure to physical demands or other hazards injury, because younger workers are more likely to work (Breslin 2008). Additional evidence suggests that experi- in more hazardous jobs (Breslin and Smith 2005) and are enced (and older) workers plan ahead in order to limit overrepresented in small enterprises, which have limited fatigue and avoid stressful emergency situations much OSH resources (Eakin, Champoux, and MacEachen more than do their less experienced coworkers (Pueyo, 2010; Headd 2000). Moreover, almost half of working Toupin, and Volkoff 2011). Experienced workers also adolescents receive no safety training, suggesting that engage in more verbal communication with their workplaces where young workers are employed pay less colleagues. attention to OSH (Knight, Castillo, and Layne 1995). These findings illustrate the potential benefit to be As life expectancy increases, the population ages, and gained from targeted job training. many workers extend their working life beyond tradi- tional retirement age, interest in the consequences of Gender injury among older workers has grown (Smith and The gender gap in labor force participation is closing others 2014). Although older workers may have a lower globally. Women’s participation has held steady at risk for injury (Chau and others 2014; Smith and others roughly 52 percent for the past few decades, and men’s 2014), they may suffer worse consequences if they are participation has declined from 81 percent in 1990 to injured, requiring longer periods for recovery and higher 77 percent in 2010. A wide gap remains in some regions. associated costs (Pransky, Loisel, and Anema 2011; Women’s participation has fallen well below 50 percent Silverstein 2008). in Northern Africa, Western Asia, and Southern Asia However, little is known about the changing OSH (UN DESA 2010). Although women predominantly and needs of workers beyond ages 55–60 years, because most increasingly work in the services sector, the proportion occupational analyses denote 55+ as the oldest age of women employed in traditionally male-dominated category (Farrow and Reynolds 2012). More research is sectors such as manufacturing has risen. crucial to inform OSH for aging workers. Injury and fatality statistics by industry suggest that women are at lower risk for workplace injury (Lin, Chen, Health and Luo 2008, 2011). However, many of these reports fail A few reports have linked chronic health problems to to account for the differential distribution of men and occupational injury, but substantial gaps remain in the women among jobs or even tasks within jobs. Evidence evidence. Several reports suggest that hearing impair- is emerging that women are at elevated risk for acute ment increases the risk for occupational accidents and injury and MSD, controlling for job and individual con- injury (Cantley and others 2015; Girard and others founders (Taiwo and others 2009; Tessier-Sherman and 2009). Diabetes, chronic heart disease, and depression others 2014). Qualitative research also suggests that male may also confer increased risk for acute occupational workers—in traditionally male- and female-dominated injury, although the evidence is more limited (Kubo and jobs—have more control over their job and often receive others 2014; Palmer, Harris, and Coggon 2008). more safety training than do their female coworkers (Kelsh and Sahl 1996; Turgoose and others 2006). Further, research examining gender differences in the Employment Sector performance of repetitive tasks suggests that identical, Globalization has subjected the manufacturing sector in force-demanding tasks may be considerably more HICs to intense international competition. As a result, strenuous for females than for males (Nordander and high-hazard technologies have moved to LMICs, and the others 2008), which could increase the risk for injury services sector has become increasingly important. and MSD among women. Although occupational injury is a risk in the services These findings, combined with the increasing labor sector, the riskiest sectors are agriculture, forestry, force participation of women globally and the large pro- fishing, construction, manufacturing, and transporta- portion of women in precarious employment, suggest tion. These sectors account for approximately half of the Occupation and Risk for Injuries 101 serious accidents at work and the largest share of fatal concentrated periods of work that lead to time pressures, accidents. According to estimates from the Survey of stress, and fatigue, which are linked to increased risk Occupational Injuries and Illnesses conducted by the for accidents and injuries. Leading risk factors for U.S. Bureau of Labor Statistics, among the 2.8 million agriculture-related injuries and fatalities include opera- nonfatal occupational injuries reported by private indus- tion of farm equipment and machinery, work with try in 2012, 75 percent occurred in service-providing animals, work performed at heights, and falling objects industries, which employed approximately 82 percent (Pfortmueller and others 2013). Tractor use is associated of the private industry workforce. The remaining with a large number of fatalities. 25 percent occurred in goods-producing industries, which employed 18 percent of the private industry Construction workforce (BLS 2013). Construction is one of the most physically demanding and dangerous sectors in both LMICs and HICs. Workers Agriculture, Forestry, and Fishing are regularly exposed to ergonomic and safety hazards Work in the agricultural, forestry and fishing sector is from manual handling, power tools and equipment, among the most hazardous, and comparatively weaker noise, confined spaces and electricity, work performed health and safety regulations, in combination with from heights, excavation, irregular work hours, and growing numbers of immigrant workers and a paucity exposure to weather extremes. A construction worksite is of surveillance data, have resulted in widespread under- also complex and dynamic. Often it comprises multiple recognition of worker injury risk. And although occupa- employers with potentially divergent safety cultures and tional health and safety research is limited overall for this a high proportion of self-employed workers, adding to sector, research for the forestry and fishing subsectors is the challenge of effectively disseminating safety informa- particularly sparse. Recognizing the dual needs for tion and interventions for effective uptake. This sector more accurate surveillance data and development and also employs a disproportionate number of immigrants, implementation of effective OSH interventions, a formal independent contractors, on-call or day laborers, contract research and public health practice agenda for this sector workers, temporary workers, and young workers—all is underway (NIOSH 2008). subgroups with higher injury risk, which presents major Agricultural production not only employs the largest challenges for OSH in this sector (CPWR 2013). number of workers worldwide (about 1.3 billion), but In HICs, construction workers have higher-than- also consistently ranks as one of the most hazardous average risk for injury and MSD, and the leading causes sectors It has high rates of both fatal and nonfatal inju- of injury involve contact with objects and overexertion. ries and fatality rates several times higher than the More than half of injuries sustained by self-employed average for all industries combined in the EU and North workers require five or more days away from work com- America (BLS 2014; Frank and others 2004; Vijayvergiya, pared to only a quarter of the injuries sustained by Bohra, and Jhanwar 2012). workers employed by firms (HSE 2014a). Agricultural injuries are less well documented in In HICs, construction workers have a three- to four- developing countries, where the vast majority of this fold risk for a fatal accident at work compared to workers workforce is located (Lehtola and others 2008). Even in in other sectors, while in LMICs the risk is as much as HICs, which have only 9 percent of the global agricul- sixfold (ILO 2014b). Even among HICs, however, fatality tural workforce, many agricultural accidents and injuries rates differ, although the reasons for this disparity are are not captured because of the high rates of self- poorly understood (Mendeloff and Staetsky 2014). employment and large number of small farms, tempo- rary workers, and migrant workers. Manufacturing The unique nature of many farms helps explain the Although the manufacturing sector comprises a diverse increased risk for injury among agricultural workers. array of industries worldwide, the majority of manufac- Many farms are small and family owned, with economic turing jobs are labor intensive. Workers who are engaged pressures fostering use of less-expensive methods and in transforming materials, substances, or components equipment that may increase injury risk. Many farms are into products are at risk for injury from physical also family homes, where children and young adults live exertion; contact with machinery and equipment; long and work at least part-time and safety training is likely work hours; changing work shifts; slips, trips, and falls learned through personal experience and from family (STFs); and new methods or organization that may members rather than through more structured processes. increase job strain. Farm work is seasonal and labor intensive; workers are The manufacturing sector employs approximately exposed to adverse weather conditions and subjected to 10 percent of the workforce in both the United Kingdom 102 Injury Prevention and Environmental Health and the United States but accounts for a disproportionate risks associated with the handling of dangerous sub- number of injuries. In the United Kingdom, the manu- stances, the performance of physical jobs in isolation, facturing sector accounts for 18 percent of nonfatal long and variable working hours, frequent need for workplace injuries requiring more than seven days away vigilance, and psychosocial and organizational factors— from work and 17 percent of major specified injuries. that combine to increase injury risk among workers in STFs on the same level (29 percent), contact with this sector are less recognized. Exposure to whole-body machinery (14 percent), and a blow by an object vibration and prolonged sitting or standing, interspersed (13 percent) are the most frequently cited causes of with the physically strenuous work of loading major or specified injury (HSE 2014b). and unloading goods, increase the risk for MSDs, espe- cially back disorders. The shift in manufacturing from Wholesale and Retail Trade inventory-based systems to leaner, just-in-time produc- Although workers in the wholesale and retail trade tion processes has created very narrow margins for sector are generally perceived as having lower risks for timely delivery. Transport workers are also exposed to injury than are workers in other sectors, many trade jobs high levels of noise in and around vehicles. are physically demanding, which places workers at risk for back and upper-extremity disorders. Given the large number of workers and continued growth in this sector, a wide range of workplace hazards may pose a risk for OCCUPATIONAL INJURY IN LMICs injury among a considerable number of workers. In According to the latest estimates from the 2013 Global addition, psychosocial and organizational factors may Burden of Disease (GBD) study, more than 80 percent contribute to the burden of injury. Historically, the of occupational injury–related deaths in the world occur causes of and potential interventions for safety hazards in LMICs (Murray and others 2014), where the death within this sector have received little attention, but this rate is higher than in HICs. Occupational risk factors is one of eight sectors for which a research agenda has for injuries are also the leading cause of occupational been developed to address existing gaps (Anderson and fatalities among men and women ages 15–49 years others 2010). (figure 6.1) and the leading cause of disability-adjusted life years (DALYs) resulting from occupational risks Health Care and Social Services in LMICs. Almost half of these deaths and DALYs are The health care and social services sector is a large attributable to transport injuries. employer in HICs, with projections suggesting continued Covering only unintentional injuries, these estimates growth as populations age. Workers in this sector are exclude workplace violence, which is increasingly mainly female and experience high rates of injury, being documented in LMICs. Table 6.2 summarizes the especially musculoskeletal injury. Injuries resulting from literature on the incidence of physical violence in the overexertion and STFs are particularly problematic workplace. (Bell and others 2008; Collins, Bell, and Gronqvist 2010). LMICs clearly need to address the alarmingly pro- Frequent lifting, transferring, and repositioning of tracted trend in occupational injuries. However, intri- patients are leading causes of musculoskeletal injury cately related factors—economic adversity, competing among health care workers in both acute and longer- priorities, resource mismanagement, workforce migra- term care settings, whereas STFs are particularly preva- tion, conflict and internal displacement, and urban-to- lent among facility support workers and community rural influx, among other factors—foster a dearth of health workers (Drebit and others 2010). Because sharps financial and human resources devoted to such efforts, injuries are extremely common among these workers, while crafting an industrial landscape that features a targeted preventive efforts have been undertaken in vulnerable workforce. Those factors, in turn, underpin recent years. most of the challenges facing implementation of OSH in LMICs. Although some of these challenges are not Transportation and Warehousing unique to LMICs, their pervasiveness impedes a simplis- The transportation and warehousing sector, which tic transfer of potentially successful occupational health enables the movement of passengers and goods via air, interventions from HICs. An overview of these chal- rail, water, and road, is vitally important to the econo- lenges follows. mies in HICs and encompasses a very diverse group of Evidence on the proximal work- and worker-related workers, jobs, and job-related hazards. The risk for risk factors in LMICs suggests that they are not substan- accidents causing human injury and fatality is widely tially different from those documented in HICs. Table 6.3 recognized. However, the range of hazards—including gives some examples of risk factors. Occupation and Risk for Injuries 103 Figure 6.1 Mortality Attributable to Occupational Risks in Low- and Many LMICs have health and safety legislation in Middle-Income Countries, 2010 place, but capacity for enforcement is weak (ILO 2012b). A common feature in several LMICs is a slower increase a. Males and sometimes a decrease in the ratio of trained labor inspectors to workers. This situation suggests inadequate Injuries institutional capacity to enforce provisions for safe work Particulate matter, environments, although comparisons over time are gases, and fumes limited by the poor quality of the data. Moreover, under- funding of government departments mandated with Asthmagens statutory inspection, as reported in Zambia (ILO 2012a), weakens the infrastructure required for carrying out Carcinogens inspections. The World Health Organization (WHO) and its 0 2 4 6 partners have called for integrating occupational health Deaths per 100,000 persons ages 15–49 with primary health care (WHO 2001, 2012). However, health care workers are too few in number and poorly b. Females distributed, particularly in LMICs, where the crisis is fed Injuries by inadequate infrastructure, insufficient investment in health care and training, and outmigration of health care Particulate matter, workers (Chen and others 2004). gases, and fumes Trade unions can be strong partners and advocates for worker safety. Recent agreements have been signed in Asthmagens the Asia-Pacific region to increase the participation of trade unions in promoting safe workplaces by orga- Carcinogens nizing campaigns and participating in national OSH strategies and plans, among other activities (ILO 2014a). 0 2 4 6 However, trade union density is low in some countries in Deaths per 100,000 persons ages 15–49 Latin America and the Caribbean, and Sub-Saharan Africa (Hayter and Stoevska 2011). This situation limits Source: Based on Global Burden of Disease Study 2013 estimates (Murray and others 2014). Table 6.2 Exposure to Physical Violence in the Workplace in Some Low- and Middle-Income Countries Region and Number location Study Study population exposed Year Percentage per year Asia Taiwan, China Chen and others 2008 Nurses, nurse aides, and 222 2003 35 clerks in a psychiatric hospital Pai and Lee 2011 Registered clinical nurses 521 — 19.6 Thailand Kamchuchat and others 2008 Nurses 545 2005 3.1 Sripichyakan, Thungpunkum, Health care workers 1,090 2001 11.0 and Supavititpatana 2003 Middle East and North Africa Egypt, Arab Rep. Abbas and others 2010 Nurses in Ismailia 970 2010 3.0 governorate Latin America and the Caribbean Brazil Alonso Castillo and Working women ages 109 — 39.0 others 2006 18–60 Palácios and others 2003 Health care workers 1,569 2001 6.0 table continues next page 104 Injury Prevention and Environmental Health Table 6.2 Exposure to Physical Violence in the Workplace in Some Low- and Middle-Income Countries (continued) Region and Number location Study Study population exposed Year Percentage per year Mexico Alonso Castillo and Working women ages 669 — 16.0 others 2006 18–60 Peru Schlick and others 2014 Children and adolescents 375 2010 3.0 attending night school in Cusco Sub-Saharan Africa Kenya El Ghaziri and others 2014 Nurses and midwives 227 2007 8.8 Mozambique Couto, Lawoko, and Svanstrom Conductors and drivers 504 — 32.0 2009 Caldas and others 2003 Health care workers 396 2001 8.0 Nigeria El Ghaziri and others 2014 Nurses and midwives 159 2008–10 32.1 Azodo, Ezeja, and Ehikhamenor Oral health care 175 2009 6.0 2011 professionals in Southern Nigeria South Africa Steinman 2003 Health care workers 1,018 2001 13.0 Tanzania El Ghaziri and others 2014 Nurses and midwives 146 2009 22.0 Other Sub-Saharan El Ghaziri and others 2014 Nurses and midwives 85 2007 21.0 African countriesa Note: — = not available. a. Central African Republic, Eritrea, Ethiopia, Malawi, Namibia, Togo, Uganda, Zambia, Zimbabwe, and others. Table 6.3 Risk Factors for Workplace Injury and Violence in Various Worker Populations in LMICs Population Risk factors of occupational injury Study Miners in Zimbabwe Underground work, long working hours, targets per shift Chimamise and others 2013 (work pressure), inadequate personal protective equipment Machine operators in Ethiopia Work pressure, malfunctioning machines, unfamiliar Ahmed 2008 techniques, unfamiliar tasks, failure to wear gloves Textile factory workers in Ethiopia Lack of training, sleep disturbance, job stress, long hours, Aderaw, Engdaw, and Tadesse 2011; manual work, work requiring visual concentration Yessuf Serkalem, Moges Haimanot, and Ahmed Ansha 2014 All workers in a representative household Regular or daily alcohol consumption Phung and others 2008 sample in Vietnam Health care workers in Cairo University Time on the job Zawilla and Ahmed 2013 hospitals Nurses in Mulago National Hospital in Lack of training on needlestick injuries, long working Nsubuga and Jaakkola 2005 Uganda hours, recapping of needles, failure to wear gloves Commercial motorcyclists in Nigeria Lack of formal education, alcohol consumption Adogu, Ilika, and Asuzu 2009 Workers in Thailand Low income, long working hours, heat stress Berecki-Gisolf and others 2013; Tawatsupa and others 2013 Workers in a commune in Vietnam Overlapping employment (full-time in industry, part-time Marucci-Wellman and others 2011 in agriculture) table continues next page Occupation and Risk for Injuries 105 Table 6.3 Risk Factors for Workplace Injury and Violence in Various Worker Populations in LMICs (continued) Population Risk factors of occupational injury Study Nurses in the Philippines (needlestick Night shifts de Castro and others 2010 injuries) Nurses and midwives in Sub-Saharan Risky client characteristics, long working shifts El Ghaziri and others 2014 Africa (workplace violence) Frontline workers in 60 factories in China Educational level, mental stress, previous injury, Yu and others 2012 working hours Cleaners in the city council’s health Lack of preemployment training Gonese and others 2006 services department in Zimbabwe Note: LMICs = low- and middle-income countries. the scope for workers to engage in organized action to Western Asia (Charmes 2012). Informality levels also promote their safety (although high union density does vary, ranging from one-half of nonagricultural employ- not necessarily mean strong bargaining power). ment in North Africa to almost three-quarters in Sub- Employers in SMEs can be an important informal Saharan Africa (ILO 2002). Some regional differences source of information on OSH for their employees. In are evident in the contribution of the informal sector to Ghana, a positive perception of organizational support GDP. The highest level is in Sub-Saharan Africa, where was associated with a positive perception and practice of the informal sector contributes almost two-thirds of safety and a lower rate of injury among workers (Gyekye GDP and one-half of nonagricultural gross value added. and Salminen 2007). Yet some employers, especially in Agriculture alone comprises up to 60–65 percent of total small enterprises, have little knowledge of OSH and informal employment except in Latin America and the rarely adopt OSH practices, although the evidence is Middle East and North Africa, where construction and limited (Hu and others 1998). manufacturing constitute the bulk of nonagricultural Estimating and tracking the burden of occupational informal employment (Charmes 2012). injuries in LMICs is crucial but challenging. Many coun- The relation between informal employment and tries lack an acceptably complete system for registering occupational injury risk is difficult to characterize fatalities. The informal sector is usually excluded from mainly because of scarce data on occupational injuries the mandatory reporting of occupational injuries, and working conditions in the informal setting. Many and injuries often are not reported even where formal workers in informal employment are poor, work in channels of reporting exist, for example, sharps injuries adverse physical conditions, and lack social insurance among health care workers (Hanafi and others 2011; protection and benefits that could mitigate the conse- Mbaisi and others 2013; Shiao and others 2009). Health quences of occupational injury (Muntaner and others facility data and some national surveys in LMICs collect 2010). Table 6.4 highlights some of the risks in informal data on injuries, but these data often are of poor quality employment situations. or incomplete, failing either to identify their relation to The few studies on this topic have found that informal work or to analyze and report on this dimension. sector workers did not have higher injury rates than Attempts to estimate the global burden are useful for formal sector workers. For example, a study in Costa highlighting the problem internationally, but they may Rica reported a relative risk of only 1.06 (Mora and not be fit for supporting local decision making or for others 2011), while in Nicaragua, injury rates in the tracking trends at the subnational level. informal sector were half those in the formal sector. The informal sector grew tangibly over the past four These reports likely reflect high levels of underreporting decades in many LMICs. In most countries in Sub- among informal workers rather than lower injury rates Saharan Africa with time-series data, the share of infor- (Noe and others 2004). In Vietnam, work-related inju- mal sector employment (including both informal and ries were 72 percent more frequent among informal formal jobs in the informal sector and informal jobs in workers (falling to 45 percent when adjusting for socio- the formal sector) in total nonagricultural employment demographic factors) than among formal workers, but is rising. A similar trend is seen in Latin America and in the role of chance could not be ruled out statistically Southern and South-East Asia. Notable exceptions are (Phung and others 2008). Severe injuries were even Mexico, South Africa, and Thailand and countries in found to be more frequent among formal workers. 106 Injury Prevention and Environmental Health Table 6.4 Occupational Hazards Identified in Various Informal Industries in LMICs Informal industry Hazard Study Agriculture Hand tools (spades and sickles), machinery (harvester and threshers), Loewenson 1998; Mohan and Patel 1992 venomous animals, pesticides, falls, exposure to sun and heat Street vending Traffic injuries, fire, assault, weather extremes Alfers 2009; Alfers and Abban 2011 Textile Fire, chemicals Regoeng 2003 Auto repair Fire, chemical solvents and acids, mechanical injury Buhlebenlosi and others 2013; Regoeng 2003 Waste management Medical waste including syringes, broken glass Cunningham, Simpson, and Keifer 2012 and recycling Welding Radiation Buhlebenlosi and others 2013 Carpentry Sharp tools Regoeng 2003 Manufacturing Sharp tools, exposure to sun and heat, chemicals Loewenson 1998 Domestic work Violence (verbal, physical, sexual); repetitive strain, household Alfers 2011 chemicals Note: LMICs = low- and middle-income countries. However, the severity of injury was defined by type of equipment interferes with work precision (Chepkener health care received. Thus, the finding probably reflects 2013). Similarly, 90 percent of welders in Southwestern differential access to health care, because case fatality was Nigeria were aware of protective eye gear, but less than higher among informal workers. These studies did half owned them and only 10 percent actually used them, not account for differences in the duration of actual citing similar reasons for nonuse (Ajayi and others 2011). exposure to injury risk. Among vegetable farmers in Ghana who use chemicals OSH in the informal sector is particularly challenging. in farming, almost three-quarters did not use protective The magnitude of the injury problem is difficult to cover when handling insecticides and about 80 percent measure because of lack of routine mandatory reports. disposed of empty containers unsafely (Ntow and others This prohibits injury from taking its deserved position on 2006). Sugarcane crushers in India thought that hand the list of priorities for policy makers and, in turn, con- injuries sustained during work were just “bad luck” or tributes to the shortage of resources devoted to addressing “God’s will,” more than 60 percent blamed injuries on it. Minimal contact with the health and safety authorities carelessness, a minority thought that safe machines were suggests that monitoring and enforcing safe working con- needed, and less than 33 percent indicated that they ditions are almost impossible tasks (Muntaner and others would use protective equipment even if it were provided 2010). The informal economy, dominated by agriculture, for free (David and Goel 2001). construction, and manufacturing, often relies on low-cost Economic hardship and conflict have long fueled manual technologies. Potentially effective control mea- cross-border labor migration and rural-urban migration sures such as engineering controls and elimination or in Africa, Asia, and Latin America. Temporary workers, substitution of hazards may be too costly and thus prob- farm laborers, female traders, and professionals have been lematic to implement in LMICs. moving between countries of West Africa along gradients Personal safety measures, such as the use of personal of economic opportunity, mainly to Côte d’Ivoire and protective equipment, can therefore be a principal means Ghana and more recently to Nigeria, slowing their pace of OSH in such settings. Workers in the informal sectors at times of economic and political crises in those coun- reportedly have high levels of knowledge about hazards tries (Adepoju 2005). Political turmoil and poverty in and personal protection methods, but compliance Zimbabwe led many Zimbabweans to enter neighboring with health and safety precautions can be exceptionally South Africa, where they found work as farm laborers challenging in the absence of enforcement. For example, (Vigneswaran 2007). In Latin America, Argentina attracts 73 percent of the Jua Kali informal sector in Kenya knew the most significant amount of immigration from within that eye injury could be prevented by the use of personal the region. The most significant migration corridors are protective equipment. However, only about 12 percent Paraguay–Argentina, Bolivia–Argentina, and Colombia– said that they actually used such equipment, mainly República Bolivariana de Venezuela. Other corridors of because of a perception of low risk and a sense that the lesser importance are Peru–Argentina and Peru–Chile Occupation and Risk for Injuries 107 (Texido and Warn 2013). In China, rural-urban migra- interventions from HICs to LMICs. Therefore, progress tion has been on the rise and forms the bulk of internal in OSH is inseparable from overall progress in many migration (Wang 2008). domains: economy, governance systems, data systems, Similar to migrant workers in HICs, migrant workers education, employment prospects, and social justice. in LMICs are at risk for social exclusion, exploitation by Addressing the problem will require community-based employers, informal employment, and poor working public health initiatives that transcend the traditional conditions (ILO 2013c; Marilda and Maciel 2014; Texido well-demarcated workplace, scale up participatory inter- and Warn 2013; Vigneswaran 2007). In China, where vention, invest in the local workforce, and improve the access to public services relies on the hukou system of quality and use of data. household registration, internal migrants who leave their place of registration cannot use health insurance THE GLOBAL SUPPLY CHAIN and other benefits and services (Mou and others 2013). Rural migrants may lack proper training for the job Over the past three decades, globalization has changed (Pringle and Frost 2003) and work long hours in the way goods are manufactured and exchanged interna- poor conditions with low wages and limited benefits tionally. International supply chains offer a flow of materi- (ILO 2011b). Studies suggest that migrant workers in als from natural resource to final product in a manner China have a higher rate of injury than do registered that is cost-efficient and easily scalable to produce urban residents and are more likely to engage in hazard- high-quality goods for affordable prices. The supply chain ous occupations (Xia and others 2012). is governed by a focal company, often a multinational, The percentage of children at work—a special subset which receives goods and materials from suppliers that of informal workers—in LMICs ranges widely from use subcontractors to develop raw materials into finished 2.5 percent of children ages 7–14 years in Costa Rica and products (Ustailieva, Eeckelaery, and Nunes 2012). India to as high as 74 percent in Benin (World Bank 2014). Firms in the EU and the United States have developed The percentage of working children who do not attend global supply chains spanning a variety of industries, school also varies widely, from around 2 percent to 89 ranging from apparel and toys to electronics (Locke 2013). percent, but does not exceed 50 percent in most LMICs. Lambert and Cooper (2000, 70) define the members of a The vast majority of working children are employed supply chain as “all companies/organizations with whom in agriculture. Others are employed in a range of the focal company interacts directly or indirectly through industries—fisheries, domestic work, mining and its suppliers or customers, from point of origin to point manufacturing, and construction—where adults are also of consumption.” Many foreign manufacturers of U.S. employed, and some are soldiers in war-affected zones. goods (contract manufacturers) are not just subcontrac- The problem of child workers is waning worldwide but is tors; they are supply-chain facilitators, providing U.S. still alarming in Sub-Saharan Africa, where one in five firms with everything from production facilities and children is working (ILO 2013b). Child workers are at engineering expertise to logistics (Eltschinger 2007). One higher risk for injuries and may suffer greater conse- advantage of using foreign factories is the unmatched quences because their bodies are still growing. Working scalability of their labor. A sense of how globalized the children often emulate unsafe behaviors of adults, lack supply chain has become can be gained by considering a adequate safety training, are at higher risk for exploitation, Nike cross-trainer shoe: the outer rubber sole is refined in and endure long working hours and minimal pay, often the Republic of Korea; processed into large rubber sheets superimposed on a background of deprivation (ILO in Taiwan, China; and shipped to an assembly plant in 2011a). Eliminating child labor remains the ultimate Indonesia, where it is attached to the shoe (Locke 2013). international goal, but a child’s work in LMICs can be Intermediate exports grew from US$2.867 trillion in instrumental for the family’s livelihood and sometimes 2000 to US$7.723 trillion in 2012, and the share of inter- the child’s own education. Until child labor has been elim- mediate goods as a percentage of total nonfuel exports is inated, efforts to mitigate hazardous working conditions 55 percent, the highest in global history (WTO 2013). for children are needed (Siddiqi and Patrinos 1995). The number of intermediate goods integrated into global In summary, LMICs bear the brunt of global occupa- supply chains has reached a record high. tional fatality and disability, yet many lack the resources Industrial nations (prominently the United States and infrastructure to tackle them effectively. Challenges and EU countries) continue to outsource a significant related to workforce shortages, compliance of employees portion of the goods and services of their primary (agri- and employers with OSH, lack of sound data, and culture, fishing, and mining) and secondary (manufac- predominance of a vulnerable and hard-to-reach infor- turing) sectors to LMICs, creating economic benefits mal workforce preclude the direct transfer of successful and reducing the number of fatal occupational accidents 108 Injury Prevention and Environmental Health occurring in HICs (Takala and Hämäläinen 2009). Their among large manufacturing plants in China, where integration into global supply chains has provided 15,000–20,000 workers are at risk for injury at any LMICs with promising economic opportunities and a given moment (Brown 2007). Moreover, these employ- prominent place in the world economy, while providing ees are hired and fired on short notice. This rise in HICs with low-cost labor (Rivoli 2009). nonpermanent labor has been associated with poorer Focal companies have a clear profit incentive to OSH outcomes in developing areas such as Bangladesh, acquire low-cost ready-for-market goods. In turn, sup- China, Lebanon, South Africa, and Thailand, as well as pliers contracted by the focal company strive to keep Central America (Baldwin 2011). Finally, forced over- costs low and to deliver products quickly and reliably time in excess of 6-day, 72-hour work weeks is all because of volatility in production orders and intense too common (Locke and Romis 2007), resulting in competition with rival firms. Because the contracted increased risk for accidents and repetitive motion factories in LMICs tend to be under intense economic injuries (Brown 2007). pressure to deliver intermediate products at a low cost, Another issue of concern is the limited availability work-related injuries and occupational hazards persist. and accessibility of OSH training and education in In order to respond to large fluctuations in the quantity source countries undergoing rapid industrialization of production in a short span of time, temporary (Ahasan 2003). The occupational health services cover- employment has become very common in factories in age in China, for example, is estimated to be in the LMICs (Smith, Sonnenfeld, and Pellow 2006). 10 percent range, whereas in HICs, the average is in the The majority of workers integrated into global supply 20–50 percent range (Barboza, Lattman, and Rampell chains are migrants who move from poor rural areas to 2012). Moreover, the intergovernmental organization the rapidly developing and industrializing urban areas overseeing the occupational health services of manufac- within their country seeking employment with higher turers in LMICs—the International Labour Organization pay, because rural job opportunities are often sparse and (ILO)—has reported less than 1 percent of occupational pay below subsistence-level wages (Welford and Frost accidents in nations such as China and India because 2006). From 1995 to 2000, 79 million migrant workers in occupational records in these countries are either nonex- China alone moved to prominent manufacturing cities in istent or inconsistently maintained (Hämäläinen, Takala, search of higher-paying employment (Wang and others and Saarela 2006). 2011). Most of these migrant workers are young (ages 17–39 years), have little to no formal education, and lack experience in an industrial environment. Consequently, Regulation of Occupational Health they tend to have a poor understanding of workplace Fragmented production in the global supply chain has risks and labor rights. They are placed in the highly resulted in the establishment of a flurry of regulatory hazardous workplaces of construction, mining, and bodies meant to control the OSH and labor rights of each manufacturing often without training, resulting in high supplier. However, the complexity of the chain often rates of injury, sickness, and death at the workplace. The prohibits effective regulation. In the twentieth century, excessive competition for temporary work means that OSH and labor rights were regulated at the national level they are reluctant to report minor injuries on the job for for many LMICs. The ILO and WHO provided similar fear of losing employment. These subcontracted laborers regulation at the international level, publishing reports have one and a half times the occupational accident risk on worker conditions and employment demographics of their full-time counterparts (Quinlan 1999). for each sector globally. However, a previous effort to This growth of contract work is not a temporary include social clauses within global trade agreements was trend. Rather, the number of workers with precarious struck down by WHO when both LMICs and HICs employment is rising in LMICs, leading to a global gap voiced their strong dissent (Locke 2013). Thus, the advent in OSH standards and regulation (Nossar, Johnstone, of the global supply chain left a void in the regulation of and Quinlan 2003; Quinlan, Mayhew, and Bohle 2001). OSH and labor rights, because authority was dispersed in In China, 80 percent of the recorded on-site deaths were a complex web of buyers and sellers. Compounding this of migrant workers (Wang and others 2011). Currently, situation are the economic incentives for governments to 50–70 percent of workers in LMICs define their employ- ignore factory OSH violations in order to keep the cost of ment as “vulnerable”—work that is low in pay, lacks production low (Locke 2013). security and safety, and provides few to no labor rights To bridge the gap in OSH oversight, nongovernmen- (Locke 2013). tal organizations (NGOs)—privately owned labor- Additionally, there is little to no professional OSH watch groups—have emerged as so-called regulators of oversight to ensure the safety of workers, especially the working conditions of factories by articulating Occupation and Risk for Injuries 109 international expectations for OSH conditions, wages, electronic products spanning the “6 Cs”: computers, and gender equality. These organizations conduct ran- communications equipment, consumer products, car dom audits of factories and publish reports on their parts, content (e-book readers), and health care products performance. (Ngai and Chan 2012). Today, 75 percent of computer NGOs have been unable to regulate suppliers or the products are manufactured by contract manufacturers focal company itself. Monitoring alone has had margin- as opposed to original equipment manufacturers (Brown ally small results, according to a case study by Locke and 2009). By 2000, the most successful of these companies Romis (2007) that analyzed audits from more than had production facilities in as many as seven countries, 800 factories in Nike’s 51-country supply chain. mostly LMICs (Locke 2013). A newer development in global supply chain regula- The magnitude of employment that these contract tion is the notion of corporate social responsibility; manufacturers manage can be illustrated by Foxconn: the focal company has a direct responsibility to protect in 2013, it employed 1.4 million workers in China alone the interests of society by upholding OSH and labor (Chan, Pun, and Selden 2013). The common structure equality standards throughout its supply chain. Many of a factory in the global supply chain is a large assem- multinational corporations, such as Adidas, Apple, bly line consisting of hundreds of workers performing a Gap, Nike, and Walmart, have defined criteria that single 20- to 30-second operation repetitively until their their suppliers must meet and then conducted factory shift ends (Sandoval and Bjurling 2014). This level audits as part of a yearly report analyzing current labor of repetition is mentally taxing (Butollo, Kusch, and conditions in their supply chain (Apple 2014; Burke, Laufer 2009) and is accompanied by reports of suicides Scheuer, and Meredith 2007; Gap 2012; Verbeek and attempted suicides (Ngai and Chan 2012). The and Ivanov 2013; Yu 2007). Little research has been common OSH risks in the electronics manufacturing conducted on the effectiveness of these initiatives industry include fatigue resulting from long work shifts (Ustailieva, Eeckelaery, and Nunes 2012). The paradox- and physically demanding work with very few or no ical demand for both high OSH standards and low-cost breaks and no proper safety equipment (Sandoval and labor creates conflicting incentives for recording Bjurling 2014). occupational injuries (Brown 2007). Serious occupational injuries are common in elec- tronics factories. According to a study that gathered 500 audit reports from 276 factory suppliers of Hewlett- Common OSH Risks Packard from June 2004 to January 2009, 59 percent of The distinctive economic incentives and pressures the factories were in violation of legal working hours, operating in the global supply chain give rise to numerous 40 percent were in violation of emergency preparation, OSH risks. Although the specific risks are not unique 32 percent were in violation of hazardous material to the global supply chain, the combination of hazards storage, and 22 percent were in violation of occupa- presents particular challenges. To examine the OSH risks tional safety (Locke 2013). Of these suppliers’ workers, engendered by the global supply chain, this section reviews 95 percent were performing repetitive tasks while the common sources of risk in the microelectronics and standing in an assembly line, and most were female textile goods workplaces, prominent industries using the migrants (Locke 2013). A similar study published by a supply chain of LMICs. coalition of NGOs, referred to as ProcureITfair, found that workers for the Excelsior Electronics plant in Electronics Industry Dongguan, China—a computer and digital electronics The electronics industry is one of the fastest-growing manufacturing facility for Apple, Intel, and Sony sectors in the world, with a vast network of suppliers components at the time of investigation—were working (Locke 2013). Beginning in the 1980s, prominent multi- for 10–12 hours on a poorly ventilated shop floor and national firms in Canada and the United States, such as inhaling industrial alcohol, cleaning agents and thinners Apple, IBM, Lucent, Maxtor, and 3Com, gradually at the printed circuit board processing area (Butollo, adapted their supply chain management to a new form Kusch, and Laufer 2009; ProcureITfair 2008). According of outsourcing that granted licenses to suppliers from to statistics gathered from Shenzhen factories that were factories across the globe, prominently China, Malaysia, released by mainland authorities, an average of 13 Mexico, and Singapore (Locke 2013). The contract workers lose a finger or an arm daily, and 1 worker per- manufacturers that received the most business were ishes onsite every 4.5 days (Murdoch and Gould 2004). Flextronics, Hon Hai/Foxconn Technology Group, and Another serious issue is the unsafe handling of Jabil, all of which have factories in LMICs around the flammable materials, leading to deadly factory fires. On world (Locke 2013). Many of these companies generate Foxconn’s campus in Chengdu, three workers perished 110 Injury Prevention and Environmental Health in a polishing department fire (Sandoval and Bjurling textile products, into the global supply chain, export- 2014). In Zhejiang Province, a factory fire killed five ing more than US$5 billion worth of products workers (Murdoch and Gould 2004). each year (Akhter and others 2010; Rahman 2004). In 2009–10, Bangladesh’s knitted and plain-weave Textile, Clothing, and Footwear Industry garment industries grew 46 and 40 percent, respec- On April 24, 2013, the Rana Plaza factory building tively, as a direct result of rising production costs in collapsed in Dhaka, Bangladesh, killing 1,129 workers China for brands such as H&M and Walmart (Jun and and injuring more than 2,000, making it the deadliest others 2012). industrial disaster to date (Adler-Milstein, Champagne, A recent study gathered data from audit reports of and Haas 2014). Just five months before the Rana Plaza 210 factories supplying a major global apparel firm that disaster, 112 workers were killed in a factory fire in span Bangladesh, China, the Dominican Republic, Dhaka at Tazreen Fashions, an apparel supplier for Honduras, and India (Locke 2013). The study revealed Disney, ENYCE, Sean Combs, Sears/Kmart, and that the apparel firm’s compliance program, regarded by Walmart (Adler-Milstein, Champagne, and Haas 2014). private regulatory programs as the most effective in the Between 1990 and 2010, 33 major fires occurred in gar- industry, had an overall compliance rate of 51 percent. ment factories, as well as 200 smaller fires in Bangladesh The criteria for compliance covered compensation, alone, injuring more than 5,000 workers (Brown 2010). working conditions, and overtime hours. The worst These textile factories—stationed in poorly structured compliance was found in the factories in South and East high-rise buildings—lack safety exits and proper elec- Asia, where 56 and 72 percent of the factories, respec- trical wiring (Adler-Milstein, Champagne, and Haas tively, were not approved. 2014). Table 6.5 shows a select number of fire accidents China has the strongest presence in the supply chain of that occurred in Bangladeshi garment factories from the footwear industry; 86 percent of all footwear sold in 2000 to 2006. the United States comes from factories in Southern China The ready-made garment sector in Bangladesh (Locke 2013). Guangdong is the hub of the athletic foot- accounts for 78 percent of the country’s export wear industry because of its well-designed ports, access earnings (Ahamed 2013). Since liberalizing its econ- to large numbers of cheap laborers, and lack of govern- omy in the 1980s, Bangladesh has integrated its ready- ment regulation (Frenkel 2001). According to recent made garment industry, which includes mass-produced publications, work in the Chinese shoe manufacturing Table 6.5 Selected Fire Accidents in Garment Factories in Bangladesh, 2000–06 Date Place Number killed Number injured Cause of fire Cause of death February 23, 2006 KTS Textiles, Chittagong 91 400 Electric short circuit Only exit locked; fire, suffocation, stampede March 6, 2006 Industry, Gazipur 3 — Fire panic Only exit blocked by boxes, smoke, stampede March 2006 Salem Fashion Wear Ltd. 3 50 Unknown Stampede May 3, 2004 Misco Super Market, 9 50 False fire alarm Stampede Dhaka August 1, 2001 Kafrul 26 76 Unknown Smoke, stampede August 8, 2001 Mico Sweater Ltd., 28 100 Unknown Single exit locked Mirpur 2000 Near the capital 48 70 Burst boiler Trapped in locked, burning building 2000 Chowdhury Knitwear, 53 100 Short circuit Fire, smoke, stampede Norshingdi August 28, 2000 A garment, Banani 12 45 Unknown Suffocation, stampede Source: Akhter and others 2010. Note: — = not available. Occupation and Risk for Injuries 111 industry is fraught with excessive overtime, managerial controls—elimination, engineering, administration, neglect of OSH conditions, and sexual harassment or personal protection (in order of decreasing (Locke 2013). effectiveness)—may not be feasible except perhaps in Much like trends in the electronics industry, large, well-resourced enterprises. Very few studies have contracted temporary work in the textile, clothing, and evaluated the effectiveness of technical measures in footwear industries has been growing, leading to danger- LMICs. In India, the use of protective eye equipment ous OSH conditions for workers (Nossar, Johnstone, and reduced the incidence of eye injury among agricultural Quinlan 2003). In a study examining the comparative workers (Chatterjee and others 2012), while the installa- dangers of contingent work in the clothing and manu- tion of mirrors above tandoor ovens showed potential facturing industries (Mayhew and Quinlan 1999), for reducing burns among oven operators in Pakistan contracted employees in the clothing industry had three (Nasrullah and Awan 2012). These remain very isolated times the number of occupational injuries as did con- islands of evidence. tracted workers in the manufacturing sector. One possi- ble explanation for this disparity in injury experience is that garment workers are often paid by an incentive Injury Care and Return to Work system that pushes them to work faster than their Prehospital, hospital, and ambulatory care for occupa- manufacturing counterparts, who are paid by the hour, tional injury in some LMICs is part of the general and increases the risk for injury. capacity for trauma care. Although basic health units in the workplace or the community can manage minor trauma, a sophisticated prehospital and hospital trauma OCCUPATIONAL INJURY INTERVENTIONS care system is crucial for saving lives and mitigating the Because working conditions significantly influence effect of severe occupational injuries. Much room exists worker performance and productivity, optimizing the for improvement in such systems in LMICs (Baker and conditions for improved health and safety has far- others 2013; Dunser, Baelani, and Ganbold 2006; Goosen reaching implications for individuals, employers, and and others 2003), but not without resources that could economies globally. Identifying and implementing be beyond reach in such settings. In a nonrandom con- effective health and safety interventions at the policy trol study, Murad, Larsen, and Husum (2012) reported a level and in individual workplaces to foster sustainable lower injury mortality rate among patients managed by and safe work environments are important. However, field-trained first responders than among those not there is no one-size-fits-all strategy for reducing the managed by first responders. Applying a similar approach risk for occupational injury. LMICs are especially in workplaces or communities may be an affordable diverse in their type and amount of resources, strength alternative for improving outcomes, particularly in the of their regulatory institutions, industrial profile, and informal sector. levels of informality, among other relevant features. Access to rehabilitation services following injury Therefore, the range of viable options for a country in cannot be dissociated from the general problem of lim- Sub-Saharan Africa, for example, may not be the same ited access to quality health services in LMICs. Similar as that for a country in Latin America. Nonetheless, to emergency medical services, there is much room considering comprehensive solutions that integrate for building and improving rehabilitation services in multiple strategies for improving not only primary LMICs (Haig and others 2009; Tinney and others 2007). prevention, but also injury care, rehabilitation, work- Data on the duration of disability after injury in LMICs force training, and data systems is important. are very scarce, and such data are critically needed for an evaluation of initiatives that aim to minimize disability and enhance early return to work. Primary Prevention Given the challenges in LMICs, prevention of occupa- tional injury should consider two distinct but related Capacity Building and Retention questions: What is known to work? How can it be Scaling up training programs to develop a competent applied successfully and sustainably? These questions occupational health workforce, including primary health can be rephrased as technical measure effectiveness care workers, needs to be coupled with simultaneous versus implementation or program effectiveness. and serious retention efforts. The disproportionate con- Because of resource constraints in LMICs, applying centration of the health workforce in urban areas is a the more effective, but also more expensive, technical global phenomenon (Chen and others 2004), but it is measures from the top of the hierarchy of hazard accentuated in LMICs by the rural-to-urban influx as a 112 Injury Prevention and Environmental Health result of poor investment in rural development. The most surveillance performed better than passive surveillance, common reasons for “brain drain” are better remunera- such an approach could be expensive and difficult to tion, safer environment, and better living conditions in monitor and sustain. As a middle ground, passive surveil- the receiving country and lack of facilities in the sending lance that is enhanced with data on place and activity country. Adjusting training to local needs—for example, during injury, supplemented with active surveillance in by providing enough training to enable locals to serve high-risk settings, has been suggested. their own populations—coupled with efforts to improve Determining whether prevention programs that rely working and living conditions, may curb outmigration on participatory approaches, particularly those overseen in LMICs (WHO 2006a). by primary health care workers, could be a suitable To overcome OSH resource scarcity, models have platform for active surveillance will be useful. Ensuring highlighted the use of intermediary organizations the quality and completeness of data is crucial for the designed to bridge the gaps between public health and success of such approaches. safety organizations and SMEs and to deliver occupa- tional health and safety services (Soares and others 2012). Because SMEs are diverse and often insular, Regulation and Enforcement straightforward information is needed regarding OSH Governments in HICs protect workers against health initiatives that can offer specific, tangible benefits and and safety risks through OSH legislation, regulation, and be readily adapted to their organizational structure. enforcement via workplace inspections that may result Through regular interactions with individual SMEs, in citations and penalties. These approaches are often intermediary organizations may offer the best opportu- considered the cornerstone of workplace safety and nity to influence OSH decision making, providing health risk management (Mischke and others 2013; short-, medium-, and long-range benefits (Gervais and Tompa, Trevithick, and McLeod 2007). However, these others 2009). policies are lacking or inconsistently applied in many parts of the developing world. Studies examining the effect of OSH inspections, Data for Planning, Monitoring, and Evaluation citations, and penalties have shown varying results Effective and targeted prevention efforts are impossible (Foley and others 2012; Friedman and Forst 2007; without viable local data. A recent review of an audit of Levine, Toffel, and Johnson 2012). Some evidence shows suppliers to Apple revealed that fewer than one in seven that inspections resulting in penalties are associated with recorded any injury or health events in the past year lower rates of lost workday injuries in the years immedi- (Apramian and Cullen 2015). The ILO guidelines for ately following inspections (Gray and Mendeloff 2005). improving national reporting of occupational injuries Other evidence suggests that penalty inspections extend acknowledge the challenges of expanding reporting their influence beyond the injuries closely related to the to cover small enterprises, migrant workers, the self- specific regulations for which citations and penalties employed, and the informal sector (Ehnes 2012). Among were levied (Gray and Mendeloff 2002; Mendeloff and the recommended solutions are legalizing migrant work Gray 2005); that is, penalties may prompt employers to and creating administrative connections through which enhance general safety efforts and to respond to cited small enterprises are obliged to report to a national data- deficiencies (Haviland and others 2010). Verbeek and base, for example, in the same way they report informa- Ivanov (2013) appraised systematic evidence for effec- tion for tax purposes or social insurance. tiveness of basic OSH interventions in settings similar to However, more pragmatic solutions may be needed those prevailing in LMICs and found that enforcement for these hard-to-reach groups. One possible solution is of regulations reduced injury rates. ensuring that occupational injury modules are part of In the most comprehensive report to date, a Cochrane periodic household or establishment surveys (Taswell review assessed evidence on the enforcement of OSH and Wingfield-Digby 2008). Enhancing routine health regulations and the prevention of occupational diseases information data with identifiers of the relation to work and injuries. Mischke and others (2013) found that and the occupation and industry of the injured person inspections likely reduce the risk for injury in the long is a promising approach. Marucci-Wellman and others term, although the magnitude of effect remains unclear. (2013) tested an active surveillance system that builds Further, focused inspections appear to have greater effect on the health information system in one commune in than more general inspections, although the current evi- Vietnam and compared its outcomes with those from a dence is low quality. Unfortunately, because inspections range of unenhanced and enhanced passive surveillance are costly and resources are limited, the enforcement or models also based on the existing system. Although active threat of enforcement fails to reach all workplaces equally. Occupation and Risk for Injuries 113 Additionally, the changing political, economic, and legal positive effects on worker safety and health behaviors, landscape of work is creating new stressors and potential but insufficient evidence that training alone improves hazards with consequences that not yet understood. health or safety outcomes (Amick and others 2010; Although regulations and enforcement activities are Robson and others 2012). A review of evidence on the designed to protect worker safety and health, employer effectiveness of OSH interventions in agriculture, obligations have not yet been fully realized (Niskanen SMEs, and informal sector settings did not find and others 2010). Given the limited number and unequal educational interventions to be effective in reducing distribution of labor inspectors worldwide, more effective injury risk (Verbeek and Ivanov 2013). However, worker mechanisms are needed to translate OSH regulations perceptions of safety training may positively affect into widespread practice. safety by increasing worker recognition of potential Achieving sustainably safe work environments within risks, thereby enhancing workers’ ability to identify the organizational structure of supply chains will require near misses and increasing the likelihood that they both private voluntary and public mandatory regulation. will report injuries at all levels of severity (Lauver and The involvement of local government can be crucial to Lester 2007). Contemporary learning theory suggests upholding proper labor standards and, historically, it has that incorporating structured dialogue and action- underused its own capacity to impose worker standards focused reflection into OSH training may enhance the on foreign investors in global supply chains (Amengual effect of training on workers’ engagement in safe work 2011). However, the fluid, fast-evolving structure of inter- behaviors and confidence in their ability to handle national supply chains means that static governmental unanticipated events safely (Burke and others 2006; law alone cannot sufficiently protect worker rights and Burke, Scheuer, and Meredith 2007). health. Given the increasingly diverse workforce and demo- Rather, a joint effort with NGOs and government graphic disparities in injury risk, OSH training programs intervention is necessary. An example of such joint need to target workers with language barriers and low regulation involves the protection of the rights of work- literacy and incorporate cultural and societal aspects to ers in maquiladoras—Mexican manufacturing plants be effective. Failing to address these aspects can deepen that operate in a free-trade zone (Locke 2013). The NGO the OSH disparities (Steege and others 2014). CEREAL (Centro de Reflexión y Acción Laboral, or Centre for Reflection and Action on Labour Issues), the Guadalajara Chamber of Commerce, and electronics Safety Climate and Safety Culture suppliers in this zone developed a dispute system known A general belief holds that management commitment as the Accord to handle issues regarding worker com- plays a fundamental role in developing a strong safety pensation and benefits. Workers file complaints within climate and culture and that strong management com- the courts of the Accord, and cases are resolved directly mitment to and support of safety enhance employee with the factory’s human resources department, bypass- adherence to safe work practices and ultimately reduce ing the slow, often ineffective, Mexican judicial system workplace injuries. (Locke 2013). Before the Accord, workers were subject to However, evidence showing a direct link between government neglect largely because of a weak union safety culture and climate and injury outcomes is lim- presence in maquiladoras and a general lack of under- ited. Having a strong safety culture may have a positive standing of Mexican law. Since the advent of the Accord effect on workers’ use of safe behaviors, injury and system, workers can file labor violation cases in court illness rates, or reporting of injuries and illnesses, but and have them resolved within a few months. the evidence is mixed or inconclusive (GAO 2012). A meta-analysis found support for an association between safety climate and safety performance, but also Worker Training found a much weaker link between safety climate and Training workers as well as managers in hazard recogni- injury (Clarke 2006). tion and control, safe work practices to reduce risk, Subsequent work examined the effect of government proper use of personal protective equipment, safety and subsidies designed to improve occupational safety by health information, and emergency procedures is a changing safety culture. Research found that only half of widely recognized, essential component of OSH pro- the subsidized interventions evaluated were deemed grams (Burke, Scheuer, and Meredith 2007; Redinger and successful in improving reporting of hazards, reducing Levine 1998). unsafe behaviors, or reducing accidents, indicating the Systematic reviews of research pertaining to the challenge of promoting organizational culture change effectiveness of OSH training found strong evidence for (Hale and others 2010). 114 Injury Prevention and Environmental Health Factors associated with successfully improving safety participatory ergonomics may be an effective strategy for culture included a planned, systematic approach that identifying and addressing workplace biomechanical, generates sufficient energy and support for deployment psychophysical, and psychosocial risk factors of multiple safety interventions, engagement and empow- (Niu 2010); reducing injury and MSD risk (Cantley and erment of workers in the learning and change process, others 2014); maximizing the involvement of workplace and training and motivation of managers at all levels stakeholders; improving production, worker perceptions, (Hale and others 2010). Evidence from the restaurant worker morale, and job satisfaction (Dennerlein and industry suggests that employees’ perceptions of manage- others 2012; Vink, Koningsveld, and Molenbroek 2006); ment’s commitment to safety and safety training are and embedding ergonomics within organizational pro- separate dimensions of the work environment, the former cesses (Driessen and others 2011; Pehkonen and others a proximal predictor of future injury and the latter a more 2009; Törnström and others 2008). distal predictor (Huang and others 2012). However, critical prerequisites for a successful ergo- nomics program are a well-established system for identifying and assessing risk factors and implementing Safety Incentive Programs solutions; communicating effectively with workers and Safety incentive programs are popular and widespread, management; and actively engaging workers, manage- yet little research has been done regarding their effect on ment, and technical personnel in the ergonomics process the occurrence and reporting of injuries (GAO 2012). (Broberg, Andersen, and Seim 2011; Niskanen and Such rate-based and behavior-based incentive programs others 2010; Pehkonen and others 2009; Zink, Steimle, are intended to entice workers to work safely, but safety and Schröder 2008). incentive programs may discourage injury reporting. Evidence showing a positive effect of ergonomic Studies evaluating incentive programs have reported hazard control on reducing injury risk across groups varying conclusions about their effect on workplace of workers and jobs has been rather limited and con- safety. Some have reported that rate-based safety incen- flicting (Fujishiro and others 2005; McSweeney and tive programs have no effect on injury reporting (Brown others 2002; Palmer and others 2012). However, several and others 2005). Others have concluded that safety studies have reported an association between reduced incentive programs reduce injuries (Alavosius and others risk for MSDs and acute injuries associated with 2009; Gangwar and Goodrum 2005), and still others have manual handling and ergonomic job modification shown that workers whose employers enact policies (Carrivick and others 2005; Marras and others 2000; involving discipline as a consequence of injury are less van der Molen and others 2005). A recent report illus- likely to report injuries for fear of punishment than trated the benefits of identifying ergonomic hazards are those whose workplaces have no such programs and controlling risk for any type of acute injury or (Lipscomb and others 2013). The bulk of evidence is MSD among a population of manufacturing workers, equivocal. These discrepancies may be due, in part, to the and risk was reduced further with each hazard control widely varying components of safety incentive programs. implemented (Cantley and others 2014). Furthermore, Some offer incentives for reporting near-miss incidents, the application of an ergonomics process for identify- reporting other safety concerns, or wearing protective ing and mitigating organizational and psychosocial equipment, and others reward work groups for having demands that contribute to both injury and MSD risk fewer injuries. at work has been the subject of some recent research Behavioral interventions such as monetary incentives, (Bentley 2009). praise and feedback, and team competition may reduce The body of scientific evidence supports the financial injuries in settings similar to those prevailing in LMICs, case for ergonomic programs. Ergonomic programs but the evidence is limited (Verbeek and Ivanov 2013). have been shown to be cost-effective, particularly in Effective implementation of behavioral interventions manufacturing, and ergonomics best practices focus on may require a greater degree of organizational regulation integrated approaches to hazard control rather than on than currently exists in LMICs. specific ergonomic tools and procedures (Amick and others 2009). Ergonomic Interventions Because ergonomic hazards vary markedly between Other Participatory Approaches industries and jobs within specific industries, the optimal Some countries in Asia are increasingly using action- means to mitigate those hazards likewise vary. Despite oriented participatory approaches to deliver OSH this variation, ever-growing evidence suggests that interventions, particularly in difficult-to-reach or Occupation and Risk for Injuries 115 difficult-to-regulate settings such as small enterprises government officials, farmers were trained to use and the informal economy (Kawakami 2007). Also illustrated checklists to propagate examples of good widely deployed in HICs, this approach involves the practice among established networks of their peers. target population in identifying hazards and developing The approach could potentially be used to address and implementing safety interventions (figure 6.2), thus child labor in hazardous agricultural work (ILO 2012d). ensuring a more appropriate fit between an intervention The program has expanded since the launch of and a particular workplace setting. Interventions using a Vietnam’s first national OSH program in 2006. participatory approach could even help eliminate A few studies have employed an uncontrolled pre- hazards and substitute them with appropriate, low-cost, and post intervention design to test the effectiveness of and safe alternatives. This approach promotes owner- these approaches. In Thailand, reductions in toluene and ship, improving the potential for compliance and carbon monoxide levels were recorded following the sustainability. application of participatory training in the informal In Cambodia, the Work Improvement in Safe Home sectors of artificial flower making and batik processing program focused on home workers and small businesses. (Manothum and others 2009). Similar benefits were OSH trainers mobilized by government, worker, and observed after applying the same approach in the infor- employer organizations assisted participants in identify- mal weaving, ceramic, and blanket-making industries ing practical safety solutions using a simple action (Manothum and Rukijkanpanich 2010). Knowledge, checklist (Kawakami and others 2011). attitude toward occupational safety, and use or provision In a slightly different version in Thailand, primary of personal protective equipment were improved after care unit (national hospital system) staff members participatory training of 525 welding workers in 25 were retrained as OSH service providers, assessing SMEs in China. However, improvements in implementa- OSH risk and giving low-cost improvement advice tion of engineering controls were inadequate (Fu and through participatory group discussions (Kawakami others 2013). 2007). In Vietnam, the Worker Improvement in Neighborhood Development program has been exten- sively applied in agriculture. Supported by provincial Prevention of Falls from Heights and Slips, Trips, and Falls Falls from heights are a serious hazard for many workers, especially in the construction sector. Same-level STFs Figure 6.2 Steps for Promoting Participatory OSH Training in pose a substantial hazard for workers in nearly every Workplaces in the Informal Economy sector, but especially in health care, food services, and wholesale and retail trade. Step 1 Identifying local partners Much has been learned about the causes and preven- tion of injuries resulting from these hazards, and resources have been developed to assist employers and workers in recognizing and controlling them. However, Step 2 Collecting local good examples barriers continue to limit the dissemination of knowl- edge and use of interventions in the field. For example, gaps in knowledge persist regarding how to use fall Step 3 Organizing pilot training protection measures (Committee to Review the NIOSH Construction Research Program 2008). Management systems designed to ensure the use of Step 4 Training local trainers fall protection measures have been shown to reduce falls among construction contractors (Becker and others 2001). Implementing comprehensive STF pre- Step 5 Training by local trainers vention programs that include analysis of common causes of STFs, general awareness campaigns, work- place hazard assessments, changes in housekeeping products and procedures, consistent removal of ice Step 6 Sustaining networks through follow-up actions and snow, changes in flooring, and provision of slip-resistant footwear for high-risk employees can Source: Kawakami 2007. substantially reduce the risk for STF injuries among Note: OSH = occupational safety and health. hospital workers (Bell and others 2008). These results 116 Injury Prevention and Environmental Health should be readily transferable to other sectors at strategies combining the provision of safe injection similar risk for STFs. devices, development and circulation of national guide- lines, education campaigns, and training of health care workers (WHO 2011). Sector-Specific Interventions Health Care Construction Worksite Safety Mechanical patient-lifting devices have been shown to A great deal has been learned about engineering solutions reduce the back compressive forces on nursing personnel to address major safety issues, including fall protection by approximately 60 percent, reduce the lifting required equipment, nail gun safety, and protection from contact during patient transfers, and improve patient perceptions with overhead power lines. However, overcoming of comfort (Garg and Owen 1992; Zhuang and others the barriers to knowledge dissemination is needed to 2000). Working with national and international research- facilitate wider recognition of hazards and implementa- ers, the National Institute for Occupational Safety and tion of available solutions (Committee to Review the Health in the United States conducted research into NIOSH Construction Research Program 2008). Efforts comprehensive safe patient-handling policies involving to increase awareness of construction worksite hazards the use of mechanical lifts and repositioning aids, a zero- should target vulnerable work groups, including immi- lift policy, and employee training on the use of lifting grant workers, young workers, and contract workers, devices. This approach was highly effective in reducing who are at substantially higher risk for injuries and fatal- back injury risk among health care workers of all ages ities in this sector. and lengths of work experience, regardless of the type of Little research has systematically examined the effect facility (Collins and others 2004). Additionally, clinical of safety culture and OSH management systems on use and scientific studies have documented the effective- reducing injury and improving working conditions in ness of blunt-tip suture needles in reducing needlestick the construction industry. However, evidence suggests injuries, and multiple resources have been developed to that when owners, contractors, contractor associations, disseminate this information to health care workers insurance carriers, and appropriate unions collaborate (CDC 2010; NIOSH 2007). to establish and promote a safety culture, the risk for Very few controlled studies demonstrate the effective- injury and fatalities can be reduced. ness of different measures in preventing sharps inju- To enhance the uptake of safety training among the ries in LMICs. Cross-sectional comparative effectiveness many immigrant workers in this sector, using training analysis conducted in the Alexandria University hospitals workers from representative cultural groups to deliver in the Arab Republic of Egypt found that factors such safety training may be more effective than using profes- as access to safe injection devices, adherence to infection sional trainers. To promote effective integration of control guidelines, access to written protocols on prompt health and safety management into construction project reporting, and training on safe injection practice were planning, communication, and control, a project funded associated with a lower probability of needlestick injury by the U.K. Health and Safety Executive developed (Hanafi and others 2011). several integrated tools for use. These include a respon- Studies on the effectiveness of educational programs sibility chart, an option evaluation chart, health and elicited inconsistent results. In contrast to an educational safety hazard workshops, drawings presenting safety program, an official imperative program with monetary information, red-amber-green lists, health and safety penalty for unsafe practice reduced syringe recapping milestones, and a process for controlling design changes among nurses in the Islamic Republic of Iran (Dianati (Cameron and Hare 2008; Hale and others 2010). and others 2012). A controlled study in China and an uncontrolled study in Taiwan, China, among student Agriculture and intern nurses showed improved practices and Various safety interventions have been implemented reporting of sharps injuries following education inter- within the agriculture sector in HICs, but their effect ventions (Wang and others 2003; Yang and others 2007). remains poorly understood. Farm safety hazards in In Cairo University hospitals, education combined with LMICs, where the vast majority of agricultural workers improved access to safe injection devices in intensive are employed, have received much less attention. care units was associated with a reduction in the inci- Results from a meta-analysis found no evidence that dence of sharps injuries (Zawilla and Ahmed 2013). educational interventions were effective at reducing Countries such as the Syrian Arab Republic also report injury risk, which is consistent with evidence suggesting improvements in injection practice and reductions in that training alone is insufficient to prevent injuries. In sharps injuries after implementation of multifaceted contrast to earlier reports, limited evidence was found Occupation and Risk for Injuries 117 that rollover protective structures on tractors reduced understanding of strategies to mitigate its causes and injuries (Lehtola and others 2008). However, the dearth adverse health and safety effects (Noy and others 2011; of high-quality studies illustrates the continued need to Williamson and others 2011). Because both work and develop and evaluate farm injury interventions, particu- nonwork activities can greatly affect fatigue, fatigue risk larly in LMICs. management requires the involvement of employers and National coordination of efforts across agriculture workers alike. sector stakeholders is under development in several Regulatory limits on hours worked are evolving into countries, including New Zealand, Sweden, and the multidimensional fatigue risk management systems that United States (Lundqvist and Svennefelt 2012, 2014). In incorporate additional risk mitigation strategies, such as addition, initiatives to address injuries and fatalities education and training; regular monitoring of fatigue associated with tractors and other farm machinery have levels; and systematic assessment of the role of fatigue in been developed, including an electronic system to warn accidents and injuries in many sectors, including trans- farm equipment drivers when someone enters the portation and health care (Gander and others 2011). hazardous zone near their vehicle (EU-OSHA n.d., “Case Development of effective and comprehensive fatigue risk Studies”). management systems will require enhanced understand- Pesticide poisoning is a major hazard in the agriculture ing of the complex relationship between fatigue and sector in LMICs. Occupational exposure causes uninten- safety, because risk may be greatest at intermediate levels tional harm, while easy access is a key factor in intentional of fatigue, when operators may be less attentive (Folkard self-harm. In accordance with expert opinion, the WHO and Åkerstedt 2004). recommends a range of promising interventions regard- ing the safe storage of pesticides in households, commu- Workplace Violence nal storage, training of farm workers and schoolchildren, International commitment to preventing workplace training of pesticide vendors, and use of community violence is mirrored in the ILO’s code of practice for leaders to disseminate information on safe handling preventing violence in the services sector (ILO 2003). of pesticides (WHO 2006b). The effectiveness of these Most of the proposed interventions apply to service measures is yet to be demonstrated. organizations such as health care institutions. Evidence Social marketing—the use of commercial marketing from areas such as government policy and strategy strategies to raise awareness and promote safe behavior— development could apply to other sectors where risks are has been advocated for tackling unsafe behaviors among not clearly defined, such as street vending and domestic young people (Lavack and others 2008; Monaghan and work. However, awareness-raising and prevention initia- others 2008; Smith 2006). Although they are difficult to tives still need to be part of participatory occupational evaluate, social marketing campaigns have been associ- health approaches that target all sectors. ated with a reduction in occupational injuries in Germany (Mustard 2007). The effectiveness of such campaigns in LMIC settings remains to be seen, but the Psychosocial Risk Management campaigns are probably suitable as components of par- The association between baseline psychosocial work- ticipatory programs. place stressors and ill health and injury was documented recently among construction and municipal utility Machine Safety workers before the performance of a random control Worker contact with machinery or equipment presents a trial of a workplace safety intervention (Bodner and major risk for severe injury, particularly in the manufac- others 2014). The results from this trial will be useful for turing and mining sectors. Some recent risk mitigation informing future interventions. efforts include the application of capacitance-sensing An analysis of data from the European Survey of technology (Powers Jr., Anmons, and Brand 2009) and Enterprises on New and Emerging Risks identified six the use of intelligent video technology to detect worker aspects associated with psychosocial risk management. presence in hazardous locations near machinery. Establishment size, industry, and country predicted the Although these efforts show promising results for pre- degree of psychosocial risk management. Larger enter- venting future injuries (Ruff 2010), they are less readily prises were associated with better psychosocial risk transferable to LMICs. management. Manufacturing and construction sectors were associated with the least comprehensive psychoso- Fatigue cial risk management, while the education, health, and Although fatigue is widely accepted as a public and social work sectors were associated with the most com- workplace safety concern, critical gaps exist in the prehensive. Countries in Northern Europe offered 118 Injury Prevention and Environmental Health more psychosocial risk management than did countries Implementation of a participatory ergonomics pro- in Southern and Eastern Europe (van Stolk and others cess consisting of both proactive and reactive components 2012). Although room exists for improvement globally, yielded a benefit-to-cost ratio of 10.6 for an auto parts this trend toward taking systemic approaches to manufacturer; although the number of worker compen- managing workplace psychosocial risks in Europe is sation claims was not reduced, the duration of claims encouraging. was shortened, suggesting a reduction in injury severity (Tompa, Dolinschi, and Laing 2009). Similarly, economic evaluation of a participatory ergonomics process imple- Diffusion of OSH Interventions to SMEs mented in a clothing manufacturing plant, which Much effort has been focused on developing OSH introduced primarily low-cost, low-tech interventions, management systems for large enterprises, but similar showed a benefit-to-cost ratio of 5.5; significant reduc- development, implementation, and evaluation have been tions in first aid incidents, modified-duty episodes, lacking for smaller enterprises (Hale and others 2010; and short- and longer-term sickness absences; and Robson and others 2007). Given the barriers to dissemi- improvements in efficiency and product quality (Tompa, nating OSH information to SMEs across many employ- Dolinschi, and Natale 2013). ment sectors, smaller businesses may require assistance Economic evaluations of OSH interventions in SMEs from external organizations and other sources to fulfill are difficult to find, although SMEs represent a significant their responsibility to protect the health and safety of force in the global economy and the cost of injuries to their workers. these establishments could be catastrophic (Targoutzidi To address this concern, several EU member states and others 2013). Economic evaluations may be lacking, have launched programs to support OSH in SMEs. in part, because smaller enterprises rarely have a separate Adequate risk assessment is a key component of OSH, but budget for OSH and routinely collect only very limited it can be costly and difficult for small businesses that lack OSH data, which makes complete economic analyses resources for the task. To address this problem, EU-OSHA quite a challenge (Lahiri, Gold, and Levenstein 2005). In introduced the Online Interactive Risk Assessment light of research suggesting that each near-miss or non- (OiRA) project to encourage and help micro and small injury accident costs at least 654 euros (US$740), the businesses assess their workplace risks via a cost-free Web infrequent collection of information by SMEs on these application designed for ease of use. Sector-specific OiRA incidents suggests inadequate diffusion of information, tools are being developed (EU-OSHA n.d., “Safeguarding making the business case for OSH interventions and Europe’s Micro and Small Enterprises”). another gap to be filled (Binch and Bell 2007; Targoutzidi and others 2013). To identify and address factors relevant to SMEs and encourage investment in OSH interventions, Cost-Effectiveness of OSH Interventions Targoutzidi and others (2013) recently developed several Although some evidence highlighting the economic case study examples as a tool for SMEs to calculate the benefits of OSH interventions has emerged in recent costs and consequences of OSH interventions. years, substantial gaps remain. Existing analyses focus If one recognizes that the economic benefits derived primarily on benefits achieved via ergonomic interven- from OSH interventions may extend beyond traditional tions in four sectors: administrative and support outcomes to include improved profitability and worker services, health care, manufacturing and warehous- engagement, then considering a range of outcomes ing, and transportation. Strong evidence supports the when evaluating OSH interventions is critical to ensure financial merits of ergonomic interventions for that the true benefits of interventions are understood the manufacturing and warehousing sector, moderate (Verbeek, Pulliainen, and Kankaanpää 2009). SMEs often evidence was found for the administrative and sup- perceive that performing work more safely would be port services and health care sectors, and limited evi- costly. For this reason, highlighting the business case for dence was found for the transportation sector OSH could facilitate the broader diffusion of OSH inter- (Guimaraes, Ribeiro, and Renner 2012; Tompa and ventions globally. others 2010). Studies have found that ergonomic and other musculoskeletal injury prevention interventions OCCUPATION AND RISK FOR ACUTE AND in manufacturing and warehousing environments are cost-effective and improve health and safety outcomes, CHRONIC MEDICAL DISEASES and paybacks are realized over a time period ranging There is a vast literature on the contributions that from just over three months to slightly more than two physical, chemical, and biologic exposures at work may years (Tompa and others 2009). make to the occurrence of acute and chronic medical Occupation and Risk for Injuries 119 conditions (Rosenstock and others 2005). Indeed, almost • Chemical hazards include pesticides, organic solvents, every pathological condition has one or more possible heavy metals, fibrogenic dusts, and plastics and other occupational causes. Some conditions may have myriad reactive chemicals. causes, although deciphering the role of workplace expo- • Pesticides include several classes of intentionally sures in each case is problematic. For others, the conse- biocidal materials. Their poor control is an established quences of workplace exposures are unique, such as the cause of substantial physiologic effect on persons poisonings caused by lead, mercury, arsenic, and other regularly exposed and may cause permanent neuro- heavy metals; the pneumoconioses caused by coal, silica, logical and other effects. or asbestos; or the rare syndromes caused by dioxins and • Organic solvents remove oils and grease and dissolve vinyl chloride. These latter causes can, at least in theory, chemicals. Ubiquitous in every metal-working be diagnosed individually. or plastics operation, they are potent liver toxins. Other consequences present as very common, such as Isolated members of the group can cause leukemia chronic obstructive lung disease caused and exacerbated (benzene), severe neuropathy (n-hexane), and heart by exposure to irritant dusts, coronary artery disease disease (carbon disulfide). caused by exposure to particulate matter or workplace • Heavy metals include both naturally occurring stressors, or the more common forms of cancer caused (arsenic) and human-extracted (lead, mercury) by exposure to several dozen common carcinogens. In materials. Many of these materials are extremely individuals with these common ailments, the contribu- toxic, and human epidemics associated with exposure tion of work can be assessed only by looking at patterns are well described (Dorne and others 2011; Jaishankar of disease in the population in aggregate. and others 2014). The degree to which modest expo- However, broad debate continues about the burden of sures contribute to overall morbidity and mortality, chronic disease that may be attributed to work even in including cardiovascular disease, cancer, and renal well-studied HICs, ranging from as little as 1–2 percent diseases, among others, is uncertain. to perhaps as high as 10 percent (Sorensen and others • Fibrogenic dusts include silica—present in every ore 2011). Far fewer data are available on LMICs. Moreover, and stone—coal, and asbestos. Although coal has although acute overexposures to toxins are common- proved less hazardous than the others, millions of place (WHO 2004), their role in morbidity and mortality workers have been exposed to these lethal dusts, which is unlikely to approach that of injury, which is the reason cause lung scarring and cancer. Part of their impor- this vast subject is relegated to a brief overview. tance is the enormous economic scope of mineral mining, refining, and fabrication and the general absence of any acute reaction to exposure, leaving Exposure Classes many with the erroneous impression that the reac- More than 100,000 chemical substances and numerous tions are little more than a nuisance. In the aggregate, biologic hazards are found in workplaces around the these dusts have contributed measurably to the world’s globe. The control of occupational disease has focused burden of chronic lung disease and cancer, despite on three major categories of hazard—physical, biological, progress in control of exposure in most HICs. and chemical: • Plastics and other reactive chemicals, including dyes, explosives, and pharmaceuticals, are becoming • Physical hazards include noise, heat and cold, increasingly problematic in LMICs because many ionizing and nonionizing radiations, and stressors workers move from rural agricultural areas to cities such as vibrations. Although the range of effects of and manufacturing environments. The biologic these factors remains incompletely studied, hearing consequences of exposure to this large and heteroge- loss (noise), cardiovascular disease (noise, heat), and neous class of materials have been studied, but their cancer (ionizing radiation) are the major concerns potential for adverse effects is difficult to ascertain. even in HICs. Although a few members of this class of agents, such • Biologic hazards include both infectious and aller- as formaldehyde, are relatively ubiquitous, most are genic microbes, highly allergenic plant and animal associated with specific processes and tasks. For this products such as latex rubber, and agents with direct reason, population-level risks tend to be modest. biologic activity such as tobacco leaves and some pharmaceuticals. In general, biologic hazards pose significant risks for the small fraction of any workforce Prevention sensitive to them. They probably contribute little to Although the fraction of chronic diseases to which overall disease burden in either HICs or LMICs. workplace factors contribute remains unknown, there is 120 Injury Prevention and Environmental Health compelling reason to believe that some progress on pre- injury control, for which the need is more immediate, vention has been made in HICs. All evidence suggests the risks are well documented and recognized, and the that high-level exposures of the sort historically associ- available information is at least slightly more tractable. ated with acute disease have become far less common in the past four to five decades. This improvement is likely the result of a combination of regulations enforced CONCLUSIONS in all HICs; successful litigation against manufacturers The most effective means to prevent occupational of products known to cause harm, such as asbestos; and injuries globally is far from certain, and universally effec- a far higher level of awareness of workplace hazards by tive intervention strategies are improbable. Nevertheless, employers and employees alike, especially in unionized sufficient evidence exists to recommend widespread sectors of the workforce. Jurisdictions with reporting implementation of several approaches. laws, including the United Kingdom and many U.S. Developing and retaining a competent health care states, have documented the marked decline in these workforce is critical for LMICs. Community-based sentinel cases (Stocks and others 2015). Although expo- initiatives to promote OSH in conjunction with public sure to noise, heat, and fine particulate matter continues health have the potential for broader reach in regions to be very widespread, potentially adding to the still-high with few resources for health care workers. Empowering burden of cardiovascular disease, average exposures to workers to advance change through wider implementa- metals, pesticides, and carcinogens have improved tion of participatory approaches could speed the identi- (Kauppinen and others 2013; Symanski, Kupper, and fication and mitigation of hazards across many regions Rappaport 1998). and sectors. Targeted and effectively delivered training The same cannot yet be said for the situation in most for vulnerable workers, such as young workers, inexperi- LMICs. The picture is clouded by severe impediments to enced workers, immigrant workers, and workers in control, with strong parallels to the issues affecting SMEs, is needed to reduce the burden of occupational injury control: injury among these subgroups of workers worldwide. Efforts to formalize segments of the informal workforce • Because HICs have banned certain harmful materials, are also needed to protect these vulnerable workers. such as asbestos or polychlorinated biphenyl, stock- piles have made their way into countries without such regulations, a cycle affectionately referred to as NOTE chemical “dumping.” • Few professionals are trained in occupational expo- World Bank Income Classifications as of July 2014 are as sure or disease control, and those with training often follows, based on estimates of gross national income (GNI) per capita for 2013: take advantage of the better employment prospects at multinational companies and in HICs. • Low-income countries (LICs) = US$1,045 or less • Few employers and workers receive training. • Middle-income countries (MICs) are subdivided: • Personal protective equipment is either too expensive a) Lower-middle-income = US$1,046 to US$4,125 or unavailable. b) Upper-middle-income (UMICs) = US$4,126 to US$12,745 • Regulations, including social programs such as • High-income countries (HICs) = US$12,746 or more. workers’ compensation, are weak, because govern- ments try to woo foreign investors with a so-called “race to the bottom.” REFERENCES • Even where regulations exist, they are often unen- forceable because of political reasons or the absence Abbas, M. A., L. A. Fiala, A. G. Abdel Rahman, and A. E. of trained inspectors and laboratories, among Fahim. 2010. “Epidemiology of Workplace Violence against others. Nursing Staff in Ismailia Governorate, Egypt.” Journal of the Egyptian Public Health Association 85 (1–2): 29–43. • The absence of unions or laws protecting the rights Adepoju, A. 2005. “Migration in West Africa.” Paper prepared of workers creates difficulty in enforcing safer work for the Policy Analysis and Research Programme of the practices. Global Commission on International Migration, Human Resources Development Centre, Lagos, September. Combined with the dearth of research, the sparseness Aderaw, Z., D. Engdaw, and T. Tadesse. 2011. “Determinants of of strategies for controlling exposure or preventing acute Occupational Injury: A Case Control Study among Textile and chronic diseases is not surprising. For this reason, Factory Workers in Amhara Regional State, Ethiopia.” the discussion combines control of these conditions with Journal of Tropical Medicine 2: 657275. Occupation and Risk for Injuries 121 Adler-Milstein, S., J. Champagne, and T. Haas. 2014. “The Right Anderson, V. P., P. A. Schulte, J. Sestito, H. Linn, and L. S. Nguyen. to Organize, Living Wage, and Real Change for Garment 2010. “Occupational Fatalities, Injuries, Illnesses, and Workers.” In Lessons for Social Change in the Global Economy, Related Economic Loss in the Wholesale and Retail Trade edited by S. Garwood, S. Croeser, and C. Yakinthou, 13–36. Sector.” American Journal of Industrial Medicine 53 (7): Plymouth, U.K.: Lexington Books. 673–85. Adogu, P. O., A. L. Ilika, and A. L. Asuzu. 2009. “Predictors Apple. 2014. “Supplier Responsibility 2014 Progress Report.” of Road Traffic Accident, Road Traffic Injury, and Death Apple, Cupertino, CA. among Commercial Motorcyclists in an Urban Area of Apramian, S., and M. Cullen. 2015. M.R. (unpublished review) Nigeria.” Nigerian Journal of Medicine 18 (4): 393–97. Azodo, C. C., E. B. Ezeja, and E. E. Ehikhamenor. 2011. Ahamed, F. 2013. “Improving Social Compliance in Bangladesh’s “Occupational Violence against Dental Professionals in Ready-Made Garment Industry.” Labour and Management Southern Nigeria.” African Health Sciences 11 (3): 486–92. in Development 13: 1–26. Baker, T., E. Lugazia, J. Eriksen, V. Mwafongo, L. Irestedt, and Ahasan, R. 2003. “Education and Training in Developing others. 2013. “Emergency and Critical Care Services in Countries.” Work Study 52 (6): 290–96. Tanzania: A Survey of Ten Hospitals.” BMC Health Services Ahmed, E. 2008. “The Risk Factors for Machine Injury of Research 13: 140. the Upper Limb Case-Crossover Study in Tikur Anbessa Baldwin, R. 2011. “Trade and Industrialisation after University Hospital, Addis Ababa, Ethiopia.” Ethiopian Globalisation’s 2nd Unbundling: How Building and Joining Medical Journal 46 (2): 163–72. a Supply Chain Are Different and Why It Matters.” Working Ajayi, I. A., A. O. Adeoye, C. O. Bekibele, O. H. Onakpoya, and Paper 17716, National Bureau of Economic Research, O. J. Omotoye. 2011. “Awareness and Utilization of Protective Cambridge, MA. Eye Device among Welders in a Southwestern Nigeria Barboza, D., P. Lattman, and C. Rampell. 2012. “How the U.S. Community.” Annals of African Medicine 10 (4): 294–99. Lost Out on iPhone Work.” New York Times, January 21. Akhter, S., A. Salahuddin, M. Iqbal, A. Malek, and N. Jahan. 2010. Becker, P., M. Fullen, M. Akladios, and G. Hobbs. 2001. “Health and Occupational Safety for Female Workforce of “Prevention of Construction Falls by Organizational Garment Industries in Bangladesh.” Journal of Mechanical Intervention.” Injury Prevention 7 (Suppl 1): i64–67. Engineering 41 (1): 65–70. Bell, J. L., J. W. Collins, L. Wolf, R. Grönqvist, S. Chiou, and Alavosius, M., J. Getting, J. Dagen, W. Newsome, and B. Hopkins. others. 2008. “Evaluation of a Comprehensive Slip, Trip, 2009. “Use of a Cooperative to Interlock Contingencies and and Fall Prevention Programme for Hospital Employees.” Balance the Commonwealth.” Journal of Organizational Ergonomics 51 (12): 1906–25. Behavior Management 29 (2): 193–211. Benavides, F. G., J. Benach, C. Muntaner, G. L. Delclos, Alfers, L. 2009. Occupational Health and Safety for Informal N. Catot, and others. 2006. “Associations between Temporary Workers in Ghana: A Case Study of Market and Street Traders Employment and Occupational Injury: What Are the in Accra. Durban, South Africa: School of Development Mechanisms?” Occupational and Environmental Medicine 63 Studies, University of KwaZulu-Natal. (6): 416–21. ———. 2011. Occupational Health and Safety and Domestic Bentley, T. 2009. “The Role of Latent and Active Failures Work: A Synthesis of Research Findings from Brazil and in Workplace Slips, Trips, and Falls: An Information Tanzania. London: Women in Informal Employment: Processing Approach.” Applied Ergonomics 40 (2): 175–80. Globalizing and Organizing (WIEGO). Berecki-Gisolf, J., B. Tawatsupa, R. McClure, S. A. Seubsman, Alfers, L., and R. Abban. 2011. Occupational Health and Safety and others. 2013. “Determinants of Workplace Injury for Indigenous Caterers in Accra, Ghana. London: WIEGO. among Thai Cohort Study Participants.” BMJ Open 3 (7): Alonso Castillo, M. M., F. Y. Musayon Oblitas, H. M. David, e003079. and M. V. Gomez Meza. 2006. “Drug Consumption and Binch, S., and J. L. Bell. 2007. The Cost of Non-Injury Accidents: Occupational Violence in Working Women, a Multicenter Scoping Study. Research Report RR585. Suffolk, U.K.: Study: Mexico, Peru, Brazil.” Revista Latino-Americana de Health and Safety Executive. Enfermagem 14 (2): 155–62. BLS (Bureau of Labor Statistics). 2013. “Workplace Injuries and Amengual, M. 2011. “Enforcement without Autonomy: Illnesses, 2012.” Bureau of Labor Statistics, U.S. Department The Politics of Labor and Environmental Regulation in of Labor, Washington, DC. Argentina.” PhD dissertation, Massachusetts Institute of ———. 2014. “Census of Fatal Occupational Injuries Summary, Technology, Cambridge, MA. 2013.” Bureau of Labor Statistics, U.S. Department of Amick, B. C. III, S. Brewer, J. M. Tullar, D. Van Eerd, D. C. Cole, Labor, Washington, DC. and others. 2009. “Musculoskeletal Disorders.” Professional Bodner, T., M. Kraner, B. Bradford, L. Hammer, and D. Truxillo. Safety 54 (3): 24–28. 2014. “Safety, Health, and Well-Being of Municipal Utility Amick, B. C. III, C. Kennedy, J. Dennerlein, S. Brewer, S. Catli, and and Construction Workers.” Journal of Occupational and others. 2010. “Systematic Review of the Role of Occupational Environmental Medicine 56 (7): 771–78. Health and Safety Interventions in the Prevention of Upper Bongers, P. M., S. Ijmker, S. van den Heuvel, and B. M. Blatter. Extremity Musculoskeletal Symptoms, Signs, Disorders, 2006. “Epidemiology of Work-Related Neck and Upper- Injuries, Claims, and Lost Time.” Journal of Occupational Limb Problems: Psychosocial and Personal Risk Factors Rehabilitation 20 (2): 127–62. (Part I) and Effective Interventions from a Bio Behavioural 122 Injury Prevention and Environmental Health Perspective (Part II).” Journal of Occupational Rehabilitation Cantley, L. F., D. Galusha, M. R. Cullen, C. Dixon-Ernst, 16 (3): 272–95. P. R. Rabinowitz, and others. 2015. “Association between Breslin, F. C. 2008. “Educational Status and Work Injury Ambient Noise Exposure, Hearing Acuity, and Risk of among Young People: Refining the Targeting of Prevention Acute Occupational Injury.” Scandinavian Journal of Work, Resources.” Canadian Journal of Public Health 99 (2): Environment, and Health 41 (1): 75–83. 121–24. Cantley, L. F., O. A. Taiwo, D. Galusha, R. Barbour, M. D. Slade, Breslin, F. C., D. Day, E. Tompa, E. Irvin, S. Bhattacharyya, and others. 2014. “Effect of Systematic Ergonomic Hazard and others. 2007. “Non-Agricultural Work Injuries among Identification and Control Implementation on Musculoskeletal Youth: A Systematic Review.” American Journal of Preventive Disorder and Injury Risk.” Scandinavian Journal of Work, Medicine 32 (2): 151–62. Environment, and Health 40 (1): 57–65. Breslin, F. C., and P. Smith. 2005. “Age-Related Differences in Carrivick, P. J., A. H. Lee, K. K. Yau, and M. R. Stevenson. 2005. Work Injuries: A Multivariate, Population-Based Study.” “Evaluating the Effectiveness of a Participatory Ergonomics American Journal of Industrial Medicine 48 (1): 50–56. Approach in Reducing the Risk and Severity of Injuries ———. 2006. “Trial by Fire: A Multivariate Examination from Manual Handling.” Ergonomics 48 (8): 907–14. of the Relation between Job Tenure and Work Injuries.” CDC (Centers for Disease Control and Prevention). 2010. Occupational and Environmental Medicine 63 (1): 27–32. Workbook for Designing, Implementing, and Evaluating a Breslin, F. C., P. M. Smith, and I. Moore. 2011. “Examining Sharps Injury Prevention Program. Atlanta, GA: CDC. the Decline in Lost-Time Claim Rates across Age Groups Chan, J., N. Pun, and M. Selden. 2013. “The Politics of Global in Ontario between 1991 and 2007.” Occupational and Production: Apple, Foxconn, and China’s New Working Environmental Medicine 68 (11): 813–17. Class.” New Technology, Work, and Employment 28 (2): Broberg, O., V. Andersen, and R. Seim. 2011. “Participatory 100–15. Ergonomics in Design Processes: The Role of Boundary Charmes, J. 2012. “The Informal Economy Worldwide: Trends Objects.” Applied Ergonomics 42 (3): 464–72. and Characteristics.” Margin: The Journal of Applied Brown, G. 2007. “Corporate Social Responsibility: Brings Economic Research 6 (2): 103–32. Limited Progress on Workplace Safety in Global Supply Chatterjee, S., D. Agrawal, A. Sahu, and D. Dewangan. 2012. Chain.” Occupational Hazards 69: 16–21. “Primary Prevention of Ocular Injury in Agricultural ———. 2009. “Global Electronics Industry: Poster Child of 21st Workers with Safety Eye Wear.” Paper prepared for the Century Sweatshops and Despoiler.” EHSToday, September 1. Seventieth All India Ophthalmological Society Conference ———. 2010. “Fashion Kills: Industrial Manslaughter in the (AIOC), Cochin, India, February 2–5. Global Supply Chain.” EHSToday, September 1. Chau, N., A. Bhattacherjee, B. M. Kunar, and L. Group. 2009. Brown, J. G., A. Trinkoff, K. Rempher, K. McPhaul, B. Brady, and “Relationship between Job, Lifestyle, Age, and Occupational others. 2005. “Nurses’ Inclination to Report Work-Related Injuries.” Occupational Medicine 59 (2): 114–19. Injuries: Organizational, Work-Group, and Individual Chau, N., D. Dehaene, L. Benamghar, E. Bourgkard, J.-M. Mur, Factors Associated with Reporting.” AAOHN Journal 53 (5): and others. 2014. “Roles of Age, Length of Service, and Job in 213–17. Work-Related Injury: A Prospective Study of 63,620 Person- Buhlebenlosi, F., N. Sibanda, P. Chaurura, and O. Chiwara. Years in Female Workers.” American Journal of Industrial 2013. “Occupational Safety in the Urban Informal Sector of Medicine 57 (2): 172–83. Gaborone, Botswana: A Situational Analysis.” International Chen, L., T. Evans, S. Anand, J. I. Boufford, H. Brown, and Journal of Scientific and Technology Research 2 (12): 293. others. 2004. “Human Resources for Health: Overcoming Burke, M. J., S. A. Sarpy, K. Smith-Crowe, S. Chan-Serafin, the Crisis.” The Lancet 364 (9449): 1984–90. R. O. Salvador, and others. 2006. “Relative Effectiveness of Chen, W. C., H. G. Hwu, S. M. Kung, H. J. Chiu, and J. D. Wang. Worker Safety and Health Training Methods.” American 2008. “Prevalence and Determinants of Workplace Violence Journal of Public Health 96 (2): 315–24. of Health Care Workers in a Psychiatric Hospital in Taiwan.” Burke, M. J., M. L. Scheuer, and R. J. Meredith. 2007. “A Dialogical Journal of Occupational Health 50 (3): 288–93. Approach to Skill Development: The Case of Safety Skills.” Chepkener, A. C. 2013. “Knowledge, Attitude, and Practice of Human Resource Management Review 17 (2): 235–50. Eye Safety among Jua Kali Industry Workers in Nairobi, Butollo, F., J. Kusch, and T. Laufer. 2009. “Buy It Fair: Guideline Kenya.” University of Nairobi. for Sustainable Procurement of Computers.” World Chimamise, C., N. T. Gombe, M. Tshimanga, A. Chadambuka, Economy, Ecology, and Development (WEED), Berlin; Local G. Shambira, and others. 2013. “Factors Associated with Governments for Sustainability Europasekretariat (ICLEI), Severe Occupational Injuries at Mining Company in Freiburg. Zimbabwe, 2010: A Cross-Sectional Study.” Pan African Caldas, A., Z. Aly, P. S. Capece, and Y. Adam. 2003. “Violence Medical Journal 14: 5. against Personnel in Some Healthcare Units in Maputo Clarke, S. 2006. “The Relationship between Safety Climate and City.” International Council of Nurses, International Labour Safety Performance: A Meta-Analytic Review.” Journal of Organization, and World Health Organization, Geneva. Occupational Health Psychology 11 (4): 315–27. Cameron, I., and B. Hare. 2008. “Planning Tools for Integrating ———. 2011. “Accident Proneness: Back in Vogue.” In Health and Safety in Construction.” Construction Occupational Health and Safety, edited by R. Burke, C. Cooper, Management and Economics 26 (9): 899–909. and S. Clarke, 95–118. London: Gower. Occupation and Risk for Injuries 123 Collins, J. W., J. L. Bell, and R. Gronqvist. 2010. “Developing Program Versus an Official-Imperative Method on Needle Evidence-Based Interventions to Address the Leading Causes Disposal Behaviour of Nurses Working in Kashan, Iran.” of Workers’ Compensation among Healthcare Workers.” Nursing and Midwifery Studies 1 (1): 3–6. Rehabilitation Nursing Journal 35 (6): 225–35, 261. Dorne, J. L., G. E. Kass, L. R. Bordajandi, B. Amzal, U. Bertelsen, Collins, J. W., L. Wolf, J. L. Bell, and B. Evanoff. 2004. “An and others. 2011. “Human Risk Assessment of Heavy Evaluation of a ‘Best Practices’ Musculoskeletal Injury Metals: Principles and Applications.” Metal Ions in Life Prevention Program in Nursing Homes.” Injury Prevention Sciences 8: 27–60. 10 (4): 206–11. Drebit, S., S. Shajari, H. Alamgir, S. Yu, and D. Keen. 2010. Committee to Review the NIOSH Construction Research “Occupational and Environmental Risk Factors for Falls Program. 2008. Construction Research at NIOSH: Reviews of among Workers in the Healthcare Sector.” Ergonomics Research Programs of the National Institute for Occupational 53 (4): 525–36. Safety and Health. Washington, DC: National Academies Driessen, M. T., K. I. Proper, J. R. Anema, D. L. Knol, Press. P. M. Bongers, and others. 2011. “The Effectiveness of Costa, G. 2010. “Shift Work and Health: Current Problems Participatory Ergonomics to Prevent Low-Back and Neck and Preventive Actions.” Safety and Health at Work 1 (2): Pain: Results of a Cluster Randomized Controlled Trial.” 112–23. Scandinavian Journal of Work, Environment, and Health Costa, G., T. Åkerstedt, F. Nachreiner, F. Baltieri, J. Carvalhais, 37 (5): 383–93. and others. 2004. “Flexible Working Hours, Health, and Dunser, M. W., I. Baelani, and L. Ganbold. 2006. “A Review and Well-Being in Europe: Some Considerations from a SALTSA Analysis of Intensive Care Medicine in the Least Developed Project.” Chronobiology International 21 (6): 831–44. Countries.” Critical Care Medicine 34 (4): 1234–42. Costa, G., and L. Di Milia. 2010. “Introductory Overview—19th Eakin, J. M., D. Champoux, and E. MacEachen. 2010. “Health International Symposium on Shiftwork and Working Time: and Safety in Small Workplaces: Refocusing Upstream.” Health and Well-Being in the 24-H Society.” Chronobiology Canadian Journal of Public Health 101: S29–33. International 27 (5): 889–97. Ecorys. 2012. “EU SMEs in 2012: At the Crossroads.” Report Couto, M. T., S. Lawoko, and L. Svanstrom. 2009. “Exposure to prepared for the European Commission, Ecorys, Rotterdam, Workplace Violence and Quality of Life among Drivers and the Netherlands. Conductors in Maputo City, Mozambique.” International Ehnes, H. 2012. “Improvement of National Reporting, Data Journal of Occupational and Environmental Health 15 (3): Collection, and Analysis of Occupational Accidents and 299–304. Diseases.” International Labour Organization, Geneva. CPWR (Center for Construction Research and Training). El Ghaziri, M., S. Zhu, J. Lipscomb, and B. A. Smith. 2014. 2013. The Construction Chart Book: The U.S. Construction “Work Schedule and Client Characteristics Associated Industry and Its Workers. Silver Spring, MD: CPWR. with Workplace Violence Experience among Nurses and Cunningham, R. N., C. D. Simpson, and M. C. Keifer. 2012. Midwives in Sub-Saharan Africa.” Journal of the Association “Hazards Faced by Informal Recyclers in the Squatter of Nurses in AIDS Care 25 (Suppl 1): S79–89. Communities of Asunción, Paraguay.” International Journal Eltschinger, C. 2007. Source Code China: The New Global Hub of Occupational and Environmental Health 18 (3): 181–87. of IT Outsourcing. Hoboken, NJ: John Wiley and Sons. da Costa, B. R., and E. R. Vieira. 2010. “Risk Factors for EU-OSHA (European Agency for Safety and Health at Work). Work-Related Musculoskeletal Disorders: A Systematic n.d. “Case Studies: Driver Assistant System.” EU-OSHA, Review of Recent Longitudinal Studies.” American Journal Luxembourg. http://archive.beswic.be/data/case-studies of Industrial Medicine 53 (3): 285–323. /driver-assistant-system/Driver-assistant-system.pdf. David, S., and K. Goel. 2001. “Knowledge, Attitude, and Practice ———. n.d.“Safeguarding Europe’s Micro and Small Enterprises: of Sugarcane Crushers towards Hand Injury Prevention The Development of the Online Interactive Risk Assessment Strategies in India.” Injury Prevention 7 (4): 329–30. Tool.” EU-OSHA, Luxembourg. http://www.oiraproject.eu. de Castro, A. B., K. Fujishiro, T. Rue, E. A. Tagalog, L. P. ———. 2007. Expert Forecast on Emerging Psychosocial Risks Samaco-Paquiz, and others. 2010. “Associations between Related to Occupational Safety and Health. Luxembourg: Work Schedule Characteristics and Occupational Injury Office for Official Publications of the European Communities. and Illness.” International Nursing Review 57 (2): 188–94. Eurofound. 2010. “European Working Conditions Survey Dembe, A. E., R. Delbos, and J. B. Erickson. 2008. “The Effect (EWCS): 2010.” European Foundation for the Improvement of Occupation and Industry on the Injury Risks from of Living and Working Conditions, Dublin. Demanding Work Schedules.” Journal of Occupational and Fabiano, B., F. Currò, and R. Pastorino. 2004. “A Study of the Environmental Medicine 50 (10): 1185–94. Relationship between Occupational Injuries and Firm Size Dennerlein, J. T., K. Hopcia, G. Sembajwe, C. Kenwood, and Type in the Italian Industry.” Safety Science 42 (7): A. M. Stoddard, and others. 2012.“Ergonomic Practices within 587–600. Patient Care Units Are Associated with Musculoskeletal Pain Farrow, A., and F. Reynolds. 2012. “Health and Safety of the and Limitations.” American Journal of Industrial Medicine Older Worker.” Occupational Medicine 62 (1): 4–11. 55 (2): 107–16. Foley, M., Z. J. Fan, E. Rauser, and B. Silverstein. 2012. “The Dianati, M., N. M. Ajorpaz, S. Heidari-Moghaddam, and Impact of Regulatory Enforcement and Consultation Visits M. Heidari. 2012. “Effect of a Face-to-Face Education on Workers’ Compensation Claims, Incidence Rates, and 124 Injury Prevention and Environmental Health Costs, 1999–2008.” American Journal of Industrial Medicine Tracing the Role of Hearing Status and Noise Exposure.” 55 (11): 976–90. Occupational and Environmental Medicine 66 (5): 319–24. Folkard, S., and T. Åkerstedt. 2004. “Trends in the Risk of Glasscock, D. J., K. Rasmussen, O. Carstensen, and O. N. Hansen. Accidents and Injuries and Their Implications for Models of 2006. “Psychosocial Factors and Safety Behaviour as Fatigue and Performance.” Aviation, Space, and Environmental Predictors of Accidental Work Injuries in Farming.” Work Medicine 75 (3): A161–67. and Stress 20 (2): 173–89. Folkard, S., and D. A. Lombardi. 2006. “Modeling the Impact Gonese, E., R. Matchaba-Hove, G. Chirimumba, Z. Hwalima, of the Components of Long Work Hours on Injuries and J. Chirenda, and others. 2006. “Occupational Injuries among ‘Accidents.’” American Journal of Industrial Medicine 49 Workers in the Cleansing Section of the City Council’s Health (11): 953–63. Services Department: Bulawayo, Zimbabwe, 2001–2002.” Folkard, S., and P. Tucker. 2003. “Shift Work, Safety, and Morbidity and Mortality Weekly Report 55 (Suppl 1): 7–10. Productivity.” Occupational Medicine 53 (2): 95–101. Goosen, J., D. M. Bowley, E. Degiannis, and F. Plani. 2003. Frank, A. L., R. McKnight, S. R. Kirkhorn, and P. Gunderson. “Trauma Care Systems in South Africa.” Injury 34 (9): 2004. “Issues of Agricultural Safety and Health.” Annual 704–8. Review of Public Health 25 (1): 225–45. Gray, W., and J. Mendeloff. 2002. “The Declining Effects of Frenkel, S. J. 2001. “Globalization, Athletic Footwear Commodity OSHA Inspections in Manufacturing, 1979–1998.” Working Chains, and Employment Relations in China.” Organization Paper 9119, National Bureau of Economic Research, Studies 22 (4): 531–62. Cambridge, MA. Friedman, L. S., and L. Forst. 2007. “The Impact of OSHA ———. 2005. “The Declining Effects of OSHA Inspections of Recordkeeping Regulation Changes on Occupational Injury Manufacturing Injuries, 1979–1998.” Industrial and Labor and Illness Trends in the U.S.: A Time-Series Analysis.” Relations Review 58 (4): 571–87. Occupational and Environmental Medicine 64 (7): 454–60. Green, F. 2005. Demanding Work. Princeton, NJ: Princeton Fu, C., M. Zhu, T. S. Yu, and Y. He. 2013. “Effectiveness of University Press. Participatory Training on Improving Occupational Health Guimaraes, L. B., J. L. Ribeiro, and J. S. Renner. 2012. “Cost- in Small and Medium Enterprises in China.” International Benefit Analysis of a Socio-Technical Intervention in a Journal of Occupational and Environmental Health 19 (2): Brazilian Footwear Company.” Applied Ergonomics 43 (5): 85–90. 948–57. Fujishiro, K., J. L. Weaver, C. A. Heaney, C. A. Hamrick, and Gyekye, S. A., and S. Salminen. 2007. “Workplace Safety W. S. Marras. 2005. “The Effect of Ergonomic Interventions Perceptions and Perceived Organizational Support: Do in Healthcare Facilities on Musculoskeletal Disorders.” Supportive Perceptions Influence Safety Perceptions?” American Journal of Industrial Medicine 48 (5): 338–47. International Journal of Occupational Safety and Ergonomics Gander, P., L. Hartley, D. Powell, P. Cabon, E. Hitchcock, and 13 (2): 189–200. others. 2011. “Fatigue Risk Management: Organizational Haig, A. J., J. Im, D. Adewole, V. Nelson, and B. Krabak. 2009. Factors at the Regulatory and Industry/Company Level.” “The Practice of Physical and Rehabilitation Medicine in Accident Analysis and Prevention 43 (2): 573–90. Sub-Saharan Africa and Antarctica: A White Paper or a Gangwar, M., and P. Goodrum. 2005. “The Effect of Time on Black Mark?” Journal of Rehabilitation Medicine 41 (6): Safety Incentive Programs in the U.S. Construction Industry.” 401–5. Construction Management and Economics 23 (8): 851–59. Hale, A. R., F. W. Guldenmund, P. L. C. H. van Loenhout, GAO (Government Accountability Office). 2012. “Workplace and J. I. H. Oh. 2010. “Evaluating Safety Management Safety and Health: Better Guidance Needed on Safety and Culture Interventions to Improve Safety: Effective Incentive Programs.” U.S. GAO, Washington, DC. Intervention Strategies.” Safety Science 48 (8): 1026–35. Gap. 2012. “Gap Inc. 2011 / 2012 Social and Environmental Hämäläinen, P., J. Takala, and K. L. Saarela. 2006. “Global Responsibility Report.” Gap Inc., San Francisco, CA. Estimates of Occupational Accidents.” Safety Science 44 (2): Garg, A., and B. Owen. 1992. “Reducing Back Stress to Nursing 137–56. Personnel: An Ergonomic Intervention in a Nursing Home.” Hämmig, O., M. Knecht, T. Läubli, and G. F. Bauer. 2011. Ergonomics 35 (11): 1353–75. “Work-Life Conflict and Musculoskeletal Disorders: A Gervais, R. L., Z. Pawlowska, A. Kouvonen, M. Karanika- Cross-Sectional Study of an Unexplored Association.” BMC Murray, K. Van den Broek, and others. 2009. “Occupational Musculoskeletal Disorders 12 (1): 60–71. Safety and Health and Economic Performance in Small Hanafi, M. I., A. M. Mohamed, M. S. Kassem, and M. Shawki. and Medium-Sized Enterprises: A Review.” Working 2011. “Needlestick Injuries among Health Care Workers of Environment Information Working Paper, European University of Alexandria Hospitals.” Eastern Mediterranean Agency for Safety and Health at Work, Bilbao, Spain. Health Journal 17 (1): 26–35. Ghaffari, M., A. Alipour, A. A. Farshad, I. Jensen, M. Josephson, Hasle, P., and H. J. Limborg. 2006. “A Review of the Literature and others. 2008. “Effect of Psychosocial Factors on Low on Preventive Occupational Health and Safety Activities in Back Pain in Industrial Workers.” Occupational Medicine Small Enterprises.” Industrial Health 44 (1): 6–12. 58 (5): 341–47. Haviland, A., R. Burns, W. Gray, T. Ruder, and J. Mendeloff. Girard, S. A., M. Picard, A. C. Davis, M. Simard, R. Larocque, 2010. “What Kinds of Injuries Do OSHA Inspections and others. 2009. “Multiple Work-Related Accidents: Prevent?” Journal of Safety Research 41 (4): 339–45. Occupation and Risk for Injuries 125 Hayter, S., and V. Stoevska. 2011. “Social Dialogue Indicators: a Major and Direct Impact on the Health and Well- International Statistical Inquiry 2008–09; Technical Brief.” Being of Workers.” ISSA, Geneva. http://www.issa.int/topics International Labour Organization, Geneva. /occupational-risks/introduction. Headd, B. 2000. “The Characteristics of Small Business Jaishankar, M., T. Tseten, N. Anbalagan, B. B. Mathew, and Enterprises.” Monthly Labor Review 123 (4): 13–18. K. N. Beeregowda. 2014. “Toxicity, Mechanism, and Health HSE (Health and Safety Executive). 2014a. “Health and Safety Effects of Some Heavy Metals.” Interdisciplinary Toxicology in Construction in Great Britain, 2014.” HSE, London. 7 (2): 60–72. ———. 2014b. “Health and Safety in Manufacturing in Great Jansen, N. W. H., D. C. L. Mohren, L. G. P. M. van Amelsvoort, Britain, 2014.” HSE, London. N. Janssen, and I. Kant. 2010. “Changes in Working Time Hu, S. C., C. C. Lee, J. S. Shiao, and Y. L. Guo. 1998. “Employers’ Arrangements over Time as a Consequence of Work-Family Awareness and Compliance with Occupational Health Conflict.” Chronobiology International 27 (5): 1045–61. and Safety Regulations in Taiwan.” Occupational Medicine Jun, M., W. Jingjing, M. Collins, W. Malei, S. Orlins, and others. (London) 48 (1): 17–22. 2012. “Sustainable Apparel’s Critical Blind Spot.” Friends Huang, Y.-H., S. K. Verma, W.-R. Chang, T. K. Courtney, D. A. of Nature, Institute of Public and Environmental Affairs, Lombardi, and others. 2012. “Management Commitment Green Beagle, Envirofriends, and Nanjing Green Stone, to Safety vs. Employee Perceived Safety Training and Beijing. Association with Future Injury.” Accident Analysis and Kamchuchat, C., V. Chongsuvivatwong, S. Oncheunjit, Prevention 47: 94–101. T. W. Yip, and R. Sangthong. 2008. “Workplace Violence ILO (International Labour Organization). 2002. “Women and Directed at Nursing Staff at a General Hospital in Southern Men in the Informal Economy: A Statistical Picture.” ILO, Thailand.” Journal of Occupational Health 50 (2): 201–7. Geneva. Kauppinen, T., S. Uuksulainen, A. Saalo, and I. Mäkinen. 2013. ———. 2003. “Code of Practice on Workplace Violence in “Trends of Occupational Exposure to Chemical Agents in Services Sectors and Measures to Combat the Phenomenon.” Finland in 1950–2020.” Annals of Occupational Hygiene 57 ILO, Geneva. (5): 593–609. ———. 2011a. “Children in Hazardous Work: What We Know, Kawakami, T. 2007. “Participatory Approaches to Improving What We Need to Do.” ILO, International Programme on Safety, Health, and Working Conditions in Informal the Elimination of Child Labour, Geneva. Economy Workplaces: Experiences of Cambodia, Thailand, ———. 2011b. “Promoting Decent Employment for Rural and Vietnam.” Subregional Office for East Asia, International Migrant Workers.” ILO, Geneva. Labour Organization, Bangkok. ———. 2012a. “Decent Work Country Profile: Zambia.” ILO, Kawakami, T., L. Tong, Y. Kannitha, and T. Sophorn. 2011. Geneva. “Participatory Approach to Improving Safety, Health, and ———. 2012b. “Decent Work Indicators in Africa: A First Working Conditions in Informal Economy Workplaces in Assessment Based on National Sources.” ILO, Geneva. Cambodia.” Work 38 (3): 235–40. ———. 2012c. “Global Employment Trends 2012: World Faces Kelsh, M. A., and J. D. Sahl. 1996. “Sex Differences in Work- a 600 Million Jobs Challenge, Warns ILO.” ILO, Geneva. Related Injury Rates among Electric Utility Workers.” ———. 2012d. “Practices with Good Potential: Towards the American Journal of Epidemiology 143 (10): 1050–58. Elimination of Hazardous Child Labour.” International Kim, H.-C., J.-Y. Min, K.-B. Min, and S.-G. Park. 2009. “Job Programme on the Elimination of Child Labour (IPEC), Strain and the Risk for Occupational Injury in Small- to ILO, Geneva. Medium-Sized Manufacturing Enterprises: A Prospective ———. 2013a. “ILO Calls for Urgent Global Action to Fight Study of 1,209 Korean Employees.” American Journal of Occupational Diseases.” ILO, Geneva. Industrial Medicine 52 (4): 322–30. ———. 2013b. “Marking Progress against Child Labour: Knight, E. B., D. N. Castillo, and L. A. Layne. 1995. “A Detailed Global Estimates and Trends 2000–2012.” IPEC, ILO, Analysis of Work-Related Injury among Youth Treated in Geneva. Emergency Departments.” American Journal of Industrial ———. 2013c. “Rural-Urban Migrants Employed in Domestic Medicine 27 (6): 793–805. Work: Issues and Challenges.” Paper prepared for ILO forum Kraemer, B. 2010. “Impact of Economic Crisis on Occupational “Making Decent Work a Reality for Domestic Workers Health and Safety Management.” Institute of Economic and in Africa: A Regional Knowledge Sharing Forum,” 1–5. Social Research (WSI), Dusseldorf. International Labor Office, Dar es Salaam, May 28–30. Kubo, J,. M. R. Cullen, L. Cantley, M. Slade, B. Tessier-Sherman, ———. 2014a. “ILO-ACTRAV Symposium: Role of Trade and others. 2013. “Piecewise Exponential Models to Assess Unions in the Promotion of OSH in Asia-Pacific.” ILO, the Influence of Job-Specific Experience on the Hazard of Geneva. Acute Injury for Hourly Factory Workers.” BMC Medical ———. 2014b. “Safety and Health in the Construction Sector: Research and Methodology 13: 89. Overcoming the Challenges.” Webinar, ILO, Geneva, Kubo, J., B. A. Goldstein, L. F. Cantley, B. Tessier-Sherman, November 7. http://www.ilo.org/empent/Eventsandmeetings D. Galusha, and others. 2014. “Contribution of Health /WCMS_310993/lang--en/index.htm. Status and Prevalent Chronic Disease to Individual Risk ISSA (International Social Security Association). 2014. “Towards for Workplace Injury in the Manufacturing Environment.” a Global Culture of Prevention: Working Conditions Have Occupational and Environmental Medicine 71 (3): 159–66. 126 Injury Prevention and Environmental Health Lahiri, S., J. Gold, and C. Levenstein. 2005. “Net-Cost Model Lombardi, D. A., S. Folkard, J. L. Willetts, and G. S. Smith. for Workplace Interventions.” Journal of Safety Research 2010. “Daily Sleep, Weekly Working Hours, and Risk of 36 (3): 241–55. Work-Related Injury: U.S. National Health Interview Survey Lambert, D. M., and M. C. Cooper. 2000. “Issues in Supply (2004–2008).” Chronobiology International 27 (5): 1013–30. Chain Management.” Industrial Marketing Management Lundqvist, P., and C. A. Svennefelt. 2012. “Health and Safety 29: 65–83. Strategy in Swedish Agriculture.” Work 41 (Suppl 1): Lapointe, J. M., C. E. P. Dionne, C. P. Brisson, and S. P. Montreuil. 5304–7. 2009. “Interaction between Postural Risk Factors and Job ———. 2014. “Swedish Strategies for Health and Safety in Strain on Self-Reported Musculoskeletal Symptoms among Agriculture: A Coordinated Multiagency Approach.” Work Users of Video Display Units: A Three-Year Prospective 49 (1): 33–37. Study.” Scandinavian Journal of Work, Environment, and Magnavita, N., M. Elovainio, I. De Nardis, T. Heponiemi, and Health 35 (2): 134–44. A. Bergamaschi. 2011. “Environmental Discomfort and Lauver, K. J., and S. W. Lester. 2007. “Get Safety Problems to Musculoskeletal Disorders.” Occupational Medicine 61 (3): the Surface: Using Human Resource Practices to Improve 196–201. Injury Reporting.” Journal of Leadership and Organizational Manothum, A., and J. Rukijkanpanich. 2010. “A Participatory Studies 14 (2): 168–79. Approach to Health Promotion for Informal Sector Workers Lavack, A. M., S. L. Magnuson, S. Deshpande, D. Z. Basil, in Thailand.” Journal of Injury and Violence Research 2 (2): M. D. Basil, and others. 2008. “Enhancing Occupational 111–20. Health and Safety in Young Workers: The Role of Social Manothum, A., J. Rukijkanpanich, D. Thawesaengskulthai, Marketing.” International Journal of Nonprofit and Voluntary B. Thampitakkul, C. Chaikittiporn, and others. 2009. “A Sector Marketing 13 (3): 193–204. Participatory Model for Improving Occupational Health Lehtola, M. M., R. H. Rautiainen, L. M. Day, E. Schonstein, and Safety: Improving Informal Sector Working Conditions J. Suutarinen, and others. 2008. “Effectiveness of in Thailand.” International Journal of Occupational and Interventions in Preventing Injuries in Agriculture: A Environmental Health 15 (3): 305–14. Systematic Review and Meta-Analysis.” Scandinavian Journal Marilda, M., and C. Maciel. 2014. “Strikes in Sugarcane of Work, Environment, and Health 34 (5): 327–36. Mills: The Forms of Resistance of Migrant Workers in Leigh, J. P., and J. P. Marcin. 2012. “Workers’ Compensation Brazil.” Paper prepared for the Fourth RUFORUM Biennial Benefits and Shifting Costs for Occupational Injury Regional Conference, Maputo, Mozambique, July 19–24. and Illness.” Journal of Occupational and Environmental Marras, W. S., W. G. Allread, D. L. Burr, and F. A. Fathallah. 2000. Medicine 54 (4): 445–50. “Prospective Validation of a Low-Back Disorder Risk Model Levine, D. I., M. W. Toffel, and M. S. Johnson. 2012.“Randomized and Assessment of Ergonomic Interventions Associated Government Safety Inspections Reduce Worker Injuries with Manual Materials Handling Tasks.” Ergonomics 43 (11): with No Detectable Job Loss.” Science 336 (6083): 907–11. 1866–86. Lin, Y.-H., C.-Y. Chen, and J.-L. Luo. 2008. “Gender and Age Marucci-Wellman, H., T. B. Leamon, J. L. Willetts, T. T. Binh, Distribution of Occupational Fatalities in Taiwan.” Accident N. B. Diep, and others. 2011. “Occupational Injuries in a Analysis and Prevention 40 (4): 1604–10. Commune in Rural Vietnam Transitioning from Agriculture ———. 2011. “Statistical Analysis of Fatalities in Construction to New Industries.” American Journal of Public Health Workers.” Journal of Occupational Safety and Health 19: 101 (5): 854–60. 75–85. Marucci-Wellman, H., D. H. Wegman, T. B. Leamon, Lippel, K. 2005. “Precarious Employment and Occupational T. T. Binh, N. B. Diep, and others. 2013. “Work-Related Health and Safety Regulation in Quebec.” In Precarious Injury Surveillance in Vietnam: A National Reporting Employment: Understanding Labour Market Insecurity in System Model.” American Journal of Public Health 103 (11): Canada, edited by L. F. Vosko, 241–55. Montreal: McGill- 1989–96. Queen’s University Press. Mayhew, C., and M. Quinlan. 1999. “The Effects of Outsourcing Lipscomb, H. J., J. Nolan, D. Patterson, V. Sticca, and D. J. Myers. on Occupational Health and Safety: A Comparative Study 2013. “Safety, Incentives, and the Reporting of Work- of Factory-Based Workers and Outworkers in the Australian Related Injuries among Union Carpenters: ‘You’re Pretty Clothing Industry.” International Journal of Health Services Much Screwed If You Get Hurt at Work.’” American Journal 29 (1): 83–107. of Industrial Medicine 56 (4): 389–99. Mbaisi, E. M., Z. Ng’ang’a, P. Wanzala, and J. Omolo. 2013. Locke, R. M. 2013. The Promise and Limits of Private Power: “Prevalence and Factors Associated with Percutaneous Promoting Labor Standards in a Global Economy. New York: Injuries and Splash Exposures among Health-Care Workers Cambridge University Press. in a Provincial Hospital, Kenya, 2010.” Pan African Medical Locke, R. M., and M. Romis. 2007. “Improving Work Conditions Journal 14: 10. in a Global Supply Chain.” MIT Sloan Management Review McKay, S., M. Craw, and D. Chopra. 2006. Migrant Workers in 48 (Winter): 54–62. England and Wales: An Assessment of Migrant Worker Health Loewenson, R. H. 1998. “Health Impact of Occupational Risks and Safety Risks. London: Health and Safety Executive. in the Informal Sector in Zimbabwe.” International Journal McSweeney, K. P., B. N. Craig, J. J. Congleton, and D. Miller. of Occupational and Environmental Health 4 (4): 264–74. 2002. “Ergonomic Program Effectiveness: Ergonomic and Occupation and Risk for Injuries 127 Medical Intervention.” International Journal of Occupational Nasrullah, M., and S. Awan. 2012. “Burns and Use of Mirror to and Safety Ergonomics 8 (4): 433–49. Prevent These Injuries among Young and Adult Workers at Mendeloff, J., and W. B. Gray. 2005. “Inside the Black Box: How Hot Clay Ovens, Tandoors: An Innovative, Cost-Effective Do OSHA Inspections Lead to Reductions in Workplace Solution in a Low-Income Setting.” Burns 38 (7): 1089–90. Injuries?” Law and Policy 27 (2): 219–37. NFIB (National Federation of Independent Business). 2002. Mendeloff, J., and L. Staetsky. 2014. “Occupational Fatality “National Small Business Poll.” Workplace Safety 2 (1). Risks in the United States and the United Kingdom.” Ngai, P., and J. Chan. 2012. “Global Capital, the State, and American Journal of Industrial Medicine 57 (1): 4–14. Chinese Workers: The Foxconn Experience.” Modern China Mischke, C., J. H. Verbeek, J. C. Job, T. C. Morata, A. Alvesalo- 38 (4): 383–410. Kuusi, and others. 2013. “Occupational Safety and Health NIOSH (National Institute for Occupational Safety and Health). Enforcement Tools for Preventing Occupational Diseases 2002. “The Changing Organization of Work and the Safety and Injuries.” Cochrane Database of Systematic Reviews 8: and Health of Working People: Knowledge Gaps and CD010183. Research Directions.” NIOSH, Cincinnati, OH. Mohan, D., and R. Patel. 1992. “Design of Safer Agricultural ______. 2008.“NORA AgFF Sector Council National Agriculture, Equipment: Application of Ergonomics and Epidemiology.” Forestry, and Fishing Agenda.” NIOSH, Washington, DC. International Journal of Industrial Ergonomics 10 (4): 301–9. ———. 2007. “Use of Blunt-Tip Suture Needles to Decrease Monaghan, P. F., C. A. Bryant, J. A. Baldwin, Y. Zhu, Percutaneous Injuries to Surgical Personnel.” NIOSH, B. Ibrahimou, and others. 2008. “Using Community-Based Cincinatti, OH. Prevention Marketing to Improve Farm Worker Safety.” Niskanen, T., J. Lehtelä, R. Ketola, and E. Nykyri. 2010. “Results Social Marketing Quarterly 14 (4): 71–87. of Finnish National Survey on EU Legislation Concerning Mora, A. M., M. G. Mora-Mora, T. Partanen, and C. Wesseling. Computer Work.” Applied Ergonomics 41 (4): 542–48. 2011. “Registration of Fatal Occupational Injuries in Costa Niu, S. 2010. “Ergonomics and Occupational Safety and Rica, 2005–2006.” International Journal of Occupational and Health: An ILO Perspective.” Applied Ergonomics 41 (6): Environmental Health 17 (3): 243–50. 744–53. Morassaei, S., F. C. Breslin, M. Shen, and P. M. Smith. Noe, R., J. Rocha, C. Clavel-Arcas, C. Aleman, M. E. Gonzales, 2013. “Examining Job Tenure and Lost-Time Claim Rates and others. 2004. “Occupational Injuries Identified by an in Ontario, Canada, over a 10-Year Period, 1999–2008.” Emergency Department–Based Injury Surveillance System Occupational and Environmental Medicine 70 (3): 171–78. in Nicaragua.” Injury Prevention 10 (4): 227–32. Mou, J., S. M. Griffiths, H. Fong, and M. G. Dawes. 2013. “Health Nordander, C., K. Ohlsson, I. Balogh, G.-Å. Hansson, of China’s Rural-Urban Migrants and Their Families: A A. Axmon, and others. 2008. “Gender Differences in Review of Literature from 2000 to 2012.” British Medical Workers with Identical Repetitive Industrial Tasks: Exposure Bulletin 106 (1): 19–43. and Musculoskeletal Disorders.” International Archives of Muntaner, C., O. Solar, C. Vanroelen, J. M. Martinez, M. Vergara, Occupational and Environmental Health 81 (8): 939–47. and others. 2010.“Unemployment, Informal Work, Precarious Nossar, I., R. Johnstone, and M. Quinlan. 2003. “Regulating Employment, Child Labor, Slavery, and Health Inequalities: Supply-Chains to Address the Occupational Health and Pathways and Mechanisms.” International Journal of Health Safety Problems Associated with Precarious Employment: Services 40 (2): 281–95. The Case of Home-Based Clothing Workers in Australia.” Murad, M. K., S. Larsen, and H. Husum. 2012. “Prehospital Working Paper 21, Australian National University, Trauma Care Reduces Mortality: Ten-Year Results from a Canberra. Time-Cohort and Trauma Audit Study in Iraq.” Scandinavian Noy, Y. I., W. J. Horrey, S. M. Popkin, S. Folkard, H. D. Howarth, Journal of Trauma, Resuscitation, and Emergency Medicine and others. 2011. “Future Directions in Fatigue and Safety 20: 13. Research.” Accident Analysis and Prevention 43 (2): 495–97. Murdoch, H., and D. Gould. 2004. “Corporate Social Nsubuga, F. M., and M. S. Jaakkola. 2005. “Needle Stick Injuries Responsibility in China: Mapping the Environment.” among Nurses in Sub-Saharan Africa.” Tropical Medicine Report prepared for the Global Alliance for Workers and and International Health 10 (8): 773–81. Communities, Baltimore, MD. Ntow, W. J., H. J. Gijzen, P. Kelderman, and P. Drechsel. Murray, C. J. L., K. F. Ortblad, C. Guinovart, S. S. Lim, T. M. Wolock, 2006. “Farmer Perceptions and Pesticide Use Practices in and others. 2014. “Global, Regional, and National Incidence Vegetable Production in Ghana.” Pest Management Science and Mortality for HIV, Tuberculosis, and Malaria during 62 (4): 356–65. 1990–2013: A Systematic Analysis for the Global Burden of Pai, H. C., and S. Lee. 2011. “Risk Factors for Workplace Disease Study 2013.” The Lancet 384 (9947): 1005–70. Violence in Clinical Registered Nurses in Taiwan.” Journal Mustard, C. 2007. “Can Social Marketing Campaigns Prevent of Clinical Nursing 20 (9–10): 1405–12. Workplace Injury and Illness?” At Work 49 (Summer): 3. Palácios, M., M. L. dos Santos, M. B. do Val, M. Medina, Nakata, A., T. Ikeda, M. Takahashi, T. Haratani, M. Hojou, and M. de Abreu, and others. 2003. “Workplace Violence in the others. 2006. “Impact of Psychosocial Job Stress on Non- Health Sector: Country Case Study, Brazil.” World Health Fatal Occupational Injuries in Small and Medium-Sized Organization, Geneva. Manufacturing Enterprises.” American Journal of Industrial Palmer, K. T., E. C. Harris, and D. Coggon. 2008. “Chronic Medicine 49 (8): 658–69. Health Problems and Risk of Accidental Injury in the 128 Injury Prevention and Environmental Health Workplace: A Systematic Literature Review.” Occupational Rahman, S. 2004. “Global Shift: Bangladesh Garment Industry and Environmental Medicine 65 (11): 757–64. in Perspective.” Asian Affairs 26 (1): 75–91. Palmer, K. T., E. C. Harris, C. Linaker, M. Barker, W. Lawrence, Redinger, C. F., and S. P. Levine. 1998. “Development and and others. 2012. “Effectiveness of Community- and Evaluation of the Michigan Occupational Health and Safety Workplace-Based Interventions to Manage Musculoskeletal- Management System Assessment Instrument: A Universal Related Sickness Absence and Job Loss: A Systematic OHSMS Performance Measurement Tool.” American Review.” Rheumatology 51 (2): 230–42. Industrial Hygiene Association Journal 59 (8): 572–81. Pehkonen, I., E.-P. Takala, R. Ketola, E. Viikari-Juntura, P. Leino- Regoeng, K. 2003. “Safety and Health in the Informal Sector Arjas, and others. 2009. “Evaluation of a Participatory and Small-Scale Industries: The Experience of Botswana.” Ergonomic Intervention Process in Kitchen Work.” Applied African Newsletter on Occupational Health and Safety 13 (1): Ergonomics 40 (1): 115–23. 10–12. Pfortmueller, C. A., D. Kradolfer, M. Kunz, B. Lehmann, Rivoli, P. 2009. The Travels of a T-Shirt in the Global Economy: G. Lindner, and others. 2013. “Injuries in Agriculture: Injury An Economist Examines the Markets, Power, and Politics of Severity and Mortality.” Swiss Medical Weekly 143: w13846. World Trade. Hoboken, NJ: John Wiley and Sons. Phung, D. T., H. T. Nguyen, C. Mock, and M. Keifer. 2008. Robson, L. S., J. A. Clarke, K. Cullen, A. Bielecky, C. Severin, “Occupational Injuries Reported in a Population-Based and others. 2007. “The Effectiveness of Occupational Injury Survey in Vietnam.” International Journal of Health and Safety Management System Interventions: A Occupational and Environmental Health 14 (1): 35–44. Systematic Review.” Safety Science 45 (3): 329–53. Powers, J. R. Jr., D. E. Anmons, and I. Brand. 2009. “Machine Robson, L. S., C. M. Stephenson, P. A. Schulte, B. C. Amick III, Safety.” Professional Safety 54 (11): 28–31. E. L. Irvin, and others. 2012. “A Systematic Review of the Pransky, G. S., P. Loisel, and J. R. Anema. 2011. “Work Disability Effectiveness of Occupational Health and Safety Training.” Prevention Research: Current and Future Prospects.” Scandinavian Journal of Work, Environment, and Health Journal of Occupational Rehabilitation 21 (3): 287–92. 38 (3): 193–208. Premji, S., P. Duguay, K. Messing, and K. Lippel. 2010. Rosenstock, L., M. Cullen, C. A. Brodkin, and C. A. Redlich. “Are Immigrants, Ethnic and Linguistic Minorities Over- 2005. Textbook of Clinical Occupational and Environmental Represented in Jobs with a High Level of Compensated Medicine, second edition. Philadelphia: Elsevier WB Risk? Results from a Montréal, Canada Study Using Census Saunders. and Workers’ Compensation Data.” American Journal of Ruff, T. 2010. “Innovative Safety Interventions.” IEEE Industry Industrial Medicine 53 (9): 875–85. Applications Magazine 16 (3): 45–49. Premji, S., K. Lippel, and K. Messing. 2008. “‘We Work by SafeWork. 2012. “Improvement of National Reporting, Data the Second!’ Piecework Remuneration and Occupational Collection, and Analysis of Occupational Accidents and Health and Safety from an Ethnicity- and Gender-Sensitive Diseases.” Programme on Safety and Health at Work and the Perspective.” Pistes 10 (1): 1–22. Environment, International Labour Organization, Geneva. Pringle, T. E., and S. D. Frost. 2003. “The Absence of Rigor and Sandoval, M., and K. A. Bjurling. 2014. “Challenging Labor: the Failure of Implementation: Occupational Health and Working Conditions in the Electronics Industry.” In Lessons Safety in China.” International Journal of Occupational and for Social Change in the Global Economy: Voices from the Environmental Health 9 (4): 309–16. Field, edited by S. Garwood, S. Croeser, and C. Yakinthou, ProcureITfair. 2008. “The Dark Side of Cyberspace.” World 99–124. Plymouth, U.K.: Lexington Books. Economy, Ecology, and Development and Students and Schlick, C., M. Joachin, L. Briceno, D. Moraga, and K. Radon. 2014. Scholars against Corporate Misbehaviour, Berlin. http:// “Occupational Injuries among Children and Adolescents www2.weed-online.org/uploads/press_release_dark_side in Cusco Province: A Cross-Sectional Study.” BMC Public _of_cyberspace.pdf. Health 14: 766. Pueyo, V., C. Toupin, and S. Volkoff. 2011. “The Role of Schulte, P. A., S. Pandalai, V. Wulsin, and H. Chun. 2012. Experience in Night Work: Lessons from Two Ergonomic “Interaction of Occupational and Personal Risk Factors in Studies.” Applied Ergonomics 42 (2): 251–55. Workforce Health and Safety.” American Journal of Public Quinlan, M. 1999. “The Implications of Labour Market Health 102 (3): 434–48. Restructuring in Industrialized Societies for Occupational Shiao, J. S., M. L. McLaws, M. H. Lin, J. Jagger, and C. J. Chen. Health and Safety.” Economic and Industrial Democracy 20 2009. “Chinese EPINet and Recall Rates for Percutaneous (3): 427–60. Injuries: An Epidemic Proportion of Underreporting in the Quinlan, M., and P. Bohle. 2008. “Under Pressure, Out of Taiwan Healthcare System.” Journal of Occupational Health Control, or Home Alone? Reviewing Research and Policy 51 (2): 132–36. Debates on the Occupational Health and Safety Effects of Siddiqi, F., and H. A. Patrinos. 1995. “Child Labor: Issues, Outsourcing and Home-Based Work.” International Journal Causes, and Interventions.” Human Capital Development of Health Services 38 (3): 489–523. and Operations Policy Working Paper 56, World Bank, Quinlan, M., C. Mayhew, and P. Bohle. 2001. “The Global Washington, DC. Expansion of Precarious Employment, Work Disorganisation, Silverstein, M. 2008. “Meeting the Challenges of an Aging and Occupational Health: A Review of Recent Research.” Workforce.” American Journal of Industrial Medicine 51 (4): International Journal of Health Services 31 (2): 335–414. 269–80. Occupation and Risk for Injuries 129 Sinclair, R. C., T. R. Cunningham, and P. A. Schulte. 2013. Tak, S., and G. M. Calvert. 2011. “The Estimated National “A Model for Occupational Safety and Health Intervention Burden of Physical Ergonomic Hazards among U.S. Diffusion to Small Businesses.” American Journal of Workers.” American Journal of Industrial Medicine 54 (5): Industrial Medicine 56 (12): 1442–51. 395–404. Smith, P., A. Bielecky, M. Koehoorn, D. Beaton, S. Ibrahim, Takala, J., and P. Hämäläinen. 2009. “Globalization of Risks.” and others. 2014. “Are Age-Related Differences in the African Newsletter on Occupational Health and Safety 19 Consequence of Work Injury Greater When Occupational (3): 70–73. Physical Demands Are High?” American Journal of Industrial Targoutzidi, A., T. Koukoulaki, E. Schmitz-Felten, K. Kuhl, Medicine 57 (4): 438–44. K. M. O. Hengel, and others. 2013. “The Business Case Smith, T., D. A. Sonnenfeld, and D. N. Pellow. 2006. Challenging for Safety and Health at Work: Cost-Benefit Analyses of the Chip: Labor Rights and Environmental Justice in the Global Interventions in Small and Medium-Sized Enterprises.” Electronics Industry. Philadelphia: Temple University Press. European Agency for Safety and Health at Work, Luxembourg. Smith, W. A. 2006. “Social Marketing: An Overview of Approach Taswell, K., and P. Wingfield-Digby. 2008. Occupational Injury and Effects.” Injury Prevention 12 (Suppl 1): i38–43. Statistics from Household Surveys and Establishment Surveys: Soares, M. M., K. Jacobs, K. Olsen, S. Legg, and P. Hasle. An ILO Manual on Methods. Geneva: International Labour 2012. “How to Use Programme Theory to Evaluate the Organization. Effectiveness of Schemes Designed to Improve the Work Tawatsupa, B., V. Yiengprugsawan, T. Kjellstrom, J. Berecki- Environment in Small Businesses.” Work 41 (Suppl 1): Gisolf, S. A. Seubsman, and others. 2013. “Association 5999–6006. between Heat Stress and Occupational Injury among Thai Sorensen, G., P. Landsbergis, L. Hammer, B. C. Amick III, Workers: Findings of the Thai Cohort Study.” Industrial L. Linnan, and others. 2011. “Preventing Chronic Disease in Health 51 (1): 34–46. the Workplace: A Workshop Report and Recommendations.” Tessier-Sherman, B., L. F. Cantley, D. Galusha, M. D. Slade, American Journal of Public Health 101 (Suppl 1): S196–207. O. A. Taiwo, and others. 2014. “Occupational Injury Risk Sørensen, O. H., P. Hasle, and E. Bach. 2007. “Working in Small by Sex in a Manufacturing Cohort.” Occupational and Enterprises: Is There a Special Risk?” Safety Science 45 (10): Environmental Medicine 71 (9): 605–10. 1044–59. Texido, E., and E. Warn. 2013. “Migrant Well-Being and Sripichyakan, K., P. Thungpunkum, and B. Supavititpatana. Development: South America.” Working Paper for the 2003. “Workplace Violence in the Health Sector: A Case World Migration Report 2013, International Organization Study in Thailand.” International Labour Organization, for Migration, Geneva. Geneva. Tinney, M. J., A. Chiodo, A. Haig, and E. Wiredu. 2007. “Medical Steege, A. L., S. L. Baron, S. M. Marsh, C. C. Menéndez, Rehabilitation in Ghana.” Disability and Rehabilitation 29 and J. R. Myers. 2014. “Examining Occupational Health (11–12): 921–27. and Safety Disparities Using National Data: A Cause for Tompa, E., R. Dolinschi, C. de Oliveira, B. C. Amick III, Continuing Concern.” American Journal of Industrial and E. Irvin. 2010. “A Systematic Review of Workplace Medicine 57 (5): 527–38. Ergonomic Interventions with Economic Analyses.” Journal Steinman, S. 2003. “Workplace Violence in the Health Sector: of Occupational Rehabilitation 20 (2): 220–34. Country Case Study in South Africa.” World Health Tompa, E., R. Dolinschi, C. de Oliveira, and E. Irvin. 2009. Organization, Geneva. “A Systematic Review of Occupational Health and Stocks, S. J., R. McNamee, H. F. van der Molen, C. Paris, Safety Interventions with Economic Analyses.” Journal of P. Urban, and others. 2015. “Trends in Incidence of Occupational and Environmental Medicine 51 (9): 1004–23. Occupational Asthma, Contact Dermatitis, Noise-Induced Tompa, E., R. Dolinschi, and A. Laing. 2009. “An Economic Hearing Loss, Carpal Tunnel Syndrome, and Upper-Limb Evaluation of a Participatory Ergonomics Process in an Musculoskeletal Disorders in European Countries from Auto Parts Manufacturer.” Journal of Safety Research 40 (1): 2000 to 2012.” Occupational and Environmental Medicine 41–47. 72 (4): 294–303. Tompa, E., R. Dolinschi, and J. Natale. 2013. “Economic Strong, L. L., and F. J. Zimmerman. 2005. “Occupational Evaluation of a Participatory Ergonomics Intervention in a Injury and Absence from Work among African American, Textile Plant.” Applied Ergonomics 44 (3): 480–87. Hispanic, and Non-Hispanic White Workers in the National Tompa, E., S. Trevithick, and C. McLeod. 2007. “Systematic Longitudinal Survey of Youth.” American Journal of Public Review of the Prevention Incentives of Insurance and Health 95 (7): 1226–32. Regulatory Mechanisms for Occupational Health and Symanski, E., L. L. Kupper, and S. M. Rappaport. 1998. Safety.” Scandinavian Journal of Work and Environmental “Comprehensive Evaluation of Long-Term Trends in Health 33 (2): 85–95. Occupational Exposure: Part 1. Description of the Database.” Törnström, L., J. Amprazis, M. Christmansson, and J. Eklund. Occupational and Environmental Medicine 55 (5): 300–309. 2008. “A Corporate Workplace Model for Ergonomic Taiwo, O. A., L. F. Cantley, M. D. Slade, K. M. Pollack, S. Vegso, Assessments and Improvements.” Applied Ergonomics 39 (2): and others. 2009. “Sex Differences in Injury Patterns among 219–28. Workers in Heavy Manufacturing.” American Journal of Turgoose, C., L. Hall, A. Carter, and C. Stride. 2006. Epidemiology 169 (2): 161–66. “Encouraging an Increase in the Employment of Women 130 Injury Prevention and Environmental Health Returners in Areas of Skill Shortage in Traditionally Male Wang, D. 2008. Rural-Urban Migration and Policy Responses Industries.” Institute of Work Psychology, University of in China: Challenges and Options. Bangkok: ILO Asian Sheffield, Sheffield, U.K. Regional Programming on Governance of Labour Migration, UN DESA (United Nations Department of Economic and International Labour Office. Social Affairs). 2010. The World’s Women 2010: Trends and Wang, H., K. Fennie, G. He, J. Burgess, and A. B. Williams. 2003. Statistics. New York: UN DESA. “A Training Programme for Prevention of Occupational Ustailieva, E., L. Eeckelaery, and I. Nunes. 2012. Promoting Exposure to Bloodborne Pathogens: Impact on Knowledge, Occupational Safety and Health through the Supply Chain Behaviour, and Incidence of Needle Stick Injuries among Literature Review. Luxembourg: European Agency for Student Nurses in Changsha, People’s Republic of China.” Safety and Health at Work. Journal of Advanced Nursing 41 (2): 187–94. Valcour, M. 2007. “Work-Based Resources as Moderators of Wang, X., S. Wu, Q. Song, L.-A. Tse, I. T. S. Yu, and others. 2011. the Relationship between Work Hours and Satisfaction “Occupational Health and Safety Challenges in China: with Work-Family Balance.” Journal of Applied Psychology Focusing on Township-Village Enterprises.” Archives of 92 (6): 1512–23. Environmental and Occupational Health 66 (1): 3–11. Vandergrift, J. L., J. E. Gold, A. Hanlon, and L. Punnett. 2012. Welford, R., and S. Frost. 2006. “Corporate Social Responsibility “Physical and Psychosocial Ergonomic Risk Factors for in Asian Supply Chains.” Corporate Social Responsibility and Low Back Pain in Automobile Manufacturing Workers.” Environmental Management 13 (3): 166–76. Occupational and Environmental Medicine 69 (1): 29–34. WHO (World Health Organization). 2001. Occupational van der Molen, H. F., J. K. Sluiter, C. T. J. Hulshof, P. Vink, and Health: A Manual for Primary Health Care Workers. Geneva: M. H. W. Frings-Dresen. 2005. “Effectiveness of Measures WHO. and Implementation Strategies in Reducing Physical Work ———. 2004. Guidelines on the Prevention of Toxic Exposures: Demands due to Manual Handling at Work.” Scandinavian Education and Public Awareness Activities. Geneva: International Journal of Work, Environment, and Health 31 (Suppl 2): Programme on Chemical Safety, WHO. 75–87. ———. 2006a. “Managing Exits from the Workforce.” In The Van Dongen, H. P. A., J. A. Caldwell Jr., and J. L. Caldwell. World Health Report 2006: Working Together for Health, 2011. “Individual Differences in Cognitive Vulnerability 97–118. Geneva: WHO. to Fatigue in the Laboratory and in the Workplace.” In ———. 2006b. Safer Access to Pesticides: Community Progress in Brain Research, edited by A. K. Gerard, 145–53. Interventions. Geneva: International Association for Suicide Amsterdam: Elsevier. Prevention, WHO. van Stolk, C., L. Staetsky, E. Hassan, and C. W. Kim. 2012. ———. 2011. “Safe Injection Global Network: Summaries of “Management of Psychosocial Risks at Work: An Analysis Injection Safety Country Success Stories.” WHO, Geneva. of the Findings of the European Survey of Enterprises on ———. 2012. “Connecting Health and Labour: Bringing New and Emerging Risks (ESENER).” European Agency for Together Occupational Health and Primary Care to Improve Safety and Health at Work, Luxembourg. the Health of Working People.” Executive Summary of the Verbeek, J., and I. Ivanov. 2013. “Essential Occupational Safety WHO Global Conference “Connecting Health and Labour: and Health Interventions for Low- and Middle-Income What Role for Occupational Health in Primary Health Countries: An Overview of the Evidence.” Safety and Health Care,” The Hague, November 29–December 1, 2011. at Work 4 (2): 77–83. Williamson, A., D. A. Lombardi, S. Folkard, J. Stutts, T. K. Verbeek, J., M. Pulliainen, and E. Kankaanpää. 2009. Courtney, and others. 2011. “The Link between Fatigue and “A Systematic Review of Occupational Safety and Health Safety.” Accident Analysis and Prevention 43 (2): 498–515. Business Cases.” Scandinavian Journal of Work, Environment, Wirtz, A., F. Nachreiner, and K. Rolfes. 2011. “Working on and Health 35 (6): 403–12. Sundays: Effects on Safety, Health, and Work-Life Balance.” Vigneswaran, D. 2007. “Special Report: Fact or Fiction? Chronobiology International 28 (4): 361–70. Examining Zimbabwean Cross-Border Migration to South World Bank. 2014. World Development Indicators: Children at Africa.” Forced Migration Studies Programme and Musina Work. Washington, DC: World Bank. http://wdi.worldbank Legal Advice Office, Johannesburg. .org/tables. Vijayvergiya, S. C., A. K. Bohra, and P. Jhanwar. 2012. “Pattern WTO (World Trade Organization). 2013. International Trade of Agriculture-Related Injuries Presented in a Tertiary Statistics 2013. Geneva: WTO. Care Hospital of South-Eastern Rajasthan.” Journal of Xia, Q. H., Y. Jiang, N. Yin, J. Hu, and C. J. Niu. 2012. “Injury Pharmaceutical and Biomedical Sciences 17 (17). among Migrant Workers in Changning District, Shanghai, Vink, P., E. A. P. Koningsveld, and J. F. Molenbroek. 2006. China.” International Journal of Injury Control and Safety “Positive Outcomes of Participatory Ergonomics in Terms Promotion 19 (1): 81–85. of Greater Comfort and Higher Productivity.” Applied Yang, Y. H., S. H. Liou, C. J. Chen, C. Y. Yang, C. L. Wang, and Ergonomics 37 (4): 537–46. others. 2007. “The Effectiveness of a Training Program Virtanen, P., U. Janlert, and A. Hammarström. 2011. “Exposure to on Reducing Needlestick Injuries/Sharp Object Injuries Temporary Employment and Job Insecurity: A Longitudinal among Soon Graduate Vocational Nursing School Students Study of the Health Effects.” Occupational and Environmental in Southern Taiwan.” Journal of Occupational Health 49 (5): Medicine 68 (8): 570–74. 424–29. Occupation and Risk for Injuries 131 Yessuf Serkalem, S., G. Moges Haimanot, and N. Ahmed Ansha. Zawilla, N. H., and D. Ahmed. 2013. “Sharps Injuries 2014. “Determinants of Occupational Injury in Kombolcha among Health Care Workers in Cairo University Hospitals.” Textile Factory, North-East Ethiopia.” International Journal International Journal of Risk and Safety in Medicine 25 (2): of Occupational and Environmental Medicine 5 (2): 84–93. 79–92. Yu, W., I. T. Yu, Z. Li, X. Wang, T. Sun, and others. 2012. Zhuang, Z., T. J. Stobbe, J. W. Collins, H. Hongwei, and G. R. “Work-Related Injuries and Musculoskeletal Disorders Hobbs. 2000. “Psychophysical Assessment of Assistive Devices among Factory Workers in a Major City of China.” for Transferring Patients/Residents.” Applied Ergonomics 31 Accident Analysis and Prevention 48 (September): 457–63. (1): 35–44. Yu, X. 2007. “Impacts of Corporate Code of Conduct on Labor Zink, K. J., U. Steimle, and D. Schröder. 2008. “Comprehensive Standards: A Case Study of Reebok’s Athletic Footwear Change Management Concepts: Development of a Supplier Factory in China.” Journal of Business Ethics 81 Participatory Approach.” Applied Ergonomics 39 (4): (3): 513–29. 527–38. 132 Injury Prevention and Environmental Health Chapter 7 Household Air Pollution from Solid Cookfuels and Its Effects on Health Kirk R. Smith and Ajay Pillarisetti INTRODUCTION Household air pollution (HAP) is now understood to Since the earliest human times, humans have used be a major risk factor for health. According to the 2013 wood as fuel for fires to cook their food. Indeed, learn- Global Burden of Disease Study (GBD), HAP is ranked ing to control fire is considered the defining moment as the single most significant environmental health risk between the pre-human and human condition factor globally. In poor countries where many households (Wrangham 2009). With the agricultural revolution rely on biomass for cooking (such as in Sub-Saharan some 10,000 years ago, agricultural residues (including Africa), HAP is ranked among the top risk factors exam- animal dung) were brought to the hearth as well. ined in the GBD assessments. Depending on which set Around 1,000 years ago, coal became used in areas of estimates is used, some 3 million to 4 million premature where it was mined easily—for example, the British deaths are thought to be caused annually by HAP. Between Isles and China (Smil 1994). These three fuels—wood, 3 and 5 percent of the GBD in terms of disability-adjusted agricultural residues, and coal—constitute the solid life years (DALYs) is attributed to it, about one-third in cooking fuels used by about 40 percent of humanity children younger than age five years and the rest divided today (Bonjour and others 2013). Typically burned in between adult men and women (for background on DALYs, simple cookstoves, these fuels produce smoke that is see Salomon 2014). now understood to cause a large burden of disease This chapter relies on two major reviews published in (Smith and others 2014). recent years. One was done as part of the Comparative Cleaner fuels (coal gas, natural gas, liquefied petroleum Risk Assessment (CRA) of the GBD project (Lim and oth- gas [LPG], and electricity) began to make inroads only in ers 2012; Lozano and others 2012; Smith and others 2014), the late nineteenth century. Although today 60 percent of and the other was done as background documentation for the world’s population uses these modern fuels (which are the World Health Organization’s (WHO) Indoor Air relatively clean in household use, even in simple cook- Quality Guidelines (IAQGs) (WHO 2014b). This chapter stoves), growth in their use has never kept up with global summarizes what is known about effective and cost- population growth, primarily because of the persistence effective interventions to reduce the health effects of expo- of biomass use among the poor. Today, almost 3 billion sure to HAP from solid cooking fuels1 and then explores people use solid cookfuels, which probably is more than at some of the issues regarding framing, interactions, and any time in world history (Bonjour and others 2013) and viable interventions. The discussion follows the classic more than the entire world population before 1960. environmental health pathway described in box 7.1. Corresponding author: Kirk R. Smith, School of Public Health, University of California, Berkeley, California, United States; krksmith@berkeley.edu. 133 Box 7.1 The Environmental Health Pathway This chapter relies loosely on the classic environmen- deposition and exfiltration of pollutants through tal health pathway for describing and understanding openings. Concentrations are not necessarily equiv- pollution risks (figure B7.1.1), which starts with alent to exposures; for example, a monitor that sources and emissions of pollution, moves to envi- measures pollution in a kitchen (defined here as the ronmental levels, then to human exposures, then to built environment around the cooking area, whether doses within the body, and finally to health impacts indoors or outdoors) for 24 hours does not reflect a (Smith 1987). In the case of household air pollution, person’s exposure to that pollution unless he or she, a source could be any type of biomass combustion, too, is in the kitchen for 24 hours. but we focus here primarily on biomass combustion Exposures are a result of the spatiotemporal rela- used in cooking. Different kinds of evidence come tionship between individuals and the pollution in to bear at each stage of the pathway, and each stage their immediate surroundings. An individual’s daily offers different avenues for control. Because some of exposure is affected by the number, type, and dura- the terminology in the pathway is discipline specific, tion of contact with all sources he or she comes into we must briefly clarify what we mean by emissions, contact with, either directly (for example, one’s own concentration, exposure, biomarkers of exposure, household cooking fire) or indirectly (for example, and biomarkers of effect. local traffic sources or a neighbor’s household cook- Emissions refer to the rate of release of a pollutant ing fire). Exposure can be assessed either through per unit of time or per unit of fuel (the “source” in personal measurement, in which an individual wears figure B7.1.1). Measurements of emissions require a monitor, or through exposure reconstruction, in sampling directly from the source of combustion. which time-activity information (for example, a Emissions samples often are taken during a cooking diary of time spent in various locations and time cycle—either actual or simulated—and rarely are spent in proximity to potential sources) is coupled captured for the entire day. Experience shows that with area monitors measuring concentration in vari- lab measurements or simulated measurements in ous microenvironments. Personal exposure typically homes usually underestimate actual emissions in is assessed for 24 or 48 hours. the field (Johnson and others 2008; Johnson and Biomarkers of exposure are measurable metabolites or others 2011). products of an interaction between an external agent Concentrations (generally measured in mass of pol- and a target molecule, cell, or organ. Biomarkers of lutant per volume of air) are a function of emis- effect are chemical, biological, or physical alterations sions, conditions in the room of interest (such resulting from an exposure that can be associated as the room’s ventilation rate), and processes like with a health endpoint or disease (WHO 1993). Figure B7.1.1 Classic Environmental Health Pathway Health Source Emissions Concentration Exposure Dose effects Biomarkers of exposure Biomarkers of effect 134 Injury Prevention and Environmental Health Most of the older literature and even some modern more typical levels are 5–10 percent (Naeher and others studies refer to the problem as one of indoor air pollu- 2007; Zhang and others 2000). By mass, the largest PIC tion, but the CRA (Lim and others 2012) carefully component by far is carbon monoxide (CO), but thou- redefined it as HAP for several reasons (Smith and sands of other compounds have been measured in wood others 2014): smoke, including nontrivial levels of dozens of well- known toxic chemical species, such as polycyclic aro- • Much of the health-relevant exposure to air pollution matic hydrocarbons, benzene, formaldehyde, and even from cooking fuel occurs in the environment around dioxin (Naeher and others 2007; Northcross and others households, not just indoors. 2012). In broad terms, the mixture is similar to the PIC • Solid cooking fuel is sufficiently polluting to affect produced from combustion of the most well-studied widespread ambient (outdoor) air pollution levels form of biomass: tobacco. Indeed, despite their differ- appreciably and, thus, to cause ill health far from the ences, exposures to these two forms of smoke have many source. similar health effects. • The term indoor implies that an effective chimney As with tobacco smoke, the risks of different diseases or other venting would solve the problem entirely, resulting from exposure to HAP probably depend in when the basic problem is dirty combustion near different ways on the landscape of components. However, people. insufficient evidence exists to pin specific diseases on • In some parts of the world, incompletely combusted particular components of wood smoke. Indeed, given solid fuels are commonly used for space heating or the many decades, more controlled conditions, and lighting, as well as for cooking, thus confusing the extensive resources devoted to studying tobacco smoke, attribution of risk and assessment of appropriate still without being able to distinguish differences in interventions unless the household uses being con- detail, the issue of wood smoke mixtures is unlikely to be sidered are specified. resolved in the foreseeable future. Therefore, like tobacco • The term indoor air pollution overlaps with much researchers, HAP researchers rely on two main indicator research on indoor pollution from other sources pollutants for measurement and risk assessment: PM2.5 (for example, from household furnishings and con- (particulate matter with an aerodynamic diameter of less sumer products). For example, the CRA now sepa- than 2.5 micrometers, called tar in tobacco smoke, the rately includes risks from indoor exposure to radon. most well-studied component of air pollution correlated with adverse health risk) and CO. Unlike tobacco smoke, This chapter focuses on the evidence base for health smoke from other types of biomass does not contain effects, because the causality between HAP and ill health measurable nicotine. However, smoke resulting from is only now being firmly established. This is unlike con- biomass combustion contains a vast range of other com- taminated water and poor sanitation, for which the ponents for which PM2.5 and CO are just indicators.2 connection to ill health was established in the nineteenth In terms of PM2.5, a typical wood fuel cookstove used century. The causality and scale of the effects from HAP by a single family for cooking household meals produces have only recently received recognition in health effects substantial pollution by any comparison. In laboratory studies, which are now appearing in large numbers. This simulations, the wood-fired three-stone stove (the most recent appearance perhaps explains why there are rela- common stove used worldwide) produces some 6 grams tively few evaluations of large-scale interventions to date. or about 400 cigarettes worth of PM2.5 per hour Initiatives presently under way provide excellent oppor- (figure 7.1) (Jetter and others 2002; Jetter and others tunities to do so. 2012). To put it into another context, one year of cook- ing on a three-stone stove emits particles equivalent to the emissions of 20 diesel trucks driving 50,000 kilome- ters a year and meeting Euro 6 standards, the standard SOURCES AND EMISSIONS planned for India in 2020. Considering that 170 million Burning biomass completely in simple stoves is extremely households in India use biomass cooking fuel today, the difficult. Even though wood and most other types of emissions are roughly equivalent to those of a mixed biomass have few intrinsic contaminants (unlike coal), fleet of 400 million diesel trucks meeting 2010 stan- substantial fractions of the fuel carbon are not com- dards (Euro 4), far more emissions than are expected in pletely oxidized to carbon dioxide; instead, they are India. In practice, field-based measurements of both converted to a vast range of products of incomplete biomass stoves and diesel trucks likely record even more combustion (PIC). As much as 20 percent of the fuel pollution than is indicated by these numbers (which are carbon can be diverted into these products, although based on laboratory evidence). Household Air Pollution from Solid Cookfuels and Its Effects on Health 135 Figure 7.1 Emissions of PM2.5 in Grams per Hour for Common Types of Stoves Showing Range of Reported Lab Measurements Three-stone stove, minimally tended Three-stone stove, carefully tended Rocket stove I Rocket stove II Charcoal stove II Charcoal stove I Fan-assisted stove Pellet stove 0 2 4 6 PM2.5 (g/hour) Source: Adapted from the comprehensive stove performance database in Jetter and others 2012; adapted with permission from Jetter, Zhao, Smith, Khan, Yelverton, Decarlo, and Hays. “Pollutant Emissions and Energy Efficiency under Controlled Conditions for Household Biomass Cookstoves and Implications for Metrics Useful in Setting International Test Standards.” Environmental Science and Technology 46 (19): 10827–34. Copyright 2012. American Chemical Society. Note: Data displayed are for dry fuel during hot start tests. g/hour = grams per hour; PM2.5 = particulate matter with an aerodynamic diameter of less than 2.5 micrometers. Toxicology studies of biomass particulates find some thousands of micrograms per cubic meter of PM2.5 and effects on cells and animals to be stronger than those causing much eye and throat irritation, particularly in produced by typical ambient air pollution or diesel par- persons unaccustomed to such levels (Diaz and others ticles and some to be weaker, with no clear trends 2007).4 The iconic blackening of walls and ceilings in vil- (Naeher and others 2007; Zelikoff and others 2002). lage kitchens using such fuels is testimony to these levels. Growing epidemiological evidence suggests that diesel Although few systematic survey data are available, particles are likely to be more hazardous than average including those from the Demographic and Health ambient particles or wood smoke particles, but all major Survey, the World Health Organization (WHO 2014a), assessments to date—for example, those of the WHO and the World Bank, worldwide only a small fraction of (2014b) and the U.S. Environmental Protection households using biomass for cooking have working Agency—conclude that, at present, insufficient evidence chimneys. The exception is China, where most rural exists to treat PM2.5 of different origins differently with kitchens have chimney stoves, partly because of the regard to control priorities.3 success of the largest stove dissemination program in history, the National Improved Stove Program (NISP), which operated from the early 1980s to the mid-1990s (Edwards and others 2007; Smith and others 2014; Zhang CONCENTRATIONS and Smith 2007). Unlike India’s National Program on In indoor kitchens, PM2.5 concentrations can Improved Chulhas, which operated during roughly the become extremely high when cooking with solid fuels same period (Venkataraman and others 2010), all stoves (Balakrishnan and others 2011), often reaching many disseminated under the NISP had chimneys. 136 Injury Prevention and Environmental Health Good chimney stoves lower peak levels of indoor intrusive exercise given the available technology. Early pollution, but they lower long-term average exposures studies commonly measured exposure only during peri- only by a factor of two, at most, because even good chim- ods of cooking, when exposure rates often are highest. ney stoves do not intrinsically reduce emissions; they These levels are hard to interpret, because relative risks merely move emissions out of the immediate room and and exposure-response relationships typically exist for into the surrounding household and village environment, annual average exposures, not for exposures only during where people also spend time. They further require regu- cooking, heating, or other activities. lar maintenance and proper use to function correctly. Evaluation of exposure is made more difficult by high Of course, a chimney does nothing to decrease out- within- and between-household variability (McCracken door air pollution, which is now understood to be and others 2009; Pillarisetti and others 2016). Several heavily influenced by household sources in some coun- parameters can influence both concentration and expo- tries. In India, for example, an estimated 25–50 percent sure, including (1) the cooking location, with some of population-weighted outdoor PM2.5 exposure results households cooking indoors, while others cook out- from emissions of primary particles from cookstoves doors, in an open area, or in a separate cooking house; (Chafe and others 2014; Guttikunda 2016; Lelieveld (2) cooking habits and type of cuisine, with some cui- and others 2015). Outdoor PM2.5 levels also include sines requiring constant attention during cooking, while secondary particles from gaseous precursors, such as others can be left unattended; and (3) the use of multiple sulfur oxide, nitrogen oxide, and semivolatile com- fires for cooking. Each of these parameters influences pounds; though not yet well quantified, these com- exposure and complicates exposure assessment. pounds also are emitted from households, as well as In the past, researchers generally assumed that as from vehicles, power plants, and other more traditional long as measurement days were typical of patterns sources of outdoor air pollution. If one considers throughout the year, then one or a few days of mea- primary particle emissions alone, household cooking surements would provide reasonable estimates of long- is responsible for an estimated 370,000 premature term averages. In recent years, however, because of high deaths globally from its contribution to general out- intrinsic intrahousehold variability, researchers have door air pollution on top of the mortality produced demonstrated that reliable estimates of long-term aver- from exposure in the household environment itself ages can be achieved only with multiple days of mea- (Chafe and others 2014). surement (McCracken and others 2009; Pillarisetti and others 2016). Although studies have detected effects even with one or a few measurements, investigators risk not being able to do so even when effects exist EXPOSURE because of the high degree of exposure misclassifica- Household cooking is nearly universally done by women, tion that occurs. who often also are responsible for the care of young chil- Additional methods of measuring exposure involve dren. These two groups generally have the highest HAP measurements of “surrogate” pollutants, such as CO, or exposure, because they tend to be near the stove during reconstruction of exposures using area measurements in combustion. As cooking fire smoke permeates the house- multiple microenvironments and time-activity diaries. hold environment, men and older children also may Exposure surrogates may be chosen because they facili- have significant exposure. However, studies have not tate more rapid or less difficult measurement of a spe- characterized these exposures nearly as well. The impor- cific pollutant. However, the decision to measure a tance of focusing on exposure (as opposed to just indoor surrogate in place of the pollutant of interest requires air pollution) is illustrated by the fairly high exposure of local, field-based validation of the surrogate as a proxy women cooking on open fires outdoors. for the pollutant. Exposure reconstruction using micro- Because monitors placed in a kitchen or living area environmental models relies on area measurements of cannot capture actual human exposure from a single pollutant concentrations in multiple environments in location (particularly for different family members), the which people spend time (for instance, kitchens, the growing practice in epidemiological and other health- outdoors, and living quarters), as well as recall or oriented research on HAP is to measure personal expo- sensor-based data on the time spent in each location. sures. This is generally done by asking participants to Individual exposures are then estimated by estimating wear portable monitoring devices for a day or in 24-hour time-weighted average pollutant concentrations (see increments for several days (Baumgartner and others Balakrishnan and others [2011] for a database of HAP 2011; Ni and others 2016; Smith and others 2010; Van studies using proxy measures and time-activity Vliet and others 2013), an expensive and somewhat methods). Household Air Pollution from Solid Cookfuels and Its Effects on Health 137 BIOMARKERS AND OTHER SIGNS OF found similar ranges of effects for each of various health HAP EFFECT outcomes in different populations, providing consider- able confidence that a degree of effect likely is real. Recent reviews (Smith and others 2014; Tolunay and A brief description of each category of disease for which Chockalingam 2012; WHO 2014b) discuss studies that there is epidemiological evidence follows. For a detailed have found biomarkers of HAP exposure (CO breath, literature review, see Smith and others (2014). carboxyhemoglobin, urinary metabolites, DNA [deoxyri- bonucleic acid] adducts) and biomarkers of HAP effect Acute Lower Respiratory Infection in Children (eye opacity, lung function, blood pressure, electrocardio- Acute lower respiratory infection (ALRI) is a leading gram ST-segment). These findings are consistent with the killer of children younger than age five years (GBD Risk disease endpoints documented for HAP and provide Factors Collaborators 2015). Smith and others (2014) support for interpolating between ambient air pollution identified 24 studies that met their inclusion criteria and smoking exposures for cardiovascular outcomes. during a systematic review and meta-analysis. Very few of the studies directly measured exposure to HAP, and many used poor-quality proxies of exposure. HEALTH IMPACTS Furthermore, the case definitions of pneumonia varied among studies. All studies save one randomized control The health impacts of air pollution exposures of various trial (RCT) were observational. The pooled odds ratio sorts, including from household fuels, are based on two (OR) from their study was 1.78 (1.45, 2.18). general categories of evidence: Although several trials are near completion, results from just one RCT have been published to date: the • Direct epidemiological studies of health impacts RESPIRE study of child ALRI in Guatemala, which in populations exposed to differing categories of compared a wood-fired cookstove with a chimney to the exposure traditional open wood-fired cookstove (Smith and others • Interpolation of risks taken from integrated exposure- 2010; Smith and others 2011). Results (summarized in response (IER) functions derived by combining the figure 7.2) show a significant effect for severe forms of results of epidemiological studies of a wide range of ALRI, but only marginally significant effects for all cases air pollution exposures in different situations. of ALRI. Of relevance is that the pilot studies justifying the conclusion that this stove would be an effective inter- Relying heavily on recent major reviews, this section vention focused on indoor air quality in the kitchen and summarizes the results of both kinds of evidence as they not on personal exposures. In the RCT, kitchen concen- relate to HAP and discusses their relative merits and trations dropped 90 percent, similar to the pilot results, remaining gaps and uncertainties. The focus is on out- but the actual exposure experienced by women and comes ranked as Class I, indicating multiple high-quality young children dropped only 50 percent, which was epidemiological studies from households in low- and below the power of the study. This is because babies and middle-income countries (LMICs), with consistent mothers do not spend all day in the kitchen, and the loca- results and particle exposures at both higher and lower tions where people spend time during the rest of the day exposures, and using exposure-response data across were not appreciably affected by the intervention. The several particle exposure settings. See Table 7.1, where wood-fired cookstove with a chimney moved most of the Class I is defined. smoke out of the kitchen and into the surrounding envi- ronment, where it still affected people and their expo- sures. The most important result of the RCT was the Direct Epidemiological Studies of HAP Exposures exposure-response analysis, enabled by the development Most health studies of HAP published to date have relied of a means to measure infant exposures directly and facil- on simple binary indicators of exposure, such as whether itated by a validated relationship between CO and PM2.5. a household’s primary cooking fuel is clean versus dirty fuel. Although simplistic, these indicators are more sta- Chronic Obstructive Pulmonary Disease ble over a year than a single measurement of personal Chronic obstructive pulmonary disease (COPD), the exposure or area of concentration. Most of the evaluated fourth leading cause of death globally (GBD Risk studies are cross-sectional in design, which poses the risk Factors Collaborators 2015), is characterized by persis- of bias by unmeasured confounders (such as socioeco- tent airflow limitation associated with chronic inflam- nomic status, smoking, and fuel/stove stacking). Many mation of the airway and lungs in response to exposure dozens of studies done by different investigators have to particles and gases (GOLD 2016). A previous 138 Injury Prevention and Environmental Health systematic review and meta-analysis evaluating the risk Figure 7.2 Relationships between Carbon Monoxide Exposure in of adult COPD from exposure to HAP identified 24 Children and Pneumonia and Severe Pneumonia from the RESPIRE studies from 12 countries as suitable for inclusion Trial in Guatemala (Smith and others 2014). The majority were cross- a. All cases of pneumonia sectional (17), 6 were case-control studies, and 1 was a retrospective cohort. All but two studies had positive pneumonia rate (per 100 child-years) risk ratios. Stratifying by gender indicated a stronger effect in women (OR, 2.30; 1.73, 2.06) than in men (OR, 80 Physician-diagnosed 1.90; 1.15, 3.13); a subanalysis of duration of exposure indicated a stronger summary effect when comparing 60 the longest to the shortest duration of exposure. All studies used proxy measures of exposure. The pooled OR reported was 1.94 (1.62, 2.33). 40 Lung Cancer 20 Lung cancer (LC) is the seventh leading cause of death globally (IHME 2016). While the use of coal for heating 0 1 2 3 4 5 6 7 and cooking is recognized as a group I carcinogen by Child carbon monoxide exposure (parts per million) the International Agency for Research on Cancer, use of biomass for cooking is considered only a probable carcino- b. Severe cases of pneumonia gen because of weaker epidemiological evidence, even pneumonia rate (per 100 child-years) 60 though several chemicals with group I status are found in Physician-diagnosed severe wood smoke. Smith and others (2014) identified 14 studies, 50 providing 13 individual estimates in a review of the rela- tionship between biomass use for cooking and LC (Bruce 40 and others 2015). Ten studies were focused in Asia, with the 30 remaining four spread across Canada, Europe, and the United States. Exposure assessment relied on survey-based 20 recall of the type of fuel used for cooking or heating, along with the duration and period of life for which biomass was 10 used in a subset. The overall OR was 1.17 (1.01, 1.37) for biomass used for cooking or heating and 1.15 (0.97, 1.37) 0 1 2 3 4 5 6 7 for cooking only. ORs were 1.21 (1.05, 1.39) and 1.95 (1.16, Child carbon monoxide exposure (parts per million) 3.27) for men and women, respectively, for studies with Group mean CO Intervention Control adequate adjustment and a reference category. Source: Adapted from Smith and others 2011; reprinted from The Lancet, Vol. 378, Smith, McCracken, Cataracts Weber, Hubbard, Jenny, Thompson, Balmes, Diaz, Arana, and Bruce, “Effect of Reduction in Household Air Pollution on Childhood Pneumonia in Guatemala (RESPIRE): A Randomised Controlled Trial.” 1717–26, Cataracts (the clouding of the lens of the eye, preventing 2011, with permission from Elsevier. the passage of light) are a leading cause of blindness glob- Note: CO = carbon monoxide. Shaded areas represent 95% confidence bounds. During the RESPIRE ally and account for approximately 0.12 percent of all trial, CO was validated for this study population as a surrogate for particulate matter exposure, which is thought to be the best metric of hazard. The dashed lines represent the mean exposure levels. DALYs (GBD Risk Factors Collaborators 2015). Toxicological evidence from animal models and epidemi- ological evidence from smokers indicated a potential chimneys were introduced rapidly in areas with no relationship between cooking with solid fuels and cata- chimneys and in one county starting around 1980 also racts. Smith and others (2014) identified seven eligible are an important part of the evidence base (Chapman studies providing eight estimates for review, all from India and others 2005; Seow and others 2014; Shen and others and Nepal. The pooled OR was 2.64 (1.74, 3.50); however, 2009). As they relate to coal smoke, however, their direct evidence for men was deemed insufficient for cataracts to relevance to the much more prevalent use of biomass be listed as a class I outcome. Therefore, only the estimate fuel worldwide is not clear, although they do show sig- for women of 2.47 (1.61, 3.73) was deemed reliable. Table nificant reductions in LC as well as COPD and adult 7.1 summarizes the ORs for primary disease outcomes. pneumonia. Unfortunately, too little exposure assess- Although not RCTs, a set of retrospective studies of a ment was conducted to include these results in the devel- “natural experiment” in China in which coal stoves with opment of IER functions. Household Air Pollution from Solid Cookfuels and Its Effects on Health 139 Table 7.1 Summary of Odds Ratio for Primary Outcomes Derived from the Systematic Review and Meta-Analysis Performed for the 2010 Comparative Risk Assessment of the Global Burden of Disease 2010 systematic review and Outcome Group studieda meta-analysis estimates 2004 CRA estimates Acute lower respiratory infection Children 1.78 (1.45, 2.18) 2.3 (1.9, 2.7) Chronic obstructive pulmonary disease Females 2.30 (1.73, 2.06) 3.2 (2.3, 4.8) Males 1.90 (1.15, 3.13) 1.8 (1.0, 3.2) Lung cancer Coal Females 1.98 (1.16, 3.36) 1.94 (1.09, 3.47) Males 1.31 (1.05, 1.76) 1.51 (0.97, 2.46) b Biomass Females 1.95 (1.16, 3.27) — b Males 1.21 (1.05, 1.39) — Cataracts Females 2.47 (1.63, 3.73) — Note: — = not available; CRA = comparative risk assessment; OR = odds ratio. a. Children younger than age five years; females and males ages 15 years and older. b. ORs from Bruce and others 2015. Interpolation of Risks Using Integrated-Exposure others 2009). The complete list of data points used to Response Functions create the model is in Burnett and others (2014, supple- IER functions were created spanning the range of global mentary material table S1). exposures to PM2.5 by separately modeling the relation- In using a wide range of concentrations from a vari- ship between exposure from four sources (ambient air ety of sources, the IERs assume that risk associated with pollution, secondhand smoke, HAP, and active tobacco these disparate sources is only a function of exposure, smoking) and five health endpoints (COPD, stroke, not smoke type, enabling the creation of a continuous heart disease, LC in adults, and ALRI in children younger response function that spans many orders of magnitude than age five years) (Burnett and others 2014; Pope and and is bounded on the low end by ambient exposure to PM2.5 and on the high end by active tobacco smoking Figure 7.3 Integrated-Exposure Response Curves Relating Exposure (Burnett and others 2014; Pope and others 2009). The to PM2.5 to Health Endpoints Associated with Exposure to Air Pollution modeled relative risks are thus a function of PM2.5 expo- sures in terms of mass concentration; all PM2.5 particles 4 are considered equally damaging to health. The resulting functions are highly nonlinear for all outcomes except LC (figure 7.3). Use of the IERs enabled estimation of the risk associ- 3 Relative risk ated with exposures at levels common in households that use solid fuel for which there are no or very few HAP studies, but that have intermediate exposures 2 between passive and active smoking. Additionally, it enabled use of the same idealized counterfactual level of approximately 7 micrograms per cubic meter for calcu- 1 lating the burden of disease attributable to HAP and ambient air pollution. Finally, it enabled comparison of 0 125 250 375 500 625 750 875 1,000 IER-modeled risk estimates with estimates backed by 3 PM2.5 exposures (μg/m ) evidence based on epidemiological studies (Smith and ALRI COPD IHD LC Stroke others 2014). A brief description of the modeled risk estimates for cardiovascular disease (CVD) (including Source: Adapted from Burnett and others 2014. stroke and heart disease) follows, along with a compari- Note: PM2.5 = particulate matter with an aerodynamic diameter of less than 2.5 micrometers; μg/m3 = micrograms per cubic meter. Includes ischemic heart disease (IHD), stroke, chronic obstructive pulmonary son of IER-modeled and epidemiological-study-based disease (COPD), and lung cancer (LC) in adults and acute lower respiratory infection (ALRI) in children. estimates for ALRI. 140 Injury Prevention and Environmental Health Cardiovascular Disease Because children do not smoke, the upper bound of Although evidence exists linking exposure to HAP with exposures in the IER for ALRI are from RESPIRE. biomarkers with known links to cardiovascular outcomes (including blood pressure and heart rate variability), few Uncertainties and Emergent Issues studies have specifically addressed CVD directly. The The health effects literature contains both uncertainties strong evidence of impacts at lower (ambient) and higher as well as new understandings with regard to exposure (active tobacco smoking) levels is good evidence for an patterns that are influencing both research and inter- effect at the intermediate levels of HAP exposure, how- vention policies. ever. Figure 7.4 indicates that, for both stroke and ische- mic heart disease (IHD), risk flattens as exposure increases, Categories of Evidence: Exposure-Response although this effect is more pronounced for stroke. RCTs have substantial cachet in international health, and their results are beginning to inform the evidence base Acute Lower Respiratory Infection in Children for HAP effects as well. However, RCTs are not as valu- Unlike CVD outcomes, both exposure-response and able or needed for HAP assessments as perhaps they are many categorical analyses found that exposure to HAP for other risk factors. Unlike the important risk factors in was associated with child ALRI. The IER for ALRI was this volume that otherwise have many conceptual informed by studies of ambient air pollution, second- similarities—poor water, sanitation, and hygiene—HAP hand smoke, and HAP. Unlike other IERs, the one for has a measurable exposure metric linked directly to ALRI contains directly measured risk and exposure data health. Exposure units in, for example, micrograms per from RESPIRE, based on repeated measures of child per- cubic meter annual levels, can thus be translated across sonal exposure to CO, which were then converted to PM populations and interventions. Indeed, this is the con- (McCracken and others 2013; Smith and others 2010). cept on which the IAQGs are based (WHO 2014b). Figure 7.4 Integrated-Exposure Response Curves for Cardiovascular Outcomes a. Ischemic heart disease 2.0 Relative risk 1.5 1.0 0 250 500 750 1,000 3 Annual average PM2.5 exposure (μg/m ) b. Stroke 2.0 Relative risk 1.5 1.0 0 250 500 750 1,000 3 Annual average PM2.5 exposure (μg/m ) Source: Adapted from Burnett and others 2014. Note: PM2.5 = particulate matter with an aerodynamic diameter of less than 2.5 micrometers; μg/m3 = micrograms per cubic meter. Shaded areas are model-based uncertainty bounds. Large uncertainties in areas approximating household air pollution exposures (300–1,000 micrograms per cubic meter) indicate a lack of evidence in those exposure ranges. Household Air Pollution from Solid Cookfuels and Its Effects on Health 141 HAP RCTs alone, however, are idiosyncratic to the ends of the exposure spectrum and bolstered by results local situation and intervention and cannot meet the for environmental tobacco smoke thus seems justifiable. full requirements of an RCT, particularly the require- It is not credible that exposures that produce CVD effects ment to have placebo controls. Most important, unlike at both higher and lower levels would not also produce exposure-response relationships, the results do not CVD effects at the levels found for HAP. Extrapolation translate easily to any other population—that is, they do beyond the available data is fraught with potential prob- not relate directly to exposure. This is one reason that lems, but a major reason to do graphs is to be able to do exposure assessment is so fundamental to environmen- interpolation. Direct HAP studies of CVD risk factors, tal health. No RCTs have been done or likely will be such as blood pressure and heart rate variability, further done for ambient air pollution, for example, but the support the existence of CVD effects, but they do not effects are known and the benefits of interventions for a themselves allow an estimate of the total CVD effect. place that has never had a health study can be estimated Although they are a major advance, the IERs include with reasonable confidence if exposures are known. This assumptions and show relationships that still need inves- is because multiple large-scale exposure-response results tigation. Three issues bear mention here. First, although are available across the world. RCTs are most valuable three of the types of pollution are composed almost for establishing causality if it is still in doubt, but they entirely of combustion particles, and although ambient are rather poor at informing policy for HAP. What mat- air pollution typically is composed mostly of combus- ters is how clean the fuel has to be to make a difference tion particles, different types of combustion, fuels, and and how much the clean cooking technology displaces mixtures of other pollutants are involved with each. old, polluting technologies; that is, how much does Diesel exhaust is different from tobacco smoke, for exposure have to be reduced to achieve a meaningful example, although both can be measured using PM2.5. health benefit, which is best informed by exposure- Second, the typical exposure patterns reflecting exposure response analysis (Peel and others 2015). to these different sources (both daily and over a lifetime) RCTs can greatly improve exposure-response results, are quite different, even if they can be reduced to a com- however. Although randomizing exposure itself in real mon metric of an annual average. Third, studies use a populations is essentially impossible, randomizing one different measure of exposure for each category of pol- important cause of variability lessens the burden of lution. Ambient air pollution studies use ambient con- potential confounders, increasing confidence in the centrations measured in central locations, such as on the results. In addition, the intervention spreads out the roof of a building in a major metropolitan area. Measured exposures more than occurs naturally and thus increases changes of this type are found to reflect changes in actual the chance of seeing effects. Exposure-response results exposures but are poor representations of absolute expo- have also been improved by the introduction of new sures. People do not live on top of buildings (where these means to measure the sources of high intrahousehold central site monitors are located), but they do live near variability—in particular, the recent wide-scale introduc- small sources that may not affect widespread ambient tion of stove use monitors (Pillarisetti and others 2014; levels but that do affect individual exposures. Few stud- Ruiz-Mercado, Canuz, and Smith 2012; Ruiz-Mercado ies of secondhand tobacco smoke and HAP in adults and others 2013). These and other technical advances have measured personal exposure, but some have tried promise to reduce exposure misclassification further and to estimate levels based on fixed monitors or models. to enhance the ability to detect effects. Studies of active tobacco smoking use inhaled smoke An additional advantage of framing HAP effects in levels or nominal dose (as measured by smoking terms of exposure is the ability to combine effects across machines) to estimate “exposure” per cigarette. other major sources of air pollution into IERs. The same effects are found in a monotonically increasing trend with Other Endpoints estimated exposure, and this provides a new class of evi- Smith and others (2014) carefully assessed the evidence dence that supports results in all the other categories of base for each outcome (disease) associated with HAP. As exposure (ambient air pollution, secondhand tobacco shown in table 7.2, three classes were established. smoke, and active tobacco smoking), but particularly Diseases in Class I were considered to have sufficient HAP. Indeed, compared to ambient air pollution and evidence to be included as formal outcomes in the CRA. active tobacco smoking, many fewer studies are available Class II diseases had a sufficient number of epidemio- for all adult outcomes, and almost none are available for logical studies to conduct meta-analyses, which are two important CVD outcomes—IHD and stroke. found in Smith and others (2014), but the evidence was Interpolation along the IER function that is fixed by active not considered consistent or otherwise convincing tobacco smoking and ambient air pollution at the two enough to be put forward as part of the formal burden 142 Injury Prevention and Environmental Health Table 7.2 Evidence Classes Evidence class Description Criteria Class IA Quantified primary outcome, based Multiple epidemiological studies of good quality in households in lower-income on binary exposures in systematic countries sufficient for meta-analyses; consistent results as well as significant and reviews and meta-analyses positive summary estimate; supporting epidemiological studies of other particle exposures, both at higher and lower exposures Class IB Quantified primary outcome, Exposure-response data available from several particle exposure settings, allowing continued development of integrated exposure-response function covering (1) child ALRI, where studies have found that active tobacco smoking does not contribute, but studies have been conducted in the other three exposure settings (outdoor air pollution, secondhand smoke, and household air pollution); (2) CVD outcomes, where studies for outdoor air pollution, secondhand smoke, and active tobacco smoking exist, allowing estimates to be interpolated for HAP Class II Quantified secondary outcome Multiple epidemiological studies in households in lower-income countries sufficient for meta-analyses; unconvincing adjustment for confounding or exposure assessment; inconsistent results or nonsignificant positive result; supporting epidemiological studies from other particle exposures Class III Nonquantified secondary outcome Still thought likely to be causal; weak or insufficient epidemiological studies from households in lower-income countries for meta-analyses; some support from other particle exposure categories Source: Adapted from Smith and others 2014. Note: ALRI = acute lower respiratory infection; CVD = cardiovascular disease; HAP = household air pollution. All evidence classes have plausible physiological mechanisms based on toxicology. of HAP. These included adult ALRI; tuberculosis; and active smoking risks for the five main diseases asso- nasopharyngeal carcinomas; tumors of the larynx, oro- ciated with each. pharynx, and hypopharynx; cervical cancer; and still- birth. Diseases in Class III were considered to have suggestive, but insufficient, evidence for quantification. INTERVENTIONS These diseases included asthma and preterm birth. This section examines the traditional paradigms that The CRA Expert Group found sufficient evidence to have dominated thinking over the past half century consider low birth weight as an outcome for HAP expo- regarding how to accelerate the transition to clean cook- sures, but the GBD project itself removed low birth ing technologies in poor populations. It then discusses weight as an outcome, focusing instead on preterm more recent paradigms.5 births. However, too few HAP studies had separated preterm birth from low birth weight for these outcomes to be included in the official CRA. (See Smith and others Old Paradigms [2014] for a discussion of the available literature.) Let Development Take Care of It A class of impacts not considered in the CRA consists Because the rich use clean fuels and the poor use dirty of neurocognitive outcomes in children, although evi- fuels (Bonjour and others 2013), one may be tempted dence of such effects has been growing, as has evidence simply to let development take care of the problem. of such effects with other pollution exposures (Smith Unfortunately, this has not worked. About the same and others 2014). This is an active area of HAP research. number of people (almost 3 billion) are using dirty fuels All of the outcomes in Classes I, II, and III are firmly today as 25 years ago, in spite of the considerable devel- associated with tobacco smoking, a much more thor- opment that has occurred in that time. More people are oughly studied source of exposure to biomass smoke. using clean fuels (gas and electricity), but the absolute Thus, the fact that the same diseases that are associated burden of exposure has not changed appreciably world- with tobacco smoking have also been found to be associ- wide. However, the trends in the absolute numbers using ated with HAP exposures (albeit at lower risk levels) is solid fuels varies by region: going up in Sub-Saharan not surprising. Indeed, the IERs provide quantitative Africa, going down in East Asia (China), and remaining evidence of the consistent relationship between HAP level in South Asia (Bonjour and others 2013). Household Air Pollution from Solid Cookfuels and Its Effects on Health 143 Make the Available Clean were made possible only by development of the IERs. Since the large national stove programs were initiated in Now a base of epidemiological evidence exists to support China and India in the early 1980s, perhaps a dozen standards that quantify what emissions level is clean other national efforts and hundreds (if not thousands) enough for good health (WHO 2014b). As mentioned, of community and nongovernmental organization emissions reductions alone do not guarantee exposure (NGO) programs, small and large, have been initiated reductions; rather, interventions must be adopted, main- worldwide to promote better stoves using the same local tained, and used regularly to achieve meaningful expo- fuels (mainly different forms of biomass).6 Although sure reductions to protect health. initially focused on fuel efficiency, many of these pro- As part of the evidence review for the IAQGs (WHO grams are also attempting to lower smoke levels—that is, 2014b), systematic reviews and meta-analyses were per- to make the available fuels clean through better combus- formed of the international literature on interventions tion, chimneys, and others approaches. As described in (Dherani and others 2014). The methods and results are WHO (2014b), this improvement has been extremely summarized in box 7.2. This review found that solid- elusive, and finding interventions that have reduced fuel stoves with chimneys delivered the largest reduc- health-related exposures substantially and sustainably tions in PM and CO concentrations, with CO levels for a large population is difficult. Nevertheless, much often reaching WHO air quality guidelines. However, progress has been made, and investments are needed to none achieved PM levels close to the guidelines. One key continue upgrading the engineering, business, and social issue is the degree of heterogeneity between studies. For marketing required to reach this goal. this reason, referring to the circumstances and results of One promising development is the parallel work of individual stove and fuel evaluations is important for the International Standards Organization and the WHO appropriate interpretation of these results. Continued to develop standards and guidelines for promoting only efforts are needed to standardize the methods used for the cleanest devices in the future. Quantitative guidelines field evaluation. Box 7.2 Assessment of Improved Biomass Stove Interventions To assess the potential health benefits that can be do postintervention (in-use) levels compare with expected following the introduction of improved the WHO air quality guidelines? solid-fuel stoves, one must examine the reductions in HAP and personal exposure—and the absolute Methods. A search was conducted of electronic levels achieved—when these interventions are in databases and specialist websites. Eligible studies everyday use. Although the results of laboratory included randomized trials, quasi-experimental and emissions tests provide valuable information on the before-and-after studies, as well as observational potential reductions in exposure, field evaluations designs and reported daily mean (24- or 48-hour) provide a more realistic assessment of exposure small PM (most reported PM2.5, but two studies when such interventions are adopted and used reported PM4) or CO, with standard deviations at scale. The key questions for the review were as or 95 percent confidence intervals. Interventions follows: were categorized as standard combustion solid-fuel stoves with and without chimneys, advanced com- • Are improved solid-fuel stoves in everyday use bustion solid-fuel stoves, clean fuels (LPG, biogas, (compared to traditional solid-fuel stoves) effec- ethanol, electricity, solar), and mixed interventions. tive for reducing average concentrations of, Studies were selected, extracted, and assessed using or exposure to, PM and CO in households in standardized procedures and forms. Baseline and LMICs? postintervention values, differences, and percent- • By what amount (in absolute and relative terms) age changes from baseline were tabulated for each do the interventions reduce PM and CO, and how study, and weighted average values were calculated box continues next page 144 Injury Prevention and Environmental Health Box 7.2 (continued) for all studies contributing data to each category Despite these large percentage reductions, post- of stove or fuel intervention. Subject to sufficient intervention levels of PM remained well above studies, meta-analysis of absolute changes in the two the WHO guidelines for group-weighted means at pollutants for each category of solid-fuel stove and around 400 micrograms per cubic meter, although clean fuel was carried out using the generic inverse- the few personal exposure studies had a considerably variance method, and publication bias was assessed. lower weighted mean of 70 micrograms per cubic Narrative summaries were provided for intervention meter. In contrast, many interventions reduced categories with very few eligible studies. CO to levels below the WHO 24-hour air quality guideline, with weighted mean values of 4–5 parts Results. A total of 38 eligible studies, some with per million for stoves with chimneys, but almost multiple estimates, was included: 27 studies that 7 parts per million for stoves without chimneys. provided data on kitchen PM, 3 on personal PM, Postintervention personal exposure in the set of 26 on kitchen CO, and 5 on personal CO. Only one chimney-stove studies was 1.7 parts per million. or two studies were available for each intervention Sensitivity analyses (conducted where the number of (LPG, electricity, charcoal, mixed). Baseline levels estimates was sufficient), including by study design, of PM and CO were variable, but all exceeded the analytic approach (that is, comparing controls with annual WHO guideline for PM2.5 of 35 micrograms only stoves in actual use or with all stoves allocated), per cubic meter by a factor of 10–100 times, and and duration of use, did not find strong effects. CO varied from just below to 6 times greater than Among the larger sets of studies, clear evidence of the 24-hour air quality guideline for CO of 7 milli- publication bias existed. Evidence from studies of grams per cubic meter (5.68 parts per million). After improved wood stoves in high-income rural set- intervention, reductions in pollutants were reported tings found, as expected, PM2.5 levels much lower for almost all individual studies; when grouped, than those of improved wood stoves in developing large reductions in the range of 38–82 percent were countries (ranging from 13 to 54 micrograms per found for kitchen PM and CO levels, with the largest cubic meter) and an association between improved reductions for solid-fuel stoves with chimneys and solid-fuel stoves (all of which were vented with some the lowest for solid-fuel stoves without chimneys. having advanced emissions control technology) and Studies reporting impacts on personal exposure were emissions reductions in a majority of households. identified only for solid-fuel chimney stoves, but Source: Based directly on Dherani and others (2014), which also contains lengthy reductions in the range of 47–76 percent were found. tables describing the published studies to date. New Paradigms attractive modern cooking appliances being an important Based on the new evidence that exposures must be sales advantage in most settings. They are not available to brought to low levels to achieve major health benefits, the populations using biomass, however, not only because and the poor performance of “improved” biomass stoves of their cost but also because of unreliable or unavailable to date, new paradigms are emerging in the field, public and private infrastructure. Any kind of gas burns although they have been operating on their own in the cleanly, including biogas and natural gas, but LPG is usu- modern energy sector all along. ally the first to reach rural areas. Rather than simply wait- ing passively for people to shift to clean fuels, there is clear Make the Clean Available need to find ways to promote these fuels to poor house- How one achieves clean cooking is no mystery. Gas and holds in a more systematic and aggressive manner. electricity are used by 60 percent of humanity, and these Several large and innovative initiatives for promoting fuels cook every cuisine without problem (although with LPG began in India in 2015. Although initiated by the taste changes compared to traditional methods for some Ministry of Petroleum and Natural Gas in collaboration foods). Unlike typical biomass stoves, gas and electric with the three national oil companies that market LPG, stoves cannot be made dirty at the household level (even these initiatives were driven by a desire to reduce the with nonoptimal use), and they do not require any special health impacts of solid fuel use for cooking. As of March attention or training. They also are aspirational, with 2016, more than US$1 billion had been committed Household Air Pollution from Solid Cookfuels and Its Effects on Health 145 to expand LPG to 50 million low-income households One major reason that clean fuel interventions do not in three years, reaching perhaps 300 million people show greater reductions is the remaining use of polluting (Ministry of Finance 2016). This ambitious initiative has fuels either in the same household or nearby, which has several innovative features designed to target LPG subsi- not been monitored well in past studies. More and dies much more precisely to poor households and away better-designed studies are needed for all kinds of clean from middle- and upper-class and commercial consum- fuel interventions, as well as new intervention modes ers. These features involve the use of modern digital that promote usage and initial adoption. technology, including bank accounts, cell phones, and biometric identification cards. In addition, widespread Embrace Leap-Frog Technologies integrated use of formal and social media to promote the Highly advanced, electronic devices are now available for effort exists, including text messages, billboards, televi- cooking. Depending on the task, electric induction stoves sion and radio, Internet, and athletic events. are 50 percent more efficient and 50 percent faster (as well The most well-known of the programs is the Give It Up as safer and longer-lived) than old-style electric stoves. scheme, through which middle- and upper-income con- They are so different as to provide a new entry into the sumers are asked to give up their LPG subsidy to house- cookstove landscape. Sales are booming in Asia, and holds below the poverty line. Households who give up prices are dropping, reaching as low as US$10 each in their subsidy are listed on a Scroll of Honor on the website some markets. Most of the sales growth is occurring and can see which family benefited from their contribu- among customers now using gas, as cooking with induc- tion.7 Some 30,000 households a day were doing so at the tion stoves is sometimes cheaper than cooking with sub- height of the program. As a result, the government was sidized LPG. How far might induction stoves be pushed able to focus new resources on providing the up-front costs into rural areas when electricity supply becomes more (stove and cylinder) to enable poor households to take on reliable? Ecuador, for example, is replacing every stove in LPG, and oil companies were incentivized to expand fuel the country with an induction stove, and other countries access substantially and to improve the reliability of supply. with excess hydropower are considering taking such an In addition, new modes of distributing LPG are being approach. Could induction stoves be linked to local power tried, including promotion and sales by women’s groups. made from renewable energy sources? This is an exciting Importantly, the government has specified that all new prospect. Even when linked to coal power, induction LPG connections since early 2016 are to be in the name of stoves create substantially less pollution exposure and the woman of the house wherever possible, a significant only minor increases in greenhouse gases (Smith 2014). movement toward improving gender engagement. Because Synthetic liquid or gaseous fuels such as the bio- shifting the subsidy from one income group to another ethanol discussed earlier and synthetic LPG made from does not entail additional government expenditure, the coal, which are clean at the household, also show promise cost-effectiveness of this effort depends only on the addi- but require additional study and evaluation of system tional expenditures for up-front costs.8 requirements. Synthetic natural gas from coal is also being Another approach is to promote clean fuels that have promoted in China and Mongolia but requires extensive not been widely adopted in high-income countries and pipeline infrastructure that makes it cost-prohibitive in thus have no established operational viability. The most most rural areas. prominent of these is biogas; although attractively clean and made renewably from animal dung, biogas is limited Target the Community Level in scope by climate, capital cost, and the need for at least Ongoing research and modeling show that, in many cir- two large animals in each household. Second is ethanol, cumstances, changing one household in a village to clean which burns cleanly and can be made renewably from fuels reduces exposure less than one might expect (Desai several crops, including sugarcane and sorghum. 2016; Smith 1987). This is because of a coverage or com- Unknown, however, is whether large-scale production munity effect—that is, even if you cook using LPG (or do would trigger demand in other sectors (for example, as a not cook at all), you are affected by all of your neighbors petroleum enhancer or beverage) that would dominate who still cook on biomass stoves. Although varying by its availability and price as a fuel. geography and meteorology, most of humanity lives in The review described in box 7.2 also examined avail- fairly close quarters, whether in cities or villages, and the able studies for clean fuel interventions. It found none for community effect is common. For this reason, the most electric cooking and too few for LPG or biogas to make effective interventions are likely to occur at the commu- an assessment. However, it found several for ethanol that nity level. This has two other advantages: providing fuels, indicated a reduction in overall exposure, but not enough stoves, and service at the community scale usually is to reach WHO air quality guidelines (WHO 2014b). more efficient, lowering costs and increasing reliability, 146 Injury Prevention and Environmental Health and it is possible to unleash social pressure to change world to address household and ambient air pollution social norms—for example, creating a smokeless village together by proposing a program to manage exposure, designation to encourage neighbors to work together not concentrations, in particular locations (Sagar and (put pressure on each other) to avoid producing smoke others 2016). If implemented, this approach would focus in their village. Indeed, these other benefits of commu- more on pollution sources that are in close proximity to nity interventions are likely to be the most critical. people (stoves and vehicles) and less on sources that are The LPG initiatives in India are promoting “smoke- far from people (power plants and industries). less villages” designed to develop LPG connections by village rather than by household. As of mid-2016, at least Household Air Pollution as a Health Problem 4,000 smokeless villages (defined as 100 percent of Part of the poor progress of previous attempts to reduce households being connected) had been certified, with HAP may be due to their origins in the technology sector thousands more being planned. rather than the health sector and the heavy emphasis the As with much of the rapid changes in the “make the technology sector places on simple local technologies and clean available” agenda, however, evaluation of smoke- community groups or NGOs. In contrast, the health less villages and other modes of LPG expansion have not sector taps the very best advanced scientific, technological, yet been subjected to high-quality evaluation, something and manufacturing techniques to develop effective vac- clearly needed. cines, antibiotics, and surgery tools; then, after those tech- niques have been proved worthwhile in highly structured field trials, the health sector makes them available through Recent Innovations prepurchase, royalty agreements, and mass manufacture New ways of thinking have emerged from the literature to reduce the cost. It then uses NGOs and other commu- but have not yet been well integrated into interventions. nity groups to bring the vaccines to vulnerable popula- Among these is growing recognition of the following. tions. Unlike the technology sector, the health sector treats everyone the same; it does not promote less effective Impact on Outdoor Pollution antibiotics in rural areas because the people there are poor. A major reason that the field has moved away from the Unequal treatment may be satisfactory when addressing term indoor to household air pollution is the realization fuel efficiency or meeting local labor and materials goals, that, although pollution may start in the kitchen, it which are important issues in their own right. The tech- moves throughout the household, then into the commu- nology sector is less effective at achieving health goals, and nity environment outside, where it adds to general ambi- its priorities raise disquieting ethical issues. ent air pollution. The degree to which this matters One reason that often is given for continued cooking depends on the situation; in India, for instance, as noted, on open fires is the taste of the food, but the health sector an estimated 25–50 percent of primary ambient PM2.5 would ask, is taste worth nearly one million lives a year comes from household cooking. Estimates are similar for in India? The health sector does not stop its programs China, although household use of solid fuel for heat is because people like the taste of tobacco or dislike wear- also seasonally important in much of the country (Liu ing seatbelts or dislike using condoms. It recognizes the and others 2016). Cleaning up household fuels clearly is importance of personal preferences, however, and brings a necessary step in dealing with outdoor pollution. social pressure to bear in an effort to change those tastes. Because outdoor air pollution has become a serious pol- The health sector has already recognized the importance icy and public concern in many countries that still have of various kinds of “herd” effects—for example, with significant household use of solid fuel, this connection sanitation and mosquito protection. First is effectiveness provides a potential impetus for control programs and a on a large scale, which has often been promoted in bio- framework for evaluation. mass stove programs. Next is the household business In late 2015, India’s Ministry of Health and Family model, which may come later. Finally, the health sector is Welfare released a white paper proposing a pioneering not afraid of subsidies but provides the evidence needed approach to air pollution (Ministry of Health and to prove that expenditures on the health of the poor are Family Welfare 2015). It was the first ministry of health cost-effective social investments. in the world to consider air pollution in the context of other health priorities, with the idea of using the health sector’s unique assets to address it (air pollution gener- Common Challenges with Interventions ally has been handled by environmental agencies, which Although many relatively small-scale, low-cost interven- have a different agenda). India’s Ministry of Health and tions (such as the provision of better-burning biomass Family Welfare is also the first government agency in the stoves) will continue, efforts to reach households at a Household Air Pollution from Solid Cookfuels and Its Effects on Health 147 large scale using existing infrastructure in the petroleum Compliance and Stacking and power sectors are growing, but in ways that better The second issue is stacking, which refers to the common focus on health. India is leading the way, but other coun- observation that people often do not switch to a new tech- tries have programs or are planning them.9 nology immediately, even if it is better in many ways and With LPG or electricity, little HAP concern exists, eventually takes over. In the case of cooking, people often because the appliances that use these cooking fuels can continue to use their traditional fuel stove even if they also stand up to variations in user behavior, and the appliances’ use an advanced biomass stove or LPG. It may take years, performance is well known, with billions in use over many in the case of LPG, before they switch entirely, a process decades. Even with the most advanced biomass stoves, that has a generational component—young women often however, good field performance is difficult to maintain, do not continue what their mothers find hard to give up. even when the stoves are used regularly. Two common As a result, with a new clean fuel alternative in the difficulties remain, however, with both approaches: cost home, all of the HAP exposure is due to continued and continued use of traditional polluting cookstoves. use of the traditional stove, and the exposure can be substantial. This is a familiar situation in health inter- Cost and Subsidy ventions: simply providing access and affordability does An advanced biomass stove with a chimney and blower not guarantee high compliance (for example, in using (characteristics likely to be required for good health) is bed nets, condoms, latrines, tuberculosis drugs, low-salt not cheap by low-income-country standards. The cost of foods, and nicotine substitutes). In most of these exam- the stove alone is likely to be more than US$100 and ples, as with HAP, a high rate of compliance is needed to probably closer to US$200, as seen in successful chimney reduce risk adequately (seemingly more than 90 percent stove programs in China and Mexico, for example. The in the case of latrines and bed nets, for example). costs of dissemination, maintenance, repair, and replace- Accordingly, as with every other health intervention that ment add to this. In addition, to date, only biomass pellet must be accompanied by behavioral change, incentives stoves are reliably clean enough to come close to the must be found and implemented to enhance compliance IAQGs for emissions. Pellet stoves require users to forfeit (Fernald, Gertler, and Neufeld 2009; Fernald, Hou, and the greatest advantage of today’s biomass fuel, which is Gertler 2008; Lim and others 2010) or, in stove parlance, that it can be gathered at no direct financial cost. to reduce the degree and duration of stacking. Financing the stoves and pelletizing infrastructure in Additional research is needed to find ways to promote ways that are sustainable for poor populations is a major reduced use of the old and increased adoption and use of challenge. the new. Recent systematic reviews of adoption and barri- The same is true of providing LPG or electric power ers to adoption of clean stoves (Puzzolo and others 2016) reliably and sustainably. Electrification offers a way to and of clean fuels and electricity (Rehfuess and others spread costs, given its many social and other benefits in 2014) highlight this need. They also challenge dissemina- addition to health. Up-front costs are substantially lower tion approaches that only market the new, as might be than for nearly equivalently clean biomass stoves, but adequate for economic sustainability. Imagining a business electricity entails recurring costs and access issues. model for eliminating the old polluting stove, however, is The very poor are unlikely to be able to afford any difficult, a phenomenon that is not uncommon with truly clean cooking technology. If significant progress is household or individual health interventions. to occur, some form of public support likely will be Approaches for triggering community pressure (for needed for some years. This is not unusual: public sup- example, conditional cash transfers and cell phone mes- port is accepted for many health-protective interven- saging) have been applied successfully in other situations tions for the poor, including vaccines, antenatal care, and and could successfully reduce HAP as well. In addition, basic antibiotics. The term subsidy often is applied to some innovations show promise even if they have never public support for fuels, usually in a pejorative way. been applied—for example, linking the use of HAP- However, subsidies for nuclear power, the coal industry, reducing cooking technology, such as LPG, to national and the solar industry are not intended to target health life insurance and rural employment schemes, as is being protection for the poor, unlike support for HAP-reducing considered in India. technologies such as LPG and advanced biomass stoves. Thus, if public expenditures can be shown to be as well targeted and effective as other expenditures on CONCLUSIONS health-protective interventions, they may be considered The health impacts of HAP have been suspected for social investments rather than subsidies, with a substan- decades, beginning with a few isolated studies more than tially different political and developmental connotation. a half century ago (Padmavati and Pathak 1959); only 148 Injury Prevention and Environmental Health recently, sufficient evidence has been marshaled to make health research, the two together can help to find ways to a systematic case for HAP’s ill health effects across a provide the world with clean household environments range of diseases. This evidence is substantiated best in effectively and steadily. the two detailed reviews used so extensively in this chap- Providing empirical evidence of the cost-effectiveness ter (Smith and others 2014; WHO 2014b). The concep- of alternative interventions is difficult, although there is tual and empirical connection between active and passive movement in this direction (Newcombe and others tobacco smoking and ambient air pollution provided by 2016; Pillarisetti, Mehta, and Smith 2016). The long- the IERs gives rise to a completely different and in itself term solution is clear: clean fuels (although they will not compelling set of arguments for HAP’s ill health effects, be available for the very poorest populations for some in addition to the growing base of epidemiological and years). Until then, however, the evident popularity of toxicological evidence. Although the evidence is insuffi- such fuels could be a model for how improved biomass cient to pin down a precise risk for all diseases now stoves are designed and disseminated.10 Only now are we attributed to HAP exposures or to establish a firm base beginning to understand how to bring clean fuels to the for diseases that have some, but insufficient, evidence to poor (but not the poorest) populations much faster than include on the list, it seems likely that HAP will remain development alone has brought, while simultaneously on the list of severe health risks affecting the world’s accelerating the movement away from traditional prac- poorest populations. tices during the transition. HAP will continue to constitute a major risk factor as long as billions of households worldwide use solid fuels. However, simply believing it to be a major risk is NOTES not sufficient to bring solutions. As noted in the intro- duction, fecal matter in the household environment World Bank Income Classifications as of July 2014 are as fol- lows, based on estimates of gross national income (GNI) per was confirmed as a major risk factor for ill health in the capita for 2013: late 1800s, but it still kills millions today in spite of considerable efforts to reduce this health burden. Both • Low-income countries (LICs) = US$1,045 or less of these risk factors share uncomfortable similarities: • Middle-income countries (MICs) are subdivided: they are significant, operate in poor populations, and a) lower-middle-income = US$1,046 to US$4,125 require behavioral and engineering innovations and b) upper-middle-income (UMICs) = US$4,126 to US$12,745 interventions. They both also seem to be refractory to • High-income countries (HICs) = US$12,746 or more. cheap solutions. How can we be sure then that HAP (which only passed the threshold of acceptability in 1. This chapter focuses almost entirely on wood fuel, which 2010 or so, and even still perhaps not as completely as dominates world use and research. Agricultural residues, fecal contamination) is not still killing millions a including animal dung, are far less consistent and less century from now? well characterized. Coal pollution is even more difficult Although basic epidemiological, toxicological, and to summarize because of wide variations in the quality of coal around the world, including the content of toxic spe- exposure research continues, HAP’s threshold of accept- cies, such as sulfur, arsenic, lead, mercury, ash, and others. ability has been passed, and serious research is needed to For a good discussion, see WHO (2014b). determine what works on a large scale. Regarding poor 2. Few studies have been conducted on the impact of sanitation, the failure to move in this direction is per- chronic CO exposures on health, and CO in wood haps partly responsible for the long delay between recog- smoke rarely causes acutely toxic exposures because of nition of the problem and its solution. Considering the the warning of extreme irritation from other wood smoke question of scale—at the household level and the insti- components. Therefore, this chapter does not explore tutional level, in terms of the agencies and organizations CO exposure further, though we note observed links that can operate on a large scale and perform careful between exposure to CO during pregnancy and adverse monitoring and evaluation of natural interventions—is outcomes. However, low-volatile solid fuels, particularly another way to frame this effort. This mode of thinking charcoal and some coals, can produce acutely hazardous CO exposures. Indeed, despite the dearth of systematic is particularly salient for those efforts now under way assessments, thousands of deaths likely occur globally with clean fuels (such as LPG and electricity) in India, each year (some even in high-income countries) as a Bhutan, Paraguay, Ecuador, and elsewhere. As HAP has result of CO exposure (for example, from charcoal grills the advantage of a measurable exposure metric, much of used indoors). this research can proceed more quickly and with less cost 3. Kerosene, another middle distillate like diesel, is still used because exposure outcomes can be used as endpoints. If, for household cooking in some countries and is widely in parallel, exposure-response is emphasized in the used for lighting in hundreds of millions of households Household Air Pollution from Solid Cookfuels and Its Effects on Health 149 without adequate electricity. Growing evidence suggests submitted to the Environmental Health Sciences Graduate that, by mass, PM2.5 from kerosene combustion is more Group, University of California, Berkeley, 88–147. toxic than PM2.5 from biomass combustion. To date, how- Dherani, M., K. Jagoe, L. P. Naeher, and C. Noonan. 2014. ever, the WHO has been unable to do more than recom- “Review 6: Impacts of Interventions on Household Air mend that kerosene be discouraged as a household fuel. Pollution Concentrations and Personal Exposure.” In WHO (See WHO 2014b.) Indoor Air Quality Guidelines: Household Fuel Combustion, 4. The WHO guideline for annual average PM2.5 concentra- edited by E. Rehfuess, D. Pope, and N. Bruce. Geneva: tions is 10 micrograms per cubic meter (WHO 2014b). World Health Organization. 5. This section draws on Smith and Sagar (2014) and Smith Diaz, E., T. Smith-Sivertsen, D. Pope, R. T. Lie, A. Diaz, (2015), as well as on WHO (2014b). and others. 2007. “Eye Discomfort, Headache, and Back 6. See http://cleancookstoves.org/. Pain among Mayan Guatemalan Women Taking Part 7. See http://mylpg.in/. in a Randomised Stove Intervention Trial.” Journal of 8. In 2016, the Give It Up program was folded into an even Epidemiology and Community Health 61 (1): 74–79. larger program to promote a total of 50 million LPG con- Edwards, R. D., Y. Liu, G. He, Z. Yin, J. Sinton, and others. 2007. nections in India in three years. “Household CO and PM Measured as Part of a Review 9. See http://www.cooking-for-life.org/. of China’s National Improved Stove Program.” Indoor Air 10. IAQGs has a section on the needs of the very poor in the 17 (3): 189–203. transition to clean fuels for all. Fernald, L. C., P. J. Gertler, and L. M. Neufeld. 2009. “10-Year Effect of Oportunidades, Mexico’s Conditional Cash Transfer Programme, on Child Growth, Cognition, Language, and Behaviour: A Longitudinal Follow-Up REFERENCES Study.” The Lancet 374 (9706): 1997–2005. Balakrishnan, K., G. Thangavel, S. Ghosh, S. Sambandam, Fernald, L. C., X. Hou, and P. J. Gertler. 2008. “Oportunidades K. Mukhopadhyay, and others. 2011. The Global Household Program Participation and Body Mass Index, Blood Air Pollution Measurement Database. Geneva: World Health Pressure, and Self-Reported Health in Mexican Adults.” Organization. http://www.who.int/indoorair/health_impacts Preventing Chronic Disease 5 (3): A81. /databases/en/. GBD Risk Factors Collaborators. 2015. “Global, Regional, and Baumgartner, J., J. J. Schauer, M. Ezzati, L. Lu, C. Cheng, and National Comparative Risk Assessment of 79 Behavioural, others. 2011. “Patterns and Predictors of Personal Exposure Environmental, Occupational, and Metabolic Risks to Indoor Air Pollution from Biomass Combustion among or Clusters of Risks in 188 Countries, 1990–2013: A Women and Children in Rural China.” Indoor Air 21 (6): Systematic Analysis for the Global Burden of Disease 479–88. Study 2013.” The Lancet 386 (10010): 2287–323. Bonjour, S., H. Adair-Rohani, J. Wolf, N. G. Bruce, S. Mehta, GOLD (Global Initiative for Chronic Obstructive Lung and others. 2013. “Solid Fuel Use for Household Cooking: Disease). 2016. Global Strategy for the Diagnosis, Country and Regional Estimates for 1980–2010.” Management and Prevention of COPD. GOLD. http://www Environmental Health Perspectives 121 (7): 784–90. .goldcopd.org/. Bruce, N., M. Dherani, R. Liu, H. D. Hosgood III, A. Sapkota, and Guttikunda, S. 2016. Urban Emissions. http://www.urbanemissions others. 2015. “Does Household Use of Biomass Fuel Cause .info/. Lung Cancer? A Systematic Review and Evaluation of the IHME (Institute for Health Metrics and Evaluation). 2016. Evidence for the GBD 2010 Study.” Thorax 70 (5): 433–41. “GBD Compare Data Visualization.” IHME, University Burnett, R. T., C. A. Pope III, M. Ezzati, C. Olives, S. S. Lim, and of Washington, Seattle. http://vizhub.healthdata.org/gbd others. 2014. “An Integrated Risk Function for Estimating -comparevizhub.healthdata.org/gbd-compare. the Global Burden of Disease Attributable to Ambient Jetter, J. J., Z. Guo, J. A. McBrian, and M. R. Flynn. 2002. Fine Particulate Matter Exposure.” Environmental Health “Characterization of Emissions from Burning Incense.” Perspectives 122 (4): 397–403. Science of the Total Environment 295 (1–3): 51–67. Chafe, Z. A., M. Brauer, Z. Klimont, R. Van Dingenen, S. Mehta, Jetter, J., Y. Zhao, K. R. Smith, B. Khan, T. Yelverton, and others. and others. 2014. “Household Cooking with Solid Fuels 2012. “Pollutant Emissions and Energy Efficiency under Contributes to Ambient PM2.5 Air Pollution and the Burden Controlled Conditions for Household Biomass Cookstoves of Disease.” Environmental Health Perspectives 122 (12): and Implications for Metrics Useful in Setting International 1314–20. Test Standards.” Environmental Science and Technology Chapman, R. S., X. He, A. E. Blair, and Q. Lan. 2005. 46 (19): 10827–34. “Improvement in Household Stoves and Risk of Chronic Johnson, M., T. Bond, N. Lam, C. Weyant, Y. Chen, and Obstructive Pulmonary Disease in Xuanwei, China: others. 2011. “In-Home Assessment of Greenhouse Gas and Retrospective Cohort Study.” BMJ 331 (7524): 1050. Aerosol Emissions from Biomass Cookstoves in Developing Desai, M. A. 2016. “Model of Postulated Coverage Effect Countries.” In Greenhouse Gas Strategies in a Changing from Clean Cooking Interventions.” In Multiscale Drivers Climate Conference 2011, 530–42. Pittsburgh: Air and Waste of Global Environmental Health. Doctoral dissertation Management Association. 150 Injury Prevention and Environmental Health Johnson, M., R. Edwards, C. A. Frenk, and O. Masera. 2008. Intervention Study.” Environment International 94: 449–57. “In-Field Greenhouse Gas Emissions from Cookstoves doi:10.1016/j.envint.2016.05.029. in Rural Mexican Households.” Atmospheric Environment Northcross, A. L., S. K. Hammond, E. Canuz, and K. R. Smith. 42 (6): 1206–22. 2012. “Dioxin Inhalation Doses from Wood Combustion in Lelieveld, J., J. S. Evans, M. Fnais, D. Giannadaki, and A. Pozzer. Indoor Cookfires.” Atmospheric Environment 49 (March): 2015. “The Contribution of Outdoor Air Pollution Sources 415–18. to Premature Mortality on a Global Scale.” Nature 525 Padmavati, S., and S. N. Pathak. 1959. “Chronic Cor (7569): 367–71. Pulmonale in Delhi: A Study of 127 Cases.” Circulation Lim, S. S., L. Dandona, J. A. Hoisington, S. L. James, 20 (3): 343–52. M. C. Hogan, and others. 2010. “India’s Janani Suraksha Peel, J. L., J. Baumgartner, G. A. Wellenius, M. L. Clark, and Yojana, A Conditional Cash Transfer Programme to K. R. Smith. 2015. “Are Randomized Trials Necessary Increase Births in Health Facilities: An Impact Evaluation.” to Advance Epidemiologic Research on Household Air The Lancet 375 (9730): 2009–23. Pollution?” Current Epidemiology Reports 2 (4): 263–70. Lim, S. S., T. Vos, A. D. Flaxman, G. Danaei, K. Shibuya, and Pillarisetti, A., L. W. H. Alnes, J. P. McCracken, E. Canuz, and others. 2012. “A Comparative Risk Assessment of Burden of K. R. Smith. 2016. “Long-Term PM2.5 in Kitchens Cooking Disease and Injury Attributable to 67 Risk Factors and Risk with Wood: Implications for Measurement Strategies.” Factor Clusters in 21 Regions, 1990–2010: A Systematic Environmental Science and Technology 48: 14525–533. Analysis for the Global Burden of Disease Study 2010.” The Pillarisetti, A., S. Mehta, and K. R. Smith. 2016. “HAPIT, the Lancet 380 (9859): 2224–60. Household Air Pollution Intervention Tool, to Evaluate the Liu, J., D. L. Mauzerall, Q. Chen, Q. Zhang, Y. Song, and others. Health Benefits and Cost-Effectiveness of Clean Cooking 2016. “Air Pollutant Emissions from Chinese Households: Interventions.” In Broken Pumps and Promises: Incentivizing A Major and Underappreciated Ambient Pollution Source.” Impact in Environmental Health, edited by E. A. Thomas, Proceedings of the National Academy of Sciences 113 (28): 7756– 147–70. Cham, Switzerland: Springer International 61. http://www.pnas.org/cgi/doi/10.1073/pnas.1604537113. Publishing AG. Lozano, R., M. Naghavi, K. Foreman, S. Lim, K. Shibuya, and Pillarisetti, A., M. Vaswani, D. Jack, K. Balakrishnan, others. 2012. “Global and Regional Mortality from 235 M. N. Bates, and others. 2014. “Patterns of Stove Usage after Causes of Death for 20 Age Groups in 1990 and 2010: Introduction of an Advanced Cookstove: The Long-Term A Systematic Analysis for the Global Burden of Disease Application of Household Sensors.” Environmental Science Study 2010.” The Lancet 380 (9859): 2095–128. and Technology 48 (24): 14525–33. McCracken, J. P., J. Schwartz, N. Bruce, M. Mittleman, Pope, C. A. III, R. T. Burnett, D. Krewski, M. Jerrett, Y. Shi, L. M. Ryan, and others. 2009. “Combining Individual- and others. 2009. “Cardiovascular Mortality and Exposure and Group-Level Exposure Information: Child Carbon to Airborne Fine Particulate Matter and Cigarette Smoke: Monoxide in the Guatemala Woodstove Randomized Shape of the Exposure-Response Relationship.” Circulation Control Trial.” Epidemiology 20 (1): 127–36. 120 (11): 941–48. McCracken, J. P., J. Schwartz, A. Diaz, N. Bruce, and K. R. Smith. Puzzolo, E., D. Pope, D. Stanistreet, E. Rehfuess, and N. G. Bruce. 2013. “Longitudinal Relationship between Personal CO and 2016. “Clean Fuels for Resource-Poor Settings: A Systematic Personal PM2.5 among Women Cooking with Woodfired Review of Barriers and Enablers to Adoption and Sustained Cookstoves in Guatemala.” PLoS One 8 (2): e55670. Use.” Environmental Research 146 (April): 218–34. Ministry of Finance. 2016. Union Budget 2016–2017. Rehfuess, E. A., E. Puzzolo, D. Stanistreet, D. Pope, and New Delhi: Ministry of Finance, Government of India. N. G. Bruce. 2014. “Enablers and Barriers to Large-Scale Ministry of Health and Family Welfare. 2015. Report of the Uptake of Improved Solid Fuel Stoves: A Systematic Steering Committee on Air Pollution and Health-Related Review.” Environmental Health Perspectives 122 (2): 120–30. Issues. New Delhi: Ministry of Health and Family Welfare, Ruiz-Mercado, I., E. Canuz, and K. R. Smith. 2012. Government of India. “Temperature Dataloggers as Stove Use Monitors (SUMs): Naeher, L. P., M. Brauer, M. Lipsett, J. T. Zelikoff, Field Methods and Signal Analysis.” Biomass and Bioenergy C. D. Simpson, and others. 2007. “Woodsmoke Health 47 (December): 459–68. Effects: A Review.” Inhalation Toxicology 19 (1): 67–106. Ruiz-Mercado, I., E. Canuz, J. L. Walker, and K. R. Smith. 2013. Newcombe, K., T. Ramanathan, N. Ramanathan, and E. Ross. “Quantitative Metrics of Stove Adoption Using Stove Use 2016. “Innovations in Payments for Health Benefits of Monitors (SUMs).” Biomass and Bioenergy 57 (October): Improved Cookstoves.” In Broken Pumps and Promises: 136–48. Incentivizing Impact in Environmental Health, edited Sagar, A. D., K. Balakrishnan, S. Guttikunda, A. Roychowdhury, by E. A. Thomas, 171–79. Cham, Switzerland: Springer and K. R. Smith. 2016. “India Leads the Way: A Health- International Publishing AG. Centered Strategy for Air Pollution.” Environmental Health Ni, K., E. Carter, J. J. Schauer, M. Ezzati, Y. Zhang, and others. Perspectives 124 (7): A116–17. 2016. “Seasonal Variation in Outdoor, Indoor, and Personal Salomon, J. 2014. “Disability Adjusted Life Years.” In Air Pollution Exposures of Women Using Wood Stoves in Encyclopedia of Health Economics, edited by A. J. Cuyler, the Tibetan Plateau: Baseline Assessment for an Energy 200–03. San Diego, CA: Elsevier. Household Air Pollution from Solid Cookfuels and Its Effects on Health 151 Seow, W. J., W. Hu, R. Vermeulen, H. D. Hosgood III, Tolunay, H. E., and A. Chockalingam. 2012. “Indoor and G. S. Downward, and others. 2014. “Household Air Outdoor Air Pollution and Cardiovascular Health.” Global Pollution and Lung Cancer in China: A Review of Studies Heart 7 (2): 87–196. in Xuanwei.” Chinese Journal of Cancer 33 (10): 471–75. Van Vliet, E. D. S., K. Asante, D. W. Jack, P. L. Kinney, Shen, M., R. S. Chapman, R. Vermeulen, L. Tian, T. Zheng, and R. M. Whyatt, and others. 2013. “Personal Exposures to others. 2009. “Coal Use, Stove Improvement, and Adult Fine Particulate Matter and Black Carbon in Households Pneumonia Mortality in Xuanwei, China: A Retrospective Cooking with Biomass Fuels in Rural Ghana.” Environmental Cohort Study.” Environmental Health Perspectives 117 (2): Research 127 (November): 40–48. 261–66. Venkataraman, C., A. D. Sagar, G. Habib, N. Lam, and K. R. Smith. Smil, V. 1994. Energy and World History. Boulder, CO: Westview 2010. “The Indian National Initiative for Advanced Biomass Press. Cookstoves: The Benefits of Clean Combustion.” Energy for Smith, K. R. 1987. Biofuels, Air Pollution, and Health: A Global Sustainable Development 14 (2): 63–72. Review. New York: Plenum Publishing. WHO (World Health Organization). 1993. Biomarkers and ———. 2014. “In Praise of Power.” Science 345 (6197): 603. Risk Assessment: Concepts and Principles, Environmental ———. 2015. “Changing Paradigms in Clean Cooking.” Health Criteria. Geneva: WHO, International Programme EcoHealth 12 (1): 196–99. on Chemical Safety. Smith, K. R., N. Bruce, K. Balakrishnan, H. Adair-Rohani, ———. 2014a. Global Health Observatory (database): Population J. Balmes, and others. 2014. “Millions Dead: How Do We Using Solid Fuels, by Country. WHO, Geneva. http://apps Know and What Does It Mean? Methods Used in the .who.int/gho/data/node.main.135?lang=en. Comparative Risk Assessment of Household Air Pollution.” ———. 2014b. WHO Indoor Air Quality Guidelines: Household Annual Review of Public Health 35: 185–206. Fuel Combustion. Geneva: WHO Smith, K. R., J. P. McCracken, L. Thompson, R. Edwards, Wrangham, R. W. 2009. Catching Fire: How Cooking Made Us K. N. Shields, and others. 2010. “Personal Child and Mother Human. New York: Basic Books. Carbon Monoxide Exposures and Kitchen Levels: Methods Zelikoff, J. T., L. C. Chen, M. D. Cohen, and R. B. Schlesinger. and Results from a Randomized Trial of Woodfired Chimney 2002. “The Toxicology of Inhaled Woodsmoke.” Journal Cookstoves in Guatemala (RESPIRE).” Journal of Exposure of Toxicology and Environmental Health, Part B: Critical Science and Environmental Epidemiology 20 (5): 406–16. Reviews 5 (3): 269–82. Smith, K. R., J. P. McCracken, M. W. Weber, A. Hubbard, Zhang, J. J., and K. R. Smith. 2007. “Household Air Pollution A. Jenny, and others. 2011. “Effect of Reduction in from Coal and Biomass Fuels in China: Measurements, Household Air Pollution on Childhood Pneumonia in Health Impacts, and Interventions.” Environmental Health Guatemala (RESPIRE): A Randomised Controlled Trial.” Perspectives 115 (6): 848–55. The Lancet 378 (9804): 1717–26. Zhang J. J., K. R. Smith, Y. Ma, F. Jiang, W. Qi, and others. Smith, K. R., and A. Sagar. 2014. “Making the Clean Available: 2000. “Greenhouse Gases and Other Airborne Pollutants Escaping India’s Chulha Trap.” Energy Policy 75 (December): from Household Stoves in China: A Database for Emission 410–14. Factors.” Atmospheric Environment 34 (26): 4537–49. 152 Injury Prevention and Environmental Health Chapter 8 Health Risks and Costs of Climate Variability and Change Kristie L. Ebi, Jeremy J. Hess, and Paul Watkiss INTRODUCTION the middle of this century. Climate change is a stress multiplier, putting pressure on vulnerable systems, pop- The scientific community agrees that climate change is ulations, and regions. For example, temperature is asso- happening, is largely human induced, and will have seri- ciated with the incidence of some food- and water-borne ous consequences for human health (Field and others diseases that are significant sources of childhood mortal- 2014). The health consequences of climate variability ity (Smith and others 2014). Reducing the burden of and change are diverse, potentially affecting the burden these diseases requires improved access to safe water and of a wide range of health outcomes. Changing weather improved sanitation. Poverty is a primary driver under- patterns can affect the magnitude and pattern of mor- lying the health risks of climate change (Smith and oth- bidity and mortality from extreme weather and climate ers 2014). Poverty alleviation programs could improve events, and from changing concentrations of ozone, the capacity of health systems to manage risks and particulate matter, and aeroallergens (Smith and others reduce the overall costs of a changing climate. 2014). Changing weather patterns and climatic shifts Climate change entails other unique challenges: may also create environmental conditions that facilitate alterations in the geographic range, seasonality, and incidence of some infectious diseases in some regions, • The magnitude, pattern, and rate of climate change such as the spread of malaria into highland areas in parts over smaller spatial scales are inherently uncertain. of Sub-Saharan Africa. Changes in water availability and • Weather patterns will continue to change until mid- agricultural productivity could affect undernutrition, century, no matter to what extent greenhouse gas particularly in some parts of Africa and Asia (Lloyd, emissions (which drive climate change) are reduced Kovats, and Chalabi 2011). Although climate change will in the short term. likely increase positive health outcomes in some regions, • The magnitude and pattern of health risks past mid- the overall balance will be detrimental for health and century will be determined largely by the extent to well-being, especially in low- and lower-middle-income which emissions are reduced in coming decades and countries that experience higher burdens of climate- the extent to which health systems are strengthened sensitive health outcomes (Smith and others 2014). to manage current risks and prepare for projected The pathways between climate change and health ones in coming decades (Field and others 2014). outcomes are often complex and indirect, making attri- bution challenging. Climate change may not be the most Significant reductions in greenhouse gas emissions important driver of climate-sensitive health outcomes (mitigation) in the next few years will be critical to prevent- over the next few decades but could be significant past ing more severe climate change later in the century, but Corresponding author: Kristie L. Ebi, Department of Global Health, University of Washington, Seattle, Washington, United States; krisebi@uw.edu. 153 they will have limited effects on weather patterns in the coming decades (Smith and others 2014). Many health short term. In terms of costing, another complexity is that outcomes are affected by weather and climate, as shown these policies and technologies are associated with short- in figure 8.1. The poor and vulnerable in LMICs, partic- term health benefits (Garcia-Menendez and others 2015). ularly children, are and will continue to be affected most. Reducing and managing health risks over the next few Until mid-century, the adverse health risks of climate decades will require modifying health systems to prepare change will mainly be exacerbations of current health for, cope with, and recover from the health consequences problems, with the possibility that diseases (for example, of climate variability and change; these changes are part of vector-borne infections) may extend their geographic what is termed adaptation. Adaptation will be required range into new areas. The largest risks will occur in pop- across the century, with the extent of mitigation being a ulations that are currently most affected by climate- key determinant of health systems’ ability to manage risks related health outcomes (Smith and others 2014). projected later in the century (Smith and others 2014). No Climate change affects health through various matter the success of adaptation and mitigation, residual pathways: risks from climate change will burden health systems, par- ticularly in low- and middle-income countries (LMICs). • Changes in the frequency and intensity of extreme Given these complexities, estimating the costs of man- weather (including heat, windstorms, and heavy rain) aging the health risks of climate variability and change is • Effects mediated through natural systems (for exam- not straightforward. The wide range of health outcomes ple, changes in the geographic range and incidence of potentially affected means counting (1) costs associated infectious diseases, such as water-, food-, and vector- with increased health care and public health interventions borne diseases, and health outcomes associated with for morbidity and mortality from a long list of climate- poor air quality, such as high concentrations of sensitive health outcomes; (2) costs associated with lost ozone and aeroallergens) work days and lower productivity; and (3) costs associated • Effects heavily mediated by human systems (for with well-being. Costs could also accrue from repeated example, occupational impacts, undernutrition, episodes of malaria, diarrhea, or other infectious diseases migration, and mental stress). that affect childhood development and health in later life. Costs associated with actions taken in other sectors are Climate change will affect mean weather variables also important for health, such as access to safe water and (for example, temperature and precipitation); the fre- improved sanitation. A portion of the costs of managing quency, intensity, and duration of some extreme weather the health risks associated with migrants and environ- and climate events; and sea level. Changes in the mean mental refugees could be, but has not been, counted. and variability of weather and climate can indepen- Further, costs and benefits will be displaced over time, dently and jointly influence the burden of climate- with costs associated with increased health burdens sensitive health outcomes. For example, rising mean occurring now because of past greenhouse gas emissions temperatures can create conditions conducive to the and benefits occurring later in the century because of geographic spread of vector-borne diseases such as mitigation implemented in the next few years. A few pre- liminary estimates have been made of the costs of adapta- Figure 8.1 Impacts of Climate Change on Human Health tion. However, more work is needed to understand how climate variability and change could affect the ability of Injuries, fatalities, Asthma, mental health impacts cardiovascular disease health systems to manage risks over long temporal scales. This chapter reviews the health risks of climate variability Severe Air Malaria, dengue, Heat-related illness weather pollution encephalitis, hantavirus, and change, discusses key components of those risks, sum- and death, rift valley fever, marizes the attributes of climate-resilient health systems, cardiovascular failure g M sinratures wore Changes lyme disease, Extreme chikungunya, temp Ri e in vector provides an overview of the costs of increasing health ex ther e heat West Nile virus a trem ecology resilience that arise from other sectors, reviews temporal and e Inc o 2lev spatial scale issues, and summarizes key conclusions regard- Environ- c rea Increasing ng ls mental i i el ng Ris le allergens Respiratory ve s ing the costs of the health risks of climate change. Forced migration, degradation s e a allergies, asthma civil conflict, s mental health impacts Water Water and food supply impacts quality impacts HEALTH RISKS OF CLIMATE VARIABILITY Cholera, AND CHANGE Malnutrition, diarrheal disease cryptosporidiosis, campylobacter, leptospirosis, harmful algal blooms Climate change is affecting morbidity and mortality worldwide, with the risks projected to increase over Source: Slide courtesy of George Luber, CDC. 154 Injury Prevention and Environmental Health malaria. At the same time, heavy precipitation events change can lead to ecological disruptions that indi- can wash away breeding grounds, resulting in short- rectly affect health (for example, by reducing crop term reductions in the number of Anopheles mosquitoes yields and altering the habitat of disease vectors). that can carry malaria. As changes continue over the Social responses to climate change, such as migra- century, thresholds may be crossed that could result in tion, also can affect human health. large increases or decreases in the incidence of climate- • Until mid-century, climate change mainly will exacer- sensitive health outcomes. bate preexisting health problems. New health condi- Figure 8.2 shows the primary exposure pathways tions may emerge, and diseases such as vector-borne for the health risks of climate change. The figure shows infections may extend their geographic range into that mediating factors, including environmental, areas that currently are unaffected. The risks will social, and health factors, affect the burden of climate- be highest in populations most affected by climate- sensitive health outcomes associated with changing related health outcomes, such as in regions that cur- weather patterns. The green arrows at the bottom rently are food insecure. indicate the possibility of positive or negative feedback • Over the past few decades, climate change contrib- mechanisms. uted to the burden of climate-sensitive health out- In the human health chapter of the Working Group II comes; however, the worldwide burden of ill health Contribution to the Intergovernmental Panel on Climate caused by climate change is relatively small compared Change (IPCC) 5th Assessment Report, Smith and oth- with that caused by other stressors and is not well ers (2014) conclude the following: quantified. • The major concerns with climate change include (1) • The health of human populations is sensitive to shifts morbidity and mortality from higher ambient tem- in weather patterns and other aspects of climate peratures and intense heatwaves; (2) higher risk of change. The effects occur directly, because of changes undernutrition from reduced food production in poor in temperature and precipitation and because of the regions; (3) health consequences of lost work capacity occurrence of extreme weather and climate events and reduced labor productivity; and (4) higher risks of (heatwaves, floods, droughts, and wildfires). Climate food-, water-, and vector-borne diseases. Figure 8.2 Conceptual Diagram of the Health Risks of Climate Change Mediating factors Environmental Social infrastructure Public health capability conditions and adaptation Direct exposures • Geography • Warning systems • Baseline weather • Flood damage • Socioeconomic status • Soil/dust • Storm vulnerability • Heat stress • Health and nutrition status Climate change • Vegetation • Primary health care Health impacts • Baseline air/water • Precipitation quality Indirect exposures • Undernutrition • Heat Mediated through natural systems: • Drowning • Floods • Allergens • Heart disease • Storms • Disease vectors • Malaria • Increased water/air pollution Via economic and social disruption • Food production/distribution • Mental stress Source: Figure 11-1 from Smith, K. R., A. Woodward, D. Campbell-Lendrum, D. D. Chadee, Y. Honda, Q. Liu, J. M. Olwoch, B. Revich, and R. Sauerborn. 2014: “Human Health: Impacts, Adaptation, and Co-Benefits.” In Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Field, C. B., V. R. Barros, D. J. Dokken, K. J. Mach, M. D. Mastrandrea, T. E. Bilir, M. Chatterjee, K. L. Ebi, Y. O. Estrada, R. C. Genova, B. Girma, E. S. Kissel, A. N. Levy, S. MacCracken, P. R. Mastrandrea, and L. L. White, eds.]. Cambridge University Press, Cambridge, United Kingdom, and New York, NY, United States. Health Risks and Costs of Climate Variability and Change 155 • Impacts on health will be reduced, but not elimi- The wide range of factors that describe vulnerability nated, in populations that benefit from rapid social to climate-related hazards can be divided into environ- and economic development, particularly among the mental, social, economic, health, and other dimensions poorest. (Cardona and others 2012; Field and others 2012). • The most effective measures to reduce vulnerabil- Environmental dimensions include physical variables ity in the near term are programs that implement (location-specific context for human-environment and improve basic health system measures, such as interactions); geography, location, and place; and settle- providing safe water, improving sanitation, securing ment patterns and development trajectories. Social essential health care, strengthening the capacity for dimensions include demographic variables such as edu- disaster preparedness and response, and alleviating cation and human health and well-being; cultural vari- poverty. ables; and institutions and governance. Cross-cutting • Important research gaps remain regarding the health factors include relevant and accessible science and tech- risks of climate change, particularly in low-income nology. In the health sector, important factors include countries (LICs). the health of the population and the status of health systems (for example, the ability of health care facilities, The magnitude and pattern of risks in future decades laboratories, and other parts of the health system to will depend on actions taken to strengthen the resil- manage an extreme event). ience of health systems to prepare for, cope with, and From the perspective of the health sector, vulnera- recover from changing burdens of climate-sensitive bility is the summation of all risk and protective factors health outcomes and on actions taken to reduce emis- that determine whether an individual or subpopulation sions of greenhouse gases that are driving climate experiences adverse health outcomes from exposure change, sea-level rise, and ocean acidification. (Balbus and Malina 2009). Sensitivity to an event is a measure of the responsiveness of an individual or sub- population to an event, often for biological reasons VULNERABILITY TO THE HEALTH RISKS OF such as the presence of a chronic disease. A rich litera- CLIMATE VARIABILITY AND CHANGE ture describes factors that increase vulnerability to extreme events. Individuals who are low on the socio- The magnitude and pattern of risks from climate change economic scale, children, pregnant women, individuals are due to the characteristics of the hazards created by with chronic medical conditions, and individuals with changing weather patterns, the extent of exposure of mobility or cognitive constraints are at higher risk of human and natural systems to the hazard, the suscepti- adverse health outcomes during an extreme event bility of those systems to harm, and their ability to cope (Balbus and Malina 2009). In addition, the social deter- with and recover from exposure (Field and others 2012; minants of health influence vulnerability. These deter- Steinbruner, Stern, and Husbands 2013). Climate-related minants include access to health care services, access to hazards can alter vulnerability to future events by chang- and quality of education, availability of resources, ing the following (Field and others 2012; Steinbruner, transportation options, social capacity, and social Stern, and Husbands 2013): norms and culture. Figure 8.3 shows the framework used to explore the • Extent of exposure (for example, reducing the pres- key drivers of vulnerability to extreme weather and ence or effectiveness of coastal barriers) climate events in the health sector (Ebi and Bowen • Susceptibility of exposed human and natural systems 2016). Impacts can be categorized into those that affect (for example, making individuals and communities environmental services, social and economic factors, or more or less susceptible by affecting their access to health status and health systems: and the functioning of health care facilities or the proportion of the population vulnerable to an event) • Impacts on environmental services include availabil- • Ability of organizations and institutions to prepare ity of safe water (including quality and quantity), for and manage events effectively and efficiently. food security, and consequences that affect ecosystem services such as wildfires, coastal erosion, and salt- Understanding the magnitude and pattern of impacts water intrusion into freshwater sources. and the factors that increase or decrease susceptibility • Impacts on social and economic factors (such as and coping abilities is vital to modifying current policies community services, livelihoods, and social capital) and to implementing new policies and programs to include economic resources, infrastructure, access to increase resilience to climate change. services, and social capital. 156 Injury Prevention and Environmental Health Figure 8.3 Key Drivers of Health Vulnerability to Extreme of climate change; legal and regulatory systems that pro- Weather and Climate Events tect health; institutional mechanisms, capacities, and structures; accountability; and community participation. Extreme Changes in: weather and Indicators are needed to describe the current baseline climate Capacity Environmental events: to cope services and to measure progress as climate change is incorpo- Intensity and Extent of exposure Social and economic rated into policies and programs. scope factors Persistence Costs are associated with implementing climate- Health status and Timing and Susceptibility health systems resilient policies and programs within each component. sequencing Few efforts have been made to estimate these costs. Some costs will be limited, such as modifying five-year plans to incorporate climate change. Others will likely be signifi- cant, such as developing new products and technologies, Source: Ebi and Bowen 2016. ensuring adequate human and financial resources (par- ticularly in LMICs), or improving infrastructure to ensure that health care facilities can withstand (and • Impacts on health status and health systems include continue to function during) more frequent and intense stress, mental illness as a consequence of the floods and storm surges. Some costs will be ongoing, event or recovery, worsening chronic diseases, and such as the need for regular reassessments of current and undernutrition. projected burdens of climate-sensitive health outcomes. Such assessments require ongoing research and develop- CLIMATE-RESILIENT HEALTH SYSTEMS ment to project the magnitude and pattern of climate- sensitive health outcomes as the climate continues to Preparing for and managing the health risks of climate change, taking into account multiple drivers and adapta- variability and change require strengthening the capacity tion options to reduce risks. Investments in surveillance, of health systems to protect and improve population monitoring, and evaluation will be needed across the health in an unstable and changing climate (WHO century to continue to prepare for and manage changing 2015). To that end, the World Health Organization vulnerabilities and risks. New tools for mapping vulner- (WHO) defines a climate-resilient system as a system ability, modeling future risks, developing scenarios, capable of anticipating, responding to, coping with, evaluating the effectiveness of public health prevention, recovering from, and adapting to climate-related shocks and undertaking other activities are all components of and stresses to bring sustained improvements to the the iterative management of climate change. health of the population. Other costs will be borne primarily by other sectors, Health systems vary across and within countries, but such as developing and deploying new agricultural culti- all share common building blocks: vars that are heat, drought, or salt tolerant. These activi- ties will be critical for ensuring food security over • Leadership and governance coming decades. In the absence of these activities, the • Health workforce costs for health systems to manage risks will be consider- • Health information systems ably higher. • Essential medical products and technologies Table 8.1 lists some possible interventions within • Service delivery each of the six building blocks of health systems, provid- • Financing. ing an overview of the wide range of efforts needed to strengthen resilience. The costs associated with some of Figure 8.4 shows the 10 components for building these activities, such as establishing and maintaining a climate-resilient health systems within these building malaria treatment program, are estimated in other blocks. chapters. Within each component, specific characteristics or Limiting the magnitude of climate change risks activities are needed to achieve resilience. For example, past mid-century requires significant reductions in within leadership and governance, leadership and politi- greenhouse gas emissions and deforestation, both now cal will are needed to ensure collaboration across all rele- and in the years to come (Field and others 2014). Estimates vant departments within a ministry, such as environmental of the costs of mitigation generally do not take into health; vector control; water, sanitation, and hygiene; and account the growing evidence that some mitigation disaster risk management. Also needed are policy priori- options have extensive health co-benefits (Smith and tization and planning that explicitly incorporate the risks others 2014). Mitigation policies and technologies (such as Health Risks and Costs of Climate Variability and Change 157 Figure 8.4 Components for Building Climate-Resilient Health Systems resilien mate ce Cli Leadership and governance and te He m a ing wo alth Cli anc rkf h fin orc lt e ea h Leadership and governance Health Financing workforce Vuln nd capa ency ss a adap erabi nt asse city, eme Emerg edne tation ssme lity, and manag prepar ilding block nt Bu s Health Service of ms he information delivery a lt h s y ste Cli m a t e - d ri s k systems g a nd h ea l t h p g warnin grate itorin infor r ogra Inte mon early med ms Essential medical M products and an en ag technologies d em vir en h an rc h on de me t of alt e a ter ntal He re s mi te of nan a he ts C li m a t e r e s i li e n t c li m a lt h a n d s u st ain a bl e t ec h n o lo gi e s an d infr a s tr u ct ur e Source: WHO 2015. transitioning energy generation to reduced use of fossil Estimating the overall costs and benefits of mitigation fuels; altering policies to increase mass transit and encour- to reduce health risks associated with climate change is age active transportation such as walking and biking; and an important research need. promoting dietary changes to reduce consumption of red meat) are associated with significant health benefits COSTS ARISING IN THE HEALTH SECTOR (termed co-benefits) that primarily will be local and will accrue well before the benefits of mitigation become evi- The health effects of extreme weather and climate dent, potentially making mitigation implementation more change will lead to potential costs. Important categories politically feasible. to consider when estimating impacts and subsequent 158 Injury Prevention and Environmental Health Table 8.1 Examples of Climate-Informed Health Interventions Climate-related health risks and mechanisms Examples of interventions Extreme heat and thermal stress • Establish occupational health exposure standards. • Improve health facility design, energy-efficient cooling and heating systems. • Ensure public education to promote behavior change (in relation to clothing and ventilation). • Develop heat-health action plans (including early warning, public communication, and response plans) such as cooling centers for high-risk populations. Water- and food-borne diseases • Enhance disease surveillance systems during high-risk seasons or periods. • Establish early warning systems to anticipate outbreaks associated with extreme weather and climate events. • Strengthen food and water quality control. Zoonotic and vector-borne diseases • Expand the scope of diseases monitored and conduct monitoring at the margins of current geographic distributions. • Establish early warning systems when data are sufficient and the association between environmental variables and health outcomes is robust. • Establish vector or pest surveillance and control programs. • Enhance diagnostic and treatment options in high-risk regions or periods. • Ensure adequate animal and human vaccination coverage. Allergic diseases and • Develop exposure forecasts for air quality, allergens, and dust. cardiopulmonary health • Enforce stricter air quality standards for pollution. • Establish programs to monitor pollen levels. • Establish allergen management. • Create plans for handling increased demand for treatment during high-risk seasons or weather conditions. Nutrition • Perform seasonal nutritional screening in high-risk communities. • Scale up integrated food security, nutrition, and health programming in fragile zones. • Promote public education and food hygiene. Storms and floods • Include climate risks in siting, designing, or retrofitting health infrastructure. • Establish early warning and early action systems, including education and community mobilization. • Assess and retrofit or construct public health infrastructure (health facilities in flood-prone areas) to be resilient to extreme weather conditions, warmer temperatures, and environmental changes. Mental health and disability • Address special needs of mental health patients (as well as patients with other disabilities) by developing emergency preparedness plans. • Address mental health needs of disaster- and trauma-exposed populations. • During extreme weather conditions, establish community watch for people with mental illness. Source: Adapted from WHO 2015. responses include immediate health sector response on people’s quality of life and well-being (or in eco- costs (additional medicine and costs of treatment) when nomic terms, welfare) exist, even if these impacts are not an extreme weather event occurs or an impact arises captured by markets. Quantifying and valuing these from changing weather trends. Other costs arise for effects is possible, expressing them in monetary terms to those affected, including injuries, illnesses, and deaths capture the economic, social, and environmental costs and lost work time. Further effects relating to the impact borne by society as a whole. Health Risks and Costs of Climate Variability and Change 159 Estimating the full costs of climate variability and these opportunity costs also apply to other health change includes three components: resource costs impacts, and many studies estimate these in addition to (medical treatment costs); opportunity costs (lost resource costs. productivity); and welfare costs or disutility (pain or Another set of studies takes the resource and oppor- suffering, concern, and inconvenience to family and tunity costs estimated from these methods and inputs others). The magnitude of welfare costs usually is derived them into economy-wide economic models. This through elicitation techniques such as contingent valua- approach captures the impacts of costs on the wider tion and stated preference. This valuation is somewhat economy, the linkages across sectors, and macro- controversial, but it is included because it is a way of economic metrics such as gross domestic product (GDP). comparing impacts using a common metric. Finally, in addition to estimating resource and oppor- tunity costs, some studies derive values for the impacts on well-being (for example, the pain and suffering from Methods illness). Techniques are available to capture this compo- A wide variety of methods have been used to estimate nent, such as assessing the willingness to pay or the the costs of climate change effects on health. These willingness to accept compensation for a particular methods capture different aspects of the resource, health outcome. These are derived using survey-based opportunity, and well-being costs and can use different stated preference methods and/or revealed preferences approaches for valuation of these elements. Also, differ- methods. ences exist between the direct costs of the events and the indirect, wider costs for the economy. Many studies focus on resource costs, although dif- Review of Studies ferences emerge even in the approach used for valua- Altered weather patterns (particularly, extreme weather tion. One set of studies explored the costs of adaptation and climate events) could affect health sector costs ini- to climate change using preventative costs. As an exam- tially through resource costs for diagnosis and treat- ple, in the case of malaria, valuation includes estimating ment. As an example, six of the weather and climate the number of malaria cases from climate change, then events that struck the United States between 2000 and looking at the costs of adaptation (a proxy for impacts) 2009 included higher concentrations of ground-level based on the costs of programs and unit costs (per ben- ozone, the 2002 outbreak of West Nile virus in Louisiana, eficiary) for insecticide-treated bed-nets plus case man- the 2003 Southern California wildfires, the 2004 Florida agement and indoor residual insecticide spraying. These hurricane season, the 2006 California heatwave, and the approaches have been widely used, particularly in stud- 2009 flooding of the Red River in North Dakota. In total, ies in LMICs. Other approaches look at resource costs these events increased health care costs an estimated directly, for example, looking at the number of addi- US$819 million, reflecting more than 760,000 encoun- tional impacts, and then estimating the health care costs ters with the health care system (Knowlton and others of treatment, that is, using estimates of the health care 2011). The total health costs, including 1,689 lives lost cost per patient and the number of hospital days spent prematurely (valued using nonmarket economic values), on average for respiratory admissions. A variation on exceeded US$15.5 billion. the resource cost method can be undertaken when Health care facilities themselves can be damaged working at aggregated scale, using investment and by extreme weather and climate events, including financial flow assessments. These look at existing expen- storm surges, floods, and wildfires, which compro- ditures and then apply a mark-up (an increase) to reflect mise critical resources required to treat patients and rising impacts from climate change. One issue with all repair or replace damaged or destroyed equipment of these approaches is that they only cover one element and buildings (Carthey, Chandra, and Loosemore of the total health costs. 2009). In 2011, 139.8 million people globally—57 percent One additional cost that arises from impacts is of all disaster victims—were affected by hydrological associated with the time or productivity lost from the disasters (floods and wet mass movements). These illness—the opportunity costs. Some studies also esti- disasters were responsible for 20 percent of all people mate these, often using values based on loss, earnings, reportedly killed in disasters and 19 percent of total or, more appropriately, labor productivity. These costs damages (Guha-Sapir and others 2012). Although the are particularly relevant for studies that focus on out- proportion of individuals seeking medical treatment door worker productivity; the primary impact for these during a disaster is typically a small subset of the total workers is associated with lost time, as worker output number of persons affected, the additional burden on is reduced because of heat and humidity. However, health care facilities can be significant (Hess and 160 Injury Prevention and Environmental Health others 2009). Floods and wildfires also can require the malaria, two climate scenarios, and updated estimates of evacuation of critical care patients, with attendant the costs of malaria treatment. In 2010, the average risks. annual costs for treating diarrheal disease and malaria When these extreme events are very large, they can cases associated with climate change were estimated to affect the ability of health care systems to function be between US$4 billion and US$7 billion, with the costs properly and to care for patients with ongoing health expected to decline over time as basic health services issues that require medication or treatment. In cases improve. From 2010 to 2050, the average annual costs where these events become significantly more frequent were estimated to be around US$3 billion, with most of or intense, health facilities might need to add surge the costs related to treating diarrheal disease; the largest capacity to help them to manage such events without burden was expected to be in Sub-Saharan Africa. interrupting service (Banks, Shah, and Richards 2007; Pandey’s estimates differ from those of Ebi’s primarily Hess and others 2009). because of the assumption of a lower baseline burden of Climate change is projected to increase the burden of disease and lower costs for malaria treatment. climate-sensitive health outcomes, leading to increased These studies considered only malnutrition, diarrheal costs in the absence of mitigation and adaptation. As disease, and malaria and therefore were underestimates. well as changing the patterns of extreme events, it will According to Parry and others (2009), the studies esti- lead to shifts in climate variables that will affect health mated only 30–50 percent of the extra health burden of outcomes. This increase in health burdens will increase climate change. the demands on public health services (for example, World Bank (2010) undertook a similar analysis of surveillance and control programs) and the demands diarrheal disease and malaria and reported much lower for health care and relevant supplies (for example, anti- estimates than these earlier studies. Whereas the earlier malarials and oral rehydration). It also will increase studies fixed the baseline incidence of disease, this study opportunity and welfare costs. Studies use different incorporated a future baseline based on the WHO methods and include different components, making Global Burden of Disease projections to 2030 (plus inter-comparisons difficult. extensions through 2050). This led to a reduction in the Many earlier estimates of the additional costs of cli- baseline incidence of diarrheal disease and malaria, sig- mate change typically focused on the health care costs nificantly reducing the additional cases due to climate associated with treating additional cases of disease, not change. The World Bank analysis also incorporated the costs of providing additional health services (health updated unit costs of prevention and treatment and risk system adaptation costs) or the wider societal costs. factors. The resulting health adaptation costs in LMICs Therefore, they underestimated the total costs. Given globally were estimated at between US$1.8 billion and these limitations, the global costs of treating future US$2.4 billion a year in the period 2010–50, with most cases of adverse health outcomes from climate change of these costs in Africa. Future health outcomes depend from such studies are estimated at billions of U.S. on multiple factors beyond the level of greenhouse gas dollars annually (Ebi 2008; Pandey 2010). emissions and resulting warming. Ebi (2008) estimated the worldwide costs in 2030 of The recent Climate Impact Research and Response additional cases of malnutrition, diarrheal disease, and Coordination for a Larger Europe (CIRCLE) study used malaria due to climate change at US$5 billion to US$16 a combination of global models: a computable general billion a year, for a high-emissions scenario, assuming no equilibrium model to project effects to 2060 and the population or economic growth (undiscounted US$). AD-RICE model to project effects beyond 2060 (OECD This estimate was based on current costs of treatment 2015).1 It considered heat mortality in a stand-alone and assumed no adaptation. The costs for additional analysis, heat- and cold-related morbidity and mortality, infrastructure and health care workers were not included, and morbidity from infectious diseases such as nor were the costs of additional public health services, malaria, schistosomiasis, dengue, diarrhea, cardiovascu- such as surveillance and monitoring. The estimated costs lar disease, and respiratory disease. The changes in labor were distributed unevenly across regions. Markandya productivity from climate-sensitive diseases were taken and Chiabai (2009) used these estimates to provide a from Bosello, Eboli, and Pierfederici (2012). regional breakdown of costs, finding the highest costs in By 2060, the largest negative effects were projected to Africa and South-East Asia. take place in Africa and the Middle East (–0.6 percent for Pandey (2010) estimated global health costs of cli- South Africa, −0.5 percent for the Middle East and North mate change based on United Nations population pro- Africa, and −0.4 percent for other African countries). jections, strong economic growth, updated projections Smaller impacts were projected for Brazil, Mexico, and of the current health burden of diarrheal diseases and LMICs in Asia (−0.3 percent), as well as for Indonesia, Health Risks and Costs of Climate Variability and Change 161 the United States, South-East Asia, and most of Latin • In 25 African countries, Egbendewe-Mondzozo and America (−0.2 percent). Some regions were projected to others (2011) used a semiparametric econometric experience positive impacts on labor productivity, the model to estimate the climate change–related costs highest being the Russian Federation (+0.5 percent), for inpatient and outpatient treatments for malaria Canada (+0.4 percent), and China (+0.2 percent). In at the end of the century (2080–100). Even marginal other regions, the projected impacts were either very changes in temperature and precipitation were pro- small or nonexistent. Changes in health care expendi- jected to affect the number of malaria cases, with tures were also estimated. The costs of vector-borne dis- most countries projected to see an increase and oth- eases were based on prevention expenditures and ers a decrease. The end-of-century treatment costs treatment costs per person per month (Chima, Goodman, as a proportion of year 2000 health expenditures per and Mills 2003). 1,000 people would be higher in the vast majority Changes in health expenditure were small as a of countries, with increases of more than 20 percent percentage of GDP. In 2060, they were projected to be in the costs of inpatient treatment in Burundi, Côte highest in LMICs in Asia (0.5 percent), Brazil, and the d’Ivoire, Malawi, Rwanda, and Sudan. Middle East and North Africa (0.3 percent). Additional • In Tanzania, Traerup, Ortiz, and Markandya (2011) demands for health services were projected to be very estimated the costs of cholera cases due to climate small in other regions and to be negative in Canada and change in 2030 to be in the range of 0.32–1.4 percent large European Union economies, such as France and of GDP. Germany (−0.1 percent). • In India, Ramakrishnan (2011) estimated the costs of Several regional and country studies support or treating additional cases of diarrhea and malaria in extend these assessments: 2030 to range between Rs 3,648 lakhs and Rs 7,787 lakhs, depending on the emissions scenario.2 • In India, Chiabai and others (2010) reported • In Saint Lucia, the Economic Commission for Latin adaptation costs for malaria, diarrhea, and malnu- America and the Caribbean (ECLAC 2011) estimated trition. Using a similar prevention cost approach to the present value of treatment costs under two scenar- the studies above, costs under different development ios of greenhouse gas emissions in the period 2010–50 scenarios were in the range of US$183 million to as US$634,000 for cardiorespiratory disease, US$33,000 US$584 million with no mitigation and US$151 for malaria, US$36,000 for dengue, and US$3.5 million million to US$476 million with mitigation achieving for gastroenteritis, using a discount rate of 1 percent. stabilization at 550 parts per million. • In Paraguay, the United Nations Development • In Kenya, the Stockholm Environment Institute used Programme (UNDP 2011) applied an investment and a malaria risk model based on altitude to assess the financial flow assessment to health, estimating total national impact of future climate change (SEI 2009). costs to be US$160.5 million by 2030 (2005 US$). The model projected that, by 2055, as a result of aver- age climate warming of 4.3°F (2.3°C) across the pro- Because adverse health outcomes are projected to jections, the population annually affected by malaria occur predominantly in LICs, treatment costs will be in rural areas above 1,000 meters (63 percent of the borne primarily by families where governments provide population) would increase as much as 74 percent (in limited health care (WHO 2004). Time off from work to the absence of adaptation). It also presented results for care for sick children will have an adverse effect on scenarios with average temperature increases of 2.2°F productivity. (1.2°C) and 5.6°F (3°C). The 10 model projections Estimates of the impact of climate change on outdoor used a range of average climate warming increases worker productivity (primarily through heat stress) indi- from 36 to 89 percent. The additional economic cate that productivity has already declined during the burden of endemic malaria disease in the 2050s was hottest and most humid seasons in parts of Africa and estimated to be more than US$92 million annually Asia, with more than half of afternoon hours projected to (with a range of US$51 million to US$106 million be lost to the need for rest breaks in 2050 in South-East annually across the temperature projections) based on Asia and up to a 20 percent loss in global productivity in the clinical and economic burden of malaria. The esti- 2100 under a moderately low emission scenario mated welfare costs increased to a range of between (Representative Concentration Pathway 4.5) (Dunne, US$154 million and US$197 million annually when Stouffer, and John 2013; Kjellstrom and others 2009; disutility costs (discomfort, pain, and inconve- Kjellstrom, Lemke, and Otto 2013). Trade-offs between nience measured by survey-based willingness-to-pay worker health and productivity will be of particular con- estimates) were taken into account. cern for workers with limited control over work practices. 162 Injury Prevention and Environmental Health Kovats and others (2011) estimated the labor produc- as the increase in heat-related mortality, although recent tivity losses for Europe at between US$321 million and literature (for example, Ebi and Mills 2013) has ques- US$792 million a year in the 2080s for a high-emissions tioned whether these effects will be fully realized. scenario, falling to between US$64 million and US$160.5 A similar analysis for Europe (Kovats and others million a year under a mitigation scenario, with impacts 2011) estimated annual welfare costs for heat-related primarily in Southern Europe. For the loss of labor pro- mortality at US$32.6 billion by the 2020s (2011–40), ductivity, a value was derived from the GDP per labor US$108.4 billion by the 2050s (2041–70), and force member. This represents the loss to society, differ- US$154.8 billion by the 2080s (2071–100) under a entiating it from a loss of earnings measure that reflects high-emissions scenario. These values were more than an only the loss for the individual. order of magnitude lower when using a different valua- The CIRCLE project also considered these impacts tion approach. Under a mitigation scenario, broadly and found that the highest impacts on labor productiv- equivalent to the 3.7°F (2°C) target, these values fell sig- ity caused by occupational heat stress in 2060 likely nificantly (after 2040), to US$84.2 billion a year by the would occur in regions with relatively large proportions 2050s (2041–70). Again, including (autonomous) accli- of outdoor workers and warm climates (OECD 2015). matization reduced these impacts significantly. As these The most severely affected regions were projected to studies highlight, choices regarding the response func- experience productivity losses between 3 and 5 percent tions and valuation metrics, as well as autonomous adap- for outdoor activities for a 1.9°F (1°C) temperature tation, can have a very large impact on estimated overall increase in non–Organisation for Economic health costs, leading to order of magnitude differences. Co-operation and Development (OECD) countries, At the global level, OECD (2015) reported heat- non–European Union European countries, Latin related mortality under climate change in high-income America (including Brazil and Chile), Mexico, China, countries. Using a value of statistical life approach, they LMICs in Asia, and South Africa. Most OECD countries, projected that the economic costs of heat-related deaths including Japan, the United States, and the OECD would increase from around US$100 billion today to European Union countries, were projected to experience US$320 billion in 2030 and US$670 billion in 2050, with effects of less than 1 percent. the highest costs in Europe and North America. Nevertheless, some health impacts from climate change are likely to affect OECD countries, particularly mortality and morbidity from higher temperatures and COSTS ARISING IN OTHER SECTORS heat extremes. Studies have assessed the impacts and full economic costs (including nonmarket valuation) in Adaptations in health systems and the health sector OECD countries. Watkiss and Hunt (2012) quantified more generally are not the only climate change adapta- and valued temperature-related mortality effects, salmo- tions required to protect human health. Although nellosis, and coastal flooding–induced mental health determining the extent to which other sectors protect impacts resulting from climate change in Europe in health can be challenging, certain sectors (such as elec- 2071–100, assessing the full welfare costs. The analysis trical and water utilities) are clearly intimately tied to found that the choice of valuation metric and inclusion public health. These ties are complex. Some sectors or exclusion of acclimatization (autonomous adapta- provide health benefits via smooth and continuous tion)3 had a major impact on the results, much more so operations, while the health sector, through regulation than climate uncertainty. and other activities, minimizes the adverse health In model runs without acclimatization, economic impacts imposed by other sectors. For example, electrifi- costs in current values were estimated at US$12.6 billion cation and a sustained, reliable power supply supports to US$31.6 billion a year by the 2020s using a value of public health and health care delivery in numerous ways; statistical life metric, but US$1 billion to US$4.2 billion a power outages are associated with significant impacts year when acclimatization was included. By the 2080s, on health. Yet, power generation, particularly based on the annual values ranged from US$52.6 billion to fossil fuels, has serious adverse health consequences that US$189.5 billion (according to choice of function and regulations have only partially succeeded in limiting in climate model) without acclimatization, and US$8.4 bil- most countries. Similarly, water treatment and distribu- lion to US$84.2 billion with acclimatization. The addi- tion are fundamental to health, while certain water man- tional welfare costs for salmonellosis from climate change agement decisions (such as dam construction) can have were estimated to be several hundred million dollars significant, if localized, adverse health consequences. annually by 2071–100. They also found the potential Thus, in considering the costs associated with adapta- reduction in cold-related mortality to be at least as large tion, costs also arise in sectors other than health whose Health Risks and Costs of Climate Variability and Change 163 activities are central to protecting health or whose adap- in cities but also to affect climate change in the long term tation choices may be maladaptive from a public health when power is generated by coal-fired power plants. perspective. Promoting health impact assessments of adaptation Adaptation will vary by baseline status in these sec- activities in other sectors is a powerful means for the tors and by location, with significant efforts needed to health sector to highlight potential disbenefits of adapta- decrease exposure in a changing climate. In the water tion activities in other sectors. sector, several adaptations will likely be needed to address water scarcity, changes in water quality, and vari- ability in precipitation. In the agriculture sector, adapta- ISSUES RELATED TO SPATIAL AND tions will be needed to maintain an adequate supply of TEMPORAL SCALE protein, energy, and micronutrients. In the forestry sec- tor, adaptations will be needed to limit the incidence of Costs of preparing for, coping with, and recovering from forest fires and associated direct and pollution-driven the health risks of climate variability and change will health impacts, and to limit the socioeconomic impacts vary across temporal and spatial scales. of disruptions to ecosystem services. In some settings, adaptations will be needed to enhance ongoing activities aimed at increasing resilience to worsening climate- Spatial Scale sensitive health threats; in others, such as the water sec- Poverty is a major driver of risk, which means that low- tor in the Arctic, fundamentally new approaches and and lower-middle income countries generally will be at infrastructure will be needed. higher risk of adverse climate-sensitive health outcomes. Other adaptations may fall outside of existing sectors. Undernutrition, malaria, and diarrheal disease—among In anticipation of sea-level rise, widespread adaptation the largest health concerns related to climate change— activities will be needed to protect infrastructure that is are leading causes of morbidity and mortality in children critical to public health (such as hospitals, clinics, and younger than age five years (Liu and others 2015; Smith dialysis centers) and to prevent saltwater intrusion into and others 2014). For example, despite recent progress, groundwater sources (which can lead to hypertension, diarrhea kills 1,584 children every day, accounting for crop failure, and limitations on drinking water supply). 9 percent of child deaths. Just 15 countries in Africa and Other adaptations will be needed to protect communi- Asia account for 71 percent of childhood mortality from ties from extreme weather and climate events, such as diarrhea and pneumonia (IVAC 2014). These countries flooding, severe storms, and extreme heat. Still others include low-income (Afghanistan, Chad, the Democratic will be needed to manage population dislocation and Republic of Congo, Ethiopia, Niger, and Uganda), lower- resettlement, which can be a significant challenge to the middle-income (Bangladesh, India, Indonesia, Kenya, health sector. In many cases, adaptation activities will Nigeria, Pakistan, and Sudan), and upper-middle-income entail managing risk, including risk reduction, risk shar- (Angola and China) countries. ing through insurance and other mechanisms, and The pathways leading to higher burdens of diarrheal enhanced recovery mechanisms. diseases vary across countries, with lack of improved san- Successful adaptation will require increased commu- itation facilities a major risk; other drivers include food nication, coordination, and integration between health and water contaminated by humans or animals, improper and other sectors. The public health sector has extensive food handling, and improper hand washing. Nine of experience collaborating with other sectors to achieve its these 15 countries are among the 10 countries that are goals and will need to build on this experience to facili- home to two-thirds of the global population with limited tate intersectoral adaptation. access to improved drinking water sources: Bangladesh, Some of this coordination will focus on highlighting China, the Democratic Republic of Congo, Ethiopia, the potential adverse consequences of adaptation activi- India, Indonesia, Kenya, Nigeria, and Pakistan. Warmer ties in other sectors and, indeed, in the health sector temperatures mean faster replication of some pathogens itself. The appropriate balance between expenditures on associated with diarrheal diseases, and higher precipita- activities that protect one population at one point in tion events can wash pathogens into water sources time but that potentially lead to some harm for other (Cann and others 2012; Kolstad and Johansson 2010). populations is not always clear. For instance, the wide- Without a significant improvement in access to safe water spread use of air conditioning to protect against extreme and improved sanitation, reducing the extent to which heat events is maladaptive to the extent that it has the climate change could increase the burden of diarrheal potential not only to worsen the heat island effect locally disease will become increasingly challenging. 164 Injury Prevention and Environmental Health Temporal Scale Improvements in this development pathway will Temporally, the rate of greenhouse gas emissions reduce the burden of climate-sensitive health outcomes reductions will affect the magnitude and pattern of cli- even before considering any impacts of climate change. mate change past mid-century, with rapid and extensive Meeting the challenges of climate change will be much reductions lowering adaptation needs later in this easier in this pathway. century (Smith and others 2014). Many policies and Another development pathway describes a world technologies to reduce greenhouse gas emissions are separated into regional blocks with little coordination associated with health co-benefits; for example, reduc- between them (Ebi 2013). This world is failing to ing emissions from point sources such as coal-fired achieve global development goals, with regional blocks power plants and from mobile sources such as transpor- characterized by extreme poverty and pockets of mod- tation could provide significant health benefits by erate wealth and the bulk of countries struggling to reducing exposure to fine particulate matter (Balbus maintain living standards for their rapidly growing and others 2014). populations. This pathway includes the following Projecting how health costs could evolve as the features: climate continues to change also requires consider- • Mortality rates are high, with mortality from climate- ation of future development pathways (Ebi 2013). Five related health outcomes (particularly undernutrition, socioeconomic development pathways describe the diarrheal diseases, and malaria) increasing and life evolution of demographic, political, social, cultural, expectancy possibly falling in LMICs because of institutional, economic, and technological trends increased childhood mortality, although some sub- through this century, along axes describing worlds regions enjoy better health. All countries experience with increasing socioeconomic and environmental a double burden of climate-related infectious and challenges to adaptation and mitigation. Also consid- chronic health outcomes. ered are ecosystems and ecosystem services affected • Large regions of the world are food and water by human activities, such as air and water quality. insecure. Each development pathway has very different • Most urban growth in LMICs occurs in unplanned implications for the burdens of climate-sensitive settlements and mostly fails to improve access to safe health outcomes and health system capacities to pre- water and improved sanitation. pare for and manage risks associated with climate • Wealthier regions do not invest in research and devel- variability and change. Using these pathways facilitates opment to help less well-off regions manage health exploration of the possible impacts and costs associ- risks. Further, governance and institutions are weak, ated with mitigating greenhouse gas emissions to a international cooperation is limited, investments in certain level and the extent of efforts required to adapt public health and health care infrastructure are low, to that level. and the number of public health and health care One development pathway is a world aiming for sus- personnel is too small to address health needs. tainable development (Ebi 2013). This pathway includes the following features: In this development pathway, the challenges to • Population health improves significantly, with managing the health risks of climate variability and increased emphasis on enhancing public health and change increase over time, with rising and increas- health care functions. ingly unaffordable costs in more vulnerable countries • Coordinated, worldwide efforts through international and regions. institutions and nongovernmental organizations The other three pathways explore a world that contin- increase access to safe water, improved sanitation, ues along its current trajectory, with health improving medical care, education, and other factors in under- but at a slower rate than in the pathway aiming for sus- served populations. tainable development; a highly unequal world where • Life expectancies increase in LICs with decreas- adaptation is difficult, but technologies are developed ing burdens of key causes of childhood mortality and deployed to reduce greenhouse gas emissions; and a (undernutrition, diarrheal diseases, and malaria). world with low challenges to adaptation, but where mit- • Funding increases for public health and health care igation of greenhouse gas emissions is difficult for a organizations, and institutions enhance their capaci- range of technological and other reasons. Each has dif- ties to prepare for, respond to, cope with, and recover ferent implications for the health costs of climate vari- from climate-related health risks. ability and change. Health Risks and Costs of Climate Variability and Change 165 CHALLENGES RELATED TO ESTIMATING other possible drivers of change, such as land use changes. COSTS AND BENEFITS Similar challenges present in estimating the benefits of interventions. Estimating the costs and benefits of climate change and Other issues that arise when considering the costs of adaptation to the associated risks presents many chal- adaptation include how to limit double counting. For lenges. These include the unique nature of the threat of example, climate change is increasing the number of climate change to the incidence, geographic distribution, cases of undernutrition, malaria, and diarrheal disease in and seasonality of a wide range of health outcomes (with many regions (Smith and others 2014). However, these risks and uncertainty increasing over coming decades) health outcomes are not independent; undernutrition and the temporal displacement between the causes of increases a child’s susceptibility to malaria and diarrheal climate change (human activities leading to the release disease. It is not clear how to count the costs of prevent- of greenhouse gases and natural climate variability) and ing and treating these health outcomes accurately. the projected timing of health impacts. Further, the costs Many researchers and modelers are estimating the of proactive mitigation for managing health risks of cli- costs of various mitigation options. Although health mate change will be incurred years to decades before systems are a source of greenhouse gas emissions, the benefits in reducing climate change are evident. Precisely sector should reduce these emissions as quickly as timing investments will not always be possible given possible. Lower emissions benefit everyone later in the inherent uncertainties about the magnitude, rate, and century; unlike air pollutants, greenhouse gases do not timing of climate change. remain local. The hazards created by a changing climate will inter- Interest in calculating the loss and damage due to act with the sensitivity of populations and regions and climate change has been growing particularly in coun- with their capacity to prepare for and cope with haz- tries that are vulnerable to its adverse effects. Loss and ards as they arise. This creates complex relationships damage refers to the impacts of climate-related stressors between climate change and health outcomes that will on human and natural systems that occur despite miti- vary over temporal and spatial scales. Because LMICs gation and adaptation efforts. Climate change that have the highest sensitivity to climate-sensitive health already is locked in because of the inertia in the climate outcomes and the least ability to adapt, they will be at system could adversely affect development in particu- highest risk (Smith and others 2014). All countries, larly vulnerable locations and populations. For example, however, will experience hazards, and all countries will saltwater intrusion from sea-level rise could mean that need to adapt and mitigate. The differences across farmers can no longer grow crops or feed animals. The countries mean that the costs of adaptation will vary issue of loss and damage arose because most of the focus over time and space. of the more than 20 years of negotiations under the Given the limited capacity of health systems to man- United Nations Framework Convention on Climate age current climate variability and change, the costs of Change has been on reducing greenhouse gas emissions, adaptation are likely to be high in the longer-term, as with less attention paid to ensuring that countries that health systems incorporate climate change into policies are particularly vulnerable to climate change but who and programs. Once adaptive risk management pro- historically were responsible for only a tiny proportion cesses are established and climate change mainstreamed of atmospheric greenhouse gases and who are experi- into policies and programs, costs by mid-century will encing adverse impacts have the financial resources to depend on the health impacts associated with the mag- adapt. This issue has been contentious because some nitude and pattern of climate change, which, in turn, will observers consider it to be synonymous with liability and depend on the extent of mitigation over coming decades. compensation. This is an active area of research and Adding to these complexities are the costs associated negotiation. with adaptation in other sectors. It is not surprising that few costs of adaptation options have been estimated. Information on some CONCLUSIONS adaptation options can be estimated from other chapters in this volume, such as the costs of surveillance and Climate variability and change present significant treatment for malaria or other vector-borne diseases. challenges for the health and well-being of individuals, However, there are challenges in estimating what portion communities, and nations. Preventing, preparing for, of the costs of extending current surveillance and health and managing climate-related risks to human and natu- care systems to prepare for changes in the geographic ral systems will be a recurring theme throughout the range of malaria could be due to climate change versus 21st century. Hallegatte and others (2016, xi) explored 166 Injury Prevention and Environmental Health the intersection of climate change and poverty and 1. For information on the CIRCLE study, see http://www offered the following conclusions: .circle-era.eu/np4/home.html. 2. A lakh is 100,000 rupees. 3. Future individuals will respond to higher temperatures Without action, climate change would likely spark through physiological and behavioral adjustments. Most higher agricultural prices and could threaten food studies ignore this effect and use impact functions derived security in poorer regions such as Sub-Saharan Africa from the current climate and apply these functions to the and South Asia. And in most countries where we have future. This overestimates impacts, because it assumes that data, poor urban households are more exposed to no autonomous adaptation takes place (acclimatization). floods than the average urban population. Climate In reality, populations will adjust autonomously (that change also will magnify many threats to health, as is, without planned adaptation) to climate change, and poor people are more susceptible to climate-related indeed, mortality rates today are fairly similar in coun- diseases such as malaria and diarrhea. . . . We need tries with very different climates. Studies that build in good, climate-informed development to reduce the acclimatization show much lower future health impacts. impacts of climate change on the poor. This means, However, little information or evidence exists on which to in part, providing poor people with social safety nets base assumptions about the rate of future acclimatization. and universal health care. These efforts will need to be Estimating the rate of change of adaptation to climate coupled with targeted climate resilience measures, such change and the rate above which impacts might start to as the introduction of heat-resistant crops and disaster increase more sharply is difficult. preparedness systems. The report further concludes that, without climate- REFERENCES resilient development, climate change could force more than 100 million people into extreme poverty by 2030 Balbus, J. M., J. B. Greenblatt, R. Chari, D. Millstein, and K. L. Ebi. 2014. “A Wedge-Based Approach to Estimating Health (Hallegatte and others 2016). Rapid, inclusive develop- Co-Benefits of Climate Change Mitigation Activities in the ment could avoid most of these impacts, and immediate United States.” Climatic Change 127 (2): 199–210. reductions in emissions could avoid many of the pro- Balbus, J. M., and C. Malina. 2009. “Identifying Vulnerable jected risks later in the century. Subpopulations for Climate Change Health Effects in the Climate change underscores the urgency of strength- United States.” Journal of Occupational and Environmental ening basic public health infrastructure, particularly in Medicine 51 (1): 33–37. poor and underserved areas. To be effective, health Banks, L. L., M. B. Shah, and M. E. Richards. 2007. “Effective systems need to incorporate climate variability and Healthcare System Response to Consecutive Florida change explicitly into all climate-sensitive policies and Hurricanes.” American Journal of Disaster Medicine programs, including disaster risk management, air pol- 2 (6): 285–95. lution control, infectious disease monitoring and sur- Bosello, F., F. Eboli, and R. Pierfederici. 2012. “Assessing the Economic Impacts of Climate Change: An Updated CGE veillance, and water and food safety and security. Point of View.” Working Paper 2, Research Paper 125, Taking advantage of the growing body of knowledge Fondazione Eni Enrico Mattei, Euro-Mediterraneo per i about environmental drivers of climate-sensitive health Cambiamenti Climatici, Venice. outcomes can provide significant public health bene- Cann, K. F., D. R. Thomas, R. L. Salmon, A. P. Wyn-Jones, and fits. Continuing to take a business-as-usual approach D. Kay. 2012. “Extreme Water-Related Weather Events and to climate change will put lives and livelihoods at risk Waterborne Disease.” Epidemiology and Infection 141 (4): and result in higher health burdens that could have 671–86. been prevented. Cardona, O. D., M. G. Ordaz, E. Reinoso, L. Yamin, B. Barbat, and others. 2012. “CAPRA—Comprehensive Approach to Probabilistic Risk Assessment: International Initiative NOTES for Risk Management Effectiveness.” Paper prepared for the 15th World Conference on Earthquake Engineering, World Bank Income Classifications as of July 2014 are as fol- Lisbon, September 24–28. lows, based on estimates of gross national income (GNI) per Carthey, J., V. Chandra, and M. Loosemore. 2009. “Adapting capita for 2013: Australian Health Facilities to Cope with Climate-Related Extreme Weather Events.” Journal of Facilities Management • Low-income countries (LICs) = US$1,045 or less 7 (1): 36–51. • Middle-income countries (MICs) are subdivided: Chiabai, A., S. Balakrishnan, G. Sarangi, and S. Nischal. 2010. a) lower-middle-income = US$1,046 to US$4,125 “Human Health.” In Costing Adaptation: Preparing for b) upper-middle-income (UMICs) = US$4,126 to US$12,745 Climate Change in India, edited by A. Markandya and • High-income countries (HICs) = US$12,746 or more. A. Mishra. New Delhi: TERI Press. Health Risks and Costs of Climate Variability and Change 167 Chima, R. I., C. A. Goodman, and A. Mills. 2003. “The Kjellstrom, T., R. S. Kovats, S. J. Lloyd, T. Holt, and R. S. J. Economic Impact of Malaria in Africa: A Critical Review of Tol. 2009. “The Direct Impact of Climate Change on the Evidence.” Health Policy 63: 17–36. Regional Labor Productivity.” Archives of Environmental Dunne, J. P., R. J. Stouffer, and J. G. John. 2013. “Reductions in and Occupational Health 64 (4): 217–27. Labour Capacity from Heat Stress under Climate Warming.” Kjellstrom, T., B. Lemke, and M. Otto. 2013. “Mapping Nature Climate Change 3: 563–66. Occupational Heat Exposure and Effects in South-East Asia: Ebi, K. L. 2008. “Adaptation Costs for Climate Change–Related Ongoing Time Trends, 1980–2011 and Future Estimates to Cases of Diarrhoeal Disease, Malnutrition, and Malaria in 2050.” Industrial Health 51 (1): 56–67. 2030.” Globalization and Health 4 (1): 9. Knowlton, K., M. Rotkin-Ellman, L. Geballe, W. Max, and ———. 2013. “Health in the New Scenarios for Climate G. M. Solomon. 2011. “Six Climate Change-Related Events Change Research.” International Journal of Environmental in the United States Accounted for about $14 Billion in Lost Research and Public Health 11 (1): 30–46. doi:10.3390 Lives and Health Costs.” Health Affairs 30 (11): 2167–76. /ijerph100x000x. Kolstad, E. W., and K. A. Johansson. 2010. “Uncertainties Ebi, K. L., and K. J. Bowen. 2016. “Extreme Events as Sources of Associated with Quantifying Climate Change Impacts on Health Vulnerability: Drought as an Example.” Weather and Human Health: A Case Study for Diarrhea.” Environmental Climate Extremes 11: 95–102. Health Perspectives 119 (3): 299–305. Ebi, K. L., and D. Mills. 2013. “Winter Mortality in a Warming Kovats, S., S. Lloyd, A. Hunt, and P. Watkiss. 2011. “The Climate: A Reassessment.” WIREs Climate Change 4 (3): Impacts and Economic Costs on Health in Europe and the 203–12. Costs and Benefits of Adaptation, Results of the EC RTD ECLAC (Economic Commission for Latin America and the ClimateCost Project.” Technical Policy Briefing Note 8. In Caribbean). 2011. An Assessment of the Economic Impact The Climate Cost Project. Final Report. Volume 1: Europe, of Climate Change on the Health Sector in Saint Lucia. edited by P. Watkiss. Stockholm: Stockholm Environment Santiago: ECLAC. Institute. Egbendewe-Mondzozo, A., M. Musumba, B. A. McCarl, and Liu, J., H. Mooney, V. Hull, S. J. Davis, J. Gaskell, and oth- X. Wu. 2011. “Climate Change and Vector-borne Diseases: ers. 2015. “Systems Integration for Global Sustainability.” An Economic Impact Analysis of Malaria in Africa.” Science 347 (6225): 1258832. International Journal of Environmental Research and Public Lloyd, S. J., R. S. Kovats, and Z. Chalabi. 2011. “Climate Change, Health 8 (12): 913–30. Crop Yields, and Undernutrition: Development of a Model Field, C. B., V. R. Barros, D. J. Dokken, K. J. Mach, to Quantify the Impact of Climate Scenarios on Child M. D. Mastrandrea, and others, eds. 2014. Impacts, Undernutrition.” Environmental Health Perspectives 119 Adaptation, and Vulnerability. Part A: Global and Sectoral (12): 1817–23. Aspects. Contribution of Working Group II to the Fifth Markandya, A., and A. Chiabai. 2009. “Valuing Climate Change Assessment Report of the Intergovernmental Panel on Climate Impacts on Human Health: Empirical Evidence from the Change. New York: Cambridge University Press. Literature.” International Journal of Environmental Research Field, C. B., V. R. Barros, T. F. Stocker, D. Qin, D. J. Dokken, and Public Health 6 (2): 759–86. and others. 2012. Managing the Risks of Extreme Events and OECD (Organisation for Economic Co-operation and Disasters to Advance Climate Change Adaptation, a Special Development). 2015. “Modelling the Economic Report of Working Groups I and II of the Intergovernmental Consequences of Climate Change.” In The Economic Panel on Climate Change. New York: Cambridge University Consequences of Climate Change. Paris: OECD. Press. Pandey, K. 2010. “Cost of Adapting to Climate Change for Garcia-Menendez, F., R. K. Saari, E. Monier, and N. E. Selin. Human Health in Developing Countries.” Discussion Paper 2015. “U.S. Air Quality and Health Benefits from Avoided 11, World Bank, Washington, DC. Climate Change under Greenhouse Gas Mitigation.” Parry, M., N. Arnell, P. Berry, D. Dodman, S. Fankhauser, and Environmental Science and Technology 49: 7580–88. others. 2009. “Assessing the Costs of Adaptation to Climate Guha-Sapir, D., F. Vos, R. Below, and S. Ponserre. 2012. Annual Change: A Review of the UNFCCC and Other Recent Disaster Statistical Review 2011: The Numbers and Trends. Estimates.” International Institute for Environment and Brussels: CRED. Development and Grantham Institute for Climate Change, Hallegatte, S., M. Bangalore, L. Bonzanigo, M. Fay, T. Kane, and London. others. 2016. Shock Waves: Managing the Impacts of Climate Ramakrishnan, S. 2011. “Adaptation Cost of Diarrhea and Change on Poverty. Washington, DC: World Bank. Malaria in 2030 for India.” Indian Journal of Occupational Hess, J. J., K. L. Heilpern, T. E. Davis, and H. Frumkin. and Environmental Medicine 15 (2): 64. 2009. “Climate Change and Emergency Medicine: Impacts SEI (Stockholm Environment Institute). 2009. The Economics and Opportunities.” Academic Emergency Medicine 16 (8): of Climate Change in East Africa.” Final Report for DFID 782–94. and DANIDA, SEI, Stockholm. IVAC (International Vaccine Access Center). 2014. Pneumonia Smith, K. R., A. Woodward, D. Campbell-Lendrum, and Diarrhea Progress Report. Baltimore, MD: Johns D. D. Chadee, Y. Honda, and others. 2014. “Human Health: Hopkins University. Impacts, Adaptation, and Co-Benefits.” In Climate Change 168 Injury Prevention and Environmental Health 2014: Impacts, Adaptation, and Vulnerability; Part A: Global Climate Change (Capacity Development for Policy Makers and Sectoral Aspects. Contribution of Working Group II to Address Climate Change): Country Summaries.” UNDP, to the Fifth Assessment Report of the Intergovernmental New York. http://www.undpcc.org/en/financial-analysis Panel on Climate Change, edited by C. B. Field, V. R. /results. Barros, D. J. Dokken, K. J. Mach, M. D. Mastrandrea, Watkiss, P., and A. Hunt. 2012. “Projection of Economic and others, chapter 11, 709–54. New York: Cambridge Impacts of Climate Change in Sectors of Europe Based on University Press. Bottom-Up Analysis: Human Health.” Climatic Change 112 Steinbruner, J. D., P. C. Stern, and J. L. Husbands. 2013. (1): 101–26. Climate and Social Stress: Implications for Security Analysis; WHO (World Health Organization). 2004. World Health National Climate Assessment. Washington, DC: National Report 2004: Changing History. Geneva: WHO. Research Council. ———. 2015. Operational Framework for Building Climate Traerup, S. L. M., R. A. Ortiz, and A. Markandya. 2011. “The Resilient Health Systems. Geneva: WHO. Costs of Climate Change: A Study of Cholera in Tanzania.” World Bank. 2010. The Costs to Developing Countries of International Journal of Environmental Research and Public Adapting to Climate Change: New Methods and Estimates; Health 8 (12): 4386–405. the Global Report of the Economics of Adaptation to UNDP (United Nations Development Programme). 2011. Climate Change Study. Synthesis Report. Washington, DC: “Assessment of Investment and Financial Flows to Address World Bank. Health Risks and Costs of Climate Variability and Change 169 Chapter 9 Water Supply, Sanitation, and Hygiene Guy Hutton and Claire Chase INTRODUCTION systematic reviews and meta-analyses, evidence papers, and literature reviews. When those sources were not Safe drinking water, sanitation, and hygiene (WASH) are available, evidence was compiled from the next best fundamental to improving standards of living for sources of published research, thus using accepted crite- people. The improved standards made possible by ria of the hierarchy of evidence for studies on health WASH include, among others, better physical health, effectiveness. Unpublished and grey literature was used protection of the environment, better educational out- where no peer-reviewed published evidence exists. comes, convenience time savings, assurance of lives lived This chapter is structured as follows: with dignity, and equal treatment for both men and women. Poor and vulnerable populations have lower • Progress in improving drinking water, sanitation, and access to improved WASH services and have poorer hygiene coverage associated behaviors. Improved WASH is therefore cen- • Impacts of poor WASH, thereby summarizing the tral to reducing poverty, promoting equality, and sup- evidence on the continued decline in mortality from porting socioeconomic development. Drinking water diarrheal disease and the emerging evidence on the and sanitation were targets in the Millennium long-term developmental and cognitive effects of Development Goals (MDGs) for 2015; under the inadequate WASH on children Sustainable Development Goals (SDGs) for the post- • Effectiveness of interventions, thereby examining 2015 period, Member States of the United Nations (UN) the health effects of specific WASH interventions, aspire to achieve universal access to WASH by 2030. The the approaches to service delivery, and the key role Human Right to Safe Drinking Water and Sanitation of broader institutional policy in accelerating and (HRTWS) was adopted in 2010 under a UN resolution sustaining progress calling for safe, affordable, acceptable, available, and • Intervention costs, efficiency, and sustainability, thereby accessible drinking water and sanitation services for all.1 assessing the socioeconomic returns of improved The scope of WASH services included in this chapter WASH and considering the requirements for popula- is shown in table 9.1. The focus is on services at the tions to have continued access to WASH services household and institutional level and on services for • Challenges, opportunities, and recommendations. personal rather than productive uses. This chapter summarizes global evidence on current This chapter uses the World Health Organization WASH coverage and effects of intervention options, and (WHO) classification of superregions as follows: Africa, it recommends areas for research and policy. Evidence the Americas, South-East Asia, Europe, Eastern comes from published synthesized evidence, such as Mediterranean, and Western Pacific. Corresponding author: Guy Hutton, WASH Section, UNICEF. ghutton@unicef.org, formerly at the Water and Sanitation Program, World Bank. 171 Table 9.1 Scope of Water, Sanitation, and Hygiene Services Included in This Chapter Service Included Excluded Water supply Water for drinking Water for productive uses Other water uses in the home (cooking, hygiene, sanitation, cleaning, laundry) Treatment, safe handling, and storage of water Sanitation Toilets and onsite excreta management Separate greywater management Management of septage (fecal sludge) Industrial wastewater management Sewerage or combined sewer-drainage systems Storm water drainage Solid waste management Hygiene Handwashing and other personal hygiene practices Food hygiene Menstrual hygiene management Environmental hygiene and cleanliness measures STATUS OF DRINKING WATER, SANITATION, All populations meet water and sanitation needs in some AND HYGIENE way, but those ways are often not sufficient, reliable, safe, convenient, affordable, or dignified. To monitor the Targets MDG water and sanitation target, the UN distinguished The MDG targets called for halving the proportion of between improved and unimproved water and sanitation the population without sustainable access to safe drink- facilities at home. For the SDG targets, one indicator is ing water and basic sanitation between 1990 and 2015. proposed per target: (1) for target 6.1, the percentage of The targets were ambitious. In 1990, 76 percent of the population using safely managed drinking water services global population used an improved drinking water and (2) for target 6.2, the percentage of population using source, and 54 percent had access to safe sanitation. The safely managed sanitation services, including a hand- MDG’s drinking water target was met in 2010; yet in washing facility with soap and water. Complementing 2015, the world remained 9 percentage points short of these proposals is a broader set of indicators distinguish- achieving the sanitation target. The SDGs for 2015– ing basic and safely managed service levels (table 9.2) 2030 have broadened from the MDG period to include (WHO and UNICEF 2015a). (1) water-use efficiency across all sectors, sustainable The indicators for global monitoring need to be kept withdrawals, and supply of freshwater to people suffer- simple for feasibility and cost. However, countries, orga- ing from water scarcity; (2) integrated water resource nizations, and programs often monitor different aspects management, and (3) water-related ecosystems. The SDG of service performance, such as quantity, quality, prox- also set ambitious WASH-related targets of universal imity, reliability, price, and affordability (Roaf, Khalfan, access to safe water (target 6.1), adequate sanitation and Langford 2005). Some countries adopt more lenient and hygiene, and the elimination of open defecation definitions, and some adopt stricter definitions. (target 6.2) as well as reduced untreated wastewater The definitions in existing monitoring systems have (target 6.3). In the overall aim of access for all, the SDG several limitations. Some limitations are partially language and spirit emphasizes progressive reduction addressed by the new indicators for higher-level services. of inequalities and leaving no one behind, as well as The new indicators were informed by the five normative providing inclusive, quality, and sustainable services— criteria, as stated in the HRTWS and shown in table 9.2: thereby ensuring access for women and for poor and accessibility, acceptability, availability, affordability, and vulnerable populations. quality.2 • The Joint Monitoring Programme’s (JMP) definition Definitions of improved facilities focuses on the technology type To understand the status of drinking water, sanitation, and is an imprecise proxy for the quality of services and hygiene, one must make a distinction between dif- (Moriarty and others 2010; Onda, LoBuglio, and ferent levels of service access and population practices. Bartram 2012; Potter and others 2010). 172 Injury Prevention and Environmental Health Table 9.2 Proposed Service Level Definitions for Monitoring SDG 6 WASH Targets Service Basic services Safely managed services Water Percentage of population using an improved Percentage of population using safely managed drinking water services. drinking water source with a total collection “Safely managed” refers to an improveda drinking water source that is time of 30 minutes or less for a round trip, located on premises, available when needed, and free from fecal (E. coli) including queuing (termed “basic” water).a and priority chemical (arsenic and flouride) contamination. Sanitation and Percentage of population not practicing open Percentage of population using safely managed sanitation services, hygiene defecation. including a handwashing facility with soap and water. ”Safely managed” Percentage of population using an improved refers to an improved sanitation facility that is not shared with other sanitation facility that is not shared with households and where excreta are either safely disposed in situ or treated other households (basic sanitation).b offsite. Percentage of population with a handwashing facility with soap and water at home. Sources: Definitions of improved, WHO and UNICEF 2006; definitions of indicators, WHO and UNICEF 2015a. Note: The higher service level indicators are proposed for SDG monitoring. SDG = Sustainable Development Goal; WASH = drinking water, sanitation, and hygiene; WatSan = water and sanitation. a. Same as improved water monitored as part of the MDG target 7c: piped water into dwelling, plot, or yard; public tap and standpipe; tubewell and borehole; protected dug well; protected spring; rainwater collection. b. Same as improved sanitation monitored as part of the MDG target 7c: flush or pour-flush to piped sewer system, septic tank, pit latrine or ventilated improved pit latrine; pit latrine with slab and composting toilet. • Self-reported responses of access by household mem- Water Supply bers may be biased (Stanton and Clemens 1987). Globally, the use of improved drinking water sources • Statistics on household access provide no indication increased from 76 percent in 1990 to 91 percent in 2015 of variations in access and practices among different (WHO and UNICEF 2015b). Regional breakdowns for household members. For example, even in communi- progress between 1990 and 2015 are shown in figure 9.1. ties with high coverage rates for sanitation, children In its 2012 report presenting 2010 estimates, the UN still commonly defecate in the open.3 showed that its MDG target of halving the proportion of • Indicators do not adequately reflect accountability the population without access to safe drinking water had and sustainability, which are key elements that cut been met (WHO and UNICEF 2012b); however, such across all the service levels. global estimates mask regional disparities and inequities in access between urban and rural populations. As of The existing approach to measuring access does not 2015, 663 million people still used unimproved water provide a good indication of sustainability. The surveys sources, compared to 1.3 billion in 1990; 2.6 billion use representative sampling and do not follow individual people have gained access to improved water since 1990. households over time. Effective monitoring of higher Rural dwellers remain unserved compared with urban service levels requires regulatory data, but coverage is dwellers (16 percent and 4 percent, respectively). In Sub- poor in low- and middle-income countries (LMICs), Saharan Africa, 44 percent of rural dwellers continue to especially in rural areas. use an unimproved water supply. Water hauling costs Sub-Saharan Africans, especially women, billions of hours each year. In 2008, more than 25 percent of the Coverage of Water Supply, Sanitation, and Hygiene population in several Sub-Saharan African countries This section presents the coverage data at global spent more than 30 minutes to make one round trip to and regional levels for drinking water and sanitation collect water; 72 percent of the burden for collecting according to the JMP definitions used for monitoring water fell on women (64 percent) and girls (8 percent), MDG target 7c, thereby using the most recent update compared with men (24 percent) and boys (4 percent) and MDG assessment report (WHO and UNICEF (WHO and UNICEF 2010). 2015b). Breakdowns are provided by rural and urban Urban areas enjoy a higher level of water service, as areas.4 indicated by the use of piped water supply; in 2015, Water Supply, Sanitation, and Hygiene 173 Figure 9.1 Drinking Water Coverage Trends, by Regions and World, Using the JMP Improved Water Definition, 1990–2015 2 4 1 7 1 5 6 2 1 3 1 7 1 3 2 1 7 2 10 9 6 3 6 4 4 8 8 6 3 8 8 5 11 12 19 8 7 26 6 8 6 23 7 17 19 31 31 25 12 22 22 28 16 23 22 28 33 30 33 26 40 32 57 63 19 13 38 39 Percent 96 55 54 89 89 92 52 86 23 31 57 74 73 69 33 61 59 44 58 54 49 44 33 30 31 30 27 25 15 16 17 19 12 7 90 19 5 90 19 5 90 19 5 90 15 90 19 5 90 1915 90 19 5 90 19 5 90 19 5 90 19 5 19 5 90 19 5 90 15 90 1 1 1 1 1 1 1 1 1 1 19 20 20 20 20 19 20 20 20 20 20 20 20 20 20 Af aran ia As rn nia As rn lA d Af hern As rn ea d tri d gio g gio d ld ra an ibb n un pe re opin re lope As or he ste te ar ca a ea ica ia ia sia a ia n es ns ns co elo W es nt us h ut rt ric st Ea Oc l ve Sa No ve r W So Ce cas e C ri ev Ea De th Ame b- De t-d h- u Su Ca ut as tin So Le La Piped on premises Other improved Unimproved Surface water Source: WHO and UNICEF 2015b. Note: JMP = Joint Monitoring Programme. four of five people living in urban areas used piped have access to their own improved sanitation facility, a water, compared to two of three in rural areas. Water fact that, due to population growth, reflects no change sources classified as improved—even piped water—do in the unserved population of 1990. However, these not guarantee the safety or continuity of the water numbers mask the fact that since 1990, 2.1 billion peo- supply. Water quality surveys conducted in five coun- ple have gained access to improved sanitation. Regional tries showed that microbiological compliance with the breakdowns in progress between 1990 and 2015 are WHO guidelines varied between water sources and shown in figure 9.2. Globally, the proportion of popula- countries (Onda, LoBuglio, and Bartram 2012). On tion practicing open defecation declined from 24 percent average, compliance was close to 90 percent for piped in 1990 to 13 percent in 2015. In South Asia, 34 percent water sources, and from 40 percent to 70 percent for still defecate in the open, compared to 23 percent in other improved sources. Extrapolating to global esti- Sub-Saharan Africa. Globally, 638 million people mates, the authors estimate that in 2010, 1.8 billion (9 percent) share their sanitation facility with another people (28 percent) used unsafe water, more than family or families. Comparing rural and urban areas, twice the population of 783 million (11 percent) that 51 percent of rural dwellers have access to improved used an unimproved water supply. sanitation, compared with 82 percent of urban dwellers. Rates of improved sanitation do not reflect the amount Sanitation of fecal waste that is not isolated, transported, or treated The use of improved sanitation increased from safely; a study of 12 cities in LMICs found that whereas 54 percent in 1990 to 68 percent in 2015, but those gains 98 percent of households used toilets, only 29 percent of fell short of meeting the global MDG target (WHO and fecal waste was safely managed (Blackett, Hawkins, and UNICEF 2015b). In 2015, 2.4 billion people still did not Heymans 2014). 174 Injury Prevention and Environmental Health Figure 9.2 Sanitation Coverage Trends, by Regions and World, Using the JMP Improved Sanitation Definition, 1990–2015 7 1 3 2 8 2 1 3 2 8 3 3 2 13 12 11 17 16 2 4 7 2 16 13 17 7 10 20 24 23 31 31 36 34 7 7 2 7 11 45 40 10 6 5 12 65 10 5 17 9 15 48 48 20 25 10 27 5 7 6 3 Percent 26 12 94 96 6 94 96 90 25 89 83 17 77 80 20 4 5 7 72 71 68 14 67 6 10 62 54 48 50 47 43 35 35 38 30 24 22 20 90 19 5 90 19 5 90 19 5 90 15 90 19 5 90 1915 90 19 5 90 19 5 90 19 5 90 19 5 19 5 90 19 5 90 15 90 1 1 1 1 1 1 1 1 1 1 19 20 20 20 20 19 20 20 20 20 20 20 20 20 20 Af aran nia As rn As ast As rn ea d Af hern As rn lA d tri d gio g gio d ld ibb n ra an un pe re opin re lope or he ste te ar ca a ea E a ia ia ia n a ia sia es ns ns co elo W es h- nt us h ut rt ric ric Ea Oc l ve Sa No ve W ut So e C ri Ce cas ev De th Ame b- So De t-d u Su Ca as tin Le La Improved Shared Unimproved Open defecation Source: WHO and UNICEF 2015b. Note: JMP = Joint Monitoring Programme. Hygiene population into five equal wealth quintiles using an Although the MDG target 7c does not provide a global asset index. In 35 Sub-Saharan African countries, house- indicator for hygiene, the data on the presence of a hand- holds in the poorest wealth quintile are 6 times less likely washing facility with soap and water are increasingly to have water access compared with the richest quintile; collected as part of nationally representative surveys and the difference for sanitation is at least 2.5 times less likely will form the basis for efforts to monitor target 6.2 of the (WHO and UNICEF 2013). Figure 9.3 illustrates the SDGs. Two main sources include nationally representa- levels of disparity—between regions, between countries tive household surveys and a global review of published in a region, and at the country level—in the differences studies (Freeman and others 2014). Research studies between rural and urban areas and between wealth suggest that the global prevalence of handwashing with quintiles. Limited datasets are available on the dispari- soap after contact with excreta is 19 percent; rates are ties between population subgroups—for example, slum lower in Sub-Saharan Africa (14 percent) and South- populations, ethnic groups, women, the elderly, and East Asia (17 percent), where the most studies have been persons who have physical impairments—as the sample conducted (Freeman and others 2014). Proxy indicators size and sampling methodology in nationally represen- for handwashing practice from nationally representative tative surveys generally do not enable sufficiently robust surveys are not reliable and tend to over report hygiene comparisons. practices (Biran and others 2008). Global reporting of institutional WASH has not yet been standardized as it has for household-level WASH; Distribution of Services efforts are under way to build a global reporting system of The JMP has reported the distribution of water supply WASH in schools and health facilities for SDG monitor- and sanitation services by wealth status, breaking the ing. The Demographic and Health Survey (DHS) Service Water Supply, Sanitation, and Hygiene 175 Figure 9.3 Mozambique Example: How Average Values Mask Massive Disparities in Household Coverage 100 96 Poorest 20% Rural 80 77 South Sudan 75 Zambezia 65 Chad Open defecation prevalence (%) 60 60 Tete 57 Burkina Faso 51 Rural 50 Sofala 50 Poorest 20% 46 Sudan Urban 43 Nampula 40 38 Southern 40 Mozambique Asia 37 Manica 28 Sierra 28 Cabo Leone Delgado 25 Sub-Saharan Africa 23 Nigeria 20 20 Gaza 14 World Richest 20% 13 South-East 13 Kenya 13 Inhambane 13 15 Urban Rural Asia 7 Maputo 3 Northern Africa 2 Niassa 0 0 Mauritius 0 Maputo cidade 0 Richest 20% Urban Source: WHO and UNICEF 2014. Provision Assessment (SPA) monitors WASH in health understanding of the challenges facing the world to facilities. WASH coverage in both primary schools and meet the goal of universal access to institutional front-line health facilities is monitored and reported WASH within 15 years and to sustain that access under the Service Delivery Indicators, currently for Sub- beyond 2030. Unsustainable water extraction, along Saharan Africa. United Nations agencies collect data on with competing demands, population growth and WASH in schools (Education Management Information migration (including urbanization), and climate System operated by UNICEF), health facilities (Health change and variability, puts significant pressure on Management Information System operated by the WHO), water supply systems. In addition, new settlements and refugee camps (UN High Commissioner for Refugees). require systematic, coordinated planning, and existing In addition to enhanced monitoring efforts by settlements require retrofitting to bring sustainable UN agencies, UN member countries need greater WASH services to citizens. 176 Injury Prevention and Environmental Health IMPACTS OF INADEQUATE WASH The availability of water for drinking and household uses affects the quantity of water consumed and the time Understanding the nature and extent of the demonstrated available to care for children in the household. Reducing negative effects of inadequate WASH on individuals, the the distance required to fetch water is associated with environment, and societies is important for those design- lower prevalence of diarrhea, improved nutrition, and ing interventions and assessing benefits and efficiency. lower mortality in children under age five years Many benefits of WASH interventions are nonhealth in (Pickering and Davis 2012); these effects may be due to nature; including only health effects in impact evaluations better hygiene practices (Curtis and Cairncross 2003; can severely underestimate the intervention benefits Esrey 1996; Esrey and others 1991), as well as to addi- (Loevensohn and others 2015). tional time available for child care or income-generating activities (Ilahi and Grimard 2000), thereby resulting in healthier children. Health Consequences Inadequate quantities or consumption of water can Contaminated water and lack of sanitation lead to the also lead to dehydration, which has a number of adverse transmission of pathogens through feces and, to a lesser effects on physical and cognitive performance and bodily extent, urine. The F-diagram explained here but not functions (Popkin, D’Anci, and Rosenberg 2010). Because shown provides a basic understanding of these pathways there are no adequate biomarkers for measuring a popu- by which pathogens from feces are ingested through lation’s hydration status, such an effect remains largely transmission by fingers, flies, fluids, fields (soil), and food: undocumented (Popkin, D’Anci, and Rosenberg 2010). Safe drinking water provides the basis for oral rehydra- • Diseases transmitted by the fecal pathway include tion salts that save lives (Atia and Buchman 2009). diarrheal disease, enteric infection, hepatitis A and E, Exposure to harmful levels of arsenic in groundwater poliomyelitis, helminths, trachoma, and adenoviruses is estimated to affect 226 million people in more than (conjunctivitis) (Strickland 2000). Most of these dis- 100 countries (Murcott 2012). Arsenic exposure causes eases are transmitted through the fecal-oral pathway, skin lesions and long-term illnesses such as cancer, but some are transmitted through the fecal-skin path- neurological disorders, cardiovascular diseases, diabetes, way (for example, schistosomiasis) and the fecal-eye and cognitive deficits among children (Naujokas and pathway (for example, trachoma). These transmis- others 2013). sions occur between humans, as well as between Excess levels of water from heavy rainfall and inade- animals and humans. quate drainage lead to flooding, thus causing injuries • Pathogens carried through urine (for example, and death, as well as heightened risk of fecal-oral and leptospirosis) mainly result from animal-to-human skin diseases (Ahern and others 2005). Earthquakes, transmission. volcanic eruptions, tsunamis, and other natural disasters • Poor personal hygiene causes fungal skin infections, leave affected populations vulnerable to infection with such as ringworm (tinea) and scabies. waterborne diseases such as diarrhea, hepatitis A and E, • Lack of handwashing is associated with respira- and leptospirosis (Jafari and others 2011). tory infections (Rabie and Curtis 2006); inadequate hand hygiene during childbirth is linked to infec- Diarrheal Disease tion (Semmelweis 1983) and neonatal mortality The most recent study estimated 842,000 global deaths (Blencowe and others 2011; Rhee and others 2008). from diarrheal disease for 2012 (Prüss-Ustün and others 2014); 43 percent of these were children under age five A systematic review and meta-analysis documented years. An estimated 502,000 deaths were caused by large and significant associations between poor water, inadequate drinking water, 280,000 by inadequate sani- sanitation, and maternal mortality (Benova, Cumming, tation, and 297,000 by inadequate hand hygiene and Campbell 2014). The precise mechanism has not (table 9.3). The regional breakdowns indicate that the been well established, but it is thought to be largely major share of global burden is in South-East Asia and attributable to puerperal sepsis. Sub-Saharan Africa. Precise estimates remain elusive Children under age five years are especially vulnera- because of poor quality data on the cause of death; insuf- ble to infection. Regular exposure to environments with ficient data on hygiene practices; and poor quality evi- high fecal loads causes enteropathy5; compromises nutri- dence on the effectiveness of some water and sanitation tional status; and leads to long-term consequences, such interventions, especially onsite sanitation. This paucity as stunting and retarded cognitive development of reliable data has led to conflicting estimates of the (Humphrey 2009; Petri and others 2008). burden of disease. The Institute for Health Metrics and Water Supply, Sanitation, and Hygiene 177 Table 9.3 Diarrheal Disease Mortality Attributed to Poor Water Supply, Sanitation, and Hygiene in Low-and Middle-Income Countries, Regional and Risk Factor Breakdown Region Water supply Sanitation Hygiene WASH Africa 229,316 126,294 122,955 367,605 The Americas 6,441 2,370 5,026 11,519 Eastern Mediterranean 50,409 24,441 28,699 81,064 Europe 1,676 352 1,972 3,564 South-East Asia 207,773 123,279 131,519 363,904 Western Pacific 6,448 3,709 6,690 14,160 World 502,061 280,443 296,860 841,818 Source: Prüss-Ustün and others 2014. Note: WASH = safe drinking water, sanitation, and hygiene. Totals may not be sum of rows because of rounding. Columns 2–4 do not sum to column 5 because of overlap in risk pathways. Evaluation’s Global Burden of Disease (GBD) study con- (Hales and others 2014). These estimates may be conser- ducted a new meta-regression analysis of available vative because they do not account for diarrheal deaths experimental and quasi-experimental interventions. caused by other risk factors such as declining water It found that poor water and sanitation account for availability and undernutrition. 0.9 percent of global disability-adjusted life years (DALY) Cholera is an endemic diarrheal disease, but it is or 300,000 deaths per year (Lim and others 2012). The strongly associated with natural disasters and civil con- resulting difference between this study and the Prüss- flict. An estimated 2.9 million cases of cholera cause Ustün and others (2014) study is 542,000 deaths, possibly 95,000 deaths each year in 69 endemic countries (Ali and because the studies included in the GBD study do not others 2015). Cholera is transmitted through fecal con- differentiate between different levels of quality of water tamination of water or food. Therefore, clean water and supply and sanitation and between poor quality imple- proper sanitation are critical to preventing its spread. mentation and lack of effect. However, good evidence is lacking as to which mix of Not all diarrheal diseases are caused by pathogens interventions (including oral cholera vaccine, case man- transmitted through inadequate WASH. Over time, dif- agement, and surveillance) is most cost-effective during ferent estimates have been made for the burden of diar- outbreaks because few high-quality evaluation studies rheal disease that can be attributed to fecal-oral have been conducted (Taylor and others 2015). transmission. Earlier estimates attribute 94 percent of Institutional settings—such as schools, health diarrheal disease to poor WASH (Prüss-Ustün and facilities, prisons, and other public settings such as Corvalan 2007); the more recent study attributes refugee camps and public markets—can pose high risks 58 percent (Prüss-Ustün and others 2014). This latter if water and sanitation are not well managed. Studies estimate is closely supported by a separate review of have documented higher rates of diarrheal disease and more than 200 studies that examined the causes of diar- gastrointestinal infection in schools that lack access to rhea in inpatients and found no pathogen present in improved drinking water and sanitation facilities (Jasper, 34 percent of cases (Lanata and others 2013). Importantly, Le, and Bartram 2012). Improved hand hygiene is partic- deaths not easily preventable through WASH interven- ularly important in institutional settings, given the ease tions (for example, rotavirus spread among young chil- with which infections spread in such environments. dren and difficult to control) were excluded from the global burden of disease estimates for diarrheal disease Helminth Infections shown in table 9.3. Thus, the data in table 9.3 provide a Helminth infections are transmitted in water by fecal more realistic picture on how many deaths are consid- matter (schistosomiasis) and in soil by soil-transmitted ered preventable by WASH interventions. helminths (STH). Although routine monitoring of Rising temperatures caused by climate change are infection rates is limited, the large number of prevalence expected to exacerbate the burden of diarrheal disease. surveys permits global estimates to be made. The WHO estimates that an additional 48,000 deaths One study of helminth prevalence data for 6,091 in children under age 15 years will be caused by cli- locations in 118 countries estimated that in 2010, mate change by 2030 and 33,000 deaths by 2050 438.9 million people were infected with hookworm 178 Injury Prevention and Environmental Health (Ancylostoma duodenale), 819.0 million with roundworm that a subclinical condition of the small intestine caused (A. lumbricoides), and 464.6 million with whipworm by chronic ingestion of pathogenic microorganisms (T. trichiura) (Pullan and others 2014). Of the 4.98 million results in nutrient malabsorption. This subclinical con- years lived with disability (YLDs) attributable to STH, dition may be the primary causal pathway between poor 65 percent of those were attributable to hookworm, WASH and child growth (Humphrey 2009). 22 percent to A. lumbricoides, and 13 percent to T. trichi- The evidence on the etiology of diarrheal disease ura. Most STH infections (67 percent) and YLDs finds an association between levels of intestinal inflam- (68 percent) occurred in Asia (Central, East, South, and mation detected through fecal samples and subsequent South-East). A separate study estimated 89.9 million STH growth deficits in infants. This evidence lends support to infections in school-age children in Sub-Saharan Africa the environmental enteropathy hypothesis that stunting (Brooker, Clements, and Bundy 2006). Annual global may be an outcome of frequent enteric infection and deaths are estimated at 2,700 for A. lumbricoides and intestinal inflammation (Kotloff and others 2013). 11,700 for schistosomiasis (Lozano and others 2010). Because of the asymptomatic nature of environmental Helminth infections cause several adverse health enteropathy, the extent and seriousness of the condition outcomes, including anemia, malnutrition, growth stunt- is not known; however, it appears to be nearly universal ing, and impaired physical and cognitive development; among those living in impoverished conditions (Salazar- those outcomes result in low school attendance and Lindo and others 2004) and may be the cause of up to educational deficits, thus leading to loss of future eco- 43 percent of stunting (Guerrant and others 2013). nomic productivity (Victora and others 2008). The risk of The risks of low birth weight and stunting are height- STH infection is greatest for those in specific occupations ened in undernourished mothers (Özaltin, Hill, and and circumstances, such as people who work in agricul- Subramanian 2010), resulting in intergenerational con- ture, who live in slums, who are poor, who have poor san- sequences of undernutrition and related conditions. itation, and who lack clean water (Hotez and others 2006). Undernutrition and Environmental Enteric Social Welfare Consequences Dysfunction Improved water supply and sanitation provide individu- Undernutrition causes an estimated 45 percent of all als with increased comfort, safety, dignity, status, and child deaths (Black and others 2013) and is responsible convenience, and also have broader effects on the living for 11 percent of global disease burden (Black and others environment (Hutton and others 2014). The social wel- 2008). Inadequate dietary intake and disease are directly fare effects are difficult to quantify, given their subjective responsible for undernutrition; however, multiple indi- nature. Nevertheless, those benefits are consistently cited rect determinants exacerbate these direct causes, includ- as among the most important for beneficiaries of water ing food insecurity, inadequate child care practices, low supply and sanitation (Cairncross 2004; Jenkins and maternal education, poor access to health services, lack Curtis 2005) and may be particularly relevant for women of access to clean water and sanitation, and poor hygiene (Fisher 2006). practices (UNICEF 1990). Political, cultural, social, and economic factors play a role as well. Stunting (height- In or Near Homes for-age below minus two standard deviations from Water supply in or adjacent to homes provides greater median height-for-age of reference population) and comfort to household members, notably women and underweight (weight-for-age below minus two standard girls tasked with fetching water; water sources closer deviations from median weight-for-age of reference to home, especially piped water, are associated with population) are forms of undernutrition associated with increased use (Howard and Bartram 2003; Olajuyigbe weakened immune systems and severe long-term conse- 2010). quences that include poor cognitive development, a Data from 18 countries indicate that women are five lower rate of school attendance, a lower level of job times more likely than men to have the responsibility for attainment, and a potentially higher risk of chronic dis- collecting household water (WHO and UNICEF 2012b). ease in adulthood (Victora and others 2008). As the distance to the water source increases, the time The links between diarrhea and child undernutrition that women could spend on income-generating activities, (Fishman and others 2004; Prüss-Ustün and Corvalan household chores, and child care decreases (Ilahi and 2006) and other enteric infections (Brown, Cairncross, Grimard 2000). A regular piped water supply can intro- and Ensink 2013; Checkley and others 2008; Guerrant duce the possibility of purchasing time- and labor-saving and others 2008; Lin and others 2013) are well devices, such as washing machines and dishwashers. documented. An emerging body of evidence suggests Although access to water infrastructure does not always Water Supply, Sanitation, and Hygiene 179 translate into wage employment for women (Lokshin (Sebastian, Hoffmann, and Adelman 2013). Lack of and Yemtsov 2005), one study finds that it can provide adequate MHM is frequently described as a hindrance to time savings in water collection, thus improving gender girls’ education, but high-quality evidence is lacking equality (Koolwal and Van de Walle 2013). (Sumpter and Torondel 2013). A randomized controlled Individuals with access to on-plot sanitation benefit trial in Nepal suggests that menses, and poor menstrual from greater privacy, comfort, and convenience. hygiene technology in particular, has no effect on absen- Accompanying a child to the toilet is more convenient teeism of girls; girls miss less than one school day a year if it is nearby and safe, and mothers can comfortably on average because of menstruation (Oster and Thornton step away from household duties to practice hygiene. 2011). However, girls may avoid going to school while In Ghana, more than 50 percent of households consid- they are menstruating, not because they lack manage- ering adopting a toilet included convenience in their ment methods but because they lack proper facilities for top three reasons for investing in sanitation (Jenkins managing menses (Jasper, Le, and Bartram 2012). and Scott 2007). In six countries of South-East Asia, the rural households that owned their own latrine saved from 4 to 20 minutes of travel time per trip Environmental Consequences (Hutton and others 2014). Privacy, comfort, and con- Two major environmental consequences of poor WASH venience benefits are magnified for vulnerable groups, practices are (1) the excessive extraction of water to meet such as the elderly or persons living with disabilities or population needs and (2) the pollution caused by poorly debilitating chronic illness. managed human excreta. On-plot sanitation reduces the risk of theft or assault The water supply for domestic use represents a small (including rape and sexual harassment), especially at proportion of overall extraction, but the concept of night or in isolated locations. Improved pit latrines are virtual water trade6 has led to a greater understand- safer, less likely to collapse, and easier for small children ing of the implications of population consumption to use. On-plot water supply and sanitation help to patterns for water use. Globally, the combined effects of avoid conflicts with neighbors, landowners, or others socioeconomic growth and climate change indicate that, over the use of shared water resources and sanitation by 2050, the population at risk of exposure to at least facilities and the use of fields or rivers for open a moderate level of water stress could reach 5 billion defecation. people (Schlosser and others 2014). A population of up to an estimated 3 billion in 2050 is nearly double the Schools and Workplaces current estimate of 1.7 billion people who live in areas Access to improved WASH services in schools and work- with a high degree of water stress. The projections are places contributes to school attendance, school perfor- made on the basis of a risk metric of frequency of water mance, and choice of where to work, especially for girls shortage in reservoirs (Sadoff and others 2015). This and women. Recent evidence from India shows that a metric combines hydrological variability and water national government program to build toilets in schools usage trends, which may be mitigated by storage infra- led to an 8 percent increase in enrollment among structure. This class of water insecurity is most severe in pubescent-age boys and girls and a 12 percent increase South Asia and Northern China, although the risk of among younger children of both genders (Adukia 2014). water shortage exists on all continents. The comparably large effect of school sanitation on pri- Overextraction of groundwater and pollution of local mary school children and the robust effects for boys and surface water bodies have led many large urban popula- girls at all ages suggest that at least some of the effect of tion centers to source municipal water supplies from school sanitation is related to health (Jasper, Le, and reservoirs or rivers that are tens or hundreds of kilome- Bartram 2012). Research has seldom analyzed academic ters from the site of treatment or consumption. performance as an outcome; however, given the role that Such schemes cost tens of millions of dollars each in improved water and sanitation have on child health and reservoir construction, pipeline, and pumping costs. school attendance rates, the current evidence lacks Groundwater resources are under increasing stress from research into their role in academic performance. unsustainable agricultural practices resulting from crop choice and energy subsidies to enable farmers to pump Menstrual Hygiene groundwater. In India and Mexico, for example, subsi- Menstrual hygiene management (MHM) is a poorly dized electricity and kerosene for farmers have led to understood and underresearched area of WASH services. serious groundwater overdraft (Scott and Shah 2004). This neglect has left women in many LMICs without Poorly managed human excreta have major access to appropriate products, facilities, and services environmental consequences; excreta pollute human 180 Injury Prevention and Environmental Health settlements, groundwater, and surface water such as INTERVENTION OPTIONS AND lakes, rivers, and oceans. The degree of pollution EFFECTIVENESS depends on wastewater, sludge, and sewage manage- ment practices; climatic factors; and the population Three main categories of interventions to improve size and density in relation to the volume of water. In WASH are as follows: highly populated river basins, municipal sewage and wastewater contribute a high proportion to overall bio- • Technology options and WASH practices cover logical oxygen demand (Corcoran and others 2011; the type of hardware, equipment, and associated Rabalais and Turner 2013). behaviors of WASH services. Not all water or san- Heavily polluted surface water has serious effects on itation technologies perform the same function, ecosystems, food webs, and biodiversity (Turner and so they can be classified by the service level they Rabalais 1991). Coastal areas that are near the discharge provide. of large, polluted rivers have reported compromised • Service delivery models cover the components of fish catch, such as in Argentina (Dutto and others WASH service implementation. Those compo- 2012). In the coastal areas of the Philippines, water nents include (1) approaches to demand generation pollution was estimated to cost US$26 million per year and WASH behavior change, (2) approaches to in lost fish catch and degraded coral reefs (World Bank strengthen supply of water and sanitation goods and 2009). Water pollution of recreational areas affects the services, and (3) approaches to improve the effective- tourism industry, thus lowering visit rates or causing ness of WASH service delivery. gastrointestinal illness or both. • Strengthening the enabling environment for WASH service delivery includes (1) measures to strengthen capacity, (2) legal framework, (3) policy and plan- Financial and Economic Consequences ning, (4) resource allocation, (5) monitoring and Financial and economic studies convert the health, evaluation, and (6) other interventions to provide social, and environmental effects of poor WASH to a a stronger foundation for implementing the tech- common money metric, thereby enabling aggregation nology and service delivery models. The evidence is as well as comparison across locations and over time. provided in annex 9B. However, these estimates are often incomplete, using crude estimates of economic value or relying on the imprecise physical effects underlying the economic Effectiveness of Technologies and Practices values. Water technologies are designed to source, treat, distrib- Damage cost studies account for the broader welfare ute, and monitor the supply of water. Epidemiological and productivity consequences of poor WASH beyond studies evaluate the effectiveness of water interventions the health effects. A review of economic impacts of poor in terms of the quantity and (microbial) quality of water water and sanitation found estimates from more than 30 supplied (Waddington and others 2009). Increasing evi- countries (see annex 9A), as well as global studies. dence enables the comparison of the incremental health Studies with economic impacts expressed as a percentage benefits of different water interventions, such as of gross domestic product (GDP) are shown in figure 9.4, improved community source, piped water, higher-quality disaggregated between health and nonhealth damages. piped water, and point-of-use treatment (chlorine, solar, Although all the studies presented in figure 9.4 and filter). Utility regulators, as well as regional and present effects in monetary units, the results are not global initiatives, monitor water quality according to directly comparable. They have different base years and service standards, such as continuity, consumption, and different effects included; some include only sanitation, number of complaints (IBNET 2014). In 2010, The and others include water and sanitation. In East Asian International Benchmarking Network for Water and and Pacific and Sub-Saharan African economies, the Sanitation Utilities (IBNET) of the World Bank reported cost of poor sanitation exceeded 2 percent of total that only 16 percent of utilities in low-income countries GDP; in South Asia, it exceeded 4 percent of GDP. A supply water continuously 24 hours per day, compared global study, including the health and time losses, val- to 86 percent of utilities in middle-income countries ued the costs in LMICs at 1.5 percent of global domes- (Van den Berg and Danilenko 2010). Even a few days of tic product (Hutton 2012). These significant economic interrupted water supply can result in significant adverse effects raise awareness of the extent of the problem, but health consequences if beneficiaries revert to using they do not indicate how to address the problem in a unimproved sources of water (Hunter, Zmirou-Navier, cost-effective manner. and Hartemann 2009). Water Supply, Sanitation, and Hygiene 181 Figure 9.4 Economic Costs of Poor Water and Sanitation in Selected Countries, as a Percentage of GDP 8 7 Cost as percentage of GDP 6 5 4 3 2 1 0 Ban India 008) Lao sh (20 ) Pak R (20 ) -Sa Indo n (20 ) 11) Nic frica 009) Phi gua 08) ine 10) Iran Mo m (2 ) ) Alg p. (20 ) Mo ria (19 2) co 99) Egy Leb an (2 0) ian Ara non ( 2) epu (19 ) blic 99) isia 01) pt, Gha 1999) Pak ep. (2 4) Gu istan 03) ma 2003) eria 06) Ban Nepa 004) Tun s (20 ) h (2 5) El S Peru 002) Ho dor (2 3) Phi duras 05) Sen ia (19 ) ega 99) Pak na (2 5) ) ) ) ) Per 003) ) de 2011 11 09 08 008 0 9 ine 07 07 ista 005 Ind (2005 Ind 1995 Chi 2009 990 ic R (201 0 Jor (200 00 Ara b R 199 b R (200 0 a 200 Gh l (200 ara (20 lipp (20 Vie s (20 Tun (20 0 Nig la (20 gla l (20 0 lipp (20 2 2 (2 (2 u (1 ia ( ( an ia ( ( ( ( ia ( ia ( lam lia b R ep. Ara na n na ista des PD bod har nes tna e is roc , Is ngo d a e a alv A gla ate Cam n Syr pt, Egy Sub Country (year of cost) ESI studies Cost of environmental Country environmental analyses (World Bank/ Other studies degradation studies environment department) (World Bank/METAP) Nonhealth Health Source: See annex 9A for fuller datasets and references. Note: GDP = gross domestic product. Economics of Sanitation Initiative (ESI) studies have been implemented by the World Bank’s Water and Sanitation Program in 34 countries of Latin America and the Caribbean, East Asia and Pacific, South Asia, and Sub-Saharan Africa. These studies estimated the costs of poor sanitation, including health and nonhealth impacts (access time, costs of accessing safe water, tourism). The Mediterranean Environmental Technical Assistance Program (METAP) of the World Bank conducted studies on the costs of environmental degradation in eight Mediterranean countries from 1999 to 2002. Country environmental analyses conducted by the World Bank in more than 20 countries since 2003 have estimated the health costs of poor water and sanitation. To increase safety, drinking water can be treated either To reduce the transmission of pathogens, sanitation at the source or at the point of use through a process of technologies isolate, transport, and treat fecal waste, and filtration or disinfection or both. The greatest health they also provide users with a dignified and comfortable effects for improved water treatment technologies con- experience when going to the toilet. Different rungs on cern the piped water supply, with greater health benefits the “sanitation ladder” confer different health impacts associated with higher-quality piped water (water that is and user experiences; hence, utilization of different safe and continuously available) (Wolf and others 2014). kinds of sanitation services or facilities can vary. For Among household-level studies, filter interventions that example, communal facilities may be poorly main- also provided safe storage (for example, ceramic filters) tained, in which case they are less likely to be used by were associated with a large reduction in diarrheal dis- women, children, and individuals who are disabled or ease (Wolf and others 2014). Neither chlorine treatment infirm. Distance also decreases usage of communal nor solar disinfection shows significant impact on diar- toilets (Biran and others 2011). rhea after meta-analysis adjusted for non-blinding of the Hygiene technologies enable users to perform basic intervention (Wolf and others 2014), although an earlier personal hygiene functions. Epidemiological studies have systematic review and meta-analysis of water quality typically used the presence of a place for handwashing interventions found household-level treatment to be with soap and water as a proxy for handwashing practice; more effective than source treatment (Clasen and others however, this has been shown to be only loosely cor- 2005). Blinding participants to the intervention and lon- related with observed handwashing behavior (Ram 2013). ger follow-up periods are recommended to better under- One synthetic review and meta-analysis of health stand the impact of point-of-use water treatment impact assessments of water and sanitation interven- interventions on diarrhea (Clasen and others 2005). tions includes 61 individual studies for water, 182 Injury Prevention and Environmental Health 12 observations comparing unimproved and improved (95% CI: 0.44–0.61) for the three soil-transmitted hel- sanitation conditions, and only 2 observations com- minths combined, 0.54 (95% CI: 0.43–0.69) for A. paring unimproved sanitation and sewer connections lumbricoides, 0.58 (95% CI: 0.45–0.75) for T. trichiura, (Wolf and others 2014). and 0.60 (95% CI: 0.48–0.75) for hookworm Table 9.4 shows relative risk reductions for different (Ziegelbauer and others 2012). movements up the water supply and sanitation ladders. Access to sanitation has been associated with lower The summary risk ratio for all observations on diarrhea trachoma, as measured by the presence of trachomatous morbidity is 0.66 (95% confidence interval [CI]: inflammation–follicular or trachomatous inflammation– 0.60–0.71) for water interventions and 0.72 (95% CI: intense with odds ratio 0.85 (95% CI: 0.75–0.95) and 0.59–0.88) for sanitation interventions (Wolf and others C. trachomatis infection with odds ratio 0.67 (95% CI: 2014). An earlier review of 25 studies investigating the 0.55–0.78) (Stocks and others 2014). association between sewerage and diarrhea or other A systematic review examined the impact of related outcomes estimated an average risk ratio of 0.70 improved WASH on child nutritional status. Specifically, (95% CI: 0.61−0.79), which increased to as much as 0.40 a meta-analysis of five randomized controlled trials when starting sanitation conditions were very poor found a mean difference of 0.08 in height-for-age (Norman, Pedley, and Takkouche 2010). z-scores of children under age five years (95% CI: A meta-analysis of hygiene interventions found an 0.00–0.16) for solar disinfection of water, provision of average risk ratio for diarrhea of 0.60 for promotion soap, and improvements in water quality (Dangour and of handwashing with soap (95% CI: 0.53–0.68) and others 2013). However, the authors raised concerns 0.76 for general hygiene education alone (95% CI: about the low methodological quality of the included 0.67–0.86) (Freeman and others 2014). These results studies and the short follow-up periods; there was are summarized in table 9.4. An earlier systematic insufficient experimental evidence on water supply review found a relative risk compared to no handwash- improvement and sanitation to include in the meta- ing of 0.84 (95% CI: 0.79–0.89) for respiratory infection analysis. Since publication of the Dangour and others (Rabie and Curtis 2006). (2013) review, several additional randomized controlled A meta-analysis that combined sanitation availabil- trials of household sanitation interventions have been ity and use examined the impact of improved sanita- completed (Briceno, Coville, and Martinez 2014; tion on soil-transmitted helminths. The meta-analysis Cameron, Shah, and Olivia 2013; Clasen and others reported the following overall odds ratios:7 0.51 2014; Hammer and Spears 2013; Patil and others 2014), Table 9.4 Meta-Regression Results for Water and Sanitation Interventions: Relative Risks Compared with No Improved Water, Sanitation, or Hygiene Practice Baseline Outcomea Baseline water Outcome water Improved community Piped water, Piped water, high Filter and safe storage source noncontinuous quality in the household Unimproved source 0.89 [0.78, 1.01] 0.77 [0.64, 0.92] 0.21 [0.08, 0.55] 0.55 [0.38, 0.81] Improved community source 0.86 [0.72, 1.03] 0.23 [0.09, 0.62] 0.62 [0.42, 0.93] Basic piped water 0.27 [0.10, 0.71] 0.72 [0.47, 1.11] Baseline sanitation Outcome sanitation Improved sanitation, no sewer Sewer connection Unimproved sanitation 0.84 [0.77, 0.91] 0.31 [0.27, 0.36] Improved sanitation, no sewer 0.37 [0.31, 0.44] Baseline hygiene Outcome hygiene General hygiene education Handwashing with soap No hygiene education or handwashing 0.76 [0.67, 0.86] 0.60 [0.53, 0.68] Sources: Water and sanitation: Wolf and others 2014; hygiene: Freeman and others 2014. a. Brackets represent 95 percent confidence intervals. Water Supply, Sanitation, and Hygiene 183 most of them failing to find a significant relationship chemicals, solar and ultraviolet lamps, and flocculation) between the interventions and child health or growth and safe storage technologies with communication- and outcomes. One exception is a study in rural Mali of behavior-change techniques (Peal, Evans, and van der Community-Led Total Sanitation (CLTS), which led Voorden 2010). Despite substantial evidence pointing to to taller children on average (+0.18 height-for-age health benefits of HWTS, skepticism remains that the z-score, CI on z-score: 0.03–0.32). These children were results may largely be the result of bias; concerns remain 6 percentage points less likely to be stunted after the about the extent of uptake, use, and scalability of com- intervention (Pickering and others 2015). Econometric mercially marketed HWTS, particularly among poor studies drawing on time series data establish links populations most at risk of diarrheal disease (Schmidt between open defecation and stunting (Spears 2013), and Cairncross 2009). between open defecation and childhood diarrhea in Handwashing promotion has been tested in forma- India (Andres and others 2014), and between open def- tive research and has applied social cognitive models to ecation and cognitive development in India (Spears and determine what motivates and changes behavior. The Lamba 2013). A source of regularly updated evidence promotion has used a variety of communication reviews on WASH interventions with strict inclusion channels—such as television, radio, theater groups, criteria is the Cochrane Library.8 community meetings, and face-to-face visits—to reach target groups who typically are mothers of young chil- dren or school-age children. A pre- and post-evaluation Effectiveness of Service Delivery Models of the approach in Burkina Faso, which targeted the Effectiveness of service delivery models is measured by behavior of safe disposal of child feces and handwashing intervention uptake, change in risky behaviors, sustain- after contact, documented increases in handwashing ability, and, to a lesser extent, health outcomes. Large- (Curtis and others 2001). A similar approach to improve scale approaches that include demand raising and handwashing behavior was piloted on a large scale under behavior change are needed to achieve universal access, the Water and Sanitation Program’s Global Scaling Up but experience has shown these approaches result in Handwashing Projects in Peru, Senegal,10 Tanzania, and lower average effectiveness. Vietnam. Experimental evidence from Peru (Galiani and others 2015), Tanzania (Briceno, Coville, and Martinez Approaches to Demand Generation and WASH 2014), and Vietnam (Chase and Do 2012) suggests the Behavior Change campaigns were only marginally successful. The Peru Demand-based approaches start from the premise that study did find large changes in behavior in a subset of lasting change is brought about when individual and communities with children who participated in a school- community behaviors are affected. CLTS and its school- based handwashing promotion intervention. Effects on based counterpart, School-Led Total Sanitation (SLTS), health were not observed in any of the countries, and the promote broader changes in sanitation and hygiene sustainability of handwashing was not measured. A key behaviors at the community level. Since its founding in obstacle identified in both Tanzania and Vietnam was 1999, the CLTS approach has rapidly expanded to more the difficult trade-off between scale and intensity of than 50 developing countries, where many thousand activities. successful applications of the approach have been made; The Global Public-Private Partnership for at least 16 national governments have adopted CLTS as Handwashing (PPPHW) combines the marketing national policy.9 Rigorous evaluation of the CLTS expertise of the soap industry with government support approach has been limited, and the reliance on the and the enabling environment to trigger behavior emergence of natural leaders presents difficulties in change and reduce diarrhea. Whereas the PPPHW has testing the effectiveness of CLTS using conventional expanded globally, the coalition has not yet been subject experimental methods. One exception comes from a to rigorous effectiveness trials (Peal, Evans, and van der recent example in rural Mali, in which CLTS was well Voorden 2010). Evaluations of PPPHWs have been implemented in a random set of villages and shown to commissioned by private soap companies and involved almost double coverage of a private latrine (Pickering providing free soap to households (Nicholson and and others 2015). others 2013), thus limiting their external validity. Specific behaviors, such as household water treat- ment and storage (HWTS) and handwashing with soap, Approaches to Strengthening Supply of Water and have been the subject of behavior change campaigns. Sanitation Goods and Services HWTS combines marketing of low-cost water treatment Supply-side approaches to water and sanitation service (for example, boiling, filtration, disinfection using delivery cover the full value chain from production and 184 Injury Prevention and Environmental Health assembly of inputs, importation, sales, distribution, models could be improved to enhance the quality of installation, and maintenance of water infrastructure services as well as increase take-up of services, especially and latrines. Services range from micro and small-scale among the poorest populations. independent water resellers; network operators; well Results-based approaches11 to development that and pit diggers; operators offering masonry, pit, and offer financial or nonmonetary rewards upon demon- septic tank emptying; and public toilet operators to stration of measurable outputs or outcomes are used medium-scale sanitation markets—or sanimarts— increasingly for achieving desirable outcomes. The spe- offering a full range of sanitation goods and services. cific details differ, but such approaches share a common Small-scale operators can effectively serve rural mar- aim of shifting the overall incentive structure from kets, where the majority of people without access to financing infrastructure to delivering services. Until piped water and sanitation live. However, the existing recently, the experience using results-based approaches literature highlights several obstacles to growth and the in water and sanitation was limited. A review by the ability of such providers to effectively serve these rural World Bank in 2010 indicated that less than 5 percent of populations. its output-based aid (OBA) portfolio was in water and Rural operators often face higher per capita costs sanitation (Mumssen, Johannes, and Kumar 2010). The because they lack economies of scale enjoyed by larger use of OBA has increased under the Global Partnership utilities and therefore have lower revenue potential (Baker on Output-Based Aid (GPOBA), which lists 22 projects 2009). Investment financing needed for growth can be in water supply and sanitation whose outputs include difficult to secure, and the lack of formalization in the water, sewerage, or sanitation connections.12 Multilateral sector can result in insecure operating environments and bilateral agencies such as the World Bank, Inter- (Sy, Warner, and Jamieson 2014). The availability of alter- American Development Bank, and Department for native sources of free or low-cost water makes rural areas International Development (DfID) have shifted funding less attractive to independent operators. Low or uneven toward results-based approaches in water and sanita- demand has limited growth opportunities for small-scale tion. As of early 2016, the World Bank’s Program-for- onsite sanitation service providers. Despite these obstacles, Results Financing (PforR) has six active operations in small-scale service providers are increasingly recognized water supply, sanitation, and hygiene. as a central part of the solution to close the gap in water Microfinance or microcredit can help poor households and sanitation access, particularly among the poor. facing liquidity constraints to invest in water supply and Supply-side strengthening is predominant in the sanitation by (1) smoothing consumption over time, (2) Community Approach to Total Sanitation (CATS) pro- encouraging households to be more willing to adopt moted by the United Nations Children’s Fund and the improved services, and (3) giving those households an Total Sanitation and Sanitation Marketing (TSSM) opportunity to purchase more durable, higher levels of approach of the World Bank Water and Sanitation service. Consumer credit has been applied successfully to Program. Recent randomized control trial impact evalu- increase the installation and use of household piped ations of TSSM in Madhya Pradesh, India (which water connections (Devoto and others 2011), but exper- included a hardware subsidy to households below the imental evidence of consumer lending for sanitation poverty line); East Java, Indonesia; and 10 rural districts remains limited. However, emerging interest in the of Tanzania found the approach varied widely in its potential of microfinance for household sanitation and effectiveness across the countries, with no increase in the results of small-scale pilots are promising. A ran- improved sanitation in Indonesia (Cameron, Shah, and domized study in Cambodia found a fourfold increase in Olivia 2013) and increases of 19 and 15.7 percent in uptake when households were offered a 12-month Madhya Pradesh (Patil and others 2014) and Tanzania low-interest loan to purchase a latrine (Shah 2013). (Briceno, Coville, and Martinez 2014), respectively. Finally, interest is emerging for using large-scale Despite better sanitation coverage in Madhya Pradesh, delivery platforms for social services and poverty reduc- large numbers of adults continued to practice open tion. These platforms can help improve the targeting of defecation. WASH services and will make use of the tools and mech- anisms those programs have for improving livelihoods Approaches to Improve the Effectiveness of WASH and outcomes for the poor. Examples include the Service Delivery following: Addressing the supply- and demand-side constraints of WASH service delivery has led to large increases in • Sanitation subsidies and financing can be targeted to access. But the persistence of regional and socioeco- conditional-cash transfer (CCT) participants, many nomic disparities in access suggests that current delivery of whom lack adequate sanitation. A more ambitious Water Supply, Sanitation, and Hygiene 185 approach could make receipt of cash transfers condi- Despite its importance, cost information is not tional on a household’s use of improved sanitation. commonly tabulated in an appropriate format to sup- These programs also provide outreach and coun- port decision making. At the policy level, budgets and seling to reach target households with sanitation resource allocations are fragmented among subsectors, promotion messages that build awareness and help levels of government, and sector partners or financiers. change behavior. Considerable differences exist between budget alloca- • Community-driven development (CDD) programs tions and disbursements. WASH-BAT (bottleneck can be used as a platform to deliver CLTS and to analysis tool), developed by UNICEF, helps consolidate follow up with participatory planning and budgeting the budgetary needs so that sector bottlenecks can be to ensure that communities become free of open removed (see annex 9B) (UNICEF 2014). At the pro- defecation. gram or service delivery levels, implementers do not • Safety-net programs that build skills and strengthen easily share information on their costs, and budgets sources of livelihood can include sanitation busi- may not be structured for simple breakdowns between nesses and services such as masonry, plumbing, and software and hardware costs. Cost studies for WASH electrical skills among the list of profitable invest- technologies are more abundant, and at the local level, ments for beneficiaries. the market or subsidized price is available. However, • Many nutrition interventions already promote the price is rarely the same as the cost. The price com- handwashing with soap, safe water, and sanitation. monly contains either a profit or a subsidy; because Handwashing demonstrations are often included in both are transfer payments, they should be excluded promotions for breastfeeding and interventions for from economic analysis. However, to ease the research feeding infants and young children, which also stress burden, it is common practice in economic analysis to the use of safe water in food preparation. use prices as a proxy for cost, adjusting for any known subsidy or profit. More innovative integration approaches may use Published cost evidence is available in aggregated those same channels to discuss with the community and unit forms. Aggregated cost includes the expendi- sanitation product options and services that are avail- ture required to meet specified targets. The World able. Evidence is needed on the effectiveness and the cost Bank estimates that the global capital costs of achieving of integrated approaches. Such information may high- universal access to WASH services by 2030 are light the need for more operational research and impact US$28.4 billion per year confidence interval [CI]: evaluations to inform policy and program design. US$13.8 billion to US$46.7 billion) from 2015 to 2030 for basic WASH and $114 billion per year (CI: US$74 billion to US$166 billion) for safely managed INTERVENTION COSTS, EFFICIENCY, AND WASH (Hutton and Varughese 2016).13 Those costs SUSTAINABILITY are equivalent to 0.10 percent of global product for basic WASH and 0.39 percent of global product for Any intervention in the WASH sector requires an eco- safely managed WASH, including 140 LMICs. Those nomic rationale, thus satisfying conditions of efficiency, needs compare with 0.12 percent of its gross product affordability, and relevance. Cost-benefit analysis com- spent on meeting the MDG water target and making pares the intervention costs with the benefits, expressed progress toward the sanitation target. Universal basic in monetary units. Cost-effectiveness analysis compares access by 2030 is feasible at current spending but the intervention costs with the benefits, expressed in requires reallocations to sanitation, to rural areas, and some other common unit, such as lives gained or pollu- to off-track regions. However, substantial further tion load to the environment averted. spending is needed to meet the higher standard of safely managed services. The costs as a proportion of Costs gross regional product are shown by MDG region in The cost of interventions is one key piece of evidence for figure 9.5. Regions most challenged to reach universal decision making, because it is relatively easy to obtain access are South Asia and Sub-Saharan Africa. and is often cited as a constraint for an investment deci- Many countries also produce investment plans for sion, whether by governments, private sectors, house- meeting national targets, thereby focusing on the financ- holds, or individuals. Costs can be measured for the ing the government will provide. The Organisation for WASH technology (the hardware), the service delivery Economic Co-Operation and Development (OECD) has approach (the “software” or program management), and created FEASIBLE, a tool for developing national financ- the enabling environment. ing strategies by comparing the costs of meeting national 186 Injury Prevention and Environmental Health Figure 9.5 Costs of Basic and Safely Managed Services as a Percentage of GRP, by Region, with Uncertainty Range 3.0 2.5 2.0 2.01 Cost as a percentage of GRP 1.5 1.0 0.64 0.85 0.58 0.5 0.45 0.39 0.36 0.31 0.27 0.21 0.23 0.12 0.10 0.15 0.12 0.10 0.09 0.04 0.05 0.02 0.04 0.03 0 ca ia ld ia ia A ia ed nia A C As or As As SS fri As LA CC lop W ea nA st rn rn rn ve Oc Ea te ste he er De es h- rth ut Ea W ut So No So SDG targets 6.1 and 6.2 Basic WASH Upper and lower estimates of the cost Upper and lower estimates of the cost of meeting SDG targets 6.1 and 6.2 of achieving basic WASH Source: Hutton and Varughese 2016. Note: CCA = Caucasus and Central Asia; GRP = gross regional product; LAC = Latin America and the Caribbean; SDG = Sustainable Development Goal; SSA = Sub-Saharan Africa; WASH = water, sanitation, and hygiene. See table 9.2 for details on upper and lower values on variables varied in sensitivity analysis. GRP is based on the aggregated gross domestic product of countries in each region. An economic growth rate of 5 percent is assumed across all regions. targets with the projected financing available.14 FEASIBLE However, those studies find monthly expenditure is has been applied in at least 12 countries (OECD 2011). more similar between the two sources because of A key input to these aggregated studies is the unit higher consumption of piped water than of other costs of WASH provision at the household or water sources (Whittington and others 2009). The community level. Because of climatic, topographical, IRC WASHCost project calculated benchmark capital and socioeconomic differences, the costs of providing and recurrent costs for basic levels of water service in service vary highly between studies, contexts, and lev- Andhra Pradesh, India; Burkina Faso; Ghana; and els of service. The costs per cubic meter of water and Mozambique (Burr and Fonseca 2013). Benchmark of wastewater services, as well as average monthly capital costs ranged from US$20 per person for bore- household bills, are available for utility services holes and hand pumps to US$152 for larger water through national regulators, regional associations, and schemes. Benchmark recurrent costs ranged from global initiatives (IBNET 2014). Studies commonly US$3 to US$15 per person per year, but actual expen- compare the cost of different sources of water supply, ditures were substantially lower. Construction cost for and they find piped water to be significantly cheaper equivalent latrines varies widely between settings per unit compared with vendor-supplied water. (Hutton and others 2014). Comparison of alternative Water Supply, Sanitation, and Hygiene 187 sanitation transportation and treatment technologies economic variables are rarely captured. The majority of also provides important policy direction; in general, economic studies build models filled with data from a fecal sludge management is considerably cheaper than mixture of sources. Global studies assessing the economic sewerage, as in Dakar, Senegal, where it was found to benefits of improved water supply and sanitation include be five times cheaper (Dodane and others 2012). health economic benefits and convenience time savings Extrapolating available data from one context to (Hutton 2013; Whittington and others 2009). Country another carries risks. Therefore, simple costing tools studies have also evaluated the value of health and time and investment in evidence gathering are required so savings (Pattanayak and others 2010). Regional studies that cost estimates of specific locations can be made.15 from Southeast Asia assess the water access, reuse, and Ideally, those who determine the costs of water supply tourism benefits of improved sanitation as a proportion and sanitation services would consider the externalities of avoided damage costs (Hutton and others 2008, 2014). and the long-run cost of supply. One study provides an Willingness-to-pay (WTP) studies have estimated the illustrative example of the full costs of water supply and economic value of water quality improvements, but only sanitation (including opportunity costs and very few studies use experimental methods (Null and environmental costs) with the low costs, varying from a others 2012). Other studies have assessed WTP to avoid high of US$2.00 per cubic meter to a low of US$0.80 per health impacts (Guh and others 2008; Orgill and others cubic meter (table 9.5) (Whittington and others 2009). 2013) and to receive piped water (Whittington and From a policy perspective, the affordability and others 2002). A systematic review of those studies has willingness to pay for those costs is a critical issue. shown that the economic value derived from the WTP A global review found that water supply costs as a for improved water quality is less than the cost of pro- proportion of household income are significantly higher ducing and distributing it (Null and others 2012). Social for poorer populations (Smets 2014) and well above the benefits have been assessed, but few have been expressed benchmark of between 3 percent and 5 percent used by in money values except WTP studies, which tend to cap- some governments and international organizations. ture all benefits and make differentiating social from other benefits difficult. Economic value is associated with river cleanup Benefits that includes improved management of municipal WASH services have a large array of welfare and wastewater, as well as improved management of indus- development benefits. Table 9.6 classifies those benefits trial discharge, agricultural runoff, and solid waste. under health, convenience, social, educational, reuse, The financial viability of WASH services has been water access, and other. expressed in terms of financial returns. The most com- Those benefits have been evaluated extensively, but prehensive source of data is from projects of multilat- few studies evaluate the benefits comprehensively. The eral development banks that routinely conduct a most robust scientific studies, such as randomized or financial assessment of WASH services before project matched prospective cohort studies, have been conducted approval and that, in some cases, report on the com- on health effects. But only few of those studies exist, and pletion of project implementation. Table 9.5 Cost Estimates of Improved Water and Sanitation Services US$ per cubic meter Cost component Full cost Minimal cost Opportunity cost of raw water supply 0.05 0.00 (“steal it”) Storage and transmission to treatment plant 0.10 0.07 (minimum storage) Treatment to drinking water standards 0.10 0.04 (simple chlorination) Distribution of water to households 0.60 0.24 (PVC pipe) Collection of wastewater from home and conveyance to treatment plant 0.80 0.30 (condominial sewers) Wastewater treatment 0.30 0.15 (simple lagoon) Damages associated with discharge of treated wastewater 0.05 0.00 (“someone else’s problem”) Total 2.00 0.80 Source: Whittington and others 2009. Note: PVC = polyvinyl chloride. Discount rate used is 6 percent. Using a 3 percent discount rate, the total cost is US$1.80 per cubic meter at full cost and US$0.70 per cubic meter at minimal cost. 188 Injury Prevention and Environmental Health Table 9.6 Benefits of Improved Drinking Water Supply and Sanitation Benefit Water Sanitation Health, burden of • Averted cases of diarrheal disease • Averted cases of diarrheal disease disease • Reduced malnutrition, enteropathy, and • Averted cases of helminths, polio, and eye diseases malnutrition-related conditions (stunting) • Reduced malnutrition, enteropathy, and malnutrition-related • Less dehydration from lack of access to water conditions (stunting) • Less disaster-related health impacts • Less dehydration from insufficient water intake because of poor latrine access • Less disaster-related health impacts Health, economic • Costs related to diseases, such as health care, • Costs related to diseases, such as health care, productivity savings productivity losses, and premature mortality losses, and premature mortality Convenience • Saved travel and waiting time for water • Saved travel and waiting time from having nearby private toilet time savings collection Educational • Improved educational levels because of higher • Improved educational levels because of higher school enrollment benefits school enrollment and attendance rates from and attendance rates from school sanitation school water • Higher attendance and educational attainment because of • Higher attendance and educational attainment improved health because of improved health Social benefits • Leisure and nonuse values of water resources • Safety, privacy, dignity, comfort, status, prestige, aesthetics, and and reduced effort of averted water hauling and gender effects gender impacts Water access • Pretreated water at lower costs for averted • Less pollution of water supply and hence reduced water benefits treatment costs for households treatment costs Reuse • Soil conditioner and fertilizer • Energy production • Safe use of wastewater Economic effects • Incomes from more tourism and business • Incomes from more tourism and business investment investment • Employment opportunity in sanitation supply chain • Employment opportunity in water provision • Rise in value of property • Rise in value of property Sources: Adapted from Hutton 2012; Hutton and others 2014. Intervention Efficiency: Cost-Benefit Analysis and best-available evidence from multiple sources The discussion of efficiency should distinguish between (Hutton 2013; Whittington and others 2009) cost-benefit analysis, which uses a common money met- ric for all costs and benefits, and cost-effectiveness Given the high costs and challenges associated with analysis, which compares interventions for one type of collecting all the cost and benefit data required for the outcome. Reviewed cost-benefit studies are provided in first approach, it is common practice to combine annex 9C. site-specific values with data extrapolated from other Efficiency studies can be conducted in two ways sources (Hutton and others 2014). Table 9.7 shows the (Whittington and others 2009): most recently available global studies that have modeled selected water supply and sanitation interventions. One • By generating estimates of cost and benefit in specific important finding from these studies is that lower tech- sites or field studies for the purposes of either evalu- nology interventions have higher returns than more ating intervention performance or selecting a site for expensive networked options. a future project (Kremer and others 2011) Global studies indicate the projected overall costs • By using model costs and benefits for specific sites or and benefits from intervention alternatives, but they larger jurisdictions, such as country or global level, are not particularly useful in guiding decisions on Water Supply, Sanitation, and Hygiene 189 Table 9.7 Benefit-Cost Ratios from Global Studies they serve large, dense populations. Providing water service on a smaller scale through either communal or Study and intervention Benefit-cost ratio in-compound wells or boreholes and onsite household Whittington and others (2009): modeled approach a sanitation may be a more appropriate and cost-efficient Networked water and sewerage services 0.65 service level for sparsely populated areas (Ferro, Lentini, and Mercadier 2011). Deep borehole with public hand pump 4.64 Total sanitation campaign (South Asia) 3.00 Household water treatment (biosand filters) 2.48 Intervention Efficiency: Cost-Effectiveness Analysis The main outcomes used in cost-effectiveness studies Hutton (2013): modeled approach b are health and environmental outcomes. When used Improved water supply (JMP definition) 2.00 to compare programs in a sector, cost-effectiveness Improved sanitation (JMP definition) 5.50 can be measured by program outcomes, such as the Sources: Hutton 2013; Whittington and others 2009. number of latrines constructed, the number of water Note: All studies include the value associated with health and convenience time connections installed, or the percentage of beneficia- savings. a. Ranges on each parameter value are then used to conduct Monte Carlo simulation ries changing behavior. For water supply interventions, that enables exploration of intervention performance in a range of different settings. health cost-effectiveness studies have been conducted Hence, even interventions with a benefit-cost ratio of 2.0 or more are expected to have (see annex 9C). Studies focus on improved water a benefit-cost ratio of less than 1.0 under some runs of the model. b. Estimates indicate global averages, and regional averages are available in the paper. supply according to the JMP definition and A separate working paper provides results for each country (Hutton 2012). point-of-use treatment by households or schools. A global study compares water supply interventions at the regional level (Clasen and others 2007). Figure 9.6 shows the cost per healthy life-year which technology and service level to choose in spe- (HLY) gained for four interventions in two regions. It cific settings. One randomized implementation study shows that the selected interventions vary by a factor in India finds similar health costs between study arms. of approximately 2.5 between the most cost-effective However, it finds statistically significant savings in (chlorination) and the least cost-effective (ceramic time in the intervention group of US$7 per household filter). However, all interventions have a cost per HLY (US$5 for water and US$2 for sanitation) during the that is below the GDP of countries in these regions, dry season, or roughly 5 percent of monthly cash thereby indicating a cost-effective use of health expenditures (Pattanayak and others 2010). A study resources. Another global study found the incremental from South Africa estimates a benefit-cost ratio of costs averted of adding point-of-use water disinfection 3.1 for small-scale water schemes (Cameron and others on top of improved water supply costs resulted in a 2011). A study from Indonesia compared three waste- cost per DALY averted of less than US$25 in water treatment interventions and finds limited eco- Sub-Saharan Africa, of US$63 in India and Bangladesh, nomic rationale for the interventions (Prihandrijanti, and of less than US$210 in South-East Asia and Malisie, and Otterpohl 2008). However, a broader the Western Pacific (Haller, Hutton, and Bartram cost-benefit study at the river basin level estimated the 2007). benefits of cleaning up the Upper Citarum river in Fewer studies have conducted health cost-effectiveness Indonesia exceeded costs by 2.3 times (Hutton and analyses of sanitation and hygiene interventions. Two others 2013). global studies by the WHO and World Bank examine Targeting the poor could be justified; children from the cost-effectiveness of water supply and sanitation poorer households are at increased health risk because combined (Günther and Fink 2011; Haller, Hutton, and they live in communities with lower access to improved Bartram 2007). Using regions defined by epidemiologi- water and sanitation facilities. A study in Bangladesh, cal strata, WHO estimates that the cost in countries India, and Pakistan estimating the cost per episode for with high child and high adult mortality is less than income quintiles shows that costs of an illness represent US$530 per DALY averted in the Eastern Mediterranean a higher proportion of income for lower quintiles and Middle East, US$650 in Sub-Saharan Africa, (Rheingans and others 2012). US$1,400 in South and South-East Asia, and US$2,800 The cost efficiency of technologies depends on the in Latin America and the Caribbean. A World Bank local geological setting, population density, and num- study on child mortality reduction estimates the average ber of households to be served. Large water distribu- cost per life year saved in Sub-Saharan African countries tion and sewerage systems may only be cost efficient if is US$1,104 for basic improved water and sanitation 190 Injury Prevention and Environmental Health Figure 9.6 Cost Per HLY Gained from Four Water Supply and Water Quality Interventions in Two World Subregions, US$, 2005 a. Africa epidemiological stratum D (AFR-D) b. South Asia and South-East Asia epidemiological stratum D (SEAR-D) 350 350 300 300 Cost per HLY gained, US$ Cost per HLY gained, US$ 250 250 200 200 150 150 100 100 50 50 0 0 Chlorination Solar Ceramic water Source-based Chlorination Solar Ceramic water Source-based disinfection filter interventions disinfection filter interventions Source: Clasen and others 2007. Note: AFR-D = African Region–high child, high adult mortality countries; HLY = healthy life-year; SEAR-D = South-East Asian Region–high child, high adult mortality countries. AFR-D and SEAR-D are part of the World Health Organization’s epidemiological subregions. and is US$995 for privately piped water and flush toilets and mental development. Overall, the health impacts of (Günther and Fink 2011). poor WASH lead to economic consequences of sev- In country studies in South-East Asia, the cost per eral percent of GDP and continue to significantly affect DALY averted of basic sanitation is less than US$1,100 in quality of life and the environment. Furthermore, water selected rural areas of Cambodia, China, Indonesia, the stress is a growing phenomenon that will affect at least Lao People’s Democratic Republic, and Vietnam; the 2.8 billion people in 48 countries by 2025. Climatic fac- exception is in the Philippines, where it is US$2,500 tors are harder to control, but water scarcity can be mit- (Hutton and others 2014). Few recent country-specific igated by changing water use patterns and reducing studies are available on hygiene interventions; one study pollution of surface waters. from Burkina Faso estimates a cost of US$51 per death Important progress has been made in achieving the averted for health education to mothers (Borghi and MDG global water and sanitation targets. In September others 2002). 2015, new global targets for universal access to safe Sustainability of water supply, sanitation, and hygiene WASH were adopted. At the current rates of progress is covered in annex 9D; financing is covered in annex 9E. and using current indicators, achieving those targets will take at least 20 years for water supply and 60 years for sanitation (WHO and UNICEF 2014). Covering the CONCLUSIONS poor and marginalized populations will continue to be a Although global deaths from diarrhea have declined challenge; the remaining unserved populations are likely significantly over the past 20 years, poor water supply, to be harder to reach as universal access is approached. sanitation, and hygiene are still responsible for a signifi- The service level benchmark of targeting safely managed cant disease burden. An estimated 842,000 global deaths services will require better policy and regulatory frame- in 2012 were due to diarrhea caused by poor WASH. works and more resources. Indeed, as environmental Other less well-quantified but important long-term consequences intensify and populations demand a health consequences of poor WASH, such as helminths higher quality of service, a higher target for service level and enteric dysfunction, remain. Those diseases affect will be increasingly required. This demand will raise children’s nutritional status, thereby inhibiting growth questions about priorities; countries will face a trade-off Water Supply, Sanitation, and Hygiene 191 between (1) dedicating policy space and spending public including the measurement of cost-effectiveness to subsidies to move populations that are already served guide policy and program design. higher up the water and sanitation ladder and (2) reach- • The social welfare consequences of poor WASH are ing populations that are not served with basic WASH not well documented but are potentially very large. services. Each country will have its unique set of chal- In particular, a greater understanding is needed lenges. The human right to drinking water and sanita- of the gender effects of inadequate WASH and of tion can serve as a reminder that priority should be given how improved WASH services contribute to gender to ensuring at least a minimum level of affordable equality. WASH service for all citizens. • A large part of the remaining challenge of improving Populations are growing and moving, economies are access to sanitation and hygiene is behavioral rather developing and becoming richer, and the climate is than technical. However, little evidence exists on the changing. Each one has its challenges and opportunities. effectiveness of behavior change using conventional Population migration to greenfield sites offers a chance methods at scale or on the transferability of behavior of implementing new and appropriate technologies, and change interventions that are successful in a particu- selection of cost-effective and affordable technologies in lar context. A better understanding of habit forma- urban planning is essential. Economic growth leads to tion and what leads to sustainable behavior change greater tax revenues for local governments and increased is needed. ability to upgrade infrastructure and expand urban • Innovative delivery platforms that leverage national renewal. Climate change challenges the delivery of poverty reduction programs, such as CCT and CDD WASH services by affecting rainfall patterns, freshwater programs, have the potential to achieve wide cover- availability, and frequency of heat events, and it exacer- age at little marginal cost. Such approaches can also bates health risks. However, this new threat, when taken provide the methodology and data sources to support seriously, can be an opportunity to overhaul outdated targeting areas of poverty in WASH services. policies and technologies. Furthermore, as nutrient • A better understanding is needed on which WASH sources for chemical fertilizer become scarcer, price interventions work in slum areas and low-income increases will force suppliers to seek alternatives; the neighborhoods and under what conditions the inter- price of composted sludge is expected to increase, ventions work. thereby attracting investments. New research, data, and • A greater understanding is needed of how output- technologies are increasingly available to present new based incentives can be used to improve WASH possibilities for addressing entrenched problems in the service delivery and to lead to greater sustainability WASH sector. of services. The following research priorities are recommended • Innovations in subsidies and microfinance are needed for immediate attention: to ensure that the poor gain access to improved sanitation. Despite greater availability and lower • To adequately address equity considerations in the cost of sanitation goods and services, some people SDG era, there is a need to understand where remain too poor to afford adequate water supply and poor people live and what their levels of access sanitation. Such populations should be identified to are. Disaggregated data on the underserved— receive hardware and financial subsidies. including slum populations, ethnic groups, women, elderly, and persons with disabilities—can support prioritization. Greater focus is needed on how to ANNEXES increase access in the lagging regions of South Asia and Africa, where a large proportion of the unserved The annexes to this chapter are as follows. They are avail- live. At the country level, policy and financial incen- able at http://www.dcp-3.org/environment. tives need to be aligned and the economic arguments made for allocating resources to WASH services. • Annex 9A. Overview of Studies Presenting Damage • More evidence is needed to support the emerging Costs of Poor Water, Sanitation, and Hygiene at the understanding of the wider health effects of water, National Level sanitation, and hygiene. Multisectoral approaches • Annex 9B. Effectiveness of Enabling Environments will become more important as the complementa- • Annex 9C. Cost-Effectiveness and Cost-Benefit Studies rities among WASH, health, and nutrition are better on Water, Sanitation, and Hygiene understood. Further, rigorously designed and con- • Annex 9D. Intervention Sustainability trolled studies are needed to quantify these benefits, • Annex 9E. Intervention Financing 192 Injury Prevention and Environmental Health NOTES (same as JMP improved definition). Basic hygiene: per- centage of population with handwashing facilities with World Bank Income Classifications as of July 2014 are as soap and water at home. Safely managed water: percent- follows, based on estimates of gross national income (GNI) age of population using safely managed drinking water per capita for 2013: services. Corresponds to population using an improved drinking water source located on the premises, available • Low-income countries (LICs) = US$1,045 or less when needed, and free of fecal and priority chemical • Middle-income countries (MICs) are subdivided: contamination. Safely managed sanitation: percentage a) lower-middle-income = US$1,046 to US$4,125 of population using safely managed sanitation services. b) upper-middle-income (UMICs) = US$4,126 to US$12,745 Includes safe onsite isolation, extraction, conveyance, • High-income countries (HICs) = US$12,746 or more. treatment and disposal, or reuse. 14. For information about the OECD’s methodology and 1. United Nations Human Rights Council, Resolution FEASIBLE computer model, see http://www.oecd.org/env 18/1, “The Human Right to Safe Drinking Water and /outreach/methodologyandfeasiblecomputermodel.htm Sanitation,” adopted September 28, 2011, http://www (accessed November 11, 2015). .worldwatercouncil.org/fileadmin/wwc/Right_to_Water 15. For example, the IRC International Water and Sanitation /Human_Rights_Council_Resolution_cotobre_2011.pdf. Center has developed the WASHCost Calculator (www 2. United Nations Human Rights Council, Resolution 18/1. .ircwash.org/washcost), whereas the World Bank’s 3. Whereas no academic literature is available with such Economics of Sanitation Initiative has developed an examples, national surveys (such as the Demographic economic assessment toolkit under the Economics of and Health Survey or the Multiple Indicator Cluster Sanitation Initiative (http://www.wsp.org/esi). Survey) show that a higher proportion of households practice unsafe management of children’s feces as com- pared with overall household unimproved sanitation REFERENCES practices. 4. JMP reports for country data and additional breakdowns Adukia, A. 2014. “Sanitation and Education.” PhD thesis, are available at http://www.wssinfo.org. Harvard University, Cambridge, MA. 5. Characterized by villous atrophy, crypt hyperplasia, Ahern, M., R. Kovats, P. Wilkinson, R. Few, and F. Matthies. increased permeability, inflammatory cell infiltrate, and 2005. “Global Health Impacts of Floods: Epidemiologic modest malabsorption. Evidence.” Epidemiologic Reviews 27 (1): 36–46. 6. The hidden flow of water if food or other commodities that Ali, M., A. Nelson, A. Lopez, D. Sack. 2015. “Updated require water to be produced are traded from one place to Global Burden of Cholera in Endemic Countries.” PLoS another. Neglected Tropical Diseases 9 (6). doi:10.1371/journal 7. An odds ratio (OR) is a measure of association between .pntd.0003832. an exposure and an outcome. The OR represents the odds Andres, L., B. Briceno, C. Chase, and J. A. Echenique. 2014. that an outcome will occur given a particular exposure, “Sanitation and Externalities: Evidence from Early compared with the odds of the outcome occurring in the Childhood Health in Rural India.” Policy Research Working absence of that exposure. Paper 6737, World Bank, Washington, DC. 8. For more on the Cochrane Library, see http://www Atia, A., and A. Buchman. 2009. “Oral Rehydration Solutions .thecochranelibrary.com. in Non-Cholera Diarrhea: A Review.” American Journal of 9. For more information on CLTS, see http://www Gastroenterology 104 (10): 2596–604. .communityledtotalsanitation.org/page/clts-approach and Baker, J. L., ed. 2009. Opportunities and Challenges for Small http://cltsfoundation.org/clts-map.html. Scale Private Service Providers in Electricity and Water 10. The impact evaluation in Senegal was compromised Supply: Evidence from Bangladesh, Cambodia, Kenya and the because of contamination of the treatment group with the Philippines. Washington, DC: World Bank and the Public- handwashing with soap intervention group. Private Infrastructure Advisory Facility. 11. Examples of results-based approaches include the following: Benova, L., O. Cumming, and O. Campbell. 2014. “Systematic output-based aid (OBA), results-based financing (RBF), Review and Meta-Analysis: Association between Water and pay-for-performance (P4P), program for results (PforR), Sanitation Environment and Maternal Mortality.” Tropical and conditional-cash transfer (CCT). Medicine and International Health 19 (4): 368–87. 12. Accessed March 31, 2014, through the OBA website, https:// Biran, A., M. Jenkins, P. Dabrase, and I. Bhagwat. 2011. www.gpoba.org. “Patterns and Determinants of Communal Latrine Usage in 13. Basic water: percentage of population using a protected Urban Poverty Pockets in Bhopal, India.” Tropical Medicine community source or piped water with a total collection and International Health 16 (7): 854–62. time of 30 minutes or less for a round-trip, including Biran, A., T. Rabie, W. Schmidt, S. Juvekar, S. Hirve, and others. queuing (same as JMP improved definition except time 2008. “Comparing the Performance of Indicators of Hand- criteria has been introduced). Basic sanitation: percent- Washing Practices in Rural Indian Households.” Tropical age of population using a basic private sanitation facility Medicine and International Health 13 (2): 278–85. Water Supply, Sanitation, and Hygiene 193 Black, R., L. Allen, Z. Bhutta, L. Caulfield, M. de Onis, Clasen, T., K. Alexander, D. Sinclair, S. Boisson, R. Peletz, and and others. 2008. “Maternal and Child Undernutrition: others. 2005. “Interventions To Improve Water Quality for Global and Regional Exposures and Health Consequences.” Preventing Diarrhoea (Review). The Cochrane Library 2005 The Lancet 371 (9608): 243–60. (10). Black, R., C. Victora, S. Walker, Z. Bhutta, P. Christian, and Clasen, T., L. Haller, D. Walker, J. Bartram, and S. Cairncross. others. 2013. “Maternal and Child Undernutrition and 2007. “Cost-Effectiveness of Water Quality Interventions for Overweight in Low-Income and Middle-Income Countries.” Preventing Diarrhoeal Disease in Developing Countries.” The Lancet 382 (9890): 427–51. Journal of Water and Health 5 (4): 599–608. Blackett, I., P. Hawkins, and C. Heymans. 2014. “The Missing Clasen, T., S. Boisson, P. Routray, B. Torondel, M. Bell, Link in Sanitation Service Delivery: A Review of Fecal and others. 2014. “Effectiveness of a Rural Sanitation Sludge Management in 12 Cities.” Water and Sanitation Programme on Diarrhoea, Soil-Transmitted Helminth Program Research Brief, World Bank, Washington, DC. Infection, and Child Malnutrition in Odisha, India: Blencowe, H., S. Cousens, L. Mullany, A. Lee, K. Kerber, and A Cluster-Randomised Trial.” The Lancet Global Health others. 2011. “Clean Birth and Postnatal Care Practices 2 (11): e645–53. to Reduce Neonatal Deaths from Sepsis and Tetanus: Corcoran, E., C. Nellemann, E. Baker, R. Bos, D. Osborn, and A Systematic Review and Delphi Estimation of Mortality others. 2011. “Sick Water? The Central Role of Wastewater Effect.” BMC Public Health 11 (Suppl 3): S11. Management in Sustainable Development.” A Rapid Borghi, J., L. Guinness, J. Ouedraogo, and V. Curtis. 2002. Response Assessment, UN HABITAT, GRID-Arendal, “Is Hygiene Promotion Cost-Effective? A Case Study in United Nations Environment Programme, Arendal, Norway. Burkina Faso.” Tropical Medicine and International Health Curtis, V., and S. Cairncross. 2003. “Effect of Washing Hands 7 (11): 960–69. with Soap on Diarrhoea Risk in the Community: A Briceno, B., A. Coville, and S. Martinez. 2014. “Promoting Systematic Review.” The Lancet Infectious Diseases 3 (5): Handwashing and Sanitation: A Crossover Randomized 275–81. Experiment in Rural Tanzania.” Working Paper, Water and Curtis, V., B. Kanki, S. Cousens, I. Diallo, A. Kpozehouen, and Sanitation Program, World Bank, Washington DC. others. 2001. “Evidence of Behaviour Change following a Brooker, S., A. Clements, and D. Bundy. 2006. “Global Hygiene Promotion Programme in Burkina Faso.” Bulletin Epidemiology, Ecology and Control of Soil-Transmitted of the World Health Organization 79 (6): 518–27. Helminth Infections.” Advances in Parasitology 62: Dangour, A., L. Watson, O. Cumming, S. Boisson, Y. Che, and 221–61. others. 2013. “Interventions to Improve Water Quality and Brown, J., S. Cairncross, and J. Ensink. 2013. “Water, Supply, Sanitation and Hygiene Practices, and Their Effects Sanitation, Hygiene and Enteric Infections in Children.” on the Nutritional Status of Children.” Cochrane Database Archives of Disease in Childhood 12. doi:10.1136 of Systematic Reviews 8: CD009382. doi:10.1002/14651858 /archdischild-2011-301528. .CD009382.pub2. Burr, P., and C. Fonseca. 2013. “Applying a Life-Cycle Costs Devoto, F., E. Duflo, P. Dupas, W. Pariente, and V. Pons. 2011. Approach to Water: Costs and Service Levels in Rural and “Happiness on Tap: Piped Water Adoption in Urban Small Town Areas in Andhra Pradesh (India), Burkina Morocco.” National Bureau of Economic Research (NBER) Faso, Ghana and Mozambique.” Working Paper 8, IRC Working Paper No. 16933, NBER, Cambridge, MA. International Water and Sanitation Centre, The Hague, the Dodane, P.-H., M. Mbéguéré, O. Sow, and L. Strande. 2012. Netherlands. “Capital and Operating Costs of Full-Scale Fecal Sludge Cairncross, S. 2004. “The Case for Marketing Sanitation.” Management and Wastewater Treatment Systems in Dakar, Field note, Water and Sanitation Program, World Bank, Senegal.” Environmental Science and Technology 46 (7): Washington, DC. 3705−11. Cameron, J., P. Jagals, P. Hunter, S. Pedley, and K. Pond. 2011. Dutto, S. M., M. Lopez Abbate, F. Biancalana, A. Berasategui, “Economic Assessments of Small-Scale Drinking-Water and M. Hoffmeyer. 2012. “The Impact of Sewage on Interventions in Pursuit of MDG Target 7C.” Science of the Environmental Quality and the Mesozooplankton Total Environment 410–411: 8–15. Community in a Highly Eutrophic Estuary in Argentina.” Cameron, L., M. Shah, and S. Olivia. 2013. “Impact Evaluation ICES Journal of Marine Science 69 (3): 399–409. of a Large-Scale Rural Sanitation Project in Indonesia.” Esrey, S. A. 1996. “Water, Waste, and Well-Being: A Multicountry Policy Research Working Paper 6360, World Bank, Study.” American Journal of Epidemiology 143 (6): 608–23. Washington, DC. Esrey, S. A., J. B. Potash, L. Roberts, and C. Shiff. 1991. “Effects Chase, C., and Q.-T. Do. 2012. “Handwashing Behavior of Improved Water Supply and Sanitation on Ascariasis, Change at Scale: Evidence from a Randomized Evaluation Diarrhoea, Dracunculiasis, Hookworm Infection, in Vietnam.” Policy Research Working Paper 6207, World Schistosomiasis, and Trachoma.” Bulletin of the World Bank, Washington, DC. Health Organization 69 (5): 609–21. Checkley, W., G. Buckley, R. Gilman, A. Assis, R. Guerrant, Ferro, G., E. Lentini, and A. Mercadier. 2011. “Economies and others. 2008. “Multi-Country Analysis of the Effects of of Scale in the Water Sector: A Survey of the Empirical Diarrhoea on Childhood Stunting.” International Journal of Literature.” Journal of Water, Sanitation and Hygiene for Epidemiology 37 (4): 816–30. Development 1 (3): 179–93. 194 Injury Prevention and Environmental Health Fisher, J. 2006. “For Her It’s the Big Issue: Putting Women at the Humphrey, J. 2009. “Child Undernutrition, Tropical Centre of Water Supply, Sanitation and Hygiene.” Evidence Enteropathy, Toilets, and Handwashing.” The Lancet 374 Report, Water Supply and Sanitation Collaborative Council, (9694): 1032–35. Geneva. Hunter, P. R., D. Zmirou-Navier, and P. Hartemann. 2009. Fishman, S., L. Caulfield, M. de Onis, M. Blössner, A. Hyder, “Estimating the Impact on Health of Poor Reliability of and others. 2004. “Childhood and Maternal Underweight.” Drinking Water Interventions in Developing Countries.” In volume 1 of Comparative Quantification of Health Risks: Science of the Total Environment 407 (8): 2621–24. Global and Regional Burden of Disease due to Selected Major Hutton, G. 2012. “Global Costs and Benefits of Drinking- Risk Factors, edited by M. Ezzati, A. D. Lopez, A. Rodgers, Water Supply and Sanitation Interventions to Reach the and C. J. L. Murray. Geneva: WHO. MDG Target and Universal Coverage.” Report, WHO/HSE/ Freeman, M., M. Stocks, O. Cumming, A. Jeandron, J. Higgins, WSH/12.01, World Health Organization, Geneva. and others. 2014. “Hygiene and Health: Systematic Review ———. 2013. “Global Costs and Benefits of Reaching of Handwashing Practices Worldwide and Update of Health Universal Coverage of Sanitation and Drinking-Water Effects.” Tropical Medicine and International Health 19 (8): Supply.” Journal of Water and Health 11 (1): 1–12. 906–16. Hutton, G., S. Kerstens, A. van Nes, and I. Firmansyah. Galiani, S., P. Gertler, N. Ajzenman, and A. Orsola-Vidal. 2015. 2013. “Downstream Impacts of Water Pollution in the “Promoting Handwashing Behavior: The Effects of Large- Upper Citarum River, West Java, Indonesia: Economic scale Community and School-level Interventions.” Health Assessment of Interventions to Improve Water Quality.” Economics (October 12). doi:10.1002/hec.3273. Water and Sanitation Program Technical Paper, World Guerrant, R., M. DeBoer, S. Moore, R. Scharf, and A. Lima. Bank, Washington, DC. 2013. “The Impoverished Gut—A Triple Burden of Hutton, G., U. Rodriguez, L. Napitupulu, P. Thang, and P. Kov. Diarrhoea, Stunting and Chronic Disease.” Nature Reviews 2008. Economic Impacts of Sanitation in Southeast Asia. Gastroenterology and Hepatology 10 (4): 220–29. Water and Sanitation Program. Washington, DC: World Guerrant, R., R. Oriá, S. Moore, M. Oriá, and A. Lima. 2008. Bank. “Malnutrition as an Enteric Infectious Disease with Long- Hutton, G., U.-P. Rodriguez, A. Winara, V. Nguyen, P. Kov, Term Effects on Child Development.” Nutrition Reviews and others. 2014. “Economic Efficiency of Sanitation 66 (9): 487–505. Interventions in Southeast Asia.” Journal of Water, Sanitation Guh, S., C. Xingbao, C. Poulos, Z. Qi, C. Jianwen, and others. and Hygiene in Development 4 (1): 23–36. 2008. “Comparison of Cost-of-Illness with Willingness-to- Hutton, G., and M. Varughese. 2016. The Costs of Meeting the Pay Estimates to Avoid Shigellosis: Evidence from China.” 2030 Sustainable Development Goal Targets on Drinking Health Policy and Planning 23 (2): 125–36. Water, Sanitation, and Hygiene. Water and Sanitation Günther, I., and G. Fink. 2011. “Water and Sanitation to Program. Washington, DC: World Bank. Reduce Child Mortality: The Impact and Cost of Water and IBNET (International Benchmarking Network for Water and Sanitation Infrastructure.” Policy Research Working Paper Sanitation Utilities). 2014. Database. IBNET, World Bank 5618, World Bank, Washington, DC. Water and Sanitation Program, Washington, DC. https:// Hales, S., S. Kovats, S. Lloyd, and D. Campbell-Lendrum, eds. www.ib-net.org/. 2014. Quantitative Risk Assessment of the Effects of Climate Ilahi, N., and F. Grimard. 2000. “Public Infrastructure and Change on Selected Causes of Death, 2030s and 2050s. Private Costs: Water Supply and Time Allocation of Women Geneva: WHO. in Rural Pakistan.” Economic Development and Cultural Haller, L., G. Hutton, and J. Bartram. 2007. “Estimating Change 49 (1): 45–75. the Costs and Health Benefits of Water and Sanitation Jafari, N., A. Shahsanai, M. Memarzadeh, and A. Loghmani. Improvements at Global Level.” Journal of Water and Health 2011. “Prevention of Communicable Diseases after 5 (4): 467–80. Disaster: A Review.” Journal of Research in Medical Sciences Hammer, J., and D. Spears. 2013. “Village Sanitation and 16 (7): 956–62. Children’s Human Capital Evidence from a Randomized Jasper, C., T.-T. Le, and J. Bartram. 2012. “Water and Sanitation in Experiment by the Maharashtra Government.” Policy Schools: a Systematic Review of the Health and Educational Research Working Paper 5580, World Bank, Washington, Outcomes.” International Journal of Environmental Research DC. Public Health 9 (8): 2772–87. Hotez, P., D. Bundy, K. Beegle, S. Brooker, L. Drake, and Jenkins, M., and V. Curtis. 2005. “Achieving the ‘Good Life’: others. 2006. “Helminth Infections: Soil-Transmitted Why Some People Want Latrines in Rural Benin.” Social Helminth Infections and Schistosomiasis.” In Disease Science and Medicine 61 (11): 2446–59. Control Priorities in Developing Countries, second edition, Jenkins, M., and B. Scott. 2007. “Behavioral Indicators of edited by D. Jamison, J. Breman, A. Measham, G. Alleyne, Household Decision-Making and Demand for Sanitation M. Claeson, D. Evans, P. Jha, A. Mills, and P. Musgrove. and Potential Gains from Social Marketing in Ghana.” Washington, DC: Oxford University Press and World Bank. Social Science and Medicine 64 (12): 2427–42. Howard, G., and J. Bartram. 2003. “Domestic Water Quantity, Koolwal, G., and D. Van de Walle. 2013. “Access to Water, Service Level and Health.” Report, WHO/SDE/WSH/03.02, Women’s Work, and Child Outcomes.” Economic World Health Organization, Geneva. Development and Cultural Change 61 (2): 369–405. Water Supply, Sanitation, and Hygiene 195 Kotloff, K., J. Nataro, W. Blackwelder, D. Nasrin, T. Farag, Norman, G., S. Pedley, and B. Takkouche. 2010. “Effects of and others. 2013. “Burden and Aetiology of Diarrhoeal Sewerage on Diarrhoea and Enteric Infections: A Systematic Disease in Infants and Young Children in Developing Review and Meta-Analysis.” The Lancet Infectious Diseases Countries (The Global Enteric Multicenter Study, 10: 536–44. GEMS): A Prospective, Case-Control Study.” The Lancet Null, C., J. G. Hombrado, R. Meeks, M. Edward, and A. P. Zwane. 382 (9888): 209–22. 2012. “Willingness to Pay for Cleaner Water in Less Kremer, M., J. Leino, E. Miguel, and A. Zwane. 2011. “Spring Developed Countries: Systematic Review of Experimental Cleaning: A Randomized Evaluation of Source Water Evidence.” Systematic Review No. 6, International Initiative Quality Improvement.” Quarterly Journal of Economics 126 for Impact Evaluation, London. (1): 145–205. OECD (Organisation for Economic Co-operation and Lanata, C., C. Fischer-Walker, A. Olascoaga, C. Torres, Development). 2011. “Meeting the Challenge of Financing M. Aryee, and others. 2013. “Global Causes of Diarrheal Water and Sanitation: Tools and Approaches.” OECD, Paris. Disease Mortality in Children <5 Years of Age: A Systematic Olajuyigbe, A. 2010. “Some Factors Impacting on Quantity of Review.” PLoS One 8 (9): e72788. Water Used by Households in a Rapidly Urbanizing State Lim, S., T. Vos, A. Flaxman, G. Danaei, K. Shibuya, and Capital in South Western Nigeria.” Journal of Sustainable others. 2012. “A Comparative Risk Assessment of Burden of Development in Africa 12 (2): 321–37. Disease and Injury Attributable to 67 risk factors and Risk Onda, K., J. LoBuglio, and J. Bartram. 2012. “Global Access Factor Clusters in 21 Regions, 1990–2010: A Systematic to Safe Water: Accounting for Water Quality and the Analysis for the Global Burden of Disease Study 2010.” The Resulting Impact on MDG progress.” International Journal Lancet 380 (9859): 2224–60. of Environmental Research and Public Health 9 (3): 880–94. Lin, A., B. Arnold, S. Afreen, R. Goto, T. Mohammad Nurul Orgill, J., A. Shaheed, J. Brown, and M. Jeuland. 2013. “Water Huda, and others. 2013. “Household Environmental Quality Perceptions and Willingness to Pay for Clean Water Conditions Are Associated with Enteropathy and Impaired in Peri-Urban Cambodian Communities.” Journal of Water Growth in Rural Bangladesh.” American Journal of Tropical and Health 11 (3): 489–506. Medicine and Hygiene 89 (1): 130–37. Oster, E., and R. Thornton. 2011. “Menstruation, Sanitary Loevensohn, M., L. Mehta, K. Cuming, A. Nicol, O. Cumming, Products, and School Attendance: Evidence from a and others. 2015. “The Cost of a Knowledge Silo: A Randomized Evaluation.” American Economic Journal: Systematic Re-Review of Water, Sanitation and Hygiene Applied Economics 3 (1): 91–100. Interventions.” Health Policy and Planning 30 (5): 660–74. Özaltin, E., K. Hill, and S. V. Subramanian. 2010. “Association Lokshin, M., and R. Yemtsov. 2005. “Has Rural Infrastructure of Maternal Stature with Offspring Mortality, Underweight, Rehabilitation in Georgia Helped the Poor?” World Bank and Stunting in Low- to Middle-Income Countries.” Journal Economic Review 19 (2): 311–33. of the American Medical Association 303 (15): 1507–16. Lozano, R., M. Naghavi, K. Foreman, S. Lim, and K. Shibuya. Patil, S., B. Arnold, A. Salvatore, B. Briceno, S. Ganguly, and others. 2010. “Global and Regional Mortality from 235 Causes of 2014. “The Effect of India’s Total Sanitation Campaign on Death for 20 Age Groups in 1990 and 2010: A Systematic Defecation Behaviors and Child Health in Rural Madhya Analysis for the Global Burden of Disease Study 2010.” The Pradesh: A Cluster Randomized Controlled Trial.” PLoS Lancet 380 (9859): 2095–128. Medicine 11 (8). doi:10.1371/journal.pmed.1001709. Moriarty, P., C. Batchelor, C. Fonseca, A. Klutse, A. Naafs, Pattanayak, S., C. Poulos, J. Yang, and S. Patil. 2010. and others. 2010. “Ladders for Assessing and Costing “How Valuable Are Environmental Health Interventions? Water Service Delivery.” WASHCost Working Paper 2, IRC Evaluation of Water and Sanitation Programmes in International Water and Sanitation Centre, The Hague, the India.” Bulletin of the World Health Organization 88 (7): Netherlands. 535–42. Mumssen, Y., L. Johannes, and G. Kumar. 2010. Output-Based Peal, A., B. Evans, and C. van der Voorden. 2010. Hygiene and Aid: Lessons Learned and Best Practices. Washington, DC: Sanitation Software: An Overview of Approaches. Geneva: World Bank. Water Supply and Sanitation Collaborative Council. Murcott, S. 2012. Arsenic Contamination in the World: An Petri, W., M. Miller, H. Binder, M. Levine, R. Dillingham, International Sourcebook 2012. London: IWA Publishing. and others. 2008. “Enteric Infections, Diarrhea, and Their Naujokas, M. F., B. Anderson, H. Ahsan, H. V. Aposhian, Impact on Function and Development.” Journal of Clinical J. Graziano, and others. 2013. “The Broad Scope of Health Investigation 118 (4): 1277–90. Effects from Chronic Arsenic Exposure: Update on a Pickering, A. J., and J. Davis. 2012. “Freshwater Availability Worldwide Public Health Problem.” Environmental Health and Water Fetching Distance Affect Child Health in Sub- Perspectives 121 (3): 295–302. Saharan Africa.” Environmental Science and Technology Nicholson, J. A., M. Naeeni, M. Hoptroff, J. R. Matheson, 46 (4): 2391–97. A. J. Roberts, and others. 2013. “An Investigation of Pickering, A. J., H. Djebbari, C. Lopez, M. Coulibaly, and the Effects of a Hand Washing Intervention on Health M. L. Alzua. 2015. “Effect of a Community-Led Sanitation Outcomes and School Absence Using a Randomised Trial Intervention on Child Diarrhoea and Child Growth in in Indian Urban Communities.” Tropical Medicine and Rural Mali: A Cluster-Randomised Controlled Trial.” The International Health 19 (3): 284–92. Lancet Global Health 3 (5): E701–11. 196 Injury Prevention and Environmental Health Popkin, B., K. D’Anci, and I. Rosenberg. 2010. “Water, Salazar-Lindo, E., S. Allen, D. Brewster, E. Elliott, A. Fasano, Hydration and Health.” Nutrition Reviews 68 (8): 439–58. and others. 2004. “Intestinal Infections and Environmental Potter, A., A. Klutse, M. Snehalatha, C. Batchelor, A. Uandela, Enteropathy: Working Group Report of the Second World and others. 2010. “Assessing Sanitation Service Levels.” Congress of Pediatric Gastroenterology, Hepatology, WASHCost Working Paper 3, IRC International Water and and Nutrition.” Journal of Pediatric Gastroenterology and Sanitation Centre, the Hague, the Netherlands. Nutrition 39 (Suppl 2): S662–69. Prihandrijanti, M., A. Malisie, and R. Otterpohl. 2008. Schlosser, C. A., K. Strzepek, X. Gao, A. Gueneau, C. Fant, “Cost-Benefit Analysis for Centralized and Decentralized and others. 2014. “The Future of Global Water Stress: An Wastewater Treatment System (Case Study in Surabaya, Integrated Assessment.” Report No. 254, Massachusetts Indonesia).” In Efficient Management of Wastewater: Its Institute of Technology Joint Program on the Science and Treatment and Reuse in Water-Scarce Countries, edited by Policy of Global Change, Cambridge, MA. I. Al Baz, R. Otterpohl, and C. Wendland, 259–68. Berlin- Schmidt, W.-P., and S. Cairncross. 2009. “Household Water Heidelberg: Springer. Treatment in Poor Populations: Is There Enough Evidence Prüss-Ustün, A., J. Bartram, T. Clasen, J. Colford, O. Cumming, for Scaling Up Now?” Environmental Science and Technology and others. 2014. “Burden of Diarrheal Disease from 43 (4): 986–92. Inadequate Water, Sanitation and Hygiene in Low- and Scott, C., and T. Shah. 2004. “Groundwater Overdraft Middle-Income Countries: A Retrospective Analysis of Data Reduction through Agricultural Energy Policy: Insights from 145 Countries.” Tropical Medicine and International from India and Mexico.” International Journal of Water Health 19 (8): 894–905. Resources Development 20 (2): 149–64. Prüss-Ustün, A., and C. Corvalan. 2006. Preventing Disease Sebastian, A., V. Hoffmann, and S. Adelman. 2013. “Menstrual through Healthy Environments: Towards an Estimate of the Management in Low-Income Countries: Needs and Trends.” Environmental Burden of Disease. Geneva: World Health Waterlines 32 (2): 135–53. Organization. Semmelweis, I. 1983. The Etiology, Concept and Prophylaxis ———. 2007. “How Much Disease Burden Can Be Prevented of Childbed Fever. Madison, WI: University of Wisconsin by Environmental Interventions?” Epidemiology 18 (1): Press. 167–78. Shah, N. B. 2013. “Microfinance Loans to Increase Sanitary Pullan, R., J. Smith, R. Jasrasaria, and S. Brooker. 2014. “Global Latrine Sales: Evidence from a Randomized Trial in Rural Numbers of Infection and Disease Burden of Soil Transmitted Cambodia.” Policy Brief. International Development Helminth Infections in 2010.” Parasites and Vectors Enterprises, Phnom Penh. 7 (37). doi:10.1186/1756-3305-7-37. Smets, H. 2014. “Quantifying the Affordability Standard.” Rabalais, N. N., and R. E. Turner, eds. 2013. Coastal Hypoxia: In The Human Right to Water: Theory, Practice, and Consequences for Living Resources and Ecosystems. Coastal Prospects, edited by M. Langford and A. Russell. Cambridge, and Estuarine Studies 58. Washington, DC: American U.K.: Cambridge University Press. Geophysical Union. Spears, D. 2013. “How Much International Variation in Child Rabie, T., and V. Curtis. 2006. “Handwashing and Risk Height Can Sanitation Explain?” Policy Research Working of Respiratory Infections: A Quantitative Systematic Paper 6351, World Bank, Washington, DC. Review.” Tropical Medicine and International Health Spears, D., and S. Lamba. 2013. “Effects of Early-Life 11 (3): 258–67. Exposure to Sanitation on Childhood Cognitive Skills: Ram, P. 2013. “Practical Guidance for Measuring Handwashing Evidence from India’s Total Sanitation Campaign.” Behavior.” Water and Sanitation Program Technical Paper, Policy Research Working Paper 6659, World Bank, World Bank, Washington, DC. Washington, DC. Rhee, V., L. Mullany, S. Khatry, J. Katz, S. LeClerq, and others. Stanton, B., and J. Clemens. 1987. “Twenty-Four Hour 2008. “Maternal and Birth Attendant Hand Washing and Recall, Knowledge-Attitude-Practice Questionnaires, and Neonatal Mortality in Southern Nepal.” Archives of Pediatric Direct Observations of Sanitary Practices: A Comparative and Adolescent Medicine 162 (7): 603–08. Study.” Bulletin of the World Health Organization 65 (2): Rheingans, R., M. Kukla, A. Faruque, D. Sur, A. Zaidi, and 217–22. others. 2012. “Determinants of Household Costs Associated Stocks, M., S. Ogden, D. Haddad, D. Addiss, C. McGuire, and with Childhood Diarrhea in Three South Asian Settings.” others. 2014. “Effect of Water, Sanitation, and Hygiene on Clinical Infectious Diseases 55 (Suppl 4): S327–35. the Prevention of Trachoma: A Systematic Review and Roaf, V., A. Khalfan, and M. Langford. 2005. “Monitoring Meta-Analysis.” PLoS Medicine 11 (2). doi:10.1371/journal Implementation of the Right to Water: A Framework for .pmed.1001605. Developing Indicators.” Global Issue Paper No. 14, Heinrich Strickland, G. 2000. Hunter’s Tropical Medicine and Emerging Böll Foundation, Berlin. Infectious Diseases. 8th ed. Philadelphia, PA: W. B. Saunders Sadoff, C. W., J. W. Hall, D. Grey, J. C. J. H. Aerts, M. Ait-Kadi, Company. and others. 2015. Securing Water, Sustaining Growth: Sumpter, C., and B. Torondel. 2013. “A Systematic Review Report of the GWP/OECD Task Force on Water Security of the Health and Social Effects of Menstrual Hygiene and Sustainable Growth. Oxford, U.K.: University of Management.” PLoS One 8 (4). doi:10.1371/journal Oxford. .pone.0062004. Water Supply, Sanitation, and Hygiene 197 Sy, J., R. Warner, and J. Jamieson. 2014. Tapping the Markets: Whittington, D., S. Pattanayak, J. Yang and B. Kumar. 2002. Opportunities for Domestic Investments in Water and “Household Demand for Improved Piped Water Services in Sanitation for the Poor. Washington, DC: World Bank. Kathmandu, Nepal.” Water Policy 4 (6): 531–56. Taylor, D. L., T. M. Kahawita, S. Cairncross, and J. H. J. Ensink. WHO (World Health Organization) and UNICEF (United 2015. “The Impact of Water, Sanitation and Hygiene Nations Children’s Fund). 2006. Meeting the MDG Drinking Interventions to Control Cholera: A Systematic Review.” Water and Sanitation Target: The Urban and Rural Challenge PLoS One 10 (8). doi:10.1371/journal.pone.0135676. of the Decade. Geneva: WHO. Turner, R., and N. Rabalais. 1991. “Changes in Mississippi River ———. 2010. Progress on Drinking Water and Sanitation: 2010 Water Quality This Century and Implications for Coastal Update. Geneva: WHO. Food Webs.” BioScience 41 (3): 140–47. ———. 2012a. Rapid Assessment of Drinking-Water Quality UNICEF (United Nations Children’s Fund). 1990. Strategy (RADWQ): A Handbook for Implementation. Geneva: WHO for Improved Nutrition of Children and Women in ———. 2012b. Progress on Drinking Water and Sanitation: Developing Countries: A UNICEF Policy Review. New 2012 Update. New York: UNICEF and WHO. York: UNICEF. ———. 2013. Progress on Drinking Water and Sanitation: 2013 ———. 2014. WASH Bottleneck Analysis Tool (WASH-BAT). Update. Geneva: WHO and UNICEF. New York: UNICEF. ———. 2014. Progress on Drinking Water and Sanitation: 2014 Van den Berg, C., and A. Danilenko. 2010. The IBNET Update. Geneva: WHO and UNICEF. Water Supply and Sanitation Performance Blue Book: ———. 2015a. WASH Post-2015: Proposed Indicators for The International Benchmarking Network for Water and Drinking Water, Sanitation and Hygiene. Geneva: WHO Sanitation Utilities Databook. Water and Sanitation and UNICEF. Program. Washington, DC: World Bank. ———. 2015b. Progress on Drinking Water and Sanitation: Victora, C., L. Adair, C. Fall, P. Hallal, R. Martorell, and others. 2015 Update and MDG Assessment. Geneva: WHO and 2008. “Maternal and Child Undernutrition: Consequences UNICEF. for Adult Health and Human Capital.” The Lancet 371 Wolf, J., A. Prüss-Üstun, O. Cumming, J. Bartram, S. Bonjour, (9609): 340–57. and others. 2014. “Assessing the Impact of Drinking-Water Waddington, H., B. Snilstveit, H. White, and L. Fewtrell. 2009. and Sanitation on Diarrhoeal Disease in Low- and Middle- “Water, Sanitation and Hygiene Interventions to Combat Income Countries: A Systematic Review and Regression Childhood Diarrhoea in Developing Countries.” Synthetic Analysis.” Tropical Medicine and International Health 8 (19). Review 001, International Initiative for Impact Evaluation, World Bank. 2009. The Philippines: Country Environmental New Delhi, India. Analysis. Washington, DC: World Bank. Whittington, D., Hanemann, W. M., Sadoff, C., and M. Jeuland. Ziegelbauer, K., B. Speich, D. Mäusezahl, R. Bos, J. Keiser, and 2009. “The Challenge of Improving Water and Sanitation others. 2012. “Effect of Sanitation on Soil-Transmitted Services in Less Developed Countries.” Foundations and Helminth Infection: Systematic Review and Meta-Analysis.” Trends in Microeconomics 4 (6): 469–607. PLoS Medicine 9 (1). 198 Injury Prevention and Environmental Health Chapter 10 Interventions to Prevent Injuries and Reduce Environmental and Occupational Hazards: A Review of Economic Evaluations from Low- and Middle-Income Countries David A. Watkins, Nazila Dabestani, Rachel Nugent, and Carol Levin INTRODUCTION The regulation of air pollution usually occurs within the Collectively, unintentional injuries and interpersonal purview of the public sector environmental agency, as violence accounted for at least 8 percent of deaths and does the provision of clean water and basic sanitation 9 percent of disability-adjusted life years (DALYs) in services that are then implemented by public works low- and middle-income countries (LMICs) in 2012 agencies (Pattanayak and others 2010). Reducing the (WHO 2016). Diseases related to air pollution and inad- health risks associated with these environmental haz- equate water and sanitation measures accounted for ards involves partnerships between ministries of health 15 percent of attributable deaths and 10 percent of and ministries responsible for environment, transporta- attributable DALYs in LMICs in 2013 (IHME 2015). tion, and public works. As another example, reduction Millennium Development Goal 7 has inspired steady of occupational hazards is considered the responsibility progress on water- and sanitation-related indicators, of employers and employees alike, but it is often moni- although many countries have not yet reached target tored and regulated by ministries responsible for labor. levels of coverage (Luh and Bartram 2016) or the newer This chapter summarizes the evidence of the costs targets for Sustainable Development Goal 6 (UN 2016). and benefits of interventions to prevent injuries and Health losses from road traffic injuries (RTIs), interper- reduce occupational and environmental risks in LMICs. sonal violence, and outdoor air pollution continue to Although the interventions reviewed reflect a set of con- rise (WHO 2016). ditions and risk factors more narrow than those covered A common feature of the seemingly disparate condi- in this volume, they are the major drivers of disease tions covered in this volume is that they can be burden in these cause and risk factor groups in LMICs. addressed largely through population-based policies The overarching objective of this chapter is to summa- and regulations using intersectoral approaches. For rize the evidence on value for money to reduce the example, risks related to most types of injuries can be burden of injuries and environmental and occupa- substantially reduced through educational programs tional risks in these settings. Evidence on the costs and and legal regulations (Ditsuwan and others 2013). cost-effectiveness of treating the medical consequences Corresponding author: David A. Watkins, Department of Medicine, University of Washington, Seattle, Washington, United States; davidaw@uw.edu. 199 of injury, trauma, and environmental exposures can be Our search yielded 4,539 titles through database found in other volumes of this series. searches and 31 additional studies through expert con- Although externalities and their policy solutions, sultation. After screening titles and abstracts, we such as tradable emissions, that are associated with air reviewed 161 full-text studies. We included 42 of those and water pollution are critical for human health, this studies in the final review on the basis of criteria related chapter does not review these economic issues as they to evaluation methods used and quality of the data. relate to environmental health. Readers are referred to Additionally, we considered only those studies pub- environmental economics textbooks and manuals for lished on or after January 1, 2000 (including costs that discussions of these issues (Maler and Vincent 2005). were collected on or after that date) and pertaining to This chapter focuses exclusively on studies of costs and LMICs (annex 10A). We assessed the quality of included cost-effectiveness (including benefit-cost studies) that studies using a checklist developed by Drummond and have been conducted in LMICs. others (2015). Annex 10B provides a flow diagram of the The economic evidence is modest for injury and review process. ambient environment interventions compared to other Among the 42 studies, we identified 16 higher-quality conditions, but important lessons can be learned about studies. The majority of those studies addressed water, the types of interventions to receive the highest priority sanitation, and hygiene (WASH); one to two studies for public investment. Benefit-cost analysis (BCA) is were found in each category of injury; and no studies the standard approach in environmental economics. were available on occupational injury. We extracted Cost-effectiveness analysis (CEA) is typically applied to 41 estimates of intervention costs and 59 estimates of health sector interventions, but environmental and intervention cost-effectiveness, including incremental other intersectoral interventions—such as development cost-effectiveness ratios (ICERs), net benefits, and and education—are more suited to BCA because many benefit-cost ratios (BCRs). All ICERs, net benefits, and of the costs and benefits are likely to accrue outside the program costs were converted to 2012 U.S. dollars. Costs health sector, and many of these direct benefits can be were first converted to local currency units based on more easily valued in monetary terms. For example, midyear exchange rates in the year that the data were improved water reduces the risk of morbidity and collected. Then they were inflated to 2012 values using mortality from diarrheal disease, but it also has many the World Bank consumer price index series for the nonhealth benefits, such as its intrinsic value to the con- country of the study. Finally, they were converted to sumer and positive effect on tourism. Further, these U.S. dollars using midyear exchange rates for 2012. benefits may even be worth more in monetary terms We qualitatively summarized the remaining 26 than the health benefits. This chapter presents CEA and studies that were of lower quality or used older data BCA evidence side-by-side with some comments on sources, but we did not extract quantitative estimates differences in methods and results. of ICERs or BCRs. In general, an intervention with an ICER less than 1–3 times the per capita gross domestic product of a country was considered cost-effective. An intervention with positive net benefits or a benefit-cost METHODS ratio greater than 1 was considered a good investment. During July 2015, we systematically searched the litera- Annex 10C provides costs, cost-effectiveness ratios, and ture on costs and cost-effectiveness of interventions to descriptive information for each study and a quality prevent injury and reduce risks in the ambient environ- assessment score. ment. The search combined terms for the specific inju- ries and occupational and environmental hazards addressed in this volume, together with economic terms COST-EFFECTIVENESS OF INTERVENTIONS and names of specific LMICs and regions containing LMICs. Our search did not include studies on self-harm, Injuries and Occupational Hazards which is treated in volume 4 of this series (Patel and Economic analyses on injury prevention suggest that others 2015). Annex 10A contains the search terms and interventions to prevent RTIs, drowning, and interper- strategy used to identify relevant articles. We were inter- sonal violence are cost-effective and may even be cost ested in CEAs regardless of perspective. However, we saving (table 10.1). A quasi-experimental study by Bishai present program costs from the health system or govern- and others (2008) assessed a traffic enforcement program ment perspective. The BCAs presented here were con- in Uganda. This program, which focused on reducing ducted from a societal perspective, which is standard in speeding, demonstrated a reduction in fatal crashes at a the field. cost of US$944 per death averted (Bishai and others 2008). 200 Injury Prevention and Environmental Health Table 10.1 Results from Economic Evaluations of Injury Prevention Interventions Cost per Cost per Study author Location, outcome as Currency as outcome (year) Intervention perspectivea presented Unit of outcome presented (year) (2012 US$) Bishai and others Traffic enforcement Uganda, 603 Death averted U.S. dollar 944 (2008) government (2005) Ditsuwan and Checkpoint and media Thailand, 10,400 DALY averted Thai baht 400 others (2013) campaign government (2004) Ditsuwan and Breath testing (selective) Thailand, 13,000 DALY averted Thail baht 600 others (2013) government (2004) Ditsuwan and Breath testing (random) Thailand, 14,300 DALY averted Thail baht 600 others (2013) government (2004) Ditsuwan and Media campaign Thailand, 10,300 DALY averted Thail baht 400 others (2013) government (2004) Ditsuwan and Breath testing (selective) Thailand, 12,700 DALY averted Thail baht 500 others (2013) and media campaign government (2004) Ditsuwan and Breath testing (random) Thailand, 13,500 DALY averted Thail baht 600 others (2013) and media campaign government (2004) Rahman and Anchal drowning Bangladesh, 812 DALY averted International 256 others (2012) prevention program societal dollar (2010) Rahman and SwimSafe drowning Bangladesh, 85 DALY averted International 27 others (2012) prevention program societal dollar (2010) Rahman and Anchal–SwimSafe Bangladesh, 362 DALY averted International 114 others (2012) drowning prevention societal dollar (2010) program (combined) Jan and others Microfinance and gender South Africa, 7,688 DALY averted U.S. dollar 9,826 (2011) training (trial period) government (2004) Jan and others Microfinance and gender South Africa, 2,307 DALY averted U.S. dollar 2,948 (2011) training (scale-up period) government (2004) Note: DALY = disability-adjusted life year. a. “Perspective” refers to the perspective from which costs were estimated. A study in Thailand modeled the cost-effectiveness of to the entire population of rural Bangladesh, they cost several hypothetical interventions for RTIs, including approximately US$114 per DALY averted (Rahman and checkpoints, media campaigns, and breath testing, alone others 2012). or in combination. Compared to the null set, all interven- Interpersonal violence is the intentional use of physi- tions were very cost-effective; all interventions were cost cal force or power against other persons by an individual saving when treatment costs averted were included in the or small group of individuals (chapter 5 in this volume, ICER calculations (Ditsuwan and others 2013). Mercy and others 2017). A modeling study by Elliot and A quasi-experimental study (Rahman and others 2012) Harris (2001) estimated the costs and benefits of land- addressed drowning prevention in rural Bangladesh. The mine clearance in postconflict Mozambique. They found intervention had two components. The first program, negative net benefits and recommended against land- Anchal, focused on direct supervision of children ages mine clearance. Despite ongoing calls for and guidance 1–5 years at child care centers near bodies of water. The on conducting economic evaluations for intimate part- second program, SwimSafe, focused on teaching children ner violence (IPV) (Duvvury and others 2013), only a ages 4–12 years about basic swimming, safety, and rescue single study included in this review focused on IPV. This of others from drowning. Both Anchal and SwimSafe study, by Jan and others (2011), evaluated a trial of were very cost-effective; when their results were com- microfinance to reduce IPV in South Africa. In the South bined and extrapolated to a program that would apply African economic context, ICERs for this program were Interventions to Prevent Injuries and Reduce Environmental and Occupational Hazards 201 favorable in both the trial and a subsequent scale-up engineering and ergonomics programs in all WHO period (Jan and others 2011). regions (Lahiri, Markkanen, and Levenstein 2005). In Additional global and regional studies and modeling the second study, engineering controls were more cost- approaches using a set of assumptions and secondary effective than masks and respirators to prevent silicosis data sources provide insights into the potential eco- (a frequent cause of occupational lung injury) in the nomic costs and benefits of injury prevention. Bishai Western Pacific, including China (Lahiri and others and Hyder (2006) found that increased enforcement, 2005). Finally, a study by Guimaraes, Ribeiro, and Renner speed bumps, bicycle helmets, motorcycle helmets, and (2012) looked at ergonomic changes in footwear manu- childproof containers to prevent poisoning were cost- facturing facilities in Brazil; their intervention had a effective in most world regions. Chisholm and others BCR of 7.2. (2012) conducted an RTI prevention study using the World Health Organization–Choosing Interventions That Are Cost-Effective (WHO-CHOICE) approach. Environmental Hazards Focusing on WHO regions, the authors found that The limited evidence on economic evaluation for (1) bicycle helmets; (2) a combination of speed limits, improved air quality comes from Mexico and focuses drunk-driving laws, and motorcycle helmet use; and on reducing pollution from small-scale industry and (3) a combination of speed limits, drunk-driving laws, automobiles (table 10.2). Blackman and others (2000) seatbelt laws, bicycle helmet use, and motorcycle helmet investigated different ways of reducing pollution from use were cost-effective in high-mortality Sub-Saharan brick kilns, which are a significant contributor to African countries, such as Kenya and Tanzania. They also industrial pollution by the small-scale informal sec- found that a combination of speed limits, impaired driv- tor. Their analysis used an air dispersion model for ing laws, seatbelt laws, and motorcycle helmet use—with pollution and assumed a linear relationship between or without bicycle helmet use—were cost-effective in particulate concentration and health outcomes. Net high-mortality Asian countries, such as India and Nepal. benefits, measured in total dollars per population, Finally, using a preevaluation–postevaluation strategy, were greatest for the strategies focused on retrofitting Stevenson and others (2008) found that a seatbelt use kilns or using natural gas. Relocating kilns to less program was cost-effective in Guangzhou City, China. densely populated areas or instituting no-burn days Additionally, two regional analyses using the WHO- had lower net benefits (Blackman and others 2000). In CHOICE model provide evidence on the cost-effectiveness another study, Stevens, Wilson, and Hammitt (2005) of measures to prevent occupational injury. In the first assessed a variety of methods for retrofitting vehicles study, training programs were more cost-effective than to reduce air pollution. They used a box model with Table 10.2 Results from Economic Evaluations of Ambient Environment Interventions: Benefit-Cost Analyses Net benefit as Currency as Net benefit Study author (year) Intervention Location presented presented (year) (2012 US$) Blackman and NMSU kiln Mexico 46,810,286 U.S. dollar 73,761,533 others (2000) (1999) Blackman and Natural gas Mexico 47,063,087 U.S. dollar 74,159,885 others (2000) (1999) Blackman and Relocation Mexico 26,759,571 U.S. dollar 42,166,522 others (2000) (1999) Blackman and No-burn days Mexico 905,308 U.S. dollar 1,426,543 others (2000) (1999) Stevens, Wilson, Catalyzed DPF Mexico 400–1,700 U.S. dollar 500–2,300 and Hammitt (2005) (2000) Stevens, Wilson, Active regeneration DPF Mexico 100–8,100 U.S. dollar 100–10,800 and Hammitt (2005) (2000) Stevens, Wilson, Diesel oxidation catalyst Mexico 100–2,600 U.S. dollar 100–3,500 and Hammitt (2005) (2000) Note: DPF = diesel particulate filter; NMSU = New Mexico State University. 202 Injury Prevention and Environmental Health health outcomes informed by concentration-response in different settings demonstrated favorable BCRs coefficients from cohort studies in the United States. (ranging from 1.9 to 5.1) for a variety of interventions, Despite the cost of the retrofit, all interventions had including filters, piped water, and boreholes, as well as impressive net benefits that ranged from US$100 to private latrines and community-based sanitation prac- nearly US$11,000 per vehicle retrofitted. tices (Cameron and others 2011; Hutton and others The largest number of economic evaluations of 2014; Whittington and others 2009). The study by environmental health interventions address water and Cameron and others (2011) used preintervention and sanitation and highlight a range of interventions at the postintervention data from their study site to estimate population, community, and household levels (tables 10.3 health outcomes, whereas the other two studies used and 10.4). One quasi-experimental study that assessed a static (equation-based) models of health outcomes. hygiene education initiative in Burkina Faso (Borghi and Another study (World Bank 2013) used a water quality others 2002) found this program to be modestly cost- simulation model to assess wastewater treatment strat- effective at approximately US$1,100 per death averted. A egies in Indonesia. This study demonstrated favorable randomized, controlled trial looked at home-based educa- BCRs for all approaches except one that focused only tion and provision of filters, specifically targeting house- on large industries. Finally, two multicountry studies holds affected by human immunodeficiency virus/acquired that used static (equation-based) models of health out- immune deficiency syndrome (HIV/AIDS) in Uganda. comes demonstrated wide variation in BCRs by coun- The ICER for this program was more than US$2,000 per try. However, all the water and sanitation DALY averted, which is not considered cost-effective in the strategies—alone or in combination—were generally a Ugandan context by Shrestha and others (2006). good return on investment (Hutton 2012; Hutton, The nine remaining water and sanitation evalua- Haller, and Bartram 2007). tions are derived from public works programs affecting Further findings based on the WHO-CHOICE and environmental health. Günther and Fink (2011) devel- other models compare cost-effectiveness of various oped a model using Demographic and Health Survey community-wide and household prevention data from 38 countries that demonstrated the relative approaches for water treatment and reduction of cost-effectiveness of both basic improved water and indoor air pollution. One study found that household sanitation and privately piped water and toilets. Three water treatments were generally more cost-effective benefit-cost analyses of improved water and sanitation than piped water supply and sewage connections in Table 10.3 Results from Economic Evaluations of Ambient Environment Interventions: Cost-Effectiveness Analyses Cost per Currency as Cost per Location, outcome as Unit of presented outcome Study author (year) Intervention perspectivea presented outcome (year) (2012 US$) Borghi and Hygiene promotion Burkina Faso, 657 Death averted U.S. dollar 1,113 others (2002) government (1999) Shrestha and Improved water among Uganda, 1,252 DALY averted U.S. dollar 2,066 others (2006) households affected by government (2004) HIV/AIDS Günther and Fink Basic improved water and Multiple, 1,104 LY saved U.S. dollar n.a. (2011) sanitation in Sub-Saharan government (2007) African countries Günther and Fink Privately piped water and Multiple, 995 LY saved U.S. dollar n.a. (2011) flush toilets government (2007) Günther and Fink Improved water and Multiple, 3,281 LY saved U.S. dollar n.a. (2011) sanitation government (2007) Günther and Fink Privately piped water and Multiple, 3,188 LY saved U.S. dollar n.a. (2011) flush toilets in other LMICs government (2007) Note: HIV/AIDS = human immunodeficiency virus/acquired immune deficiency syndrome; DALY = disability-adjusted life year; LMICs = low- and middle-income countries; LY = life year; n.a. = not applicable. Costs were estimated for multiple countries and then aggregated; no cost deflator is available for this grouping. a. “Perspective” refers to the perspective from which costs were estimated. Interventions to Prevent Injuries and Reduce Environmental and Occupational Hazards 203 Table 10.4 Results from Economic Evaluations of Ambient Environment Interventions: Benefit-Cost Analyses Study author (year) Intervention Location Benefit-cost ratioa Currency used (year) Cameron and Improved water South Africa 3.1 South African rand others (2011) (2008) Hutton and Community sewage system (urban) East Asia and Pacific 1.9 U.S. dollar others (2014) (2008) Hutton and Private wet pit latrine (rural) East Asia and Pacific 5.1 U.S. dollar others (2014) (2008) Whittington and Borehole and public hand pump Multiple 3.3 U.S. dollar others (2009) (2007) Whittington and Community-led total sanitation Multiple 3.0 U.S. dollar others (2009) program (2007) Whittington and Biosand filter Multiple 3.0 U.S. dollar others (2009) (2007) Whittington and Large dam Multiple 3.7 U.S. dollar others (2009) (2007) World Bank (2013) Treatment of domestic wastewater Indonesia 2.3 Indonesian rupiah (2010) World Bank (2013) Treatment of industrial wastewater, Indonesia 0.6 Indonesian rupiah differentiating all industries or (2010) large industries only World Bank (2013) Treatment of domestic and Indonesia 2.0 Indonesian rupiah industrial wastewater (2010) World Bank (2013) Treatment of domestic and Indonesia 2.3 Indonesian rupiah industrial wastewater and recycling (2010) of industrial wastewater Hutton, Haller, and Improved water Multiple 4.4–31.6 n.a. Bartram (2007) Hutton, Haller, and Improved water and sanitation Multiple 5.5–45.5 n.a. Bartram (2007) Hutton, Haller, and Improved water and sanitation Multiple 5.2–45.0 n.a. Bartram (2007) and universal basic access Hutton, Haller, and Universal basic access and Multiple 5.7–40.7 n.a. Bartram (2007) point-of-use treatment Hutton, Haller, and Regulated piped water supply and Multiple 2.1–11.8 n.a. Bartram (2007) sewer connection Hutton (2012) Universal access to improved sanitation Multiple 2.8–8.0 n.a. Hutton (2012) Universal access to improved drinking Multiple 0.6–3.7 n.a. water sources Hutton (2012) Universal access to improved sanitation Multiple 2.0–5.3 n.a. and improved drinking water sources Note: n.a. = not applicable. Costs were estimated for multiple countries and then aggregated; no cost deflator is available for this grouping. Net benefits and benefit-cost ratios are presented from a societal perspective. a. Benefit-cost ratios are presented as given in the article because they are independent of currency inflation and exchange rates. 204 Injury Prevention and Environmental Health different world regions (Haller, Hutton, and Bartram interventions in Thailand and Uganda cost less than 2007). Within the household, two studies found that US$1 per capita. However, the combined drowning using home-based water filter techniques was ulti- prevention program in Bangladesh and the IPV preven- mately more cost-effective than boiling water (Clasen tion program cost more than US$20 per capita per par- and others 2007; Clasen and others 2008). A study by ticipant, owing to the higher resource needs and training Larsen (2003) found that providing safe sanitation required to conduct these programs. Mine clearance was facilities was more cost-effective than providing safe very expensive, at more than US$14,000 per hectare, water supplies, which was more cost-effective than which is part of the reason the program had net negative hygiene improvement strategies. Finally, Jeuland and benefits. Pattanayak (2012) developed a model for use in con- ducting multiple sensitivity analyses related to cook- Environmental Hazards stove interventions. A major finding is that uptake levels of interventions and other assumptions greatly Table 10.6 summarizes the costs of programs to reduce alter the estimated net benefits to households; in some environmental risks. Air pollution control, despite being cases, households are worse off with improved cook- cost-effective, was relatively expensive in Mexico. Water stoves (Jeuland and Pattanayak 2012). and sanitation programs also required large investments over a long time horizon, with the exception of personal point-of-use technologies. Special attention was paid in PROGRAM COSTS these studies to the recurrent costs of WASH infrastruc- ture, which in many cases were a substantial contributor Injuries and Occupational Hazards to overall costs and might be borne by households. The estimated costs of the injury prevention program However, these large costs appeared to be outweighed derived from the cost-effectiveness and benefit-cost by the economic benefits. A collection of studies on studies are presented in table 10.5. The RTI prevention water and sanitation demand not included in this Table 10.5 Program Costs of Injury Prevention Interventions Location, Cost as Currency as Cost Study author (year) Intervention perspectivea presented Unit of cost presented (year) (2012 US$) Bishai and others Traffic enforcement Uganda, 0.45 Per vehicle per U.S. dollar 0.70 (2008) government year (2005) Ditsuwan and others Sobriety checkpoints by Thailand, 0.27 Per capita per Thai baht 0.01 (2013) metropolitan police government year (2004) Ditsuwan and others Sobriety checkpoints by Thailand, 5.29 Per capita per Thai baht 0.23 (2013) traffic police government year (2004) Ditsuwan and others Media campaign on Thailand, 2.86 Per capita per Thai baht 0.12 (2013) drink-driving government year (2004) Rahman and others Anchal drowning Bangladesh, 60.50 Per capita first International dollar 19.07 (2012) prevention program societal year (2010) Rahman and others Anchal drowning Bangladesh, 50.74 Per capita second International dollar 16.00 (2012) prevention program societal year onward (2010) Rahman and others SwimSafe drowning Bangladesh, 13.46 Per capita per International dollar 4.24 (2012) prevention program societal year (2010) Elliot and Harris (2001) Mine clearance Mozambique, 6,176.50 Per hectare U.S. dollar 14,249.00 societal (1996) Jan and others (2011) Microfinance and gender South Africa, 42.93 Per capita per 2 U.S. dollar 54.87 training (trial period) government years (2004) Jan and others (2011) Microfinance and gender South Africa, 12.88 Per capita per 2 U.S. dollar 16.46 training (scale-up) government years (2004) a. “Perspective” refers to the perspective from which costs were estimated. Interventions to Prevent Injuries and Reduce Environmental and Occupational Hazards 205 Table 10.6 Program Costs of Ambient Environment Interventions Currency as Study author Location, Unit of presented Cost (year) Intervention perspectivea Cost as presented cost (year) (2012 US$) Air pollution Blackman and NMSU kiln Mexico, 175,214.00 Per year U.S. dollar 276,094.00 others (2000) societal (1999) Blackman and Natural gas Mexico, 249,553.00 Per year U.S. dollar 393,234.00 others (2000) societal (1999) Blackman and Relocation Mexico, 350,429.00 Per year U.S. dollar 552,190.00 others (2000) societal (1999) Blackman and No-burn days Mexico, 24,692.00 Per year U.S. dollar 38,909.00 others (2000) societal (1999) Stevens, Wilson, Catalyzed DPF Mexico, 270.00 Per vehicle U.S. dollar 360.00 and Hammitt societal per year (2000) (2005) Stevens, Wilson, Active regeneration DPF Mexico, 600.00 Per vehicle U.S. dollar 800.00 and Hammitt societal per year (2000) (2005) Stevens, Wilson, Diesel oxidation catalyst Mexico, 90.00 Per vehicle U.S. dollar 120.00 and Hammitt societal per year (2000) (2005) Water, Sanitation, and Hygiene Borghi and Hygiene promotion Burkina Faso, 8.11 Per capita U.S. dollar 13.73 others (2002) government (1999) Cameron and Improved water South Africa, 153.31 Per capita South African 2.58 others (2011) societal per year rand (2008) Shrestha and Improved water among Uganda, 2.19 Per capita U.S. dollar 3.61 others (2006) households affected by government per year (2004) HIV/AIDS Whittington and Borehole and public hand Multiple, 2.00 Per house U.S. dollar n. a. others (2009) pump societal per month (2007) Whittington and Community-led total Multiple, 0.39 Per house U.S. dollar n. a. others (2009) sanitation program societal per month (2007) Whittington and Biosand filter Multiple, 1.25 Per house U.S. dollar n. a. others (2009) societal per month (2007) Whittington and Large dam Multiple, 3,743,000.00 Per 75 years U.S. dollar n. a. others (2009) societal (2007) World Bank (2013) Treatment of domestic Indonesia, 888,000,000,000.00 Per 20 years Indonesian 109,000,000.00 wastewater societal rupiah (2010) World Bank (2013) Treatment of industrial Indonesia, 172,000,000,000.00 Per 20 years Indonesian 21,000,000.00 wastewater, differentiating societal rupiah all industries or large (2010) industries only table continues next page 206 Injury Prevention and Environmental Health Table 10.6 Program Costs of Ambient Environment Interventions (continued) Currency as Study author Location, Unit of presented Cost (year) Intervention perspectivea Cost as presented cost (year) (2012 US$) World Bank (2013) Treatment of domestic and Indonesia, 1,059,000,000,000.00 Per 20 years Indonesian 130,000,000.00 industrial wastewater societal rupiah (2010) World Bank (2013) Treatment of domestic Indonesia, 1,164,000,000,000.00 Per 20 years Indonesian 143,000,000.00 and industrial wastewater societal rupiah and recycling of industrial (2010) wastewater Günther and Basic sanitation (latrines) Multiple, 600.00–965.00 Per house U.S. dollar n.a. Fink (2011) government per lifetime (2007) Günther and Basic improved water Multiple, 309.00–545.00 Per house U.S. dollar n.a. Fink (2011) government per lifetime (2007) Günther and Flush toilets Multiple, 2,099.00–2,400.00 Per house U.S. dollar n.a. Fink (2011) government per lifetime (2007) Günther and Privately piped water Multiple, 1,623.00–2,509.00 Per house U.S. dollar n.a. Fink (2011) government per lifetime (2007) Note: HIV/AIDS = human immunodeficiency virus/acquired immune deficiency syndrome; DPF = diesel particulate filter; NMSU = New Mexico State University; n.a. = not applicable. Costs were estimated for multiple countries and then aggregated; no cost deflator is available for this grouping. a. “Perspective” refers to the perspective from which costs were estimated. review suggests that despite the high cost of water and studies here. Injury often leads to disability, which can sanitation infrastructure, households in LMICs are gen- be permanent. From a human capital perspective, injury erally willing to pay a substantial amount for these leads to foregone wages and hinders economic growth goods and services (Dutta and Tiwari 2005; Pattanayak and development (Nguyen and others 2016). Gender- and others 2010). based violence, in particular, can hinder women’s eco- Some recent studies can inform the costs of these nomic opportunities and impede gender equity, leading programs, although data may be incomplete or out of to consequences for overall economic growth (Duvvury date. For example, a study by Banerjee and others and others 2013). Occupational injury may have delete- (2007) found that social marketing of safe water sys- rious effects on both employees and employers, but in tems resulted in economies of scale and greater finan- practice, costs often fall exclusively on employees. In cial sustainability. A study by Crocker and Bartram countries where trauma care is financed predominantly (2014) looked at the cost of routine water quality test- by out-of-pocket funds, injuries can have catastrophic ing in seven countries. They documented great hetero- or impoverishing effects on households (Nguyen and geneity and economic inefficiencies in existing others 2016). Preventing injury in these settings can monitoring practices and proposed ways of optimizing reduce the risk of medical impoverishment (Olson and monitoring. A study by Dodane and others (2012) others 2016). demonstrated that in urban Senegal, sewer-based sys- The rationale for government intervention to pre- tems were more expensive, less feasible, and no more vent environmental exposures stems from the concept effective than more commonly used on-site waste man- of internalizing externalities or reducing the social agement systems. costs associated with private decisions. Such social costs include negative effects on the environment. The OTHER ECONOMIC ASPECTS OF INJURIES BCA approach used in most environmental health evaluations reviewed should, in principle, include an AND ENVIRONMENTAL HEALTH estimate of the environmental (nonhealth) benefits Prevention of injury is likely to have more economic of scaling up WASH and addressing air pollution. benefits than those described in the cost-effectiveness In practice, however, many of these benefits are Interventions to Prevent Injuries and Reduce Environmental and Occupational Hazards 207 undervalued. For example, households appear to place these settings will be impressive and warrant further a high intrinsic value (willingness-to-pay) on WASH investigation. services, but when these services are not present, With respect to environmental risks, the evidence is households face high coping costs related to unreliable compelling for scaling up WASH interventions, particu- water supply (Dutta and Tiwari 2005). In such cases, larly basic sanitation measures and point-of-use water public sector regulation can improve economic effi- quality interventions. Piped water and other large-scale ciency. Empirical evidence suggests that in most set- infrastructure projects can be good value for money tings, people are willing to pay substantial amounts to from a societal standpoint (that is, including the intrinsic mitigate negative externalities in the injury and envi- value individuals place on having piped water indepen- ronmental health domains (Ortuzar, Cifuentes, and dent of its health effect). However, these projects require Williams 2000). more substantial up-front investments and may be infea- sible as a first step for low-income countries. Whether WASH infrastructure should be a high priority for other LIMITATIONS IN EVIDENCE sectoral reasons (that is, besides the health benefits) for Our review has summarized the recent economic governments of low-income countries is unclear. evaluations that focus on preventing injuries and Unfortunately, little economic evidence exists on reducing environmental hazards in LMICs. The abso- control of outdoor air pollution in LMICs outside of lute number of studies on such hazards in LMICs Mexico, and there are essentially no economic studies published since 2000 is small in comparison to other published after 2000 on control of indoor air pollution. areas of health. However, there are a few additional In contrast, outdoor air pollution—which, in the near studies from LMICs published before 2000; these were future, is expected to supplant the burden of indoor reviewed in chapters 42 and 43 of Disease Control air pollution in LMICs—has been the focus of many Priorities in Developing Countries (second edition) epidemiological studies and economic evaluations in (Bruce and others 2006; Kjellstrom and others 2006). high-income countries. These analyses can serve as a Despite broad recognition of the problem and first step to informing policy in LMICs. (See, for exam- many effective interventions, there is surprisingly ple, the economic evaluations conducted by the U.S. little economic literature on the prevention of RTIs Environmental Protection Agency [2011].) Analyses of in LMICs. Evidence from high-income countries outdoor air pollution policies in LMICs will be of shows that investments in promoting or establishing utmost importance given trends in urbanization and helmet laws, vehicle inspections, seatbelt use, and, to a the potential for mitigating climate change. We note lesser extent, speed limits appear to be good value that a discussion of the economics of climate change is for money (Waters, Hyder, and Phillips 2004). Many particularly challenging and is outside the scope of this of these approaches would likely be effective in review. See chapter 8 in this volume (Ebi, Hess, and LMICs. However, the costs, affordability, and cost- Watkiss 2017) for a discussion of potential global costs effectiveness may vary, depending on country context and benefits of mitigating climate change. and the amount of government and regulatory infra- Even more difficult than evaluating health sector structure in place to implement and scale up these interventions, economic evaluations of cross-sectoral interventions. interventions with multiple types of costs and benefits Similarly, there are few economic analyses on pre- face serious methodological challenges, given the diffi- vention of falls, burns, poisoning, or injuries from culty of valuing and combining the direct and indirect forces of nature in LMICs. Given the increasing burden benefits of reducing personal injury and promoting of injuries (in particular, falls) and of interpersonal environmental health. The literature is concentrated in violence in LMICs, this area will be important for several areas, especially RTI interventions and struc- future research. Studies in high-income countries have tural WASH interventions. Both areas suggest quite demonstrated that regulatory and legal interventions positive results, but do not fully account for social ben- are cost-effective, but the social and legal differences efits. For RTIs, an accounting for downstream health compared to LMICs make these studies difficult to gen- effects of injury is lacking. For environmental expo- eralize (Waters and others 2004). Further, the most sures, the key issue is the difficulty of capturing and neglected area for economic analysis by far is occupa- measuring benefits across sectors. In these cases, we are tional injury. We found no high-quality studies in likely to be underestimating the benefits of such inter- LMICs in this area, but we anticipate that the produc- ventions and therefore underinvesting in environmen- tivity benefits of addressing occupational hazards in tal health protection. 208 Injury Prevention and Environmental Health CONCLUSIONS Blackman, A., S. Newbold, J. Shih, and J. Cook. 2000. “The Benefits and Costs of Informal Sector Pollution Control: We have identified several areas in injury prevention and Mexican Brick Kilns.” Discussion Paper 00-46, Resources environmental and occupational health where a good for the Future, Washington, DC. economic case might be made for action. Yet impor- Borghi, J., L. Guinness, J. Ouedraogo, and V. Curtis. 2002. tant methodological challenges remain, as well as unan- “Is Hygiene Promotion Cost-Effective? A Case Study in swered questions and conditions for which no literature Burkina Faso.” Tropical Medicine and International Health exists. If one considers the human capital and other 7 (11): 960–69. Bruce, N., E. Rehfuess, S. Mehta, G. Hutton, and K. Smith. social costs associated with these diseases and risk factors, 2006. “Indoor Air Pollution.” In Disease Control Priorities in more research in this area is urgently needed to inform Developing Countries (second edition), edited by D. T. Jamison, the design and implementation of intersectoral policies. J. G. Bremen, A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove. Washington, DC: World Bank. ANNEXES Cameron, J., P. Jagals, P. Hunter, S. Pedley, and K. Pond. 2011. The annexes to this chapter are as follows. They are “Economic Assessments of Small-Scale Drinking-Water Interventions in Pursuit of MDG Target 7C.” Science of the available at http://www.dcp-3.org/environment. Total Environment 410–11: 8–15. Chisholm, D., H. Naci, A. Hyder, N. Tran, and M. Peden. • Annex 10A. Search Terms Used to Identify Relevant 2012. “Cost Effectiveness of Strategies to Combat Road Literature Traffic Injuries in Sub-Saharan Africa and South East Asia: • Annex 10B. Flow Chart of Identification, Screening, Mathematical Modelling Study.” BMJ 344: e612. and Eligibility of Included Studies Clasen, T., C. McLaughlin, N. Nayaar, S. Boisson, R. Gupta, • Annex 10C. List of Included Studies, Main Findings, and others. 2008. “Microbiological Effectiveness and Cost and Quality Assessment of Disinfecting Water by Boiling in Semi-Urban India.” American Society of Tropical Medicine and Hygiene 79 (3): 407–13. Clasen, T., D. Thao, S. Boisson, and O. Shipin. 2007. NOTE “Microbiological Effectiveness and Cost of Boiling to World Bank Income Classifications as of July 2014 are as Disinfect Drinking Water in Rural Vietnam.” Environmental follows, based on estimates of gross national income (GNI) Science and Technology 42 (12): 4255–61. per capita for 2013: Crocker, J., and J. Bartram. 2014. “Comparison and Cost Analysis of Drinking Water Quality Monitoring Requirements ver- • Low-income countries (LICs) = US$1,045 or less sus Practice in Seven Developing Countries.” International • Middle-income countries (MICs) are subdivided: Journal of Environmental Research and Public Health 11: a) lower-middle-income = US$1,046 to US$4,125 7333–46. b) upper-middle-income (UMICs) = US$4,126 to US$12,745 Ditsuwan, V., J. Veerman, M. Bertram, and T. Vos. 2013. “Cost- • High-income countries (HICs) = US$12,746 or more. Effectiveness of Interventions for Reducing Road Traffic Injuries Related to Driving under the Influence of Alcohol.” All intervention costs in this chapter have been converted Value in Health 16: 23–30. into 2012 US$, using the World Bank consumer price index or Dodane, P., M. Mbeguere, O. Sow, and L. Strande. 2012. regional inflation rates, unless otherwise noted. “Capital and Operating Costs of Full-Scale Fecal Sludge Management and Wastewater Treatment Systems in Dakar, Senegal.” Environmental Science and Technology 46: 3705–11. REFERENCES Drummond, M. F., M. J. Schulpher, K. Claxton, G. L. Stoddart, Banerjee, A., D. McFarland, R. Sinh, and R. Quick. 2007. “Cost and G. W. Torrance. 2015. Methods for the Economic and Financial Sustainability of a Household-Based Water Evaluation of Health Care Programmes, fourth edition. Treatment and Storage Intervention in Zambia.” Journal of Oxford, U.K.: Oxford University Press. Water and Health 5 (3): 385–95. Dutta, V., and A. Tiwari. 2005. “Cost of Services and Willingness Bishai, D., B. Asiimwe, S. Abbas, A. Hyder, and W. Bazeyo. 2008. to Pay for Reliable Urban Water Supply: A Study from “Cost-Effectiveness of Traffic Enforcement: Case Study Delhi, India.” Journal of Water Science and Technology 5 (6): from Uganda.” Injury Prevention 14: 223–27. 135–44. Bishai, D., and A. Hyder. 2006. “Modeling the Cost Effectiveness Duvvury, N., A. Callan, P. Carney, and S. Raghavendra. of Injury Interventions in Lower and Middle Income 2013. “Intimate Partner Violence: Economic Costs and Countries: Opportunities and Challenges.” Cost Effectiveness Implications for Growth and Development.” Working and Resource Allocation 4 (2): 1–11. Paper 82532, World Bank, Washington, DC. Interventions to Prevent Injuries and Reduce Environmental and Occupational Hazards 209 Ebi, K. L., J. J. Hess, and P. Watkiss. 2017. “Health Risks and Costs of Larsen, B. 2003. “Hygiene and Health in Developing Countries: Climate Variability and Change.” In Disease Control Priorities Defining Priorities through Cost-Benefit Assessments.” (third edition): Volume 7, Injury Prevention and Environmental International Journal of Environmental Health Research 13: Health, edited by C. N. Mock, R. Nugent, O. Kobusingye, and S37–46. K. R. Smith. Washington, DC: World Bank. Luh, J., and J. Bartram. 2016. “Drinking Water and Sanitation: Elliot, G., and G. Harris. 2001. “A Cost-Benefit Analysis Progress in 73 Countries in Relation to Socioeconomic of Landmine Clearance in Mozambique.” Development Indicators.” Bulletin of the World Health Organization 94: Southern Africa 18 (5): 625–33. 111–21. Guimaraes, L., J. Ribeiro, and J. Renner. 2012. “Cost-Benefit Maler, K. G., and J. R. Vincent, eds. 2005. Handbook of Environmental Analysis of a Socio-Technical Intervention in a Brazilian Economics: Economywide and International Environmental Footwear Company.” Applied Ergonomics 43: 948–57. Issues, third edition. Amsterdam: North Holland. Günther, I., and G. Fink. 2011. “Water and Sanitation to Mercy, J. A., S. D. Hillis, A. Butchart, M. A. Bellis, C. L. Ward, Reduce Child Mortality: The Impact and Cost of Water and and others. 2017. “Interpersonal Violence: Global Impact Sanitation Infrastructure.” Policy Research Working Paper and Paths to Prevention.” In Disease Control Priorities (third 5618, World Bank, Washington, DC. edition): Volume 7, Injury Prevention and Environmental Haller, L., G. Hutton, and J. Bartram. 2007. “Estimating Health, edited by C. N. Mock, R. Nugent, O. Kobusingye, the Costs and Health Benefits of Water and Sanitation and K. R. Smith. Washington, DC: World Bank. Improvements at Global Level.” Journal of Water and Health Nguyen, H., R. Ivers, S. Jan, A. Martiniuk, L. Segal, and others. 5 (4): 467–81. 2016. “Cost and Impoverishment 1 Year after Hospitalisation Hutton, G. 2012. Global Costs and Benefits of Drinking-Water due to Injuries: A Cohort Study in Thai Binh, Vietnam.” Supply and Sanitation Interventions to Reach the MDG Injury Prevention 22 (1): 33–39. Target and Universal Coverage. Geneva: World Health Olson, Z., J. A. Staples, C. N. Mock, N. P. Nguyen, A. M. Bachani, Organization. and others. 2016. “Helmet Regulation in Vietnam: Impact Hutton, G., L. Haller, and J. Bartram. 2007. “Global Cost-Benefit on Health, Equity, and Medical Impoverishment.” Injury Analysis of Water Supply and Sanitation Interventions.” Prevention 22 (4): 233–38. Journal of Water and Health 5 (4): 481–502. Ortuzar, J. D., L. A. Cifuentes, and H. C. W. L. Williams. Hutton, G., U. Rodriguez, A. Winara, N. Anh, K. Phyrum, 2000. “Application of Willingness-to-Pay Methods to Value and others. 2014. “Economic Efficiency of Sanitation Transport Externalities in Less Developed Countries.” Interventions in Southeast Asia.” Journal of Water, Sanitation Environment and Planning A 32 (11): 2007–18. and Hygiene for Development 4 (1): 23–36. Patel, V., D. Chisholm, T. Dua, R. Laxminarayan, and M. E. IHME (Institute for Health Metrics and Evaluation). 2015. “Global Medina-Mora, eds. 2015. Disease Control Priorities (third Burden of Disease Study 2013 (GBD 2013) Data Downloads.” edition): Volume 4, Mental, Neurological, and Substance Use IHME, Seattle. http://ghdx.healthdata.org/global-burden Disorders. Washington, DC: World Bank. -disease-study-2013-gbd-2013-data-downloads. Pattanayak, S., C. Poulos, J. Yang, and S. Patil. 2010. “How Jan, S., G. Ferrari, C. Watts, J. Hargreaves, J. Kim, and others. Valuable Are Environmental Health Interventions? 2011. “Economic Evaluation of a Combined Microfinance Evaluation of Water and Sanitation Programmes in India.” and Gender Training Intervention for the Prevention of Bulletin of the World Health Organization 88: 535–42. Intimate Partner Violence in Rural South Africa.” Health Rahman, F., S. Bose, M. Linnqn, A. Rahman, S. Mashreky, Policy and Planning 26: 366–72. and others. 2012. “Cost-Effectiveness of an Injury and Jeuland, M., and S. K. Pattanayak. 2012. “Benefits and Costs Drowning Prevention Program in Bangladesh.” Pediatrics of Improved Cookstoves: Assessing the Implications of 130 (6): 1–10. Variability in Health, Forest and Climate Impacts.” Plos One Shrestha, R., E. Marseille, J. Kahn, J. Lule, C. Pitter, and others. 7 (2): e30338. 2006. “Cost-Effectiveness of Home-Based Chlorination Kjellstrom, T., M. Lodh, T. McMichael, G. Ranmuthugala, and Safe Water Storage in Reducing Diarrhea among HIV- R. Shrestha, and others. 2006. “Air and Water Pollution: Affected Households in Rural Uganda.” American Journal of Burden and Strategies for Control.” In Disease Control Tropical Medicine and Hygiene 74 (5): 884–90. Priorities in Developing Countries (second edition), edited Stevens, G., A. Wilson, and J. Hammitt. 2005. “A Benefit-Cost by D. T. Jamison, J. G. Bremen, A. R. Measham, G. Alleyne, Analysis of Retrofitting Diesel Vehicles with Particulate M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove. Filters in the Mexico City Metropolitan Area.” Risk Analysis Washington, DC: World Bank. 25 (4): 1–17. Lahiri, S., C. Levenstein, D. Nelson, and B. Rosenberg. Stevenson, M., D. Hendrie, R. Ivers, Z. Su, and R. Norton. 2008. 2005. “The Cost Effectiveness of Occupational Health “Reducing the Burden of Road Traffic Injury: Translating Interventions: Prevention of Silicosis.” American Journal of High-Income Country Interventions to Middle-Income Industrial Medicine 48: 503–14. and Low-Income Countries.” Injury Prevention 14: 2848–49. Lahiri, S., P. Markkanen, and C. Levenstein. 2005. “The Cost UN (United Nations). 2016. Sustainable Development Goals: Effectiveness of Occupational Health Interventions: 17 Goals to Transform Our World. Goal 6: Ensure Access to Preventing Occupational Back Pain.” American Journal of Water and Sanitation for All. New York: UN. http://www.un Industrial Medicine 48: 515–29. .org/sustainabledevelopment/water-and-sanitation. 210 Injury Prevention and Environmental Health U.S. EPA (United States Environmental Protection Agency). Whittington, D., M. Shanemann, C. Sadoff, and M. Jeuland. 2011. “The Benefits and Costs of the Clean Air Act from 2009. “The Challenge of Improving Water and Sanitation 1990 to 2010: Final Report–Rev. A.” Office of Air and Services in Less Developed Countries.” Foundations and Radiation, U.S. EPA, Washington, DC. Trends in Microeconomics 4 (6–7): 469–609. Waters, H. R., A. A. Hyder, and T. L. Phillips. 2004. “Economic WHO (World Health Organization). 2016. Global Health Evaluation of Interventions to Reduce Road Traffic Injuries: Estimates 2012. Geneva: WHO. http://www.who.int A Review of the Literature with Applications to Low and /healthinfo/global_burden_disease/en/. Middle-Income Countries.” Asia-Pacific Journal of Public World Bank. 2013. “Downstream Impacts of Water Pollution Health 16 (1): 23–31. in the Upper Citarum River, West Java, Indonesia: Waters, H. R., A. A. Hyder, Y. Rajkotia, S. Basu, and J. A. Economic Assessment of Interventions to Improve Rehwinkel. 2004. The Economic Dimensions of Interpersonal Water Quality.” Technical Paper 85194, World Bank, Violence. Geneva: World Health Organization. Washington, DC. Interventions to Prevent Injuries and Reduce Environmental and Occupational Hazards 211 Chapter 11 Helmet Regulation in Vietnam: Impact on Health, Equity, and Medical Impoverishment Zachary Olson, John A. Staples, Charles N. Mock, Nam Phuong Nguyen, Abdulgafoor M. Bachani, Rachel Nugent, and Stéphane Verguet INTRODUCTION to 93 percent of riders within months (Hung, Stevenson, Road traffic injury (RTI) accounts for a substantial and and Ivers 2006; Nguyen, Passmore, and others 2013). increasing burden of mortality, morbidity, and health Studies in other settings have examined the influence of care costs in developing nations. Globally, road traffic is helmet use policies on aggregate health, but the distri- responsible for 1.3 million fatal and 78 million nonfatal bution of benefits and equity improvements resulting injuries each year (WHO 2013a; World Bank and IHME from such regulatory changes remains understudied 2014). In the Western Pacific, RTI is the leading cause of and uncertain (Ngo and others 2012; Passmore, Tu, and mortality for people ages 15–49 years (WHO 2013b). others 2010). Direct economic costs are estimated to exceed US$500 Traffic injury can lead to substantial and potentially billion worldwide and are anticipated to grow in tandem impoverishing health expenditures (Wagstaff 2010). with motorization of the developing world (WHO 2004; Legislation mandating helmet use is one non–health World Bank and IHME 2014). The potentially substan- sector policy that may protect individuals against this tial out-of-pocket (OOP) medical costs associated with financial risk. In nations with universal health cover- traffic injury may result in catastrophic expenditures age, helmet regulation may also reduce government (expenditures that crowd out a significant portion of spending for traffic injuries and thus free up health household expenditures) and subsequent impoverish- spending for other conditions. Defining the magnitude ment (Wagstaff 2010). of the health and financial benefits attributable to In response to the growing burden of traffic injury, Vietnam’s comprehensive helmet policy might bolster the government of Vietnam passed comprehensive leg- the case for a similar policy in neighboring countries islation mandating the use of motorcycle helmets in (for example, Cambodia) and in other low-and- 2007. This legislation extended the mandatory use of middle-income countries. helmets to all riders on all roads, substantially increased Extended cost-effectiveness analysis (ECEA) incorpo- penalties for failure to wear a helmet, and provided for rates the dimensions of equity and financial risk protection increased enforcement (Passmore, Nguyen, and others into economic evaluations (Verguet, Laxminarayan, and 2010). As a result, helmet use increased from 30 percent Jamison 2014; Verguet and others 2013, 2015). In this Corresponding author: Zachary Olson, School of Public Health, University of California, Berkeley, CA, United States; zolson@berkeley.edu. Zachary Olson and John Staples contributed equally to the work. 213 chapter, a simulation model is used to perform an ECEA all roads was introduced in June 2007, came into force for examining the influence that Vietnam’s 2007 helmet government workers in September 2007, and came into legislation has had in four areas: force for the general public in December 2007 (Passmore, Nguyen, and others 2010). Legislation introduced in • Road traffic deaths and nonfatal injuries September 2007 increased the fines for failure to wear a • Individuals’ direct costs of acute care treatment for helmet from US$2–US$5 to US$11–US$22 per offense, motorcycle injuries the latter range representing about 30 percent of average • Individuals’ income losses from missed work monthly income per capita (Government of Vietnam • Individuals’ financial risk. 2007; Passmore, Nguyen, and others 2010). At that time, the majority of Vietnamese households were willing to pay the average market price of US$17 for a standard METHODS helmet (Pham and others 2008). Design For the period of interest, the annual number of nonfatal Variables traffic injuries reported by Vietnam’s National Traffic In the simulation, all input parameters were abstracted Safety Committee is not disaggregated by type of road from academic studies and from reports issued by gov- user and generally lacks consistency and credibility. For ernmental and nongovernmental agencies (table 11.1; see example, the number of nonfatal traffic injuries reported also annex 11A, table 11A.1). The output estimates of by police in 2007 (10,300) is drastically different from primary interest were traffic deaths averted, nonfatal traf- the number noted in health data reports for the same fic injuries averted, individuals’ OOP acute care medical year (445,000) (WHO 2009). To address this discrep- costs averted, and individuals’ income losses averted dur- ancy, the model developed for this chapter uses second- ing the one-year post-policy period. Costs were viewed ary data to simulate the benefits accruing from the 2007 from the individuals’ perspective, including both OOP comprehensive helmet policy. After ensuring consistency acute care costs and income losses. Estimation of sub- of the model with reported reductions in total road traf- acute and chronic outpatient medical costs was not pos- fic deaths (National Traffic Safety Committee 2014; sible, as reliable input parameters were not available. Passmore, Nguyen, and others 2010), an ECEA was per- All costs are in 2012 U.S. dollars and were converted formed to estimate the distribution of health benefits using consumer price indexes and exchange rates as and costs across income groups. Conceptually, the study reported by the World Bank (2012). period includes a one-year pre-policy baseline period (July 2006 to June 2007), a six-month transition period during which the majority of the helmet policy legisla- Major Assumptions tion was introduced and came into effect (June to According to the National Traffic Safety Committee December 2007), and a one-year post-policy evaluation (2014), the number of registered motorcycles in Vietnam period (January to December 2008). increased from 21 million in 2007 to 25 million in 2008. However, for simplicity, the model assumes that the number of registered motorcycles remained static at the Setting pre-policy level during the study period. This assumption At the midpoint of the study period, Vietnam was a makes the estimates more conservative but substantially lower-middle-income country with a population of improves their interpretability and generalizability. about 84 million and per capita gross domestic product Furthermore, the effectiveness of motorcycle helmets in (GDP) of about US$1,200 (World Bank 2012). About 95 Vietnam was assumed to be equivalent to the estimated percent of registered vehicles were motorized two-wheel effectiveness of helmets in high-income countries (HICs). vehicles (WHO 2013c). Prior to the 2007 legislation, the However, major concerns have been raised regarding the incidence of road traffic deaths was estimated to be 14 proliferation of substandard helmets in Vietnam (Hung, per 100,000 people per year (WHO 2009). About 55 Stevenson, and Ivers 2008; WHO 2015; Yu and others percent of health care costs were paid out of pocket 2011). Given the lack of local data regarding the effective- (Tien and others 2011; WHO 2010). ness of substandard helmets, this crucial issue was Prior to 2007, Vietnam had limited motorcycle helmet addressed in a separate sensitivity analysis. For the main legislation with incomplete implementation and enforce- analysis, the distribution of fatal and nonfatal traffic inju- ment. Comprehensive legislation that made helmet use ries across income quintiles was assumed to reflect the compulsory for all motorcycle riders and passengers on distribution of motorcycle ownership across quintiles, 214 Injury Prevention and Environmental Health Table 11.1 Model Input Parameters Parameter Estimate (range) References Population of Vietnam 84,000,000 World Bank 2012 Pre-policy RTI deaths 13,000 WHO 2009 Pre-policy nonfatal RTIs 445,000 WHO 2009 % of RTI deaths attributable to motorcycles 58 (51–73) Hoang and others 2008; Ngo and others 2012; Pham and others 2008; % of nonfatal RTIs attributable to motorcycles 59 (51–75) Hoang and others 2008; Hung, Stevenson, and Ivers 2006; Ngo and others 2012; Nguyen, Passmore, and others 2013; Pham and others 2008 % of nonfatal motorcycle RTIs with head 21 (10–32) Nguyen, Ivers, and others 2013 injury Pre-policy helmet use (%) 30 (20–40) Hung, Stevenson, and Ivers 2006; Nguyen, Passmore, and others 2013 Postpolicy helmet use (%) 93 (83–98) Nguyen, Passmore, and others 2013 Average direct acute care cost of nonfatal RTI 436 (366–506) Nguyen, Ivers, and others 2013 with a helmet (US$) Average direct acute care cost of nonfatal RTI 559 (416–702) Nguyen, Ivers, and others 2013 without a helmet (US$) Expected increase in treatment cost for each 1 Nguyen, Ivers, and others 2013 US$10 increase in income (%) Income loss (number of weeks) 32 Hoang and others 2008 Mean per capita income, by quintile (US$) 305, 530, 777, 1,185, 2,730 GSO 2010 Distribution of motorcycle ownership by 20, 35, 54, 73, 94 GSO, NIHE, and ORC Macro 2006 quintile (%) Relative risk of death, helmet vs. no helmet 0.58 (0.50–0.79) Liu and others 2008 Relative risk of injury, helmet vs. no helmet 0.31 (0.25–0.66) Liu and others 2008 Per capita cost of policy implementation (US$) 0.29 Chisholm and others 2012 (correspondence from Dan Chisholm) Note: RTI = road traffic injury. Table 11A.1 in the annex provides the detailed rationale and additional sources for selection of point estimates and ranges. as obtained from the Vietnamese Demographic and following the comprehensive helmet policy, the number Health Survey. This assumption also was explored in of deaths and head injuries averted within each quintile sensitivity analyses (GSO, NIHE, and ORC Macro 2006). during the one-year post-policy evaluation period was simulated (annex tables 11A.3–11A.4 and equations 11A.1–11A.4). Consequences for Health To simulate the impact on health consequences, the number of deaths and nonfatal head injuries attribut- Consequences for Cost and Affordability able to motorcycles in the one-year baseline period was The OOP acute care costs averted by the policy were estimated, as was the pre-policy proportion of motorcy- simulated by subtracting the expected OOP costs of cle riders using helmets. Helmet effectiveness (expressed hospitalization in the post-policy period from the as the relative risk of head injury among riders wearing expected OOP cost in the baseline period. The expected helmets compared to riders not wearing helmets) was cost was derived from published estimates on average estimated using published odds ratios. By accounting cost for injury with and without a helmet, taking into for the increase in the proportion of helmeted riders account variations in the severity and type of injury Helmet Regulation in Vietnam: Impact on Health, Equity, and Medical Impoverishment 215 based on helmet use (Nguyen, Ivers, and others 2013). Sensitivity Analysis These changes in costs were then multiplied by the esti- A univariate sensitivity analysis was performed on key mated change in incidence of motorcycle injuries (see model inputs to test their influence on the findings. Upper annex 11A, equation 11A.5 in table 11A.5). and lower bounds for the inputs were obtained from pub- Empirical research has shown variations in the lished studies wherever possible and were otherwise average direct acute care cost of treatment by income derived from available data or plausibly estimated group in Vietnam (Nguyen, Ivers, and others 2013). The (annex 11A, table 11A.1). One critical sensitivity analysis average direct acute care cost of treatment in each explored the impact of substandard helmets in Vietnam, income quintile was derived by combining the estimated accounting for less safe designs (half-head or cap style), quintile-specific monthly income per capita with the failure to meet quality standards, and inadequate fasten- reported 1 percent increase in the cost of treating a trau- ing of chin straps. Each safety deficit was assumed to halve matic brain injury for every US$10 increase in monthly the reduction in relative risk of death or injury provided income per capita (GSO 2010; Nguyen, Ivers, and others by the helmet, and this was combined with the approxi- 2013). Income losses were calculated by multiplying mate population prevalence of each deficit to estimate a monthly per capita income by the Vietnamese average lower bound of population-level helmet effectiveness (see absence from work of eight months following traumatic annex 11A, table 11A.1 and figure 11A.9). brain injury (Hoang and others 2008). Additional sensitivity analyses were performed to Two measures of financial risk protection were evaluate the influence of model input distributional calculated: cases of poverty averted and catastrophic assumptions on the distribution of health and financial health expenditures averted. Both measures reflect the benefits across income quintiles (annex 11A, table 11A.2). reduction in financial hardship that may occur when The distribution of motorcycle deaths and nonfatal inju- an injury is averted or when the cost of treatment is ries across quintiles, the distribution of pre-policy helmet reduced. Cases of poverty averted were defined as the use, and the distribution of post-policy helmet use across number of individuals who, as a result of the helmet quintiles were varied in these analyses, first alone and policy, would no longer fall below the national poverty then by multivariate sensitivity analysis. line because of a traffic injury. In the baseline model, 21 percent of the population is living in poverty (World Bank 2012). Cases of catastrophic health expen- RESULTS ditures averted were defined as the number of people who, as a result of the policy, would no longer be paying The results of the analysis are presented in table 11.2. more than 25 percent of their annual income per capita Assuming that helmets in Vietnam are as effective on direct acute care costs. The threshold for a cata- as helmets in HIC, the simulation estimates that the strophic health expenditure varies depending on the 2007 helmet policy prevented approximately 2,200 literature, but it generally lies between 2.5 percent and deaths and 29,000 head injuries, saved individuals 15 percent of household income or between 10 percent US$18 million in direct acute care costs, and averted and 45 percent of disposable income (Wagstaff and van US$29 million in individual income losses in the year Doorslaer 2003). For a population of P individuals with following its introduction (table 11.2). Countrywide a certain income distribution,1 the number of people implementation of the policy cost the government an injured before the intervention in each quintile was estimated US$24 million, although this cost was offset multiplied by the probability that they would face pov- by revenue arising from the collection of fines and erty or a catastrophic health expenditure. The same enforcement (Chisholm and others 2012). From a estimate was recalculated using the post-intervention government perspective (which accounts for imple- injury rate and costs. Subtraction yielded the number of mentation costs only), the helmet policy cost about cases of poverty or catastrophic health expenditure that US$11,000 per death averted or US$800 per nonfatal were averted in the population (see annex 11A, equation injury averted. From a societal perspective (which 11A.6 in table 11A.5). sums individuals’ OOP direct acute care cost savings, The government’s cost of implementing the compre- individuals’ averted income losses, and government’s hensive helmet legislation in Vietnam was approximated implementation costs), the policy saved approximately by multiplying the estimated costs per capita of imple- US$11,000 per death averted or US$800 per nonfatal menting the legislation in South-East Asia (including injury averted. legislation and program management, media, enforce- The main distributional analysis assumed that the ment, and helmet purchase) by the population of distribution of traffic injury reflected the distribution of Vietnam (Wagstaff and van Doorslaer 2003).2 motorcycle ownership across income quintiles and 216 Injury Prevention and Environmental Health Table 11.2 Estimated Reduction in Death, Injury, and Cost Attributable to the Mandatory Helmet Legislation in Vietnam Pre-policy estimate (attributable to Estimated absolute reduction Estimated relative reduction Indicator motorcycles) (range)a (%) (range)a Deaths 7,400 2,200 (1,000–2,700) 29 (14–37) Nonfatal head injuries 54,100 29,000 (12,700–44,500) 54 (23–82) Direct acute care costs for nonfatal head 35 18 (8–28) 52 (24–81) injuries (US$, millions) Income losses following death or nonfatal 63 29 (11–40) 46 (18–64) head injury (US$, millions) Direct acute care costs plus income 98 48 (24–72) 49 (24–73) losses (US$, millions) a. Values in parentheses represent lower and upper bounds obtained on univariate sensitivity analyses. Figure 11.1 Deaths and Nonfatal Head Injuries Averted as a Result of the Helmet Policy in Vietnam, by Income Quintile (1 = Poorest, 5 = Richest) Deaths averted Nonfatal head injuries averted 800 12,000 700 10,000 600 8,000 500 Injuries averted Deaths averted 400 6,000 300 4,000 200 2,000 100 0 0 1 2 3 4 5 1 2 3 4 5 Wealth quintile Wealth quintile found that the wealthiest quintiles own the greatest The helmet legislation likely has averted poverty for per- number of motorcycles and thus accrue a larger share of sons in the second and third income quintiles, amount- the health and financial benefits (in absolute terms) ing to nearly 11,000 cases of poverty averted. Persons in from the 2007 helmet policy (figure 11.1). With regard to the first quintile are poor already, and the cost is not so financial risk protection, traffic injury is so expensive to high that those in the fourth and fifth quintiles will be treat that any injury averted also would avert cata- thrust into poverty. strophic health expenditures (figures 11.2 and 11.3). In Table 11.2 presents the lower and upper values other words, both before and after the policy, traffic obtained in a univariate sensitivity analysis. The sen- injury leads to health expenditures that exceed 25 sitivity analysis that accounted for substandard and percent of per capita income, amounting to more than inadequately fastened helmets yielded the lowest esti- 22,000 cases of catastrophic health expenditure averted. mates of deaths and injuries averted, a finding with Helmet Regulation in Vietnam: Impact on Health, Equity, and Medical Impoverishment 217 Figure 11.2 Out-of-Pocket Costs Averted as a Result of the Helmet under conditions of perfectly equitable pre-policy Policy in Vietnam, by Income Quintile (1 = Poorest, 5 = Richest) motorcycle injury and death. This finding is amplified 2012 US$, millions when occurring in conjunction with highly inequitable pre-policy helmet use (with highest use among the 30 wealthy) and perfectly equitable post-policy helmet use 25 (annex 11A, figures 11A.6–11A.8). Costs averted 20 15 DISCUSSION 10 Assuming that helmets in Vietnam are as effective as 5 those in HIC, the 2007 comprehensive helmet policy prevented approximately 2,200 deaths and 29,000 head 0 injuries, saved individuals US$18 million in direct acute 1 2 3 4 5 care costs, and averted US$29 million in individual Wealth quintile income losses in the year following its introduction. The Indirect costs averted Direct costs averted combination of anticipated health and financial benefits makes a comprehensive helmet policy strongly prefera- ble to the pre-policy status quo. These findings suggest that similar comprehensive legislation and enforcement Figure 11.3 Financial Risk Protection Afforded as a Result of should be considered in countries where motorcycles are the Helmet Policy in Vietnam, by Income Quintile (1 = Poorest, pervasive, yet helmet use is less common. 5 = Richest) In the simulations, the relative reduction in motorcycle crash deaths fell from 29 percent to 14 percent after 8,000 accounting for the proliferation of less-effective helmets in 6,000 Vietnam. Policy makers wanting to enact an effective com- 4,000 prehensive helmet law might consider making provisions 2,000 for adequate regulatory enforcement of manufacturers, retailers, and motorcycle riders to ensure that helmets are 0 1 2 3 4 5 of adequate quality and appropriately fastened. Wealth quintile The results of the ECEA suggest that the wealthy Cases of poverty averted likely accrued a large share of the absolute health and Cases of catastrophic health expenditures averted financial benefits resulting from the helmet use legisla- tion. This finding was dependent on the assumption Note: No data are available for cases of poverty averted for quintiles 1, 4, and 5. that the risk of RTI tracked with motorcycle ownership. In contrast, under all of the conditions tested, the leg- islation was likely to have prevented a greater number clear implications for policy and enforcement. Deaths, of cases of poverty resulting from motorcycle accidents injuries, and OOP costs averted were extremely among the near poor and middle-income quintiles. sensitive to variation in the proportion of motorcycle This supports the conclusion that injury prevention injuries anticipated to cause head injury. Direct costs also is poverty prevention among individuals of lesser of acute care averted also were highly sensitive to wealth. In settings with universal health insurance, cost variation in the average costs of acute care for crash savings from a comprehensive helmet policy (poten- victims with and without helmets. The univariate tially substantial, as the wealthy are known to use a sensitivity analyses, along with those for poverty and disproportionate share of public health care) might catastrophic health expenditures averted, are pre- also be freed up for use on other health priorities sented graphically in the annex (figures 11A.1–11A.5 (Wagstaff 2010). and 11A.9). The validity of the model’s estimates are supported Distributional sensitivity analyses demonstrated by the results of prior research. The findings antici- that the distribution of health benefits is highly sensi- pate a 29 percent reduction in deaths from motorcy- tive to variation in the pre-policy distribution of cle accidents and a 17 percent reduction in deaths motorcycle injury across quintiles. Both health and from all traffic accidents. These results are similar to financial benefits accrue disproportionately to the poor the 36 percent reduction in deaths from motorcycle 218 Injury Prevention and Environmental Health accidents generally anticipated with helmet legisla- The potential for impoverishment resulting from helmet tion and the 18 percent reduction in deaths from infraction fines was assumed to be uncommon and rela- all traffic accidents reported in Vietnam in the year tively inconsequential. following introduction of the helmet legisla- tion (Passmore, Nguyen, and others 2010; Passmore, Tu, and others 2010). The results also are in har- Cost-Effectiveness mony with the results of regional evaluations of The results suggest that Vietnam’s 2007 helmet legislation helmet use legislation (Chiu and others 2000; was cost-effective. Large health and financial benefits Ichikawa, Chadbunchachai, and Marui 2003; Tsai and accrued to the wealthy, yet the policy also provided sig- Hemenway 1999). nificant health benefits and substantial financial risk protection to Vietnam’s poorest citizens. As countries develop and more individuals acquire motorcycles, we are Limitations likely to see a reversal in the distribution of benefits from Several limitations are related to the model and its helmet legislation. Increased ridership among the poor inputs. First, the modeling study estimated the antici- will increase the risk of injury, yet improved helmets pated effectiveness and cost-effectiveness of the 2007 are likely to be worn only by the wealthy. The issues of comprehensive helmet policy in Vietnam but did not road traffic safety are only going to grow as motorcycles directly measure the benefits or costs. The published become more accessible throughout the region. academic literature has not yet articulated the observed Fortunately, most countries have implemented helmet benefits and costs of this policy, despite the crucial legislation. Others, such as Cambodia, have recently importance of these values for evaluating policy success. expanded their policies to include passengers and chil- Second, many of the inputs used (including pre-policy dren. The implications of road safety policy go well deaths and injuries attributable to motorcycles, acute beyond health, as our analysis has shown. Policy makers care costs, and policy implementation costs) were not wishing to account for such effects may want to use ECEA directly available and had to be derived or estimated to understand the likely influence of policy on equity. from published reports. The use of academic and non- governmental reports rather than government surveil- lance data improves the quality of data but diminishes AUTHOR CONTRIBUTIONS AND the local applicability of the results. Third, the main analysis ignored the influence of substandard helmets in ACKNOWLEDGMENTS Vietnam because of the absence of reliable estimates of This chapter is based on the following article: Olson, Z., the relative effectiveness of the substandard helmets, J. A. Staples, C. N. Mock, N. P. Nguyen, A. M. Bachani, particularly in a setting with relatively low traffic speeds and others. 2016. “Helmet Regulation in Vietnam: impact (Ackaah and others 2013).3 on Health, Equity, and Medical Impoverishment.” Injury The analysis was also limited by several assumptions Prevention 1–6. Epub January 4, 2016. doi:10.1136 made in constructing the model. The assumption that /injuryprev-2015-041650. the number of motorcycles on the road was the same Zachary Olson, John A. Staples, Charles N. Mock, before and after the policy rendered the estimated bene- Rachel Nugent, and Stéphane Verguet were responsible fits more conservative, interpretable, and generalizable for the design of the study. Olson and Staples had full (Le and Blum 2013). Changes in the prevalence of speed- access to all of the data and take responsibility for the ing and alcohol use, increased enforcement of traffic integrity of the data and the accuracy of the data analysis. laws not related to helmets, changes in road maintenance They also prepared the initial draft of the manuscript. All and congestion, and other secular trends were ignored. authors contributed to data interpretation, revised the The cost estimates did not account for a potential manuscript, and provided approval for submission. increase in nonhead injuries among riders whose lives The study was funded with support from the Bill and were saved by helmet use. The simulated number of Melinda Gates Foundation to the Disease Control deaths averted represents less than 10 percent of the Priorities Network. Olson was funded in part by National simulated number of injuries averted, so the potential Institute of Child Health and Human Development grant increase in nonhead injuries was anticipated to be mini- T32-HD007275. No funding organization was involved mal. Lastly, insufficient information made estimating the in the design and conduct of the study; collection, man- higher costs for individuals and the higher revenue for agement, analysis, and interpretation of the data; prepa- government resulting from improved enforcement and ration, review, and approval of the manuscript; or higher fines resulting from the helmet policy impossible. decision to submit the manuscript for analysis. Helmet Regulation in Vietnam: Impact on Health, Equity, and Medical Impoverishment 219 The authors thank Dan Chisholm, Jonathan GSO, NIHE (National Institute of Hygiene and Epidemiology), Passmore, Dean Jamison, Carol Levin, Beth E. Ebel, and ORC Macro. 2006. Vietnam Population and AIDS Mara K. Hansen, Jenny Nguyen, Clint Pecenka, and Alex Indicator Survey 2005. Calverton, MD: GSO, NIHE, and Quistberg for their insightful comments. ORC Macro. Hoang, H., T. Pham, T. Vo, P. Nguyen, C. Doran, and others. 2008. “The Costs of Traumatic Brain Injury Due to ANNEX Motorcycle Accidents in Hanoi, Vietnam.” Cost Effectiveness and Resource Allocation 6 (1): 17. The annex to this chapter is as follows. It is available at Hung, D. V., M. R. Stevenson, and R. Q. Ivers. 2006. “Prevalence http://dcp-3.org/environment. of Helmet Use among Motorcycle Riders in Vietnam.” Injury Prevention 12 (6): 409–13. • Annex 11A. Supplementary Tables and Figures on the ———. 2008. “Motorcycle Helmets in Vietnam: Ownership, Effects of Vietnam’s Mandatory Helmet Legislation Quality, Purchase Price, and Affordability.” Traffic Injury Prevention 9 (2): 135–43. on Health, Equity, and Medical Impoverishment Ichikawa, M., W. Chadbunchachai, and E. Marui. 2003. “Effect of the Helmet Act for Motorcyclists in Thailand.” Accident Analysis and Prevention 35 (2): 183–89. NOTES Le, L. C., and R. W. Blum. 2013. “Road Traffic Injury among World Bank Income Classifications as of July 2014 are as fol- Young People in Vietnam: Evidence from Two Rounds of lows, based on estimates of gross national income (GNI) per National Adolescent Health Surveys, 2004–2009.” Global capita for 2013: Health Action 6: 1–9. Liu, B., R. Ivers, R. Norton, S. Boufous, S. Blows, and others. • Low-income countries (LICs) = US$1,045 or less 2008. “Helmets for Preventing Injury in Motorcycle Riders.” • Middle-income countries (MICs) are subdivided: Cochrane Database of Systematic Reviews 1: CD004333. a) lower-middle-income = US$1,046 to US$4,125 National Traffic Safety Committee. 2014. “Vietnam Road b) upper-middle-income (UMICs) = US$4,126 to US$12,745 Traffic Injury Official Estimates.” National Traffic Safety • High-income countries (HICs) = US$12,746 or more. Committee, Hanoi. Ngo, A. D., C. Rao, N. P. Hoa, D. G. Hoy, K. T. Q. Trang, and 1. A proxy for individual income was extracted from the others. 2012. “Road Traffic Related Mortality in Vietnam: income distribution of Vietnam derived from its GDP per Evidence for Policy from a National Sample Mortality capita (US$1,200 in 2012 US$) and its Gini index (0.36) Surveillance System.” BMC Public Health 12 (July): 561. (Salem and Mount 1974; World Bank 2012). Nguyen, H., R. Q. Ivers, S. Jan, A. L. Martiniuk, Q. Li, and oth- 2. This number was derived by Dan Chisholm, Jonathon ers. 2013. “The Economic Burden of Road Traffic Injuries: Passmore, and Nguyen Phuong Nam using the same model Evidence from a Provincial General Hospital in Vietnam.” cited (Chisholm and others 2012). Injury Prevention 19 (2): 79–84. 3. See also Viet Nam News 2014. Nguyen, H. T., J. Passmore, P. V. Cuong, and N. P. Nguyen. 2013. “Measuring Compliance with Viet Nam’s Mandatory Motorcycle Helmet Legislation.” International Journal of Injury Control and Safety Promotion 20 (2): 192–96. REFERENCES Passmore, J., L. H. Nguyen, N. P. Nguyen, and J. M. Olive. Ackaah, W., F. Afukaar, W. Agyemang, T. Thuy Anh, A. R. Hejar, 2010. “The Formulation and Implementation of a National and others. 2013. “The Use of Non-Standard Motorcycle Helmet Law: A Case Study from Viet Nam.” Bulletin of the Helmets in Low- and Middle-Income Countries: World Health Organization 88 (10): 783–87. A Multicentre Study.” Injury Prevention 19 (3): 158–63. Passmore, J., N. T. Tu, M. A. Luong, N. D. Chinh, and N. P. Nam. Chisholm, D., H. Naci, A. A. Hyder, N. T. Tran, and M. Peden. 2010. “Impact of Mandatory Motorcycle Helmet Wearing 2012. “Cost-Effectiveness of Strategies to Combat Road Legislation on Head Injuries in Viet Nam: Results of a Traffic Injuries in Sub-Saharan Africa and South East Asia: Preliminary Analysis.” Traffic Injury Prevention 11 (2): 202–5. Mathematical Modelling Study.” BMJ 344: e612. Pham, K. H., W. X. L. Thi, D. J. Petrie, J. Adams, and Chiu, W. T., C. Y. Kuo, C. C. Hung, and M. Chen. 2000. “The C. M. Doran. 2008. “Households’ Willingness to Pay for Effect of the Taiwan Motorcycle Helmet Use Law on Head a Motorcycle Helmet in Hanoi, Vietnam.” Applied Health Injuries.” American Journal of Public Health 90 (5): 793–96. Economics and Health Policy 6 (2–3): 137–44. Government of Vietnam. 2007. “Resolution on a Number Salem, A. B. Z., and T. D. Mount. 1974. “A Convenient of Urgent Countermeasures to Curb Traffic Safety and Descriptive Model of Income Distribution: The Gamma Alleviate Traffic Congestion.” Resolution 32/2007/NQ-CP Density.” Econometrica 42 (6): 1115–27. (June 29, 2007), Government of Viet Nam, Hanoi. Tien, T. V., H. T. Phuong, I. Mathauer, and N. T. K. Phuong. GSO (General Statistics Office). 2010. Vietnam Household Living 2011. A Health Financing Review of Viet Nam with a Focus on Standards Survey. Hanoi: Statistical Publishing House. Social Health Insurance. Geneva: World Health Organization. 220 Injury Prevention and Environmental Health Tsai, M., and D. Hemenway. 1999. “Effect of the Mandatory ———. 2009. “Road Safety Status: Country Profile, Vietnam.” Helmet Law in Taiwan.” Injury Prevention 5 (4): 290–91. WHO, Geneva. http://www.who.int/violence_injury Verguet, S., R. Laxminarayan, and D. T. Jamison. 2014. _prevention/road_safety_status/country_profiles/viet_nam.pdf. “Universal Public Finance of Tuberculosis Treatment in ———. 2010. World Health Statistics 2010. Geneva: WHO. India: An Extended Cost-Effectiveness Analysis.” Health http://www.who.int/whosis/whostat/EN_WHS10_Full Economics 24 (3): 318–32. .pdf. Verguet, S., S. Murphy, B. Anderson, K. A. Johansson, R. Glass, ———. 2013a. Global Status Report on Road Safety 2013. and others. 2013. “Public Finance of Rotavirus Vaccination Geneva: WHO. http://www.who.int/violence_injury in India and Ethiopia: An Extended Cost-Effectiveness _prevention/road_safety_status/2013/en/index.html. Analysis.” Vaccine 31 (42): 4902–10. ———. 2013b. “Road Safety Fact Sheet, Western Pacific Region.” Verguet, S., Z. D. Olson, J. B. Babigumira, D. Desalegn, K. A. WHO, Geneva. http://www.wpro.who.int/mediacentre Johansson, and others. 2015. “Health Gains and Financial /factsheets/fs_20130627/en/. Risk Protection Afforded by Public Financing of Selected ———. 2013c. “Road Safety Status: Country Profile, Vietnam.” Interventions in Ethiopia: An Extended Cost-Effectiveness WHO, Geneva. http://www.who.int/violence_injury Analysis.” The Lancet Global Health 3 (5): e288–96. _prevention/road_safety_status/2013/country_profiles Viet Nam News. 2014. “Police Struggle to Identify Substandard /viet_nam.pdf. Helmets.” Viet Nam News, July 2. http://vietnamnews.vn ———. 2015. Study on Motorcycle Helmet Quality in Viet /society/256910/police-struggle-to-identify-substandard Nam. WHO and Hanoi School of Public Health. http:// -helmets.html. www.wpro.who.int/vietnam/topics/injuries/helmet_quality Wagstaff, A. 2010. “Measuring Financial Protection in _vietnam.pdf. Health.” In Performance Measurement for Health System World Bank. 2012. World Development Indicators. Washington, Improvement: Experiences, Challenges, and Prospects, edited by DC: World Bank. http://data.worldbank.org/data-catalog P. C. Smith, E. Mossialos, I. Papanicolas, and S. Leatherman, /world-development-indicators. 114–37. Cambridge, U.K.: Cambridge University Press. World Bank and IHME (Institute for Health Metrics and Wagstaff, A., and E. van Doorslaer. 2003. “Catastrophe and Evaluation). 2014. Transport for Health: The Global Burden Impoverishment in Paying for Health Care: With Applications of Disease from Motorized Road Transport. Seattle, WA: to Vietnam 1993–1998.” Health Economics 12 (11): 921–34. IHME. WHO (World Health Organization). 2004. World Report on Yu, W. Y., C. Y. Chen, W. T. Chiu, and M. R. Lin. 2011. Road Traffic Injury Prevention. Geneva: WHO. http://www “Effectiveness of Different Types of Motorcycle Helmets .who.int/violence_injury_prevention/publications /road and Effects of Their Improper Use on Head Injuries.” _traffic/world_report/en/. International Journal of Epidemiology 40 (3): 794–803. Helmet Regulation in Vietnam: Impact on Health, Equity, and Medical Impoverishment 221 Chapter 12 Household Energy Interventions and Health and Finances in Haryana, India: An Extended Cost-Effectiveness Analysis Ajay Pillarisetti, Dean T. Jamison, and Kirk R. Smith INTRODUCTION DALYs per year (Forouzanfar and others 2015).1 An Approximately 40 percent of the world’s population estimated 4 percent of the deaths occur in children relies on solid fuels, including wood, dung, grass, crop under age five years because of pneumonia, which residues, and coal, for cooking (Bonjour and others overall accounts for 12 percent of total child deaths in 2013). Household air pollution (HAP) arising from this India. use of solid fuels results in 3 million to 4 million deaths Attempts to reduce this burden fall into two primary yearly from acute lower respiratory infection (ALRI) in categories: (1) those that seek to make biomass combus- children and chronic obstructive pulmonary disease tion cleaner and more efficient, and (2) those that seek (COPD), ischemic heart disease (IHD), stroke, and to replace biomass use with liquid fuels or electricity lung cancer in adults. This burden constitutes approxi- (Foell and others 2011; Smith and Sagar 2015). Private mately 5 percent of global mortality, ranking highest and public sector actors have taken action in India to among all environmental risk factors contributing to reduce this large burden of disease. Private sector endeav- global ill health (Forouzanfar and others 2015; Smith ors include research, development, marketing, and dis- and others 2014). tribution of biomass stoves by large multinational In India, the reliance on solid fuels and the esti- corporations, such as Philips and BP, and smaller Indian mated related burden of disease are pronounced. An and international firms, such as Envirofit, Greenway, estimated 770 million individuals—approximately First Energy, BioLite, and Prakti. In all cases, the evalua- 70 percent of the total population (Government of tions of the viability of these interventions for long-term India 2011)—living in 160 million households con- use, which would be required to reduce exposures and tinue to use solid fuels as a primary energy source for thus the health burden, have been mixed (Brooks and cooking (Venkataraman and others 2010). Among all others 2016; Pillarisetti and others 2014; Sambandam risk factors contributing to ill health in India, exposure and others 2015). to HAP from cooking ranks second for mortality, with The government of India has undertaken a number approximately 925,000 premature deaths yearly; it of policy initiatives to address HAP through improved ranks third for lost disability-adjusted life years biomass combustion, beginning in the 1980s with a (DALYs), amounting to approximately 25 million lost failed National Programme on Improved Chulhas Corresponding author: Ajay Pillarisetti, Division of Environmental Health Sciences, School of Public Health, University of California, Berkeley, CA, United States; ajaypillarisetti@gmail.com. 223 (Kishore and Ramana 2002) and continuing in 2010 intervention and its health and financial consequences with a National Biomass Cookstoves Initiative. More across income groups. Verguet, Laxminarayan and recently, two innovative programs—the Give It Up Jamison (2015), for example, looked at public finance (GIU) and Smokeless Village (SV) campaigns—are and enhanced borrowing capacity as policies to affect seeking to bring clean cooking via liquefied petroleum tuberculosis treatment in India. Verguet and others gas (LPG) to the rural poor (Smith and Sagar 2015). (2015) assessed the consequences of a policy to increase Both GIU, which encourages better-off Indian house- tobacco taxes in China. Including distributional analysis holds to voluntarily give up their LPG subsidies and by income quintile enables novel policy evaluations, as redirects those subsidies one-for-one to below-poverty- well as an evaluation of the GIU campaign. line (BPL) families, and SV, which connects every This ECEA focuses on policies to reduce exposure to household in a village to LPG, occur in close collabora- HAP in Haryana, India. This state has a population of tion with India’s three national oil companies. In mid- 20 million, about 55 percent of whom use solid fuels for 2016, Indian Prime Minister Narendra Modi introduced cooking, although significant heterogeneity exists between Pradhan Mantri Ujjwala Yojana (Ujjwala), a program to both rural and urban areas and between available datasets extend the GIU and SV campaigns by making free LPG for analyses. In addition, we benefit from the availability of connections available to all BPL households. This pol- published continuous exposure-response relationships for icy will affect approximately 50 million households.2 HAP-related diseases and a fuel gathering–based time These programs have the potential to substantially metric, allowing us to quantify the potential earnings reduce the mortality and morbidity associated with the gained by use of a stove that improves fuel efficiency.3 use of solid fuels for cooking, if one assumes near-com- plete transitions to clean fuels (Smith and Sagar 2015). This chapter describes an extended cost-effectiveness Review of Economic Analyses of Household Energy analysis (ECEA) of policies designed to promote uptake Interventions of hypothetical HAP control interventions aligned with Existing peer-reviewed literature on the costs, benefits, three national government programs: and cost-effectiveness of HAP interventions is sparse (Jeuland, Pattanayak, and Bluffstone 2015). A limited • A low-cost, mud chimney stove, as was promoted in number of global (Hutton and others 2006; Hutton, the National Programme on Improved Chulhas that Rehfuess, and Tediosi 2007; Jeuland and Pattanayak operated from about 1983 to 2002 (We evaluate this 2012; Mehta and Shahpar 2004) and geography-specific program under the same current conditions as the (Arcenas and others 2010; Aunan and others 2013; Malla other programs.) and others 2011; Pant 2011) economic evaluations exists. • An advanced combustion cookstove, like that A short review follows. being promoted in the current National Biomass Mehta and Shahpar (2004) found wide variation in Cookstoves Initiative the cost-effectiveness of improved stoves and LPG and • A transition to LPG being promoted in the national propane interventions across the World Health Give It Up campaign. Organization subregions, but their analysis did not con- sider the cost of illness and treatment or potential non- Our scenarios simplify complex behavioral issues by health benefits of transitioning to cleaner cooking. assuming full use of all intervention stoves in order to Hutton and others (2006) and Hutton, Rehfuess, and estimate best-case health and welfare benefits of clean Tediosi (2007) performed a global cost-benefit analysis cooking transitions. We evaluate the sensitivity of our of eight scenarios that reduced exposure through a tran- use assumption in annex 12A. Our goal is to indicate the sition to either clean fuels or clean biomass stoves and types of policy-relevant analyses that are possible using considered benefits including improved health, decreased ECEA and the magnitude of potential benefits of LPG emissions of climate-altering pollutants, fewer lost work adoption. days, and time savings. They found that both the clean Traditional economic cost-effectiveness analyses, such fuel transition and the improved stove transition had as that by Mehta and Shahpar (2004), focus on the U.S. favorable cost-benefit ratios of 4.3 and approximately 60, dollars spent per death or per DALY averted. ECEA also respectively. Unlike Hutton and others (2006) and Mehtha considers the financial implications of policies across and Shahpar (2004), both of which used regional scale wealth strata of a population (introduced in Verguet, inputs, Jeuland and Pattanayak (2012) modeled costs and Laxminarayan, and Jamison 2015), in this case, by income benefits from household and societal perspectives for quintile. ECEAs assess the consequences of financial or clean fuel and clean stove technologies. They found that other policies that influence the aggregate uptake of an transitions away from traditional cooking yield variable 224 Injury Prevention and Environmental Health results; some interventions have high probabilities of net As part of the GIU campaign, in addition to the redis- costs to households and societies. Their modeling indi- tribution of the subsidy to the poor, the corporate social cates that LPG, kerosene, and improved charcoal stoves responsibility funds of the three national oil companies have the highest probability of net positive benefits at were used to cover the upfront costs of the regulator and household and societal scales. They note that the findings cylinder deposits—a subsidy of approximately 2,000 are sensitive to a number of factors, including emission rupees (Rs) (approximately US$30) made available to rates and fuel costs. BPL households, an official category that varies some- Cost-benefit analysis has been applied in a number what by state. Some states also provide a stove to families of geography-specific studies, including in Nepal receiving the GIU benefit. According to the Ministry of (Malla and others 2011; Pant 2011), China (Aunan and Petroleum (2016), 10 million middle-income house- others 2013), the Western Pacific region (Arcenas and holds had given up their LPG subsidy as of May 1, 2016 others 2010), and Kenya and Sudan (Malla and others (Smith and Sagar 2015). Ujjwalla extends this by provid- 2011). Malla and others (2011) found that across three ing the same subsidy to all BPL households through use separate interventions in Nepal (smoke hood), Kenya of a new allotment of about US$1.2 billion of Indian (LPG or smoke hoods), and Sudan (LPG), benefits government funds (Times of India 2016). exceeded costs over the 10-year intervention period, although there was significant heterogeneity among study sites. They note, however, that the effect of mon- Estimation of Health Benefits of Clean Cooking etized health benefits was relatively small across all sites, Understanding improved health attributable to a HAP- compared to time and fuel savings. In China, Aunan reducing intervention, such as a transition to LPG, relies and others (2013) evaluated transitions in no-chimney on complex exposure science and behavioral processes. or chimney stove homes to either second-generation The relationship between exposure to HAP and health is improved cookstoves or community-scale pellet stoves. nonlinear and is described through a set of integrated- In all cases, benefit-cost ratios were positive (central exposure response (IER) (Burnett and others 2014) estimate range of 3.3–14.7), and the largest ratios curves that link exposure to particulate matter with an occurred by switching away from chimneyless stoves. effective diameter of less than 2.5 micrometers (PM2.5, a Only health benefits were monetized. Similarly, key component of combustion-generated air pollution) Pant (2011) and Arcenas and others (2010) used cost- with a number of health endpoints. IERs currently exist of-illness and value of a statistical life, respectively, to for ALRI, IHD, stroke, lung cancer, and COPD. The IERs assess the effect of household energy transitions by integrate exposure data from a range of PM2.5 sources, using survey data. Pant (2011) modeled the effect of a including HAP, active tobacco smoking, secondhand transition from dung fuel to biogas, noting the health tobacco smoke, and ambient air pollution. cost per household—driven by medication expenses— The continuous nature of the exposure-response to be 61.3 percent higher in dung-burning households relationships allows modeling of the potential health than the cost of fuel in biogas households. benefits of a reduction in exposure to PM2.5 attributable to a specific intervention by disease type (Pillarisetti, Mehta, and Smith 2016). However, quantifying exposure Clean Fuel Intervention Costs reductions is challenging and relies on either expensive Interventions considering either fuels—such as LPG or and intrusive monitoring of individuals or sophisticated natural gas—or electricity must contend with both modeling of pollution levels and time-activity patterns. upfront and recurrent costs. In India, before 2015, every Exposure reductions are complicated by issues of com- cylinder of LPG sold to household customers was subsi- pliance or stove stacking, the phenomenon of continuing dized at the point of sale, regardless of the income of the to use the traditional cooking technology even though a household. In 2015, the government announced that new technology or fuel has come into the household cylinders would be sold at full price to all consumers, but (Brooks and others 2016; Johnson and Chiang 2015; that households would have subsidies transferred directly Pillarisetti and others 2014; Ruiz-Mercado and others to their bank accounts—the PAHAL scheme (Tripathi, 2011; Sambandam and others 2015; Smith and others Sagar, and Smith 2015). Among others goals, this policy 2015). However, this situation is not unusual in health sought to prevent small and medium enterprises from interventions, where provision of a healthier technology being able to buy subsidized fuel intended for house- needs to be followed by policies to encourage long-term holds from the black market. Current subsidies are use and elimination of the unhealthy behavior (for approximately one-fourth of the cost of a cylinder, example, with condoms, bednets, and latrines). In a although they vary with the market price of LPG. sense, then, the analyses here represent an efficacy Household Energy Interventions and Health and Finances in Haryana, India: An Extended Cost-Effectiveness Analysis 225 approach—the best that could be achieved for each provided by the Institute for Health Metrics and intervention. Evaluation in the Global Burden of Disease Study 2010 To address these issues and others, we have developed is used to bound estimates of averted DALYs and deaths (1) an online tool that uses the IERs and relevant back- attributable to an intervention. ground data to estimate the potential effect of an inter- We also evaluated two additional modes of distribut- vention known as the Household Air Pollution ing background disease (annex 12A). In the first, dis- Intervention Tool (HAPIT) (Pillarisetti, Mehta, and ease data were split on the basis of the overall percentage Smith 2016) and (2) standard protocols to use HAPIT to of Haryana’s population in each quintile, calculated by estimate averted ill health (Smith and others 2014). multiplying the number of households per quintile by the number of people per household. In the second, we assumed (1) that all quintiles had equal populations METHODS and age distributions and (2) that solid fuel use (SFU) linearly decreased as wealth increased, beginning at Estimation of Reductions in Morbidity and Mortality 90 percent in quintile (Q) 1 and ending at 60 percent Resulting from HAP Interventions SFU in Q5. This chapter uses a modified version of HAPIT (based on the version described in Pillarisetti, Mehta, and Smith [2016] but modified to facilitate evaluation of Evaluation of the Consequences of Policy multiple scenarios at a subnational scale) to estimate We evaluate the effect of policies leading to 100 percent the averted deaths and DALYs attributable to an inter- penetration and adoption of three interventions—a vention over a five-year period. Briefly, HAPIT uses simple mud chimney stove; a fan-assisted, forced-draft national background health data and the methods and semi-gasifier, also known as a blower stove; and an databases developed as part of the Comparative Risk expansion of LPG—on exposure to PM2.5 and subse- Assessment (Lim and others 2012), a component of the quent ill health in Haryana, India (table 12.2). Although Global Burden of Disease Study 2010 (GBD 2010) each scenario is grounded in either past or ongoing (Lozano and others 2012), to determine pre- and policy initiatives (discussed in the introduction), we post-intervention population attributable fractions. focus on a simulation of potential benefits of these The burden of disease averted can then be determined policies under aspirational conditions. We did, how- by multiplying the background disease-specific burden ever, assess the sensitivity of our findings to the by the difference in population attributable fractions. assumption of full adoption (annex 12A) by modeling Notably, therefore, HAPIT incorporates exposure- a scenario with high adoption of chimney stoves (90 response functions for five separate diseases associated percent) and moderate adoption of blower stoves (65 with air pollution in recent international assessments percent) and LPG stoves (50 percent). based on synthesizing results from multiple individual Simple mud chimney stoves cost approximately epidemiological studies in a number of countries. It US$10 (Dutta and others 2007), while blower stoves estimates the effect of interventions based on the back- cost approximately US$60. We assume that chimney ground conditions of each of the diseases in the country stoves have low maintenance costs and work for one considered (in this case, India). Pillarisetti, Mehta, and year and then provide no benefit, which is consistent Smith (2016) provide a detailed explanation of HAPIT with surveys in India. Similarly, blower stoves have low and its underlying calculations. yearly maintenance costs, but they need to be replaced Background disease data for Haryana were not read- once every three years. The transition to LPG incurs a ily accessible. Instead, underlying disease burden data for number of costs, including the cost of the LPG stove India from the GBD 2010 were scaled by the proportion (approximately US$20), and the connection fee, secu- of the population living in Haryana. To estimate back- rity deposit, and administrative costs for the first cylin- ground disease characteristics by income quintile (table der (approximately US$30). Cylinder refills cost 12.1) in Haryana, we distributed premature deaths for approximately US$8.70 per cylinder unsubsidized and children and adults and DALYs according to the fraction US$6.60 per cylinder subsidized. Families use approxi- of all solid fuel–using households in Haryana residing in mately nine cylinders per year, on average, across India. a specific income quintile, as determined through analy- Total costs to the government are described in table 12.5 sis of the Indian Human Development Survey (IHDS) later in this chapter. 2005–06 (Barik and Desai 2014; Desai, Vanneman, and Using exposure models developed with data from National Council of Applied Economic Research 2005). India, we assume the pre-intervention exposure to PM2.5 The uncertainty in the background disease estimates for adults is 337 micrograms per cubic meter (μg/m3) 226 Injury Prevention and Environmental Health Household Energy Interventions and Health and Finances in Haryana, India: An Extended Cost-Effectiveness Analysis Table 12.1 Background Disease Burden in India and in Haryana, India, by Income Quintile ALRIa COPDb IHDb Lung Cancerb Strokeb DALYs DALYs DALYs DALYs DALYs Solid fuel Population Deaths (thousands) Deaths (thousands) Deaths (thousands) Deaths (thousands) Deaths (thousands) use (%) (millions) India 200,000 17,000 910,000 26,000 1,100,000 26,000 83,000 2,100 610,000 12,000 63 1,000 Haryana 4,600 400 21,000 600 26,000 600 1,900 50 14,000 280 552 20 Q1 (Poorest) 1,000 90 4,600 130 5,700 130 420 10 3,100 60 89 Q2 1,000 90 4,800 140 6,000 140 440 10 3,200 65 90 Q3 1,000 90 4,600 130 5,700 130 420 10 3,100 60 88 4 Q4 780 70 3,600 100 4,400 100 320 10 2,400 50 66 Q5 (Wealthiest) 730 60 3,400 95 4,200 100 310 10 2,200 45 62 Sources: Global Burden of Disease Study 2010, India country profile (IHME 2015) (disease burden data); IHDS 2005–06 (Desai, Vanneman, and National Council of Applied Economic Research 2005) (population data); Census of India 2011 (Government of India 2011) (solid fuel use). Note: ALRI = acute lower respiratory infection; COPD = chronic obstructive pulmonary disease; DALYs = disability-adjusted life years; IHD = ischemic heart disease; Q = quintile. a. ALRI in children under age five years. Apportioned by the percentage of all solid fuel–using households in each quintile. b. Chronic outcomes in adults. 227 Table 12.2 Potential Interventions in Haryana, India Existing Proposed Exposure Reduction in coverage,a % coverageb reduction biomass fuel Intervention Description Target population (quintile) (%) (%) use (%) Chimney stove A simple mud-brick 3.4 × 106 households 11 (Q1), 10 (Q2), 100 50 15 chimney stove with two in Haryana 12 (Q3), 24 (Q4), potholes 2.6 × 106 households 28 (Q5) Blower stove A single pothole using unclean fuelsc 100 63 42 semi-gasifier stove LPG Fuel stored as liquid 100 90 100 under slight pressure, burned as a gas Sources: IHDS 2005–06 (Desai, Vanneman, and National Council of Applied Economic Research 2015) (chimney stove, blower stove, and LPG); Census of India 2011 (Government of India 2011) (chimney stove); Bailis and others 2007 (chimney stove); Sambandam and others 2015 (blower stove). Note: LPG = liquefied petroleum gas. a. Coverage equals the percentage of households in Haryana currently using an equivalent or better technology. For LPG, this includes households using LPG, electricity, or biogas but does not indicate exclusive use of these clean cooking technologies. b. This comprises the total population (that does not currently have an equivalent or better cooking technology) to cover by a specific intervention. c. Unclean fuels include the following Census of India 2011 categories: firewood, crop residue, and cowdung cake; coal, lignite, and charcoal; and kerosene. Eighty-five percent of these households are rural. and 150 μg/m3 for children (Balakrishnan and others lower household medical expenditures. Expenditures 2013; Northcross and others 2010; Pillarisetti, Mehta, averted are based on the probability of seeking care for and Smith 2016; Smith and others 2014). We scale the acute and chronic conditions and the combined inpa- central estimate of exposure by the respective exposure tient and outpatient costs of such visits, including reduction (table 12.2) attributable to a given interven- drugs, hospital visits, and transportation to and from tion. For this analysis, we estimate intervention costs and clinics. To calculate the expenditure averted by com- benefits across five years, child health gains accrue plete adoption of an intervention, we scale the cost of instantly at the start of each year, and adult health gains hospital or doctor visits and related expenditures by the are weighted using the U.S. Environmental Protection relative reduction in DALYs attributable to an interven- Agency cessation lag (U.S. EPA 2004) model. For sim- tion separately for acute (ALRI) and chronic (COPD, plicity, averted deaths and DALYs are reported in total. IHD, stroke, and lung cancer) conditions. For example, For all evaluated interventions, we consider deployments for a hypothetical intervention that reduces DALYs only in households using biomass fuels in traditional associated with chronic diseases by 10 percent, we cookstoves. assume a 10 percent reduction in health care–related expenditure on chronic diseases. Treatment-seeking behaviors and associated costs are Household Expenditures derived from IHDS (Desai, Vanneman, and National We assume households take responsibility for replacing Council of Applied Economic Research 2005) and IHDS stoves after their useful lifetime has passed. For this anal- summary documents (Barik and Desai 2014). ysis, households replace a blower stove once during the Approximately 94 percent of households across India five years of the evaluation, a useful lifetime for this class seek treatment for short-term illnesses, defined as fever, of interventions consistent with evidence from the liter- cough, and diarrhea. This figure is consistent with ature (Pillarisetti and others 2014; Sambandam and Haryana data extracted from IHDS databases and is others 2015). We assume that a gas stove needs no applied equally for all quintiles. Similarly, we apply replacement during the five years of this assessment and national treatment-seeking percentages by quintile to that households do not replace their chimney stove after the chronic illnesses of concern. Treatment-seeking the first year, consistent with findings from the National behaviors and associated costs4 by quintile are described Programme on Improved Chulhas (Venkataraman and in table 12.3. others 2010). Additionally, we translate the increase in fuel effi- We also assume that the averted ill health attribut- ciency attributable to an intervention into weekly able to this publicly financed intervention results in time savings by multiplying the increase in fuel 228 Injury Prevention and Environmental Health Table 12.3 Treatment-Seeking Behaviors and Associated Costs Disease Behavior and cost Q1 Q2 Q3 Q4 Q5 Acute diseases (ALRI) Treatment (%) 94 94 94 94 94 Median cost (US$) 17 14 10 18 12 Chronic diseases (IHD, COPD, Treatment (%) 88 86 90 94 95 lung cancer, stroke) Median cost (US$) 19 20 22 24 38 Sources: Data extracted from IHDS data (Desai, Vanneman, and National Council of Applied Economic Research 2005) and summary documents (Barik and Desai 2014). Note: ALRI = acute lower respiratory infection; COPD = chronic obstructive pulmonary disease; IHD = ischemic heart disease; Q = quintile. Table 12.4 Intervention Financial Parameters per Unit Government Household Time savings, Stove One-time costs, Yearly costs, Stove One-time costs, Yearly costs, US$ hours per year Intervention (US$) US$ (quintile)a US$ (quintile)b (US$) US$ (quintile) (quintile)b (quintile) Chimney 10 5 n.a. n.a. n.a. 1 25 (Q1), 19 (Q2), 20 (Q3), 15 (Q4), 14 (Q5) Blower 60 10 n.a. n.a. 60 5 72 (Q1), 53 (Q2), 56 (Q3), 41 (Q4), 40 (Q5) LPG status quo 0 0 33 20 30 (Q1), 46 (Q1), 170 (Q1), 30 (Q2), 46 (Q2), 126 (Q2), 30 (Q3), 46 (Q3), 134 (Q3), 30 (Q4), 46 (Q4), 97 (Q4), 30 (Q5) 46 (Q5) 96 (Q5) c LPG–GIU 0 30 (Q1), 33 (Q1), 20 0 (Q1), 46 (Q1), 30 (Q2), 33 (Q2), 0 (Q2), 46 (Q2), 0 (Q3), 0 (Q3), 30 (Q3), 80 (Q3), 0 (Q4), 0 (Q4), 30 (Q4), 80 (Q4), d 0 (Q5) 0 (Q5) 30 (Q5) 80 (Q5) Note: GIU = Give It Up; LPG = liquefied petroleum gas; n.a. = not applicable; Q = quintile. a. For LPG scenarios, one-time costs are the LPG connection costs. Current prices are available at IndianOil Corporation, https://indane.co.in/connection_tarrifs.php; US$1.00 = 68.13 rupees (Rs). For biomass stoves, one-time costs represent the cost of implementation. b. For LPG scenarios, yearly costs are the cost of the fuel subsidy to the government and the cost of the fuel to the households. The analysis assumes that houses use nine cylinders per year at an unsubsidized cost of US$8.80 (Rs 600.00) per cylinder; that Haryana has 3.35 million homes; and that the per cylinder subsidy is approximately US$3.70 (Rs 250.00). For biomass stoves, yearly costs are stove maintenance costs borne by the household. c. India’s national oil companies cover connection costs for 60 percent of households; connection costs to the household apply only for the upper three quintiles. d. The subsidy provided to existing LPG users is redirected to the lower-income quintiles. efficiency relative to the base case scenario by the days, in Haryana.5,6 Accordingly, a household’s yearly time spent collecting fuel. We place a monetary value total averted expenditure as the result of an interven- on this gained time using the Mahatma Gandhi tion is the sum of the wage earned during time previ- National Rural Employment Guarantee Act’s guaran- ously spent collecting fuel and the avoided health teed wage of Rs 251 (US$$3.70) per day, for up to 100 expenditure, minus any cost to the household of the Household Energy Interventions and Health and Finances in Haryana, India: An Extended Cost-Effectiveness Analysis 229 intervention (for example, stove maintenance or LPG stove and for their LPG fuel. This analysis assumes replacement or fuel costs): that houses use nine cylinders per year at an unsubsi- dized cost of US$8.80 (Rs 600.00) per cylinder and ⎛ Time savings ( hrs ) ⎞ that the per cylinder subsidy is approximately US$3.70 Averted expenditure = ⎜ × Wage ⎟ ⎝ 8 hrs/day ⎠ (Rs 250.00). Table 12.4 summarizes the per unit private and public intervention costs. + Averted medical expenditure All analyses were carried out using R 3.1 statistical − Intervention cost software (R Foundation 2015); plots were generated using the ggplot2 system (Wickham 2009). Government Costs The government incurs costs from providing the inter- RESULTS vention. The provision cost of chimney and blower Under the assumptions of the analysis, the intervention stoves includes the upfront cost of the intervention and a pathways described result in reductions in ill health one-time cost of deployment. For the LPG intervention, attributable to using solid fuel for cooking. The scale of we consider a policy pathway mimicking the ongoing those reductions varies both among interventions and Give It Up campaign. In this scenario, subsidies are pro- among quintiles; all interventions show higher reduc- vided to all solid fuel–using households in only the lower tions in ill health in the poorest three quintiles two income quintiles (Q1 and Q2 in tables and figures); (figure 12.1 and table 12.5). The costs to the govern- the subsidy given to existing LPG users in the upper three ment of the five-year programs vary widely among income quintiles is redirected to the lower two quintiles. interventions: the chimney stove intervention costs Because subsidies are being retargeted, the net cost to the approximately US$39 million, and the blower stove government is zero; the upper-income quintiles absorb intervention costs approximately US$180 million. At the additional costs. However, we assume the connection these prices, a life saved by the chimney stove costs the costs borne by the corporate social responsibility funds government approximately US$20,000 and the cost of from oil companies for the lower two income quintiles an averted DALY is US$520, whereas a life saved by the could be used elsewhere by these companies, which are blower stove costs US$10,000 and an averted DALY owned largely by the government, and thus represent a costs US$275. Complete replacement of traditional cost to the government. All households pay for their own stoves in solid fuel–using households in Haryana results Figure 12.1 Averted Deaths and DALYs for Three Classes of Interventions in Haryana, India, by Income Quintile a. Annual averted deaths b. Annual averted DALYs 110 400 100 350 90 Averted deaths (thousands) Averted DALYs (thousands) 80 300 70 250 60 200 50 40 150 30 100 20 50 10 0 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Income quintile Income quintile Intervention scenario LPG stove Blower stove Chimney stove Note: DALYs = disability-adjusted life years; LPG = liquefied petroleum gas; Q = quintile. Shaded areas account for uncertainty in background disease conditions and indicate the minimum and maximum avoidable burden. The relatively constant shape of the lines for each scenario is a byproduct of high solid fuel use across all income quintiles (range 60 percent to 90 percent) and the increasing number of people per household with increasing income, despite an approximately equal number of homes in each quintile. 230 Injury Prevention and Environmental Health Table 12.5 Five-Year Government Intervention Costs, Costs to Households, Household Expenditures Averted, and Deaths and DALYs Averted for Chimney Stove, Blower Stove, and LPG Intervention Pathways Q1 Q2 Q3 Q4 Q5 N = 669,000 N = 670,000 N = 671,000 N = 671,000 N = 670,000 SFU = 595,000 SFU = 604,000 SFU = 592,000 SFU = 445,000 SFU = 418,000 Chimney stove Government costs 8,900,000 9,100,000 8,900,000 6,700,000 6,300,000 Household maintenance costs 3,000,000 3,000,000 3,000,000 2,200,000 2,000,000 Household expenditures averted 6,700,000 4,900,000 5,100,000 2,800,000 2,600,000 Deaths averted 420 450 470 340 350 DALYs averted 16,000 17,000 18,000 13,000 14,000 Blower stove Government costs 42,000,000 42,000,000 41,000,000 31,000,000 29,000,000 Household maintenance and stove 50,000,000 51,000,000 50,000,000 38,000,000 36,000,000 replacement costs Household expenditures averted 52,000,000 27,000,000 31,000,000 8,300,000 8,500,000 Deaths averted 3,700 4,100 4,200 3,000 3,200 DALYs averted 140,000 150,000 160,000 110,000 120,000 LPG–GIU pathway Government costs 18,000,000 18,000,000 0 0 0 Fuel cost to households 150,000,000 150,000,000 270,000,000 200,000,000 190,000,000 Household expenditures averted 95,000,000 36,000,000 −72,000,000 −91,000,000 −84,000,000 Deaths averted 9,100 9,900 10,000 7,000 7,300 DALYs averted 310,000 340,000 350,000 240,000 250,000 Note: DALYs = disability-adjusted life years; GIU = Give It Up; LPG = liquefied petroleum gas; Q = quintile; SFU = solid fuel use. Fuel cost to households includes the up-front, one-time stove costs and recurrent maintenance and fuel costs. Household expenditures averted are the sum of the hourly wages accrued and the medical costs averted minus the cost to the households. For the LPG-GIU pathway, subsidy retargeting has different implications for solid fuel users versus current LPG users. Solid fuel users assume the additional full cost of unsubsidized LPG, while current LPG users assume only the difference between the full, unsubsidized LPG cost and the subsidized cost. in 2,000 averted deaths and 77,000 averted DALYs for The LPG stove averts the most deaths and DALYs the chimney stove and 18,100 averted deaths and across all income quintiles, per US$100,000 spent 676,000 averted DALYs for the blower stove. (figure 12.2). The figure panels for the LPG intervention The LPG pathway, in which the government pays for evenly split the costs between all income quintiles, connection charges from the allocation of corporate though the only additional expenditure by the govern- social responsibility funds, costs approximately US$36 ment is for the bottom two quintiles. million, or approximately US$11 per home when aver- Figure 12.3 depicts the trends in expenditures averted aged across all homes or US$30 per home when aver- by households by quintile. Notably, households in the aged across only the lower two income quintiles. This poor quintiles avoid more private expenditure than do policy pathway assumes that all higher-income house- households in the upper quintiles. This finding is most holds (n ~ 2,000,000) give up their subsidy and that the pronounced for the blower and LPG stoves. The reclaimed subsidy can be targeted to solid fuel–using described LPG intervention, in which the richest house- households (n ~ 1,340,000) in the lower-income quin- holds receive no subsidy for fuel, simulates the GIU tiles, effectively not altering the cost to the government. campaign and results in a net cost to these households, Under this scheme, an averted death costs the govern- which must pay the full, unsubsidized price for their fuel ment US$825 and an averted DALY costs US$25. Over and their one-time connection costs. The national oil the five-year evaluation period, 1,484,000 DALYs and companies cover connection costs for poor households, 44,000 deaths are averted. which also receive subsidized fuel. Household Energy Interventions and Health and Finances in Haryana, India: An Extended Cost-Effectiveness Analysis 231 Figure 12.2 Averted Deaths and DALYs per US$1 Million Spent for Three Classes of Interventions in Haryana, India, over Five-Year Intervention Lifetime, by Income Quintile a. Annual averted deaths Blower stove Chimney stove LPG stove Deaths averted per US$100,000 spent 150 100 50 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Income quintile Income quintile Income quintile b. Annual averted DALYs Blower stove Chimney stove LPG stove DALYs averted per US$100,000 spent 5,500 5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Income quintile Income quintile Income quintile Note: DALYs = disability-adjusted life years; LPG = liquefied petroleum gas; Q = quintile. Panels represent intervention classes. For the LPG scenario, gas subsidies given up by income quintiles 4 and 5 result in no expense to the government for the intervention in these quintiles; the subsidy is retargeted evenly to income quintiles 1, 2, and 3. DISCUSSION • It takes into account the earning potential of individ- uals who save time by transitioning to more efficient We present results from ECEAs of policies designed to stoves, which require less fuel and less time spent achieve high uptake of three hypothetical classes of HAP collecting fuel. interventions in Haryana, India. The classes of interven- • It uses a continuous exposure-response function to tions presented match historical modes of household estimate health benefits of interventions with differ- energy programs, first attempted with chimney stoves, ent exposure reduction potentials. then with blower-assisted biomass stoves, and, most • It evaluates a current LPG policy pathway that mim- recently, with a transition to truly clean cooking using ics the ongoing retargeting of LPG fuel subsidies. LPG. By evaluating multiple types of interventions, we are able to compare cheap, poorly performing chimney stoves By considering earnings and medical expenses averted with intermediate (blower) and modern (LPG) options. as a result of these interventions, we hope to present a Our approach is novel in several ways: more rigorous and multidimensional set of options for policy makers to evaluate and consider as they seek to • It seems to be the first ECEA to date evaluating reduce the significant health burden associated with household energy policies. exposure to smoke arising from use of solid fuel 232 Injury Prevention and Environmental Health combustion for cooking. Unlike many other ECEAs, we Figure 12.3 Averted Private Expenditure for Each Class of Intervention present both averted deaths and averted DALYs, a com- over the Proposed Five-Year Intervention Lifetime bined metric of both morbidity and mortality. Our findings indicate that from the perspective of 125 avoiding ill health, the evaluated LPG scenario out- 100 performs attempts to make biomass combustion clean, 75 Private expenditures averted in although it imposes an additional fuel cost on house- 50 U.S. dollars (millions) holds. Our findings show, however, that the added cost 25 of subsidized fuel—at least for the poorest income quintiles—is more than covered by the monetary val- 0 ues of saved time and avoided healthcare. This striking −25 finding, however, is heavily influenced by the assump- −50 tion that use of time previously spent collecting fuel is −75 repurposed for productive economic return, which may not be true in these settings. The income quintile– −100 based consequences of these policy pathways are com- −125 plicated by the underlying distributions of SFU and Q1 Q2 Q3 Q4 Q5 disease burden. Using those distributional conse- Income quintile quences alone to make policy decisions would favor less effective interventions and ignore the significant health LPG stove Blower stove Chimney stove burden remaining from adoption of such technology. Note: LPG = liquefied petroleum gas; Q = quintile. Negative values indicate net costs to households. In contrast, the financial protection provided by the However, the upper quintiles are voluntarily giving up their subsidy in the GIU campaign. Annex 12A, LPG policy pathways benefits the poor, who receive figure 7 shows the per household costs and savings by income quintile and intervention scenario. subsidized fuel, free LPG connections, and reduced health care costs and who stand to gain the most in We conducted a sensitivity analysis to evaluate the wages. This inverse relationship between those quin- effect of a hypothetical scenario with an equal number tiles that receive an advantage from health benefits of people per quintile and linearly decreasing SFU as versus financial benefits is not uncommon to ECEA wealth increases. Under these conditions, the effects of (Pecenka and others 2015); it reveals the methodology’s all three scenarios were most profound for the poorer ability to highlight multiple policy-relevant facets income quintiles. masked by traditional cost-effectiveness analysis. It also Our analysis has a number of limitations. First, we indicates an area of ongoing concern. Public financing present only a small number of potential household of interventions such as those that are targeted to quin- energy intervention scenarios under ideal use scenarios, tiles and that are modeled to benefit the most may have none of which considers an additional benefit in the unintended consequences. form of reducing the herd or the neighborhood effect; Strikingly, retargeting the subsidy—as is happening that is, by swapping out entire communities, wider gains in India under the Give It Up campaign—significantly in exposure reduction could occur. We assume that increases the cost-effectiveness of interventions targeting households transition fully to the cleaner technology in the poorest income quintiles. This finding suggests that all three cases and do not revert to older technologies such an approach—especially when considered in light even partially. We address this shortcoming in part by of the considerable financial risk protection afforded by evaluating the effects of adoption rates less than targeting the lower-income quintiles—may be a way to 100 percent and the effects of variation in exposure that quickly and efficiently move resources to those most may represent suboptimal intervention performance or vulnerable to both the health and the financial effects of use of both old and new interventions. We find that the SFU for cooking. overall trends in our findings are robust to these types of Despite these clear distributional benefits, less pro- changes (annex 12A). We chose this framing to indicate found difference exists among quintiles than originally the potential effects of statewide adoption and use of anticipated, which explains the relatively constant val- LPG, realizing that such a transition will take time and ues across quintiles seen in figures 12.1, 12.2, and 12.3. must contend with issues of stove stacking. We acknowl- We believe this is due in part to relatively high SFU edge that such a framing does not contend with issues of numbers across all quintiles and an increasing number the perceived costs of fuel versus potential health savings of people per household as wealth increases, resulting and wage gains, which may be viewed independently by in a skewed distribution of background disease rates. household decision makers. Household Energy Interventions and Health and Finances in Haryana, India: An Extended Cost-Effectiveness Analysis 233 Second, our study uses older IHDS data from 2005–06. Because of the need to purchase fuel, the financial Newer data—either from IHDS itself or from other conditions of the biomass and LPG scenarios are funda- national surveys—may provide more up-to-date num- mentally different, but we attempt to explore them here bers on the penetration of LPG in Haryana and on in the same analysis. In doing so, we evaluated the cur- treatment-seeking behaviors and related costs. We also rent LPG subsidy system as a given and took only the recognize that our mapping of BPL households to IHDS extra costs of LPG connections in the GIU campaign as income quintiles 1 and 2 may not match the current real- the cost of the LPG expansion, thereby assuming that the ity. However, to our knowledge, unlike other surveys, shift of subsidy from the middle class to the poor did not IDHS has the benefit of providing a single source from itself incur any change in government expenditures; that which to gather almost all parameters needed for this is, there were no transactional costs to the government. analysis, thereby preventing potentially problematic com- The framing in a different country without any current parison among surveys with different sampling frames. LPG subsidy, however, might be quite different. Although Similarly, we use national data on treatment-seeking the funds for connections currently come from the behavior for Haryana; data from IHDS at the state level required social responsibility funds of the national oil were unrealistically homogenous across income quintiles. companies, we assign these as government expenditures Our quantification of the monetary value of time because they could have been used for other purposes. savings does not account for behavioral aspects related Finally, in the absence of information, we assumed no to job-seeking behavior or any potential rebound effects operational costs to the government to design, promote, of adoption of cleaner cooking technologies. We manage, and evaluate the large-scale disseminations that acknowledge that small daily or weekly time savings may would be needed in all three scenarios. not be large enough and (1) that a search for a job via the Solid fuel–using households in upper income Mahatma Gandhi National Rural Employment quintiles (Q3, Q4, and Q5) absorb a significant fuel Guarantee Act is warranted, (2) that employment oppor- costs in our modeled LPG intervention, as they move tunities exist for the modeled time savings, or (3) that from no LPG to full-price LPG. The large number of household members who experience time savings might solid fuel–using households in the wealthier income engage in employment versus other household or quintiles suggests that future subsidies and LPG- leisure-related activities. We assume that people will promoting programs may wish to address these successfully seek employment in the national program households using a sliding subsidy, as suggested by and that the time saved by shifting away from biomass Tripathi, Sagar, and Smith (2015). Ongoing LPG pro- can be utilized productively—findings that may not hold grams in India do not currently have a provision to and warrant further investigation. Our approach does target these households. offer an empirical money metric of time savings based Future analyses of this type should investigate on an ongoing program in Haryana and India at large. alternate methods of apportioning underlying data from Thus, it is based on a documented rural wage rate. national data to state-level data and into disease quintiles. We assume universal adoption of the cleaner cooking Furthermore, although partitioning by SFU is reason- technology in Haryana for the main scenarios described able in our example, it may mask behavioral patterns in this chapter. This assumption is highly optimistic related to solid fuel use that we did not anticipate and because older, more polluting stoves are often not aban- differences between quintiles that impact disease doned immediately when a cleaner one of any kind is distributions. adopted; studies have shown that even the best modern Beyond methodological and data limitations, we biomass stoves often do not perform well over time in assume the status quo remains constant with respect to reducing pollution exposures. However, LPG is essen- international LPG prices and related subsidies. This tially always clean. We bound our chimney stove sce- assumption ensures analytic tractability, but it may nario by assuming that it breaks down after one year and not hold, given the volatility in oil prices globally. India is not rebuilt, which has often occurred with the inex- is undergoing a rapid, policy-driven transformation that pensive models widely deployed in India. An alternative is dramatically increasing access to LPG for communities approach would be to build in regular replacement of previously reliant on solid fuels. Although these ongoing these stoves, or perhaps to move to the much more changes may alter the calculus behind the results, they expensive and robust chimney stoves that have been highlight the need for being able to perform multifac- successfully used in other countries and subregions. For eted analyses that consider more than simply basic example, in China and Mexico and in Central America, cost-effectiveness to estimate the potential effects of chimney stoves that function for a decade or more are large programs—precisely the type of evaluation facili- not uncommon, but costs are at least 15 times greater. tated by ECEA. 234 Injury Prevention and Environmental Health CONCLUSIONS well-established repair and refill system in place in the region, although it requires reliable extension to addi- Exposure to HAP from solid cookfuel, mainly as bio- tional populations. mass, causes an estimated 925,000 deaths yearly in India To be effective, however, any intervention program today. The number of people most affected—700 mil- must focus not only on providing access to the interven- lion to 800 million—has not declined in 30 years, despite tion but also on enhancing use over the long term, considerable economic development and the growth of including continuing to pay for subsidized fuel and clean fuel use for the middle class. Other approaches are repair and replacement of the stoves. Only when use of clearly needed to address this health hazard. the traditional polluting biomass stoves is greatly reduced Three types of national policies have been initiated to over time and replaced by LPG or another equally clean address the health, social, and environmental effects of alternative will full health benefits be secured. inefficient household biomass use. In the 1980s and 1990s, households relied on inexpensive stoves made locally with simple materials but without much improvement in ANNEX smoke emissions, although often including a chimney. The annex to this chapter is as follows. It is available at Around 2010, a new program was initiated that promised http://www.dcp-3.org/environment. to develop and promote biomass stoves that produced far less pollution emissions, but usually this program did not • Annex 12A. Supporting Information incorporate chimneys. Starting in 2014, the national GIU campaign began to greatly expand access to a clean modern fuel, LPG, for BPL families by using innovative ACKNOWLEDGMENTS financing and promotional modalities. Newer initiatives— including the Smokeless Village program and the recently The authors thank Rachel Nugent, Zachary Olson, announced Ujjwala program—continue this trend of Charles N. Mock, and two anonymous reviewers for making clean cooking available widely across India. This their thoughtful comments and suggestions. chapter has evaluated each of the approaches separately for their cost-effectiveness in hypothetical deployments in NOTES the same northern Indian state, Haryana. We believe these types of modeling exercises are instructive and help target World Bank Income Classifications as of July 2014 are as fol- further, field-based studies evaluating the effect of lows, based on estimates of gross national income (GNI) per capita for 2013: programs. Lowering household exposures to air pollution can • Low-income countries (LICs) = US$1,045 or less decrease both health and financial burdens not only by • Middle-income countries (MICs) are subdivided: reducing medical costs, but also by averting household a) lower-middle-income = US$1,046 to US$4,125 expenditures or avoiding lost wage earnings. The scale of b) upper-middle-income (UMICs) = US$4,126 to US$12,745 the reduction—and the amount of disease burden left • High-income countries (HICs) = US$12,746 or more. untouched by each of them—varies widely among the three intervention options, however, although cost- 1. We use the DALY that was first widely deployed in the effectiveness varies less widely. More modest reductions GBD 2010. It is not adjusted by age weighting or discount- from chimney stoves, for example, are to some extent ing. It cannot be directly compared to previous versions matched by more modest costs. that did so. The innovative policy of Ujjwala, extending out of 2. Policy measures to increase access to LPG among the rural the GIU campaign, starts by retargeting existing LPG poor in India are advancing rapidly. The programs men- subsidies away from the middle class to the poor. This tioned in this chapter are up to date as of July 2016. approach can result in highly cost-effective health 3. Our analysis does not include any assessment of out- door air pollution and the reduction in emissions from improvements in the poorest quintiles of the population. the household sector, which account for an estimated It is accompanied by some shift of costs to the mid- 25–50 percent of ambient small particle exposures in India dle class, but, notably, by their agreement and without a (Chafe and others 2014; Lelieveld and others 2015). net increase in government expenditures. By being the 4. We use IHDS questions about cough and fever in the cleanest of the options examined, LPG also has the past month as a proxy for ALRI. The total number of potential to achieve the greatest health benefits. households reporting this proxy for ALRI per quintile Compared to the other two fuels, LPG also benefits from is multiplied by 12 to obtain the number of cases per a familiar long-lived cooking technology that has a quintile per year and then divided by the total number Household Energy Interventions and Health and Finances in Haryana, India: An Extended Cost-Effectiveness Analysis 235 of households in the quintile to determine the number of Social Research, 2016-02-16. http://doi.org/10.3886/ICPSR cases per household per year. The per case cost estimate 22626.v11. is multiplied by the number of cases per household by Dutta, K., K. N. Shields, R. Edwards, and K. R. Smith. 2007. quintile to determine the yearly cost. “Impact of Improved Biomass Cookstoves on Indoor 5. See http://www.haryanarural.gov.in/guidelines/MGNREGS Air Quality near Pune, India.” Energy for Sustainable /1025MGNREGS_wage_notification_2015_16.pdf. Development 11: 19–32. 6. See http://www.haryanarural.gov.in/detail-nrega.htm#bnote. Foell, W., S. Pachauri, D. Spreng, and H. Zerriffi. 2011. “Household Cooking Fuels and Technologies in Developing Economies.” Energy Policy 39 (12): 7487–96. Forouzanfar, M., L. Alexander, H. R. Anderson, V. F. Bachman, REFERENCES S. Biryukov, and others. 2015. “Global, Regional, and Arcenas, A., J. Bojö, B. R. Larsen, and F. Ruiz Ñunez. 2010. National Comparative Risk Assessment of 79 Behavioural, “The Economic Costs of Indoor Air Pollution: New Results Environmental and Occupational, and Metabolic Risks or for Indonesia, the Philippines, and Timor-Leste.” Journal of Clusters of Risks in 188 Countries, 1990–2013: A Systematic Natural Resources Policy Research 2 (1): 75–93. Analysis for the Global Burden of Disease Study 2013.” Aunan, K., L. W. H. Alnes, J. Berger, Z. Dong, L. Ma, and The Lancet 386 (10010): 2287–323. others. 2013. “Upgrading to Cleaner Household Stoves and Government of India. 2011. Census of India 2011: Houses, Reducing Chronic Obstructive Pulmonary Disease among Household Amenities, and Assets. Registrar General and Women in Rural China: A Cost-Benefit Analysis.” Energy for Census Commissioner, New Delhi. Sustainable Development 17 (5): 489–96. Hutton, G., E. Rehfuess, and F. Tediosi. 2007. “Evaluation of Bailis, R., V. Berrueta, C. Chengappa, K. Dutta, R. Edwards, the Costs and Benefits of Interventions to Reduce Indoor and others. 2007. “Performance Testing for Monitoring Air Pollution.” Energy for Sustainable Development 11 (4): Improved Biomass Stove Interventions: Experiences of 34–43. the Household Energy and Health Project.” Energy for Hutton, G., E. Rehfuess, F. Tediosi, and S. Weiss. 2006. Sustainable Development 11 (2): 57–70. Evaluation of the Costs and Benefits of Household Energy and Balakrishnan, K., S. Ghosh, B. Ganguli, S. Sambandam, Health Interventions at Global and Regional Levels. Geneva: N. Bruce, and others. 2013. “State and National Household World Health Organization. Concentrations of PM2.5 from Solid Cookfuel Use: Results Institute for Health Metrics and Evaluation (IHME). 2015. GBD from Measurements and Modeling in India for Estimation Compare. Seattle, WA: IHME, University of Washington. of the Global Burden of Disease.” Environmental Health http://vizhub.healthdata.org/gbd-compare. 12: 77. Jeuland, M. A., and S. K. Pattanayak. 2012. “Benefits and Costs Barik, D., and S. Desai. 2014. “Determinants of Private of Improved Cookstoves: Assessing the Implications of Healthcare Utilisation and Expenditure Patterns in India.” Variability in Health, Forest and Climate Impacts.” PLoS In India Infrastructure Report 2013|14: The Road to Universal One 7 (2): e30338–15. Health Coverage, edited by S. B. A. S. Ghosh, 52–64. Jeuland, M. A., S. K. Pattanayak, and R. Bluffstone. 2015. New Delhi: Orient Blackswan Private Limited. “The Economics of Household Air Pollution.” Annual Bonjour, S., H. Adair-Rohani, J. Wolf, N. G. Bruce, S. Mehta, Review of Resource Economics 7 (1): 81–108. and others. 2013. “Solid Fuel Use for Household Cooking: Johnson, M. A., and R. A. Chiang. 2015. “Quantitative Guidance Country and Regional Estimates for 1980–2010.” for Stove Usage and Performance to Achieve Health and Environmental Health Perspectives 121 (7): 784–90. Environmental Targets.” Environmental Health Perspectives Brooks, N., V. Bhojvaid, M. A. Jeuland, J. J. Lewis, O. Patange, 123 (8): 820–26. and others. 2016. “How Much Do Alternative Cookstoves Kishore, V., and P. V. Ramana. 2002. “Improved Cookstoves Reduce Biomass Fuel Use? Evidence from North India.” in Rural India: How Improved Are They? A Critique of Resource and Energy Economics 43: 153–71. the Perceived Benefits from the National Programme on Burnett, R. T., C. A. Pope 3rd, M. Ezzati, C. Olives, S. S. Lim, Improved Chulhas (NPIC).” Energy 27 (1): 47–63. and others. 2014. “An Integrated Risk Function for Lelieveld, J., J. S. Evans, M. Fnais, D. Giannadaki, and A. Pozzer. Estimating the Global Burden of Disease Attributable to 2015. “The Contribution of Outdoor Air Pollution Sources Ambient Fine Particulate Matter Exposure.” Environmental to Premature Mortality on a Global Scale.” Nature 525 Health Perspectives 122 (4): 397–403. (7569): 367–71. Chafe, Z. A., M. Brauer, Z. Klimont, R. Van Dingenen, Lim, S. S., T. Vos, A. D. Flaxman, G. Danaei, K. Shibuya, and S. Mehta, and others. 2014. “Household Cooking with Solid others. 2012. “A Comparative Risk Assessment of Burden of Fuels Contributes to Ambient PM2.5 Air Pollution and the Disease and Injury Attributable to 67 Risk Factors and Risk Burden of Disease.” Environmental Health Perspectives 122 Factor Clusters in 21 Regions, 1990–2010: A Systematic (12): 1314–20. Analysis for the Global Burden of Disease Study 2010.” The Desai, S, R. Vanneman, and National Council of Applied Lancet 380 (9859): 2224–60. Economic Research, New Delhi. 2005. India Human Lozano, R., M. Naghavi, K. Foreman, S. Lim, K. Shibuya, and Development Survey (IHDS). ICPSR22626-v11. Ann others. 2012. “Global and Regional Mortality from 235 Arbor, MI: Inter-University Consortium for Political and Causes of Death for 20 Age Groups in 1990 and 2010: 236 Injury Prevention and Environmental Health A Systematic Analysis for the Global Burden of Disease Health Relevant Exposure Reductions? Results from Study 2010.” The Lancet 380 (9859): 2095–28. Initial Assessment of Select Commercial Models in India.” Malla, M. B., N. Bruce, E. Bates, and E. Rehfuess. 2011. Ecohealth 12 (1): 25–41. “Applying Global Cost-Benefit Analysis Methods to Indoor Smith, K. R., N. Bruce, K. Balakrishnan, H. Adair-Rohani, Air Pollution Mitigation Interventions in Nepal, Kenya J. Balmes, and others. 2014. “Millions Dead: How Do We and Sudan: Insights and Challenges.” Energy Policy 39 (12): Know and What Does It Mean? Methods Used in the 7518–29. Comparative Risk Assessment of Household Air Pollution.” Mehta, S., and C. Shahpar. 2004. “The Health Benefits of Annual Review of Public Health 35: 185–206. Interventions to Reduce Indoor Air Pollution from Smith, K. R., A. Pillarisetti, L. D. Hill, D. Charron, S. Delapena, Solid Fuel Use: A Cost-Effectiveness Analysis.” Energy for and others. 2015. “Proposed Methodology: Quantification Sustainable Development 8 (3), 53–59. of a Saleable Health Product (aDALYs) from Household Ministry of Petroleum, Government of India. 2016. “My LPG. Cooking Interventions.” Household Energy, Climate, and in.” http://mylpg.in/index.aspx. Health Research Group, University of California, Berkeley, Northcross, A., Z. Chowdhury, J. McCracken, E. Canuz, and and Berkeley Air Monitoring Group, Berkeley, for the K. R. Smith. 2010. “Estimating Personal PM2.5 Exposures World Bank. Using CO Measurements in Guatemalan Households Smith, K. R., and A. D. Sagar. 2015. “Making the Clean Cooking with Wood Fuel.” Journal of Environmental Available: Escaping India’s Chulha Trap.” Energy Policy 75: Monitoring 12: 873–78. 410–14. Pant, K. P. 2011. “Cheaper Fuel and Higher Health Costs Times of India. 2016. “PM to Launch Rs 8,000 Crore Scheme for among the Poor in Rural Nepal.” AMBIO 41 (3): 271–83. Free LPG Connections to Poor.” April 22. Pecenka, C. J., K. A. Johansson, S. T. Memirie, D. T. Jamison, and Tripathi, A., A. D. Sagar, and K. R. Smith. 2015. “Promoting S. Verguet. 2015. “Health Gains and Financial Risk Protection: Clean and Affordable Cooking.” Economic Political Weekly An Extended Cost-Effectiveness Analysis of Treatment and 48: 81–84. Prevention of Diarrhoea in Ethiopia.” BMJ Open 5: e006402. U.S. EPA (United States Environmental Protection Agency). Pillarisetti, A., S. Mehta, and K. R. Smith. 2016. “HAPIT, the 2004. “Advisory on Plans for Health Effects Analysis in Household Air Pollution Intervention Tool, to Evaluate the the Analytical Plan for EPA’s Second Prospective Analysis: Health Benefits and Cost-Effectiveness of Clean Cooking Benefits and Costs of the Clean Air Act, 1990–2020.” Interventions.” In Broken Pumps and Promises: Incentivizing EPA-SAB-COUNCIL-ADV-04-002, Advisory Council on Impact in Environmental Health, edited by E. Thomas, 147– Clean Air Compliance Analysis, U.S. EPA, Washington, DC. 70. Cham, Switzerland: Springer International Publishing Venkataraman, C., A. D. Sagar, G. Habib, N. Lam, and AG. K. R. Smith. 2010. “The Indian National Initiative for Pillarisetti, A., M. Vaswani, D. Jack, K. Balakrishnan, M. N. Advanced Biomass Cookstoves: The Benefits of Clean Bates, and others. 2014. “Patterns of Stove Usage after Combustion.” Energy for Sustainable Development 14: Introduction of an Advanced Cookstove: The Long-Term 63–72. Application of Household Sensors.” Environmental Science Verguet, S., C. L. Gauvreau, S. Mishra, M. MacLennan, and Technology 48 (24): 14525–33. S. M. Murphy, and others. 2015. “The Consequences of R Foundation for Statistical Computing. 2015. R: A Language and Tobacco Tax on Household Health and Finances in Rich and Environment for Statistical Computer. Vienna: R Foundation Poor Smokers in China: An Extended Cost-Effectiveness for Statistical Computing. http://www.R-project.org. Analysis.” The Lancet Global Health 3 (4): e206–16. Ruiz-Mercado, I., O. Masera, H. Zamora, and K. R. Smith. 2011. Verguet, S., R. Laxminarayan, and D. T. Jamison. 2015. “Adoption and Sustained Use of Improved Cookstoves.” “Universal Public Finance of Tuberculosis Treatment in Energy Policy 39: 7557–66. India: An Extended Cost-Effectiveness Analysis.” Health Sambandam, S., K. Balakrishnan, S. Ghosh, A. Sadasivam, Economics 24 (3): 318–32. S. Madhav, and others. 2015. “Can Currently Available Wickham, H. 2009. ggplot2: Elegant Graphics for Data Analysis. Advanced Combustion Biomass Cook-Stoves Provide New York: Springer. Household Energy Interventions and Health and Finances in Haryana, India: An Extended Cost-Effectiveness Analysis 237 Chapter 13 Costs and Benefits of Installing Flue-Gas Desulfurization Units at Coal-Fired Power Plants in India Maureen L. Cropper, Sarath Guttikunda, Puja Jawahar, Kabir Malik, and Ian Partridge INTRODUCTION the amount of coal burned to generate a kilowatt hour Coal-fired power plants, in addition to emitting green- (kWh) of electricity. Recent studies suggest serious house gases, are a major source of local pollution and health effects associated with SO2 emissions from Indian health damages throughout the world. China, the United power plants. Guttikunda and Jawahar (2014) estimate States, and other countries that rely on coal for electricity that Indian power plants caused more than 80,000 production regulate emissions from coal-fired power deaths in 2011; they attribute 30–40 percent of these plants, primarily for health reasons. In the United States, deaths to SO2. Cropper and others (2012) suggest that as the 1990 Clean Air Act Amendments caused many power many as 60 percent of the deaths associated with coal- plants to switch to low-sulfur coal or to install flue-gas fired power plants in India may be attributable to SO2 desulfurization units (FGD units, or scrubbers). emissions rather than to directly emitted particulate Subsequent tightening of sulfur dioxide (SO2) regulations matter or oxides of nitrogen (NOx). has caused more plants to scrub their emissions. In 2010, This chapter analyzes the health benefits and the costs power plants with FGD units accounted for 60 percent of of installing FGD units at each of the 72 coal-fired power the electricity generated from coal in the United States plants in India, plants that in 2009 constituted 90 percent (Schmalensee and Stavins 2013). By 2013, 95 percent of of coal-fired generating capacity. We estimate the health China’s coal-fired generating capacity had been fitted with benefits of one FGD unit by estimating SO2 emissions FGD units (Ministry of Environmental Protection 2014). from a plant without an FGD unit and then translating India, which relies on coal to generate 76 percent of those emissions into changes in ambient air quality. This its electricity (CEA 2015), did not regulate SO2 emissions is accomplished using an Eulerian photochemical dis- from coal-fired power plants until December 2015. That persion model (CAMx) that allows SO2 to form fine lack of regulation may have been due, in part, to the low sulfate particles (smaller than 2.5 micrometers in diam- sulfur content of Indian coal (Chikkatur and Sagar eter [PM2.5]) in the atmosphere. The impacts of PM2.5 on 2007). Indian coal is approximately 0.5 percent sulfur by premature mortality are estimated for ischemic heart weight, similar to Powder River Basin (PRB) coal in the disease, stroke, lung cancer, chronic obstructive pulmo- United States (Lu and others 2013). However, the popu- nary disease (COPD), and acute lower respiratory infec- lation exposed to SO2 emissions from power plants in tion (ALRI) using the integrated exposure response India is much greater than that in the United States, as is (IER) coefficients in Burnett and others (2014). Corresponding author: Maureen Cropper, Department of Economics, University of Maryland, College Park, MD, United States; cropper@econ.umd.edu. 239 We assume that a scrubber will reduce SO2 emissions by connected to the grid in India. The size of each circle on 90 percent. The annual reductions in premature mortal- the map is proportional to the electricity generated by ity and associated life years lost resulting from use of each plant. State governments owned 45 of the plants, scrubbers are combined with an estimate of annualized the central government owned 22, and private entities capital and operating costs to compute the cost per owned 5. Table 13.1 describes the operating characteris- statistical life saved and cost per disability-adjusted life tics of these plants in terms of installed capacity, electric- year (DALY) averted associated with each FGD unit. ity generated, and other characteristics. We analyze the Reducing SO2 emissions from coal-fired power plants impact of plant emissions in 2008–09, the year for which offers additional benefits that we do not quantify. These we have information on the sulfur content of coal.1 include improvements in visibility (which yield aesthetic Total coal-fired generating capacity in India doubled and recreation benefits) and reduced acidic deposition. between March 2009 and March 2015 (CEA 2015), from Acidic deposition can reduce soil quality (through nutri- 76 gigawatts (GW) to 164.6 GW.2 Accordingly, our anal- ent leaching), impair timber growth, and harm freshwa- ysis underestimates the health impacts of current SO2 ter ecosystems (USEPA 2011). emissions from the power sector. We note, however, that the plants we analyze remain subject to pollution control METHODS laws, and most plants would find it difficult to meet these laws without installing FGD units. Estimating the Health Impacts of SO2 Emissions from To calculate the SO2 emissions of each plant, we must Power Plants know the plant’s annual electricity generation, the Our analysis focuses on 72 coal-fired power plants amount of coal burned per kWh, and the sulfur content (shown in map 13.1) which in March 2009 constituted of the coal burned.3 We estimate the cost-effectiveness of 90 percent of the coal-fired generating capacity FGD units using the Central Electricity Regulatory Commission of India’s benchmark operating condi- Map 13.1 Coal-Fired Power Plants: All Plants in Dataset tions.4 These assume that each plant operates at 85 percent of capacity (the median operating capacity Ashgabat Dushanbe TURKMENISTAN TAJIKISTAN NATIONAL CAPITALS STATE BOUNDARIES (INDIA) for the 72 plants was 79 percent in 2008–09). We use INTERNATIONAL BOUNDARIES benchmark conditions because actual operating condi- Kabul C H I N A tions are likely to fluctuate from year to year. AFGHANISTAN Islamabad Coal consumption per kWh is based on actual coal ISLAMIC ROPAR consumption per kWh in 2008–09. On average, coal REPUBLIC OF IRAN PAKISTAN LEHRA MOHABBAT BHATINDA PANIPAT burned per kWh is much higher at Indian power plants SURATGARH New Delhi BADARPUR RAJGHAT DADRI (NCTPP) I.P. STATION F'BAD EXTN. H'GANJ B NEPAL BHUTAN than at coal-fired power plants in the United States TANDA Kathmandu Thimphu (0.77 kg/kWh versus 0.47 kg/kWh) (Malik 2013). This PANKI KOTA PARICHA UNCHAHAR BARAUNI KAHALGAON FARAKKA STPS difference is due in part to the lower heat content of 25°N RIHAND STPS VINDH'CHAL STPS OBRA THERMAL ANPARA SINGRAULI STPS TENUGHAT PATRATU MEJIA BAKRESWAR Indian coal, but it is also due to inefficiencies in plant Muscat KUTCH LIGNITE SANJAY GANDHI D.P.L. Dhaka GANDHI NAGAR WANAKBORI INDIA BOKARO B DURGAPUR CHANDRAPURA SANTALDIH BANDEL KOLAGHAT TITAGARH operation (Chan, Cropper, and Malik 2014). The sulfur KORBA WEST KORBA STPS SIKKA REPL BUDGE BUDGE SATPURA SIPAT STPS OMAN NEW COSSIPORE SOUTHERN REPL content of coal (averaging 0.53 percent sulfur by weight) I.B. VALLEY K'KHEDA II KORADI TALCHER STPS UKAI THERMAL BHUSAWAL TALCHER MYANMAR BANGLADESH comes from a survey of Indian power plants conducted CHANDRAPUR NASIK DHANU 20°N Nay Pyi Taw R'GUNDEM STPS R'GUNDEM B KOTHAGUDEM SIMHADRI by the authors. This finding corresponds closely to fig- RAICHUR VIJAYAWADA ures reported by Lu and others (2013), Garg and others Arabian Sea 15°N RAYALSEEMA 15°N (2002), and Reddy and Venkataraman (2002). Based on NORTH MADRAS ENNORE Bay of Bengal benchmark conditions, the 72 plants emit approximately METTUR NEYVELI FST EXT NEYVELI 3 million tons of SO2 annually. CAMx, an Eulerian photochemical dispersion model, NEYVELI (M CUT) 10°N TUTICORIN Andaman 10°N was run to estimate the impact of each plant’s SO2 emis- SRI LANKA S ea sions on ambient air quality.5 The model, which includes COAL-FIRED POWER PLANTS: I N D I A N O C E A N Colombo Central government–owned power plants gas-to-aerosol conversion for SO2 to sulfates, supports State government–owned power plants 5°N Privately owned power plants plume rise calculations for each power plant using Malé MALDIVES 465.375–2606.100 >2,606.100–5,174.970 three-dimensional meteorological data.6 The model was >5,174.970–8,786.280 >8,786.280–15,636.600 run separately for each plant, simulating 365 days of IBRD 42285 | December 2016 >15,636.600–24,273.960 Note: The size of the circle is proportional 0° emissions, to calculate the increase in annual average to the electricity generation under baseline conditions. fine particle concentrations corresponding to the plant’s 60°E 65°E 70°E 75°E 80° E 85°E 90°E 95°E emissions. The model was run at a 0.25˚grid resolution 240 Injury Prevention and Environmental Health Table 13.1 Operating Characteristics of Power Plants in the Dataset Summary Statistics—Actual Operations Average Standard deviation Median Minimum Maximum Nameplate capacity (MW) 948 674 840 63 3,260 Generation (GWh) 6,393 5,446 5,305 103 26,601 Capacity utilization (%) 75 20 79 11 100 Sulfur content of coal (%) 0.53 0.19 0.5 0.21 2.00 SO2 emissions (tons/yr) 37,727 31,857 30,423 778 188,010 Note: Number of observations = 72 power plants. Data based on actual operations for the year 2008–09. GWh = gigawatt hour; MW = megawatt; SO2 = sulphur dioxide. and combined with 2011 population data to calculate 25.54 YLL, a figure close to that reported in the 2013 the population-weighted increase in annual average GBD. DALYs lost because of PM2.5 are the sum of YLL PM2.5 concentrations associated with the plant. and years lived with disability (YLD). In the 2013 GBD, Epidemiological research has found consistent associ- 97 percent of DALYs associated with ambient air pollu- ations between premature mortality and PM2.5. Pope and tion are YLL. We have not calculated the YLD associated others (2002) report significant impacts of exposure to with SO2 emissions; therefore, our estimates of the PM2.5 in cities in the United States on all-cause, cardio- health benefits of emissions reductions understate total pulmonary, and lung cancer mortality. This work formed health benefits. the basis of early studies of the global burden of air pol- lution (Cohen and others 2004). More recent studies of Estimates of Health Effects Associated with SO2 the Global Burden of Disease (GBD) (Lim and others Emissions 2013) use meta-analyses of epidemiological studies from Our calculations suggest that approximately 15,500 deaths several sources to quantify the impact of a wider range of in 2013 were attributable to SO2 emissions from the PM2.5 exposures on cardiovascular and respiratory deaths, 72 plants, with stroke (7,600) and ischemic heart disease as well as deaths from lung cancer and ALRI (Burnett and (4,200) accounting for the majority of deaths. Table 13.2 others 2014). The 2013 GBD estimates that 587,000 reports the distribution of deaths and DALYs (by cause) deaths in India in 2013 were attributable to ambient air for the 72 plants. These deaths, in the aggregate, are asso- pollution (GBD 2013 Risk Factors Collaborators 2015). ciated with approximately 400,000 YLL.10 If the plants in Premature mortality associated with the increase in our study were to operate under benchmark operating annual average PM2.5 concentrations for each plant was conditions at capacity factors of 85 percent, the deaths calculated as the product of baseline deaths (by cause) attributable to SO2 emissions would increase to approxi- and the fraction of deaths attributable to sulfates. The mately 17,900 per year, with an associated 457,000 YLL. fraction of deaths attributable to sulfates for each disease The number of deaths per plant varies from more than is given by 1-exp(β*ΔC), where β is the change in the 1,300 to fewer than 20. The 30 plants with the highest relative risk attributable to a one microgram per cubic number of deaths account for 78 percent of the total meter change in PM2.5, and ΔC is the population-weighted deaths and 56 percent of the total generation capacity. The change in ambient PM2.5 concentrations associated with 20 plants with the highest number of deaths account for SO2 emissions from the plant. The β coefficients were 65 percent of total deaths. Deaths per plant are correlated calculated using the IERs for ischemic heart disease, with total emissions (r = 0.38) and also with the size of the stroke, lung cancer, COPD, and ALRI developed by exposed population. Population density in India is highest Burnett and others (2014) and reported by the Institute in the north of India, which is also the part of India with for Health Metrics and Evaluation (IHME).7 For each the highest levels of ambient PM2.5. Therefore, it is not disease, the change in relative risk (β) was evaluated at surprising that the 30 plants with the highest number of the population-weighted annual average exposures for deaths (map 13.2) are located in northern India. India used in the 2010 Global Burden of Disease (Brauer and others 2012).8 Baseline deaths by age and cause were obtained from the IHME.9 Costs and Benefits of FGD Units We also calculate the years of life lost (YLL) associ- An FGD unit is an end-of-pipe technology that ated with mortality attributable to SO2 emissions. We removes SO2 from combustion gases before they exit estimate that, on average, each death is associated with the smokestack. Flue gases are treated with an alkaline Costs and Benefits of Installing Flue-Gas Desulfurization Units at Coal-Fired Power Plants in India 241 Table 13.2 Deaths and DALYs Associated with SO2 Emissions, by Plant Deaths DALYs Cause Mean 25%ile 75%ile Mean 25%ile 75%ile COPD 18.5 5.5 20.1 351 104 392 Stroke 106 31.4 115 2,230 660 2,490 IHD 58.8 17.4 63.8 1,350 399 1,510 ALRI 29.9 8.9 32.5 1,510 449 1,690 Lung cancer 2.3 0.68 2.6 60.1 17.8 67.1 All causes 216 64 241 5,500 1,630 6,150 All causes, benchmark conditions 249 77.3 301 6,360 1,974 7,690 Note: ALRI = acute lower respiratory infection; COPD = chronic obstructive pulmonary disease; DALY = disability-adjusted life year; IHD = ischemic heart disease; SO2 = sulphur dioxide. Map 13.2 Coal-Fired Power Plants: Top 30 Sulfate Deaths of the scrubbers installed in the United States are wet scrubbers (USEPA 2004).11 Another rapidly expanding Dushanbe Ashgabat TURKMENISTAN TAJIKISTAN NATIONAL CAPITALS STATE BOUNDARIES (INDIA) technology is seawater FGD (swFGD) units. These units INTERNATIONAL BOUNDARIES use the alkalinity of seawater to remove SO2 from the Kabul C H I N A flue gases; the by-product is water, which is then treated AFGHANISTAN Islamabad and discharged back into the sea. Seawater FGD units are ISLAMIC Ropar capable of reducing SO2 emissions up to 95 percent, REPUBLIC OF PAKISTAN Lehra Mohabbat Lehra PAKISTANBhatinda depending on the technology used.12 IRAN Panipat Suratgarh I.P. Station Dadri (Nctpp) (Nctpp) BHUTAN Both scrubber technologies are in use in India. The New Delhi NEPAL Badarpur H'ganj B B Kathmandu Thimphu Indian Supreme Court required the installation of an Tanda Unchahar Kahalgaon FGD unit at the Dahanu plant in Maharashtra. FGD Barauni 25°N Obra Thermal Anpara Stps Rihand Stps Rihand Singrauli Stps Singrauli Santaldih Chandrapura Farakka Stps Stps units are also in operation at the Trombay and Udupi Bakreswar Dhaka Muscat plants (table 13.3).13 Both Dahanu and Trombay have Patratu D.P.L. D.P.L. Vindh'chal Stps Vindh'chal Stps Mejia Mejia Wanakbori OMAN INDIA Kolaghat Talcher Stps MYANMAR swFGD units. Seawater FGD units have lower capital BANGLADESH 20°N Chandrapur Chandrapur Nay Pyi Taw costs and much lower variable costs than wFGD units, R'gundem R'gundemStps Stps but they can be installed only in coastal areas. We assume that FGD units installed at plants in coastal areas are Arabian Sea 15°N 15°N swFGD units, and that FGD units installed at all other Bay of Bengal locations are wFGD units. The estimates of FGD unit costs are based on tariff orders issued by State Electricity Regulatory Commissions 10°N Andaman 10°N (SERCs) for power plants that currently operate an FGD SRI LANKA S ea unit or for plants that are planning to install one in the I N D I A N O C E A N Colombo COAL-FIRED POWER PLANTS: TOP 30 SULFATE DEATHS Central government–owned power plants near future (table 13.3). We also use information from 5°N State government–owned power plants tariff determination norms and calculations of bench- Malé MALDIVES 185.206110–226.330037 226.330038–283.269279 mark capital costs used by the Central Electricity 283.269280–375.241327 375.241328–562.857852 Regulatory Commission (CERC 2009).14 IBRD 42286 | December 2016 562.857853–1,138.496675 Note: SO2 = sulfur dioxide. The size of the circle is Table 13.4 shows the assumptions used to construct 0° proportional to the estimated deaths associated with SO2 emissions under baseline conditions. the cost estimates. We assume a capital cost of US$84,000/ 60°E 65°E 70°E 75°E 80° E 85°E 90°E 95°E MW for a swFGD unit (MERC 2009, 2011) and a cost of US$109,000/MW for a wFGD unit, based on 250 MW units.15 For smaller units, we assume that the elasticity of substance that reacts with the acidic SO2 to form a capital costs with respect to installed capacity is by-product that is removed before flue gases are emitted. −0.3 (Cichanowicz 2010). The greater costs for wFGD In a wet limestone FGD (wFGD) unit, gases are treated units reflect the expenditures for reagent handling with limestone slurry, which is sprayed on the gas in an and by-product disposal facilities. In contrast, absorber unit. Gypsum, which can be sold commercially, swFGD units discharge their water by-product back into is produced as a by-product. Approximately 85 percent the sea and do not require as much capital investment. 242 Injury Prevention and Environmental Health Table 13.3 FGD Units in India, Planned and Operational Location Company State Status of FGD unit Capacity Manufacturer Type Trombay TATA Power Maharashtra Operating Unit 5: 500 MW ABB Seawater Trombay TATA Power Maharashtra Operating Unit 8: 250 MW — Seawater Ratnagiri JSW Maharashtra Under construction 1,200 MW (4 units × Alstom Seawater 300 MW) Udupi LANCO Karnataka Operating 1,200 MW (2 units × Ducon Wet limestone 600 MW) Dahanu RELIANCE Maharashtra Operating 500 MW Ducon Seawater Bongaigaon NTPC Assam Under construction 750 MW BHEL Wet limestone Vindhyachal, stage V NTPC Madhya Pradesh Planned 500 MW BHEL Wet limestone Mundra, stage III ADANI Gujarat Planned 1,980 MW — Seawater Note: FGD = flue-gas desulfurization; MW = megawatt; — = not available. Table 13.4 Operating Characteristics for Cost reduction in deaths attributable to SO2 emissions. An Calculations: Baseline Assumptions important question is the period over which this reduction would occur. Apte and others (2015) assume Benchmark no lag between emissions reductions and the associ- Capacity utilization (%) 85 ated reductions in deaths. USEPA (2011) assumes that Capital discount rate (%) 8 80 percent of the reduction in PM2.5 mortality is Plant life (retrofit) (yrs) 20 achieved within five years of the reduction in emis- sions. We view our calculations as representing the FGD Unit Type benefits of a scrubber that has been in operation for at Wet limestone Seawater least 5 years and therefore assume that 80 percent of the reduction in mortality has been achieved in calcu- Capital costs (US$/MW) $109,091 $84,364 lating lives saved and DALYs averted. Fixed operating costs $0.473 $0.364 (US$/MWh) Electricity costs (US$/kWh) $0.0636 $0.0636 RESULTS Auxiliary consumption (%) 1.5 1.25 Our benchmark calculations suggest that requiring all FGD unit efficiency (%) 90 90 72 plants in our study to install scrubbers would save Retrofit cost factor (%) 30 30 12,890 lives and 329,000 DALYs annually, at an average Note: The capital costs above are derived from information for the Dahanu (seawater cost of US$131,000 per life saved or US$5,140 per DALY FGD unit) and Bongaigaon (wet FGD unit) power plants. In both cases, the costs reflect averted (table 13.6).17 The cost per life saved (CPLS) of installation of an FGD unit in a new plant and not a retrofit. Costs are increased by installing a scrubber, however, varies greatly across plants, 30 percent to reflect the higher costs of retrofitting a scrubber. FGD = flue-gas desulfurization; kWh = kilowatt hour; MW = megawatt; MWh = megawatt hour. from US$24,700 to US$1,244,000, depending on the mag- nitude of the plant’s emissions and the size of the exposed population. If plants are ranked by their CPLS, retrofitting As a comparison, these figures are slightly lower than scrubbers at the 30 most cost-effective plants would save wFGD unit prices in the United States prior to the post- approximately 9,200 lives at an average cost of US$67,000 2006 spike in prices.16 Capital and operating costs per per life saved or US$2,600 per DALY averted.18 Requiring plant are summarized in table 13.5. We note that the cost scrubbers at the 30 plants with the highest deaths associ- per ton of SO2 removed implied by our calculations is, ated with SO2 emissions would save more lives and DALYs on average, US$613 (2013 US$). (10,060 and 257,000, respectively) at a higher average To calculate the health benefits of installing an FGD CPLS of US$96,000. This finding is not surprising: lives unit, we assume that the FGD unit will remove saved (the denominator when calculating CPLS) is 90 percent of SO2 emissions. Because of the linearity increasing in the number of deaths associated with the of sulfate formation and the approximate linearity of plant when operating without an FGD unit; hence, CPLS relative risk for a small change in concentrations, is negatively correlated with deaths attributable to base- this reduction in emissions implies a 90 percent line SO2 emissions for each plant (r = –0.43).19 Costs and Benefits of Installing Flue-Gas Desulfurization Units at Coal-Fired Power Plants in India 243 Table 13.5 Capital and Operating Costs of FGD Unit Installation per Plant Summary Statistics: Benchmark Operations Standard Average deviation Median Minimum Maximum SO2 emissions (tons/yr) 42,678 30,344 36,405 2,704 169,192 FGD unit capital costs (US$, millions) 133 95.4 119 10.9 462 Operating costs, fixed (US$, millions) 3.3 2.4 3.0 0.22 11.5 Operating costs, variable (US$, millions) 6.7 4.8 6.0 0.4 23.2 Total annual FGD unit cost (US$, millions) 23.5 16.9 21.1 1.77 81.7 Note: Number of observations = 72 power plants. Calculations based on benchmark capacity utilization of 85 percent. FGD = flue-gas desulfurization. SO2 = sulphur dioxide. Table 13.6 Cost-Effectiveness of FGD Unit Installation, US$ Cost per Life Saved, by Plant Cost per DALY Averted, by Plant Total lives Total cost Total DALYs saved (millions) Median Minimum Maximum averted Median Minimum Maximum All plants 12,890 $1,691 $167,000 $24,700 $1,244,000 329,000 $6,540 $967 $48,713 30 plants with lowest CPLS 9,196 $615 $62,490 $24,707 $137,474 235,000 $2,447 $967 $5,383 30 plants with most deaths 10,061 $965 $111,980 $24,707 $381,676 257,000 $4,385 $967 $14,944 30 largest plants (MW) 7,910 $1,164 $251,980 $33,439 $1,244,127 202,000 $9,866 $1,309 $48,713 Note: Calculations based on benchmark operating conditions (capacity factor of 85 percent), assuming an FGD unit removes 90 percent of flue gases. The number of lives saved (or DALYs averted) is the number saved five years after installation, holding population and death rates at 2013 levels. CPLS = cost per life saved; DALY = disability-adjusted life year; FGD = flue-gas desulfurization; MW = megawatt. However, identifying plants with the lowest CPLS Our estimates are sensitive to assumptions about may be difficult from a policy perspective. A more scrubbing costs, as well as to assumptions about health likely option would be to require the largest plants to impacts. Our baseline discount rate of 8 percent is a scrub their emissions. The 30 largest plants in terms of social discount rate, equal to the rate of interest on installed capacity account for 61 percent of sulfate government bonds in India. If this is replaced by the deaths. Requiring them to be retrofitted with FGD weighted private cost of capital, which we estimate to be units would save approximately 7,910 lives and 11.2 percent, the CPLS would increase by 14.3 percent, 202,000 DALYs, at an average CPLS of US$147,000 from US$131,000 to US$150,000.21 Reducing capacity (US$5,760 per DALY averted). This approach clearly factors from the benchmark level of 0.85 to 0.68 would delivers fewer health benefits per dollar spent than increase the CPLS by approximately 20 percent. At the requiring the plants associated with the largest num- same time, our estimate of the impact of a cessation lag ber of deaths and DALYs to scrub their emissions. is quite conservative. We effectively assume that only Although economies of scale exist in scrubber instal- 80 percent of the ultimate mortality benefits of scrub- lation, and although deaths are positively correlated bing will be received. Eliminating the cessation lag would with plant size, the effectiveness of a scrubber also reduce the CPLS by 20 percent. depends on the size of the exposed population; the We also note that retrofitting power plants with largest plants are not necessarily those with the largest scrubber units would increase the cost of electricity. In exposed populations.20 Cropper and others (2012), we estimate that a swFGD To maximize the number of lives saved for a given unit would increase the levelized cost of electricity by amount spent, plants with the lowest CPLS would be approximately 9 percent. A wFGD unit could increase retrofitted first. These are not necessarily the largest the cost by up to 15 percent. plants. The benefits of installing a scrubber depend on the size of the exposed population, which depends on DISCUSSION plant location. The 30 plants with the lowest CPLS asso- ciated with SO2 emissions are primarily located in Compared to coal mined in the rest of the world, domes- densely populated northern India, primarily in Uttar tic coal in India has high ash content but low sulfur Pradesh, West Bengal, Punjab, Haryana, and Jharkhand. content. Since 1984, regulations have limited particulate 244 Injury Prevention and Environmental Health matter emitted directly from coal-fired power plants; annually at an average cost of US$131,000 per life saved however, before December 2015, no regulations existed (US$5,140 per DALY averted). However, considerable that would limit secondary particle formation by restrict- heterogeneity exists in the CPLS across plants. Targeting ing emissions of SO2 or NOx.22 Plants are, however, the retrofitting regulation to plants with lower CPLS subject to minimum stack height requirements and would be more cost-effective. plants generating 500 MW of electricity or more are For any of the policy options considered, a relevant required to leave space to allow for an FGD unit retrofit question is whether the CPLS is less than the value of the in the future. Plants generating between 210 and 500 MW associated mortality reductions, measured in terms of of electricity must have stacks at least 220 meters in what people are willing to pay for them. In the United height; units that generate more than 500 MW of elec- States and other Organisation for Economic tricity must have stacks at least 275 meters in height. Co-operation and Development (OECD) countries, Taller stacks decrease ambient SO2 concentrations by the value of mortality risk reductions is measured by the causing the particulate matter they emit to be dispersed value per statistical life (VSL)—the sum of what peo- over a larger area, but they do not eliminate exposure, ple would pay for small risk reductions that sum to one especially in densely populated areas. statistical life saved.26 Both the United States and OECD In December 2015, the Ministry of Environment, countries have adopted official values for the VSL that Forests and Climate Change issued limits on SO2 emis- are used in benefit-cost analyses of environmental poli- sions.23 Plants built before 2017 that generate more than cies. Whether FGD units pass the benefit-cost test 500 MW of electricity are restricted to SO2 emissions of requires an estimate of the VSL for India. 200 milligrams per cubic meter (mg/Nm3); plants that Estimates of the VSL for India could be based on generate less than 500 MW are restricted to SO2 emis- empirical studies conducted in India or could be sions of 600 mg/Nm3.24 A plant burning coal that con- transferred from United States and OECD values, tak- tains 0.5 percent sulfur by weight emits approximately ing into account differences in incomes. Empirical 1,350 mg/Nm3, thus violating current standards. estimates of the VSL in India range widely, from Retrofitting the plant with an FGD unit would permit US$57,000 (Bhattacharya, Alberini, and Cropper the plant to achieve the Ministry’s standards.25 Currently, 2007) to US$407,000 (Madheswaran 2007).27 three plants in India have installed FGD units—Dahanu Transferring the USEPA’s VSL from the United States (Maharashtra), Trombay (Maharashtra), and Udupi to India at current exchange rates (using an income (Karnataka). According to the Central Electricity elasticity of one) implies a VSL of US$250,000.28 This Authority, eight FGD units either are in operation or are suggests that a program to retrofit FGD units on all in the planning stages (table 13.3). 72 power plants in our study would pass the Our analysis suggests that the emphasis placed on benefit-cost test, on average. FGD unit installation SO2 controls is warranted. The historic approach— also would pass the benefit-cost test on an individual relying on tall stacks—mirrors the approach taken in plant basis at most of the plants in the study, includ- the United States in the 1980s to achieve local air ing the 30 plants with the lowest CPLS.29 quality standards. Although Indian coal has lower Because big plants are easier to target, regulations that sulfur content than coal mined in the eastern United would require the retrofitting of FGD units at the largest States, a greater amount of coal is used to produce a plants (those with the largest installed capacity) might kWh of electricity in India because of the low heating be possible. The CPLS averaged over the 30 largest plants value of Indian coal. In addition, the increase in in our sample is US$147,000, suggesting that this regula- imported coal with higher sulfur content will poten- tion would, on average, pass the benefit-cost test. tially increase the average sulfur content of coal used However, targeting the installation of FGD units to in Indian power plants. The large numbers of people plants with the highest number of deaths would save exposed combined with the magnitude of SO2 emis- more lives per dollar spent. sions from coal-fired power plants makes SO2 a key pollutant of concern from a health standpoint. ACKNOWLEDGMENTS We thank Resources for the Future and the World Bank for CONCLUSIONS funding, and Zachary Lazri and Anna Malinovskaya for Our analysis suggests that retrofitting existing plants excellent research assistance. We dedicate this chapter to with FGD units could yield significant health benefits. Shama Gamkhar, our coauthor, who died before it was Requiring all 72 plants in our sample to retrofit FGD completed. We also thank Russ Dickerson, Jeremy Schreifels, units would save almost 13,000 lives (330,000 DALYs) and two anonymous referees for helpful comments. Costs and Benefits of Installing Flue-Gas Desulfurization Units at Coal-Fired Power Plants in India 245 NOTES 12. USEPA’s AP-42 database indicates that a swFGD unit can achieve up to 95 percent SO2 removal; equipment suppliers World Bank Income Classifications as of July 2014 are as claim SO2 removal efficiencies of up to 99 percent with follows, based on estimates of gross national income additives in the flue gas stream. (GNI) per capita for 2013: 13. The only plants in table 13.3 that are in our sample are the Trombay, Udupi, and Vindhyachal plants. • Low-income countries (LICs) = US$1,045 or less 14. The CERC is responsible for tariff determination for all • Middle-income countries (MICs) are subdivided: central government–owned power plants and those selling a) lower-middle-income = US$1,046 to US$4,125 inter-state power. The guidelines established by the CERC b) upper-middle-income (UMICs) = US$4,126 to US$12,745 are also used by individual state SERCs in their tariff • High-income countries (HICs) = US$12,746 or more. calculations. All costs in Indian rupees (Rsk) have been converted to US$ using an exchange rate of US$1 = 55Rsk 1. The Indian fiscal year runs from April 1 of each year and are thus in 2013 US$. through March 31. 15. Personal communication with an NTPC (India’s largest 2. Average installed capacity of all coal-fired plants in March power utility) engineer. NTPC is involved in setting up a 2015 was 1,067 MW, and median installed capacity was new plant in Bongaigaon, Assam, that will have a wFGD 950 MW, which is slightly larger than for our 72 plants. unit installed. The FGD unit is being provided by an 3. Total emissions of SO2 are calculated using the sulfur con- Indian company, BHEL. According to online sources, tent of coal and coal consumption, as well as assumptions BHEL reports a rule of thumb cost estimate for a wFGD about the volume of flue gases per ton of coal burned. unit of US$90,700/MW. We use the more conservative 4. The CERC’s benchmark operating conditions are used in estimate. tariff setting by the central government (CERC 2009). We 16. https://www.eia.gov/electricity/annual/html/epa_09_04 also use these benchmark conditions in calculating the .html. See also Muller (2016). annualized cost of operating an FGD unit. 17. The average CPLS of requiring all plants to scrub their 5. See http:// www.camx.com. emissions is the total cost listed in table 13.6 (US$1.69 6. The meteorological data (wind, temperature, pressure, billion) divided by the lives saved. Similarly, the average relative humidity, and precipitation) are derived from cost per DALY averted is US$1.69 billion divided by the the global reanalysis database of the National Center for DALYs averted (329,000). Environmental Prediction (NCEP) and processed through 18. A ranking based on CPLS is identical to a ranking based the Weather Research and Forecasting (WRF) meteorolog- on cost per DALY. A simplifying assumption underlying ical model at a one-hour temporal resolution. the calculations (as in the 2013 GBD) is that the age distri- 7. See http://ghdx.healthdata.org/record/global-burden bution of the population and death rates by age and cause -disease-study-2010-gbd-2010-ambient-air-pollution are uniform throughout the country. -risk-model-1990-2010. 19. Twenty-one of the plants with the lowest CPLS are also the 8. Specifically, we evaluated the change in relative risk at plants with the largest number of deaths associated with the population-weighted ambient concentration of PM2.5 SO2 emissions. within a 100 kilometer radius surrounding each plant, 20. The 13th largest plant in the sample, based on installed computed using the supplementary material from Brauer capacity, has the highest CPLS (US$1,244,000). The plant and others (2012). Concentrations ranged from 15 to is located in the south of India and has a smaller exposed 46 μg/m3, with a mean of 27 μg/m3. population than plants in northern India. 9. We use death rates by age and cause reported in the 2013 21. Our estimate of the private cost of capital is based on Global Burden of Disease. https://cloud.ihme.washington a debt-equity ratio of 70–30, the private rate of return .edu/index.php/s/b89390325f728bbd99de0356 on capital allowed by the CERC (15.5 percent), and the d3be6900?path=%2FIHME%20GBD%202013%20 assumption that the plant can borrow at a rate of 9.3% Deaths%20by%20Cause%201990-2013. (the Bank of India base rate at the time of writing). 10. YLL are calculated for each cause of death by multiplying 22. Prior to December 2015, emission limits for total sus- the number of deaths by the average number of life years pended particulates called for units below 210 MW to lost based on the age distribution of deaths. YLL are then emit no more than 350 mg/Nm3 and units greater than summed across all five causes of death. 210 MW no more than 150 mg/Nm3. The use of coal 11. Dry scrubber technologies (including spray dry scrubbers with ash content exceeding 34 percent is prohibited in any and circulating fluidized bed scrubbers) have lower capital thermal power plant located more than 1,000 km from the costs than wFGD units and lower removal rates. These are pithead or in urban, sensitive, or critically polluted areas. much less commonly used than wFGD units (Carpenter http://cpcb.nic.in/Industry_Specific_Standards.php. 2014), and we have no cost data on their operation in 23. Gazette of India, December 8, 2015. Ministry of Environment, India. Therefore, we do not analyze them as a control Forests and Climate Change Notification. S.O. 3305(E). option. Environment (Protection) Amendment Rules, 2015. 246 Injury Prevention and Environmental Health 24. Plants built after 2017 may emit no more than 100 mg Cichanowicz, J. E. 2010. “Current Capital Cost and Cost- of SO2 per Nm3. These plants certainly would require Effectiveness of Power Plant Emissions Control FGD units; however the cost of installing scrubbers when Technologies.” Prepared for Utility Air Regulatory Group. plants are built is lower than the cost of retrofitting them. Cohen, A. J., H. R. Anderson, B. Ostro, K. D. Pandey, 25. A referee notes that the 600 mg/Nm3 standard could be M. Krzyzanowski, and others. 2004. “Mortality Impacts achieved by installing a dry scrubber, which would have of Urban Air Pollution.” In Comparative Quantification of lower capital costs than a wFGD unit. Health Risks: Global and Regional Burden of Disease Due to 26. To illustrate, if each of 10,000 people were willing to Selected Major Risk Factors, volume 2, edited by M. Ezzati, pay US$100 over the coming year to reduce their risk of A. D. Lopez, A. Rodgers, and C. U. J. L. Murray. Geneva: dying by 1 in 10,000 during this period, on average, one World Health Organization. statistical life would be saved and the VSL would equal Cropper, M., S. Gamkhar, K. Malik, A. Limonov, and US$100 × 10,000 or US$1,000,000. I. Partridge. 2012. “The Health Effects of Coal Electricity 27. Both values were obtained by converting Indian rupees (Rsk) Generation in India.” Discussion Paper 12–25, Resources to US$ using the average exchange rate for 2007 and then for the Future, Washington, DC. converting to 2013 US$ using the Consumer Price Index. Garg, A., M. Kapshe, P. R. Shukla, and D. Ghosh. 2002. “Large 28. USEPA’s official VSL is US$7.4 million (2006 US$), imply- Point Source (LPS) Emissions from India: Regional and ing a VSL to per capita income ratio of 159:1 (USEPA Sectoral Analysis.” Atmospheric Environment 36 (2): 2011). Applying this ratio to per capita income in India in 213–24. 2014–15 (US$1,570) yields a VSL of US$250,000. GBD 2013 Risk Factor Collaborators. 2015. “Global, Regional, 29. Forty-seven of the 72 plants have a CPLS of less than and National Comparative Risk Assessment of 79 Behavioral, US$250,000; 64 have a CPLS of less than US$407,000. Environmental and Occupational, and Metabolic Risks or Clusters of Risks in 188 countries, 1990–2013: A Systematic Analysis for the Global Burden of Disease Study 2013.” REFERENCES The Lancet 386 (10010): 2287–323. Apte, J. S., J. D. Marshall, A. J. Cohen, and M. Brauer. Guttikunda, S. K., and P. Jawahar. 2014. “Atmospheric Emissions 2015. “Addressing Global Mortality from Ambient PM2.5.” and Pollution from the Coal-fired Thermal Power Plants in Environmental Science and Technology 49: 8057–66. India.” Atmospheric Environment 92: 449–60. Bhattacharya, S., A. Alberini, and M. Cropper. 2007. “The Value Lim, S. S., T. Vos, A. D. Flaxman, G. Danaei, K. Shibuya, and of Mortality Risk Reductions in Delhi, India.” Journal of others. 2013. “A Comparative Risk Assessment of Burden of Risk and Uncertainty 34 (1): 21–47. Disease and Injury Attributable to 67 Risk Factors and Risk Brauer M., M. Amann, R. T. Burnett, A. Cohen, F. Dentener, Factor Clusters in 21 Regions, 1990–2010: A Systematic and others. 2012. “Exposure Assessment for Estimation Analysis for the Global Burden of Disease Study 2010.” of the Global Burden of Disease Attributable to Outdoor The Lancet 380 (9859): 2224–60. Air Pollution.” Environmental Science and Technology 46: Lu, Z., D. G. Streets, B. de Foy, and N. A Krotkov. 2013. “Ozone 652–60. Monitoring Instrument Observations of Interannual Burnett, R. T., C. A. Pope III, M. Ezzati, C. Olives, S. S. Lim, and Increases in SO2 Emissions from Indian Coal-Fired Power others. 2014. “An Integrated Risk Function for Estimating Plants during 2005–2012.” Environmental Science and the Global Burden of Disease Attributable to Ambient Technology 47: 13993–4000. Fine Particulate Matter Exposure.” Environmental Health Madheswaran, S. 2007. “Measuring the Value of Statistical Perspectives 122 (4): 397–403. Life: Estimating Compensating Wage Differentials among Carpenter, A. 2014. “Water-Saving FGD Technologies.” Workers in India.” Social Indicators Research 84: 83–96. Cornerstone. http://cornerstonemag.net/tag/dry-scrubbers. Malik, K. 2013. “Essays on Energy and Environment in CEA (Central Electricity Authority). 2015. Growth of Electricity India.” PhD dissertation, University of Maryland, College Sector in India from 1947–2015. New Delhi: Government Park, MD. of India. MERC (Maharashtra Electricity Regulatory Commission). 2009. CERC (Central Electricity Regulatory Commission). 2009. “MERC Order for RInfra-G for APR of FY 2009-10 and Tariff Determination Methodology for Thermal Power Plants. Determination of Tariff for FY 2010–11.” Case 99 of 2009, New Delhi: Government of India. Maharashtra Electricity Regulatory Commission, Mumbai. Chan, H. S., M. L. Cropper, and K. Malik. 2014. “Why Are ———. 2011. “MERC Order for Truing Up of FY 2009-10 Power Plants in India Less Efficient Than Power Plants in and APR of FY 2010–11.” Case 122 of 2011, Maharashtra the United States?” American Economic Review Papers and Electricity Regulatory Commission, Mumbai. Proceedings 104 (5): 586–90. Ministry of Environmental Protection, People’s Republic of Chikkatur, A. P., and A. D. Sagar. 2007. “Cleaner Power in China. 2014. List of SO2 Scrubbers in Coal-Fired Power India: Towards a Clean-Coal-Technology Roadmap.” Belfer Plants. http://english.mep.gov.cn. Center for Science and International Affairs Discussion Paper Muller, N. Z. 2016. “Environmental Benefit-Cost Analysis and the 6: 1–261. National Accounts.” NBER Working Paper, Cambridge, MA. Costs and Benefits of Installing Flue-Gas Desulfurization Units at Coal-Fired Power Plants in India 247 Pope, C. A., III, R. T. Burnett, M. J. Thun, E. E. Calle, D. Krewski, Schmalensee, R., and R. N. Stavins. 2013. “The SO2 Allowance and others. 2002. “Lung Cancer, Cardiopulmonary Trading System: The Ironic History of a Grand Policy Mortality, and Long-Term Exposure to Fine Particulate Experiment.” Journal of Economic Literature 27 (1): 103–22. Air Pollution.” Journal of the American Medical Association USEPA (United States Environmental Protection Agency). 287 (9): 1132–41. 2004. Air Pollution Control Technology Fact Sheet. Reddy, M. S., and C. Venkataraman. 2002. “Inventory of Aerosol Washington, DC: USEPA. and Sulphur Dioxide Emissions from India: I—Fossil Fuel ———. 2011. The Benefits and Costs of the Clean Air Act from Combustion.” Atmospheric Environment 36 (4): 677–97. 1990 to 2020. Washington, DC: USEPA. 248 Injury Prevention and Environmental Health DCP3 Series Acknowledgments Disease Control Priorities, third edition (DCP3) com- We thank the many contractors and consultants who piles the global health knowledge of institutions and provided support to specific volumes in the form of experts from around the world, a task that required the economic analytical work, volume coordination, chapter efforts of over 500 individuals, including volume edi- drafting, and meeting organization: the Center for tors, chapter authors, peer reviewers, advisory com- Disease Dynamics, Economics & Policy; Center for mittee members, and research and staff assistants. For Chronic Disease Control; Centre for Global Health each of these contributions we convey our acknowledg- Research; Emory University; Evidence to Policy Initiative; ment and appreciation. First and foremost, we would Public Health Foundation of India; QURE Healthcare; like to thank our 32 volume editors who provided the University of California, San Francisco; University of intellectual vision for their volumes based on years of Waterloo; University of Queensland; and the World professional work in their respective fields, and then Health Organization. dedicated long hours to reviewing each chapter, pro- We are tremendously grateful for the wisdom and viding leadership and guidance to authors, and fram- guidance provided by our advisory committee to the ing and writing the summary chapters. We also thank editors. Steered by Chair Anne Mills, the advisory com- our chapter authors who collectively volunteered their mittee ensures quality and intellectual rigor of the high- time and expertise to writing over 170 comprehensive, est order for DCP3. evidence-based chapters. The National Academies of Sciences, Engineering, We owe immense gratitude to the institutional spon- and Medicine, in collaboration with the Interacademy sor of this effort: The Bill & Melinda Gates Foundation. Medical Panel, coordinated the peer-review process for The Foundation provided sole financial support of all DCP3 chapters. Patrick Kelley, Gillian Buckley, Megan the Disease Control Priorities Network (DCPN). Many Ginivan, Rachel Pittluck, and Tara Mainero managed thanks to Program Officers Kathy Cahill, Philip Setel, this effort and provided critical and substantive input. Carol Medlin, Damian Walker and (currently) David World Bank Publishing provided exceptional guid- Wilson for their thoughtful interactions, guidance, and ance and support throughout the demanding produc- encouragement over the life of the project. We also wish tion and design process. We would particularly like to thank Jaime Sepúlveda for his longstanding support, to thank Carlos Rossel, Mary Fisk, Nancy Lammers, including chairing the Advisory Committee for the sec- Rumit Pancholi, Deborah Naylor, and Sherrie Brown ond edition and, more recently, demonstrating his for their diligence and expertise. Additionally, we thank vision for DCP3 while he was a special advisor to the Jose de Buerba, Mario Trubiano, Yulia Ivanova, and Gates Foundation. We are also grateful to the University Chiamaka Osuagwu of the World Bank for providing of Washington’s Department of Global Health and suc- professional counsel on communications and market- cessive chairs King Holmes and Judy Wasserheit for ing strategies. providing a home base for the DCP3 Secretariat, which Several U.S. and international institutions contrib- included intellectual collaboration, logistical coordina- uted to the organization and execution of meetings that tion, and administrative support. supported the preparation and dissemination of DCP3. 249 We would like to express our appreciation to the follow- • Union for International Cancer Control, cancer ing institutions: consultation (November 2013, December 2014) • Harvard T. H. Chan School of Public Health, eco- • University of Bergen, consultation on equity (June nomic evaluation consultation (September 2015) 2011) • University of California, Berkeley School of Public • University of California, San Francisco, surgery Health, and Stanford Medical School, occupational and volume consultations (April 2012, October 2013, environmental health consultations (December 2015). February 2014) • Institute of Medicine, first meeting of the Advisory Carol Levin provided outstanding governance for cost Committee to the Editors (March 2013) and cost-effectiveness analysis. Stéphane Verguet added • Harvard Global Health Institute, consultation on valuable guidance in applying and improving the policy measures to reduce incidence of noncommu- extended cost-effectiveness analysis method. Elizabeth nicable diseases (July 2013) Brouwer, Kristen Danforth, Nazila Dabestani, Shane • National Academy of Medicine, systems strengthen- Murphy, Zachary Olson, Jinyuan Qi, and David Watkins ing meeting (September 2013) provided exceptional research assistance and analytic • Center for Disease Dynamics, Economics & Policy assistance. Brianne Adderley ably managed the budget (Quality and Uptake meeting, September 2013; and project processes, while Jennifer Nguyen, Shamelle reproductive and maternal health volume consulta- Richards, and Jennifer Grasso contributed exceptional tion, November 2013) project coordination support. The efforts of these indi- • National Cancer Institute, cancer consultation viduals were absolutely critical to producing this series, (November 2013) and we are thankful for their commitment. 250 DCP3 Series Acknowledgments Volume and Series Editors VOLUME EDITORS the first hospital trauma registries in Sub-Saharan Africa and codeveloped the Kampala Trauma Score, now used Charles N. Mock in many low-income countries. Her research interests Charles N. Mock, MD, PhD, FACS, has training as both a include injury surveillance, injury severity measurement, trauma surgeon and an epidemiologist. He worked as a emergency trauma care systems, road safety, and surgeon in Ghana for four years, including at a rural hos- drowning. pital (Berekum) and at the Kwame Nkrumah University of Science and Technology (Kumasi). In 2005−07, he served as Director of the University of Washington’s Rachel Nugent Harborview Injury Prevention and Research Center. In Rachel Nugent is Vice President for Global 2007−10, he worked at the World Health Organization Noncommunicable Diseases at RTI International. She (WHO) headquarters in Geneva, where he was responsi- was formerly a Research Associate Professor and ble for developing the WHO’s trauma care activities. In Principal Investigator of the DCPN in the Department 2010, he returned to his position as Professor of Surgery of Global Health at the University of Washington. (with joint appointments as Professor of Epidemiology Previously, she served as Deputy Director of Global and Professor of Global Health) at the University of Health at the Center for Global Development, Director Washington. His main interests include the spectrum of of Health and Economics at the Population Reference injury control, especially as it pertains to low- and Bureau, Program Director of Health and Economics middle-income countries: surveillance, injury preven- Programs at the Fogarty International Center of the tion, prehospital care, and hospital-based trauma care. National Institutes of Health, and senior economist at He was President (2013−15) of the International the Food and Agriculture Organization of the United Association for Trauma Surgery and Intensive Care. Nations. From 1991–97, she was associate professor and department chair in economics at Pacific Lutheran University. Olive Kobusingye Olive Kobusingye is an accident and emergency surgeon and injury epidemiologist. She is on faculty at Makerere Kirk R. Smith University School of Public Health where she heads the Kirk R. Smith is Professor of Global Environmental Trauma, Injury, and Disability (TRIAD) Project and Health at University of California, Berkeley School of coordinates graduate training in those disciplines. Before Public Health. He is also founder and coordinator of the Makerere, Olive was the Regional Advisor on Violence, campus-wide Masters Program in Global Health and Injuries, and Disabilities at the World Health Environment. Previously, he was founder and head of Organization’s regional office for Africa (AFRO). She is the Energy Program of the East-West Center in Honolulu. the founding Executive Director of the Injury Control He serves on a number of national and international Center–Uganda, and founding Secretary General of the scientific advisory committees, including the Global Injury Prevention Initiative for Africa. She established Energy Assessment, National Research Council’s Board 251 on Atmospheric Sciences and Climate, the Executive Innovation (CIGI) Chair in Global Health Economics in Committee for WHO Air Quality Guidelines, the the Balsillie School of International Affairs there. She has Intergovernmental Panel on Climate Change, and the consulted for the World Bank, the Asian Development Global Burden of Disease. He is a member of the U.S. Bank, several United Nations agencies, and the National Academy of Sciences and holder of the Tyler International Development Research Centre, among and Heinz Prizes for environmental achievement. others, in work conducted in over 20 low- and middle- income countries. She led the work on nutrition for the Copenhagen Consensus in 2008, when micronutrients SERIES EDITORS were ranked as the top development priority. She has Dean T. Jamison served as associate provost of graduate studies at the University of Waterloo, vice-president academic at Dean T. Jamison is Emeritus Professor in Global Health Wilfrid Laurier University in Waterloo, and interim dean Sciences at the University of California, San Francisco, and at the University of Toronto at Scarborough. the University of Washington. He previously held aca- demic appointments at Harvard University and the University of California, Los Angeles. Prior to his academic Prabhat Jha career, he was an economist on the staff of the World Bank, Prabhat Jha is the founding director of the Centre for where he was lead author of the World Bank’s World Global Health Research at St. Michael’s Hospital and Development Report 1993: Investing in Health. He serves as holds Endowed and Canada Research Chairs in Global lead editor for DCP3 and was lead editor for the previous Health in the Dalla Lana School of Public Health at the two editions. He holds a PhD in economics from Harvard University of Toronto. He is lead investigator of the University and is an elected member of the Institute of Million Death Study in India, which quantifies the causes Medicine of the U.S. National Academy of Sciences. He of death and key risk factors in over two million homes recently served as Co-Chair and Study Director of The over a 14-year period. He is also Scientific Director of the Lancet’s Commission on Investing in Health. Statistical Alliance for Vital Events, which aims to expand reliable measurement of causes of death worldwide. His Rachel Nugent research includes the epidemiology and economics of tobacco control worldwide. See the list of Volume Editors. Ramanan Laxminarayan Hellen Gelband Ramanan Laxminarayan is Director of the Center for Hellen Gelband is an independent global health policy Disease Dynamics, Economics & Policy in Washington, expert. Her work spans infectious disease, particularly DC. His research deals with the integration of epidemi- malaria and antibiotic resistance, and noncommunica- ological models of infectious diseases and drug resis- ble disease policy, mainly in low- and middle-income tance into the economic analysis of public health countries. She has conducted policy studies at Resources problems. He was one of the key architects of the for the Future, the Center for Disease Dynamics, Affordable Medicines Facility–malaria, a novel financing Economics & Policy, the (former) Congressional Office mechanism to improve access and delay resistance to of Technology Assessment, the Institute of Medicine of antimalarial drugs. In 2012, he created the Immunization the U.S. National Academies, and a number of interna- Technical Support Unit in India, which has been credited tional organizations. with improving immunization coverage in the country. He teaches at Princeton University. Susan Horton Susan Horton is Professor at the University of Waterloo Charles N. Mock and holds the Centre for International Governance See the list of Volume Editors. 252 Volume and Series Editors Contributors Safa Abdalla Nazila Dabestani School of Medicine, Stanford University, Palo Alto, Department of Global Health, University of California, United States Washington, Seattle, Washington, United States Rajeev B. Ahuja Kristie L. Ebi Department of Burns and Plastic Surgery, Department of Global Health, University of Lok Nayak Hospital and Maulana Azad Medical College, Washington, Seattle, Washington, United States New Delhi, India Xiagming Fang Spenser S. Apramian College of Engineering, China Agricultural University, Stanford University, Palo Alto, California, United States Beijing, China Abdulgafoor M. Bachani G. Gururaj Department of International Health, Johns Hopkins National Institute of Mental Health and Neurosciences, University, Baltimore, Maryland, United States Bangalore, India Mark A. Bellis Centre for Public Health, Liverpool John Moores Sarath Guttikunda University, Liverpool, United Kingdom Division of Atmospheric Sciences, Desert Research Institute, Reno, Nevada, United States Alexander Butchart World Health Organization, Geneva, Switzerland Jeremy J. Hess Division of Emergency Medicine, University of Linda F. Cantley Washington, Seattle, Washington, United States School of Medicine, Yale University, New Haven, Connecticut, United States Susan D. Hillis U.S. Centers for Disease Control and Prevention, Atlanta, Claire Chase Georgia, United States Water and Sanitation Program, World Bank, Washington, DC, United States Connie Hoe Maureen L. Cropper Bloomberg School of Public Health, Johns Hopkins Energy Research Center, University of Maryland, University, Baltimore, Maryland, United States College Park, Maryland, United States Guy Hutton Mark R. Cullen WASH Section, United Nations Children’s Fund, New School of Medicine, Stanford University, Palo Alto, York, New York; formerly at the Water and Sanitation California, United States Program, World Bank, Washington, DC, United States 253 Adnan A. Hyder Zachary Olson Department of International Health, Johns Hopkins Berkeley School of Public Health, University of University, Baltimore, Maryland, United States California, Berkeley, California, United States Rebecca Ivers Ian Partridge The George Institute for Global Health, Sydney, School of Public Affairs, University of Texas, Austin, New South Wales, Australia Austin, Texas, United States Dean T. Jamison Margie Peden Global Health Sciences, University of California, Department for Management of Noncommunicable San Francisco, San Francisco, California, United States Diseases, Disability, Violence and Injury Prevention, Puja Jawahar World Health Organization, Geneva, Switzerland Urban Emissions, New Delhi, India Ajay Pillarisetti Lisa Keay Berkeley School of Public Health, University of The George Institute for Global Health, Sydney, California, Berkeley, California, United States New South Wales, Australia Fazlur Rahman Carol Levin Centre for Injury Prevention, Health Development Department of Global Health, University of and Research, Dhaka, Bangladesh Washington, Seattle, Washington, United States Mark L. Rosenberg David Mackie The Task Force for Global Health, Decatur, Georgia, Department of Intensive Care, Red Cross Hospital, United States Beverwijk, the Netherlands John A. Staples Kabir Malik Department of Medicine, University of Washington, World Bank, Washington, DC, United States Seattle, Washington, United States David Meddings Department for Management of Noncommunicable Stéphane Verguet Diseases, Disability, Violence and Injury Prevention, Department of Global Health and Population, Harvard World Health Organization, Geneva, Switzerland T. H. Chan School of Public Health, Boston, Massachusetts, United States James A. Mercy U.S. Centers for Disease Control and Prevention, Catherine L. Ward Atlanta, Georgia, United States Department of Psychology, University of Cape Town, Cape Town, South Africa Nam Phuong Nguyen World Health Organization, Vietnam Country Office, David A. Watkins Hanoi, Vietnam School of Medicine, University of Washington, Seattle, Washington, United States Robyn Norton The George Institute for Global Health, Sydney, Paul Watkiss New South Wales, Australia Paul Watkiss Associates, Oxford, United Kingdom 254 Contributors Advisory Committee to the Editors Anne Mills, Chair Amanda Glassman Professor, London School of Hygiene & Tropical Chief Operating Officer and Senior Fellow, Center for Medicine, London, United Kingdom Global Development, Washington, DC, United States Olusoji Adeyi Glenda Gray Director, Health, Nutrition and Population Global Executive Director, Perinatal HIV Research Unit, Practice, World Bank, Washington, DC, United States Chris Hani Baragwanath Hospital, Johannesburg, Kesetebirhan Admasu South Africa Former Minister of Health, Addis Ababa, Ethiopia Demissie Habte George Alleyne Chair of Board of Trustees, International Clinical Director Emeritus, Pan American Health Organization, Epidemiological Network, Addis Ababa, Ethiopia Washington, DC, United States Ala Alwan Richard Horton Regional Director Emeritus, World Health Organization, Editor, The Lancet, London, United Kingdom Regional Office for the Eastern Mediterranean, Cairo, Edward Kirumira Arab Republic of Egypt Dean, Faculty of Social Sciences, Makerere University, Rifat Atun Kampala, Uganda Professor, Global Health Systems, Harvard T. H. Chan School of Public Health, Boston, Massachusetts, United Peter Lachmann States Professor, University of Cambridge, Cambridge, United Kingdom Zulfiqar Bhutta Chair, Division of Women and Child Health, Aga Khan Lai Meng Looi University Hospital, Karachi, Pakistan Professor, University of Malaya, Kuala Lumpur, Agnes Binagwaho Malaysia Former Minister of Health, Kigali, Rwanda Adel Mahmoud Mark Blecher Senior Molecular Biologist, Princeton University, Senior Health Advisor, South Africa Treasury Princeton, New Jersey, United States Department, Cape Town, South Africa Patricia Garcia Anthony Measham Minister of Health, Lima, Peru World Bank (retired) Roger Glass Carol Medlin Director, Fogarty International Center, National Independent Consultant, Washington, DC, Institutes of Health, Bethesda, Maryland, United States United States 255 Alvaro Moncayo Richard Skolnik Researcher, Universidad de los Andes, Bogotá, Colombia Yale University School of Public Health (retired) Jaime Montoya Stephen Tollman Executive Director, Philippine Council for Health Professor, University of Witwatersrand, Johannesburg, Research and Development, Taguig City, the Philippines South Africa Ole Norheim Professor, Department of Global Public Health and Jürgen Unutzer Primary Care, University of Bergen, Bergen, Norway Professor, Department of Psychiatry, University of Washington, Seattle, Washington, United States Folashade Omokhodion Professor, University College Hospital, Ibadan, Nigeria Damian Walker Deputy Director of Data and Analytics, Bill & Melinda Toby Ord Gates Foundation, Seattle, Washington, United States President, Giving What We Can, Oxford, United Kingdom Ngaire Woods K. Srinath Reddy Director, Global Economic Governance Program, President, Public Health Foundation of India, Oxford University, Oxford, United Kingdom New Delhi, India Sevket Ruacan Nopadol Wora-Urai Dean, Koç University School of Medicine, Istanbul, Turkey Professor, Department of Surgery, Phramongkutklao Hospital, Bangkok, Thailand Jaime Sepúlveda Executive Director, Global Health Sciences, University Kun Zhao of California, San Francisco, San Francisco, California, Researcher, China National Health Development United States Research Center, Beijing, China 256 Advisory Committee to the Editors Reviewers Roberto Bertollini Marcus R. Keogh-Brown Scientific Committee on Health, Environmental, and Department of Global Health and Development, Emerging Risks of the European Commission, London School of Hygiene & Tropical Medicine, Luxembourg London, United Kingdom H. Ron Chan Patrick Kinney University of Manchester, Manchester, United Kingdom Department of Environmental Health Sciences, Columbia University, Mailman School of Public Carolyn J. Cumpsty-Fowler Health, New York, New York, United States Center for Injury Research and Policy, Bloomberg School of Public Health, Johns Hopkins University, Barry Kistnasamy Baltimore, Maryland, United States South African Department of Health, Johannesburg, Kathie L. Dionisio South Africa National Exposure Research Laboratory, U.S. Sharon Levi Environmental Protection Agency, Research Triangle, Beterem-Safe Kids Israel, Petah Tikva, Israel North Carolina, United States Jay P. Graham Leslie Morris-Iveson Milken Institute School of Public Health, The George Consultant, Oxford, United Kingdom Washington University, Washington, DC, United States Daniel Pope James Hammitt Department of Public Health and Policy, Harvard T. H. Chan School of Public Health, Boston, University of Liverpool, Liverpool, Massachusetts, United States United Kingdom Rema Hanna Saidur Rahman Harvard Kennedy School, Cambridge, Massachusetts, Bangladesh University of Health Sciences, Dhaka, United States Bangladesh Stephen Hargarten Gordon S. Smith Injury Research Center, Medical College of Wisconsin, School of Medicine, University of Maryland, Milwaukee, Wisconsin, United States College Park, Maryland, United States 257 Jukka Takala Yuan Xu Workplace Safety and Health Institute, Singapore Institute of Environmental, Energy, and Sustainability, The Chinese University of Hong Kong, Hong Kong Leonardo Trasande SAR, China School of Medicine, New York University, New York, New York, United States Hisham Zerriffi Elizabeth Ward Faculty of Forestry, University of British Columbia, Violence Prevention Alliance, Kingston, Jamaica Vancouver, British Columbia, Canada 258 Reviewers Policy Forum Participants The following individuals provided valuable insights to improve this volume’s key findings through participation in the Disease Control Priorities–World Health Organization, Regional Office for the Eastern Mediterranean policy forum on Road Traffic Injury Prevention and Trauma Care. The forum was held in Sharjah, the United Arab Emirates, on February 22, 2016, and was organized by Dr. Ala Alwan, Regional Director Emeritus and member of the DCP3 Advisory Committee to the Editors. Bahar Idris Abu Garda Mohamed Awadh Alrawas Federal Minister of Health, Khartoum, Sudan Director General of Traffic Affairs, Royal Oman Police, Muscat, Oman Faisal Al-Anaizi Director of Injury Prevention Programme, Ministry of Ala Alwan Health, Riyadh, Saudi Arabia Regional Director Emeritus, World Health Organization, Regional Office for the Eastern Mediterranean, Mohamed Saad AlKharji Cairo, Arab Republic of Egypt Head of Traffic Department, Ministry of Interior, Doha, Qatar Raed Arafat Deputy Health Minister, Ministry of Health, Bucharest, Wahid Al-Kharusi Romania Ambassador Emeritus, Muscat, Oman Osama Bahar Reda’a Al Menshawy Major, Directorate of Traffic Services, Ministry of Minister of Health, El-Beida, Libya Interior, Manama, Bahrain Nasir Baoum Yasser Al Naimi Minister of Public Health and Population, Sana’a, Consultant, Hospital Sector, Ministry of Health, Republic of Yemen Abu Dhabi, United Arab Emirates Gayle DiPietro Omran bin Mohammad Al Omran Global Road Safety Program Manager, Director General of Road Services Department, The International Federation of Red Cross and Ministry of Transport, Riyadh, Saudi Arabia Red Crescent Societies, Geneva, Switzerland Hussain Abdul Rahman Al-Rand Ahmed ElAnsary Assistant Undersecretary, Health Centers and Head of Ambulance and Emergency Medical Services, Clinics, Ministry of Health, Abu Dhabi, Ministry of Health and Population, Cairo, Arab United Arab Emirates Republic of Egypt 259 Gururaj Gopalakrishna Medicine, Johns Hopkins University School of Professor and Head, Department of Epidemiology, Medicine, Baltimore, Maryland, United States; Joint World Health Organization Collaborating Centre for Appointment, Department of International Health, Injury Prevention and Safety Promotion, Centre for Johns Hopkins Bloomberg School of Public Health; Public Health, National Institute of Mental Health and Visiting Faculty, Department of Emergency Medicine, Neurosciences (NIMHANS), Bangalore, India Aga Khan University, Karachi, Pakistan Mohammad Jalili Ramzi Salamé Associate Professor of Emergency Medicine and Vice Executive Director, Interministerial National Road Chancellor of Education, Tehran University of Medical Safety Council, Beirut, Lebanon Science, Tehran, Islamic Republic of Iran Essam Sharaf Junaid Razzak Former Prime Minister and Minister of Transport; Director, Telemedicine Program, and Senior Advisor Professor of Road Engineering, Cairo University, for Global Health, Department of Emergency Cairo, Arab Republic of Egypt 260 Policy Forum Participants Index Boxes, figures, maps, notes, and tables are indicated by b, f, m, n, and t respectively. A indoor. See household air pollution; India, acute lower respiratory infection (ALRI), 138, 139f, household air pollution case study 141, 223 interventions adaptation to climate change, 154, 163–64 essential interventions to address, 17t, 19 methods to estimate costs of, 160 lack of data on cost-effectiveness, 208 need for cost estimations, 166 regulation as effective intervention, 199 Afghanistan, diarrheal disease burden in, 164 LMICs vs. HICs, 5–6, 5t Africa. See also specific regions and countries mortality trends for ambient particle pollution, burns mortality rate in, 58 secondhand smoke, and ozone pollution, 7, climate change-related costs in, 161 10f, 31t, 32 diarrheal disease burden in, 164, 178t scrubbers, use of, 239. See also India, coal-fired migrant workers in, 107 power plants case study motorcycles and electric bikes in, 41 from solid fuels used for cooking and home heating, nontransport unintentional injuries in, 58 31t, 32 poisoning mortality rate in, 58 trends in burden of, 33 road traffic mortality rate in, 35 alcohol use aging population burns and, 59 falls, incidence of, 28, 56, 57t, 66 childhood violence, increased risk of, 76, 82 fall prevention programs, 60, 61t, 66 driving under the influence, 40, 45 risks of, 58–59, 66 drowning and, 29, 59 natural disasters, effect of, 58 reducing availability and harmful use of, 83, 84t violence against elderly persons, 74 ambient air pollution. See air pollution agricultural sector Angola, diarrheal disease burden in, 164 climate change, 164 Apte, J. S., 243 occupational injuries, 102 Arcenas, A., 225 pesticide poisoning of agricultural workers, 118 Argentina Ahuja, R. B., 65–66 Bolivia–Argentina migration corridor, 107 air conditioning and climate change, 164 Paraguay–Argentina migration corridor, 107 air pollution, 5–6 water pollution in, 181 coal-fired power plants, costs and benefits of, arsenic in groundwater, 177 239–48. See also India, coal-fired power Asia. See also specific regions and countries plants case study climate change deaths possible to avert through interventions, 20, 20t labor productivity and, 163 economic analysis of control, 14, 206t related costs in, 161, 162 261 diarrheal disease burden in, 164 labor productivity, climate change’s impact on, 163 lung cancer studies in, 139 violence in migrant workers in, 107 alcohol-related violence, strategies to reduce, 83 occupational injuries, participatory approaches to costs of child abuse, 79t interventions in, 115–16 costs of interpersonal violence, 76, 77–78t physical violence in the workplace in, 104t firearms licensing, effect of, 86 Stepping Stones program, 86 bullying, 72 Aunan, K., 225 burden of injuries, 1, 2, 21nn1–2, 25–34. See also Australia DALYs; mortality trends; specific types of injuries fall injuries, cost-effectiveness of prevention and health conditions programs, 64 burns and heat exposure, 56f interpersonal violence, costs of, 77t drowning, 56–57, 56f, 66 occupational injuries, decline in, 97 falls, 32, 56, 56f, 66 road traffic injuries in interpersonal violence, 71–72, 76–81, 207 costs of, 38 occupational injuries, 33, 103–8, 110 Safe Systems approach, 45 poisoning, 56f, 66 road traffic injuries, 35–38 B Burkina Faso, water, sanitation, and hygiene (WASH) Balakrishnan, K., 137 services in, 184 Banerjee, A., 207 cost-effectiveness, 191 Bangladesh hygiene education, cost-effectiveness of, 203 burns from kerosene lamps in, 60 IRC WASHCost project, 187 children in poor households, increased health risk Burnett, R. T., 140 of, 190 burns and heat exposure costs burden of injuries, 56f of drowning prevention interventions, 205 cooking as source of, 59 of intimate partner violence in, 80t costs and cost-effectiveness of interventions, 65–66 of road traffic injuries in, 38 DALYs trends in, 57t, 58 diarrheal disease burden in, 164 LMICs (all ages and genders), 28t, 29 migrant workers in, 109 education and increasing community Prevention of Child Injury through Social awareness, 63 Intervention and Education (PRECISE) International Society for Burn Injuries (ISBI), 58 including Anchal and SwimSafe, 62, 65, 201 interventions, 61t, 63, 67 textile, clothing, and footwear industry in, 111, 111t essential interventions to address, 16t Bartram, J., 207 kerosene stoves as factor, 59 Belize, costs of road traffic injuries in, 38 LMICs vs. HICs, 3, 4t, 59 benefit-cost analysis (BCA), 13–14b, 200 mortality trends in, 9, 12f, 55 in environmental health evaluations, 207 income and regions, 57t, 58 India, coal-fired power plants case study, 245 LMICs (all ages and genders), 27t, 29, 58 of water, sanitation, and hygiene (WASH), 14, males and age groups, 57t, 58 189–90, 190t, 203, 204t platform vs. floor-level cooking, 63 Bhutan, clean fuel initiatives in, 149 risk factors, 59–60 bicycle injuries. See road traffic injuries scalding from hot water taps, 63 biologic hazards, 120 smoke detectors, use of, 63 Bishai, D., 44, 200, 202 stove design improvements, 63 Blackman, A., 202 Burundi, climate change-related costs in, 162 Bloomberg Philanthropies Global Road Safety Program, 45, 46b C Bolivia–Argentina migration corridor, 107 Cambodia Bosello, F., 161 water, sanitation, and hygiene (WASH) services in Brazil cost-effectiveness, 191 climate change-related costs in, 161, 162 microfinance linked to household sanitation, 185 ergonomic changes in footwear manufacturing in, 15 Work Improvement in Safe Home program, 116 262 Index Cameron, J., 203 improvements, effect of, 183 Cameroon, violence against women in, 74 vulnerability of children, 177 Canada child labor, 30, 108, 116 climate change in Children and Violence Evaluation Challenge Fund, 83 related costs, 162 child sexual abuse, 74 related labor productivity improvements, 162 Chile costs of intimate partner violence in, 79t, 81 costs electronics industry in, 110 of intimate partner violence in, 79t lung cancer studies in, 139 of road traffic injuries in, 38 Youth Relationship Project, 83 of violence against women in, 81 cardiovascular disease, 141, 141f labor productivity, climate change’s impact on, 163 Caribbean. See Latin America and the Caribbean China cataracts, 139, 140t benefit-cost analysis of clean energy in, 14, 225 Centers for Disease Control and Prevention (CDC), 58 coal-fired power plants in, 239 CEREAL (Centro de Reflexión y Acción Laboral, or diarrheal disease burden in, 164 Centre for Reflection and Action on Labour household air pollution trends in Issues), 114 coal stoves with chimneys, introduction of Chad, diarrheal disease burden in, 164 (National Improved Stove Program), 136, 139, chemical contamination, 4 144, 148 essential interventions to address, 17t impact on outdoor pollution, 147 exposure to chemicals and toxins, 4, 29, 30t, 119–21 labor productivity, climate change’s impact essential interventions to address, 17t on, 162, 163 prevention, 120–21 occupational injuries in types of hazards, 120 electronics industry, 110 Chiabai, A., 161, 162 health care sector, 117 child abuse health services coverage for, 109 Children and Violence Evaluation Challenge Fund, 83 manufacturing sector, 106t costs and cost-effectiveness of interventions, 88t migrant workers’ risk, 108, 109 costs associated with, 76, 78–79t participatory discussions with health care DALYs among youth ages 10 to 24 years, 74 providers to prevent, 116 interventions, 83 textile, clothing, and footwear industry, 111 essential interventions to address, 16t, 19 road traffic injuries, costs of, 38 as risk factor for youth violence and intimate tobacco taxes in, 224 partner violence, 82 water, sanitation, and hygiene (WASH) services in street children, increased risk of, 75 cost-effectiveness, 191 child health water insecurity in Northern China, 180 acute lower respiratory infection (ALRI), 138, 139f, Chisholm, D., 44, 202 141, 223 cholera, 178 child safety caps on medications, 64 chronic obstructive pulmonary disease (COPD), diarrheal illnesses responsible for deaths, 31 138–39 drowning and, 29, 59 CIRCLE (Climate Impact Research and Response interventions, 61t Coordination for a Larger Europe) study, prevention programs, 62 161, 163 falls, risk of, 59, 66 civil conflict. See collective violence and conflicts interventions, 61, 61t, 66 climate change, 153–70 India, household air pollution case study, 223 adaptation, 154, 163–64 noncommunicable diseases, increased risk of due to methods to estimate costs of, 160 violence during childhood, 76 need for cost estimations, 166 poisoning cases, 28, 60 agriculture sector, 164 water, sanitation, and hygiene-related illnesses, air conditioning and, 164 31, 177, 178 challenges related to estimating costs and children in poor households, increased health benefits, 166 risk of as justification for targeting, 190 climate-resilient health systems, 157–58, 159t Index 263 costs of, 157, 208 by income levels, 72t defined, 157 LMICs (all ages and genders), 27t need to develop, 167 by region, 73f costs arising in health sector, 158–63 Colombia double counting for linked outcomes, 166 Colombia–Venezuela migration corridor, 107 lost opportunity costs, 160, 161 costs of interpersonal violence in, 78t methods to estimate, 160 Congo, Democratic Republic of, diarrheal disease mortality trends, 163, 178 burden in, 164 studies on previous weather-related events, construction industry 160–61 occupational injuries, 102 vector-borne disease, 162 worker falls, 61 costs arising in other sectors, 163–64 cooking forestry sector, 164 burns from. See burns and heat exposure health effects of, 153, 154–56, 154f cleaner fuels, types of, 133. See also liquefied cost estimates of managing, 154 petroleum gas (LPG) diarrheal disease, 154, 161, 164, 178 pollution from. See household air pollution; India, heat as cause of, 155 household air pollution case study primary exposure pathways, 155, 155f solid cookfuels, types of, 133 research needs, 156 Cooper, M. C., 108 undernutrition due to reduced food availability, corporate social responsibility, 110 155, 165 cost-effectiveness of interventions, 11–14, 200–211. vector-borne disease, 155, 162, 165 See also benefit-cost analysis vulnerability to, 156–57, 157f benefit-cost ratios (BCRs), 14, 200 infrastructure protection in event of, 164 burns and heat exposure, 65–66 interventions child abuse, 88t climate-resilient health systems, 157–58, 159t drowning, 11, 64–65, 65t, 200 essential interventions to address, 18t, 19 economic evaluation of, 10–15, 200 most effective measures, 156 falls, 64, 65t reductions in greenhouse gas emissions, household air pollution, 14 153–54, 166 incremental cost-effectiveness ratios (ICERs), 200, mitigation strategies, 163 201–2, 203 mortality trends, 165 injuries and occupational hazards, 200–202 heat-related mortality, 163 interpersonal violence, 87, 88–89t, 200, 201–2, 205 outdoor worker productivity, impact on, 162–63 occupational injuries, 119 power outages and, 163 lack of data, 208 public health services affected by, 161, 164 poisoning, 65t, 66 spatial scale of risk, 164 road traffic injuries, 11, 200, 202 as stress multiplier, 153 water, sanitation, and hygiene (WASH), 181–84, sustainable development pathway, 165 182f, 190–91, 191f, 200, 203, 208 temporal scale of risk, 165 costs of injuries unique challenges of, 153 child abuse, 76, 78–79t Climate Impact Research and Response Coordination general equilibrium approach, 37 for a Larger Europe (CIRCLE) study, 161, 163 human capital approach, 37, 38 coal-fired power plants case study, 239–48. See also interpersonal violence, 76–81, 77–80t India, coal-fired power plants case study occupational injuries, 97, 207 Cochrane reviews road traffic injuries, 37–38, 213 on OSH enforcement, 113 willingness-to-pay approach, 37, 38 on WASH interventions, 184 costs of interventions collective violence and conflicts burns and heat exposure, 65–66 cholera and, 178 child abuse, 88t DALYs in LMICs (all ages and genders), 28t drowning, 64–65, 65t, 205 LMICs vs. HICs, 4t falls, 64, 65t mortality trends in India, household air pollution case study, 230–32, 231t 264 Index government costs, 230, 231t economic evaluations of preventing injuries and household expenditures, 228–30, 229t, reducing environmental hazards, 208 231t, 233f fall prevention programs, 60 interpersonal violence, 87, 88–89t, 200, 201–2, 205 nontransport unintentional injuries, 55, 64 packages of interventions and, 3b occupational injuries, 4, 109, 112 poisoning, 65t, 66 agricultural safety, 117–18 road traffic injuries, 205 construction industry, 117 water, sanitation, and hygiene (WASH), 186–88, planning, monitoring, and evaluation 187f, 188t, 205, 206–7t data for, 113 Côte d’Ivoire road traffic injuries in LMICs, 35, 38, 48 climate change-related costs in, 162 dating violence, prevention of, 83 migrant workers in, 107 Decade of Action for Road Safety (2011–2020), 26, 35, Crocker, J., 207 40, 41, 49 Cropper, M., 239, 244 Democratic Republic of Congo, diarrheal disease cultural and social norms burden in, 164 household air pollution, interventions need to Demographic and Health Survey change, 147 on indoor air pollution, 136 interpersonal violence, 85t, 86 on sanitation practices, 193n3, 203 road safety, 47 Service Provision Assessment (SPA), 175–76 Vietnamese, 215 D diarrheal disease and intestinal infections, 4, 31. See also DALYs helminth infections burns and heat exposure, 57t, 58 child deaths attributable to, 31 LMICs (all ages and genders), 28t, 29 climate-change related burden of, 154, by cause in LMICs (all ages and genders), 27t, 28 161, 164, 178 collective violence and conflicts, 28t undernutrition and, 179 drowning, 57t water, sanitation, and hygiene (WASH), relationship LMICs (all ages and genders), 28t, 29 to, 177–78 falls, 56, 57t Disease Control Priorities in Developing Countries, LMICs (all ages and genders), 28, 28t history of series editions, 2–3b gender-based violence, 207 Ditsuwan, V., 44 household air pollution, 31t, 32, 133, 139 Dodane, P., 207 India, coal-fired power plants case study, 241 Dominican Republic, textile, clothing, and footwear averted DALYs, 243 industry in, 111 India, household air pollution case study, 223, 228, drinking water. See water, sanitation, and hygiene 235n1 (WASH); water supply averted DALY costs, 230, 231t, 232f driving under the influence, 40, 45 interpersonal violence drowning among youth ages 10 to 24 years, 74 alcohol as contributing factor, 59 LMICs (all ages and genders), 28t, 29 burden of injuries, 56–57, 56f, 66 occupational injuries, 29–30, 30t children as highest risk group, 59 poisoning, 57t, 58 interventions, 61t LMICs (all ages and genders), 28, 28t prevention programs, 62 road traffic injuries, 35 cost-effectiveness of interventions, 11, by age and gender, 26–27t, 37 64–65, 65t, 200 all ages and genders, 28t costs of interventions, 64–65, 65t, 205 by gender and country’s income level, 36t DALYs trends in, 57t water, sanitation, and hygiene (WASH), 31–32, 190 LMICs (all ages and genders), 28t, 29 Dangour, A., 183 environmental factors, 59 data and reporting issues human factors, 59 child abuse and youth violence, 74, 81 International Life Saving Federation World cost studies for WASH technologies, 186 Drowning Report, 59 diarrheal disease, 177 interventions, 61t, 66–67 Index 265 essential interventions to address, 16t Egypt, health care sector’s occupational risks in, evidence for prevention, 62 105t, 117 prevention challenges in LMICs, 62–63 electronics industry, 110–11 LMICs vs. HICs, 3, 4t, 63 Elliot, G., 201 mortality trends in, 9, 12f, 56–58, 56f El Salvador age group, gender, and country type, 58t interpersonal violence, costs of, 76, 78t income and regions, 56t, 57t violence against women in, 74 LMICs (all ages and genders), 27t, 29 emissions. See air pollution; household air pollution males and age groups, 57t, 58 environmental risks. See also air pollution; household Working Group on Child Drowning in LMICs, 62 air pollution; water, sanitation, and hygiene age-standardized mortality, 9f E burden of, 33 East Asia. See also specific countries drowning, 59 financial and economic consequences of inadequate economic evaluation of interventions, 10–15, WASH in, 181 202–4t, 202–5 household air pollution in, 143 approaches to inform decisionmaking, 13–14b road traffic mortality rate in, 35 costs of interventions, 205–7 textile, clothing, and footwear industry in, 111 environmental risk transition, 6 Eastern Mediterranean. See also specific countries essential interventions to address, 15–19, 17–18t burns, mortality rates from, 58 global environmental health risks, 7–9 burns in, 29 decline in, 8, 11f diarrheal disease burden in, 164 greenhouse gas emissions, 8, 11f. See also air poisoning mortality rate in, 58 pollution unintentional injuries (nontransport) in, 57t intersectoral partnerships to address, 199 Ebi, K. L., 161 modern environmental health risks, 7 Economic Commission for Latin America and the morbidity and mortality trends, 6, 31t Caribbean (ECLAC), 162 traditional environmental health risks, 6–7 economic evaluation for informed decision water, sanitation, and hygiene, risks for inadequate making, 13–14b. See also cost-effectiveness of services, 180–81 interventions epilepsy, 59 benefit-cost analysis (BCA), 13–14b, 200 ergonomic problems, 4, 29–30, 30t benefit-cost ratios (BCRs), 14, 203, 203t ergonomic problems and interventions, 4, 15, 29–30, cost-effectiveness analysis (CEA), 13–14b, 200 30t, 115 extended cost-effectiveness analysis (ECEA), Ethiopia 13–14b diarrheal disease burden in, 164 health and financial risk, 13b manufacturing sector’s occupational risks in, 105t specific health outcome, 13b European Commission on societal burden of road Ecuador traffic injuries, 38 clean fuel initiatives in, 149 European Survey of Enterprises on New and Emerging intimate partner violence, costs of, 80t Risks, 118 violence against women in, 74 European Working Conditions Survey, 99, 100 education Europe/European Union burns and heat exposure, 63 burns in, 29 cost-effectiveness of training programs in mortality rates from, 58 preventing low-back pain, 14–15 climate change’s impact interpersonal violence, teaching life skills to children Climate Impact Research and Response and adolescents to prevent, 83, 84t Coordination for a Larger Europe (CIRCLE) poisoning, community-based educational study, 161, 163 interventions for, 63 costs of, 162 preschool enrichment programs, 83 on labor productivity, 163 water, sanitation, and hygiene (WASH) services in diarrheal disease burden in, 164 schools, 176, 178, 180, 184 global supply chains of, 108 Egbendewe-Mondzozo, A., 162 lung cancer studies in, 139 266 Index occupational injuries in, 97 France road traffic mortality rate in, 35 climate change-related costs in, 162 unintentional injuries (nontransport) in, 57t vehicle safety study in, 41 Excelsior Electronics workers in China, 110 fuels. See household air pollution; India, coal-fired extended cost-effectiveness analysis (ECEA), 13–14b power plants case study; India, household air India, household air pollution case study, 14b, pollution case study 224, 232 Vietnam helmet regulation case study, 14b, 213–14 G Garg, A., 240 F gender F. Eboli, 161 child abuse of girls vs. boys, 74 factory fires falls, deaths and injuries caused by, 56, 57t, 59 Bangladesh, 111, 111t infanticide rates of female infants, 29 China, 110–11 natural disasters, effect of, 58 falls occupational injuries and, 30, 101 burden of injuries, 32, 56, 56f, 66 poisoning, 60 children at risk for, 59, 66 violence. See women, violence against interventions, 61, 61t, 66 Germany costs and cost-effectiveness of interventions, 64, 65t climate change-related costs in, 162 DALYs trends in, 56, 57t costs of child abuse in, 79t, 81 LMICs (all ages and genders), 28, 28t occupational injuries in agricultural sector in, 118 LMICs vs. HICs, 3, 4t small and medium enterprises (SMEs), 98 mortality trends in, 9, 12f, 56, 56f Ghana income and regions, 57t access to on-plot sanitation in, 180 LMICs (all ages and genders), 27t, 28 IRC WASHCost project, 187 males and age groups, 57t migrant workers in, 107 as occupational injuries, 59 occupational injuries in, 107 older people at risk for, 58–59, 66 Global Alliance for Clean Cookstoves (GACC), 58, 63 fall prevention programs, 60, 61t, 66 Global Burden of Disease 2013 Study (GBD 2013), 26, risk factors for, 58–59 56, 241 vitamin D supplementation as cost-effective globalization, 25, 97, 101 intervention, 64 global supply chains, 108–12 working age people, 59, 66 Global Partnership on Output-Based Aid (GPOBA), 185 interventions, 60–61, 61t, 66 Global Public-Private Partnership for Handwashing Fang, X., 81 (PPPHW), 184 FEASIBLE (OECD tool to develop national financing Global Road Safety Partnership, 49 strategies), 186–87 Goswami, P., 65–66 fibrogenic dusts, exposure to, 120 greenhouse gas emissions. See climate change Fink, G., 203 groundwater fire. See burns and heat exposure; factory fires; arsenic in, 177 household air pollution overextraction of, 180 firearms Guadalajara Chamber of Commerce, 114 homicide rates associated with, 73–74 Guatemala legislation restricting access to, 86 RESPIRE (randomized exposure study of pollution reducing access to, 84t, 86 indoors and respiratory effects) program in, Flextronics, 110 63, 138, 139f flue-gas desulfurization units (scrubbers). See India, violence against women in, 74 coal-fired power plants case study Guimaraes, L., 202 forces of nature. See natural disasters Günther, I., 203 forestry sector climate change, 164 H occupational injuries, 102 Haiti, child abuse in, 74 Foxconn (Chinese manufacturer), 110–11 Hallegatte, S., 166–67 Index 267 Hammitt, J., 202 home-based work, 98, 99t, 116 handwashing. See hygiene; water, sanitation, and homicide rates hygiene by age and gender, 72–73, 73t Harris, G., 201 costs associated with, 76 Haryana, India, household energy interventions in, firearms associated with, 73–74, 82 223–38. See also India, household air pollution reducing access to, 84t, 86 case study LMICs vs. HICs, 3, 4t hazardous industries by region, 73f occupational injuries in, 30 Honduras, textile, clothing, and footwear industry outsourcing of hazardous jobs, 4 in, 111 health sector Hon Hai/Foxconn Technology Group, 110 climate change, costs arising from, 158–63. See also household air pollution, 133–52. See also India, climate change household air pollution case study household air pollution interventions, development biomarkers, 138 of, 147 defined, 134b natural disasters, costs arising from, 160–61 cleaner fuels, types of, 133 heavy metals, exposure to, 120 common types of stoves, 135, 136f helmet regulation. See motorcycle helmet use; Vietnam Comparative Risk Assessment (CRA) of Global helmet regulation case study Burden of Disease Study (2010), 133, 135, helminth infections, incidence of, 31, 178–79, 183 140t, 143, 226 Hendrie, D., 44 concentrations, 136–37, 150n4 Hewlett-Packard factory suppliers, occupational defined, 134b injuries of, 110 cost-effectiveness of interventions, 14 high-income countries DALYs trends, 31t, 32, 133, 139 burns in, 63 emergent issues, 141–43 interventions, 67 emissions, 135–36 climate change’s impact defined, 134b heat-related mortality, 163 environmental health pathway, 134b on labor productivity, 163 evidence classes, 142–43, 143t costs of child abuse in, 76 exposures, 137, 149n2 drowning interventions in, 65 defined, 134b falls in, 59, 66 exposure-response, 141–42 intervention programs for older people, health impacts, 138–43, 223 60, 61t, 66 acute lower respiratory infection (ALRI), 138, firearms associated with homicides in, 82 139f, 141, 223 interventions in cardiovascular disease, 141, 141f not yet fully employed interventions, 20 cataracts, 139, 140t with proven cost-effectiveness, 18 chronic obstructive pulmonary disease (COPD), occupational injuries in, 98–103 138–39 agricultural safety, 117 direct epidemiological studies, 138 ban on harmful materials, 121 interpolation of risks using integrated-exposure construction sector, 102 response functions, 140, 140–41f health services coverage for, 109 lung cancer, 139 participatory approaches to interventions, 116 trends in burden of, 33 regulation and enforcement, 113–14 interventions, 143–48 poisoning intervention programs in, 64 assessment of improved biomass stove road traffic injuries in, 3, 4t, 36 interventions, 144–45b EU project JAMIE (Joint Action for Injury challenges with, 147–48 Monitoring in Europe), 48 cleaner cooking methods, 145–46, 150n8 mortality rate, decrease in, 36 community level initiatives, 146–47, 148 proven interventions, 42–43, 208 compliance and stacking, 148 societal burden, 38 cost and subsidy, 148 HIV/AIDS, 75, 203 cost-effectiveness of, 149 268 Index deaths possible to avert through, 20 water, sanitation, and hygiene (WASH) in essential interventions to address, 17t IRC WASHCost project, 187 health sector development of, 147 sanitation facilities in schools, 180 impact on outdoor pollution, 147, 235n3 Total Sanitation and Sanitation Marketing new paradigms, 145–47 (TSSM) pilot, 185 old paradigms, 143–44 India, coal-fired power plants case study, 239–48 LMICs vs. HICs, 5t benefit-cost analysis, 245 mortality trends in LMICs, 31t, 32, 133 Central Electricity Regulatory Commission of research needs, 148 India’s benchmark operating conditions, use risks linked to income, 6–7, 9f of, 240, 246n4, 246n14 as significant environmental health risk, 133 costs and benefits of scrubbers, 241–43, 243–44t, solid cookfuels, types of, 133, 149n1 246n18, 246n21 World Health Organization’s (WHO) Indoor Air retrofitting vs. installing at time of building plant, Quality Guidelines, 133 247n24 housing-related falls, 61 retrofitting with scrubbers, effect on cost of Human Right to Safe Drinking Water and Sanitation electricity, 244 (HRTWS), 171 DALYs from ambient air pollution, 241, 242t Hunt, A., 163 averted DALYs, 243 Hutton, G., 224 dry scrubber technology to remove sulfur dioxide, Hyder, A. A., 44, 202 246n11, 247n25 hygiene, 175, 182. See also water, sanitation, and estimating health impacts, 240–41, 242t, 246n8 hygiene introduction of regulation (December 2015) to Global Public-Private Partnership for Handwashing lower sulfur dioxide emissions, 239, 245 (PPPHW), 184 maps of all coal-fired plants in India, 240m I of top 30 sulfate deaths, 242m IBNET (International Benchmarking Network for mortality trends from sulfur dioxide emissions, 239, Water and Sanitation Utilities), 181 241, 242m, 242t, 246n9 ILO (International Labour Organization), 109, averted deaths, 243 113, 118 operating characteristics of power plants, 239–40, India 240t, 246n2 burns reliance on coal for electricity, 239 from lamp use in, 60 results of study, 243–44, 244t risk factor for women in, 60 scrubbers in use or planned, 245, 246n15, children in poor households, increased health 246nn19–20 risk of, 190 seawater technology to remove sulfur dioxide, 242, climate change-related costs in, 162 246n12 coal-fired power plants in, 239–48. See also India, State Electricity Regulatory Commissions (SERCs) coal-fired power plants case study estimating unit costs of scrubbers, 242 cost-effectiveness of road traffic injury interventions types of coal, 239, 245, 246n22 s in, 202 wet limestone technology to remove sulfur dioxide, 242 diarrheal disease burden in, 164 years of life lost (YLL), calculation of, 241, 246n10 groundwater overdraft in, 180 India, household air pollution case study, 32, 223–37 household air pollution in. See also India, household burden of disease as background to study, 226, 227t air pollution case study child mortality associated with, 223 cataracts and, 139 costs of interventions, 230–32, 231t clean fuel initiatives, 149 government costs, 230, 231t concentrations, 137 household expenditures, 228–30, 229t, 231t, 233f LPG, use of, 146, 150n8, 224 DALYs from, 223, 228, 235n1 outdoor pollution, impact on, 147 averted DALY costs, 230, 231t, 232f occupational injury in, 107, 109 estimating potential effect of intervention by using protective equipment use, 112 Household Air Pollution Intervention Tool textile, clothing, and footwear industry, 111 (HAPIT), 137, 226 Index 269 evaluation of consequences of policy, 226–28 costs of injury prevention programs, 205, 205t extended cost-effectiveness analysis (ECEA) of investments in, 13–14b control interventions, 14b, 224, 232 essential injury prevention policies, 15–16t, 15–19 Give It Up subsidy program, 146, 150n8, 224, 225, global mortality by type of injury, 35, 36f 230, 233, 235 injury risk transition, inability to show, 9, 12–13f Indian Human Development Survey (2005–06), use intersectoral benefits of inventions, 19, 199 of, 226, 228, 234, 235–36n4 LMICs vs. HICs, 3–4, 4t inventions Institute for Health Metrics and Evaluation’s Global clean fuel initiatives, 223, 225, 235 Burden of Disease (GBD) study, 177–78 distribution of biomass stoves, 223, 224 Institute of Medicine on firearms controls as violence PAHAL subsidy scheme, 225 reduction strategies, 86 by stove type, 225, 228t, 231t intentional injury. See collective violence and conflicts; limitations of study, 233–34 interpersonal violence LPG, beneficial use of, 224, 225, 228t, 231, 231t, 233, intergenerational effects of interpersonal violence, 75 235, 235n2 Intergovernmental Panel on Climate Change (IPCC) mortality associated with, 223, 235 5th Assessment Report, 155 averted mortality costs, 230, 231t, 232f International Agency for Research on Cancer, 139 National Biomass Cookstoves Initiative, 224 International Benchmarking Network for Water and National Program on Improved Chulhas, 136, 144, Sanitation Utilities (IBNET), 181 145–46, 223–24, 228 International Labour Organization (ILO), 109, 113, 118 outdoor air pollution effects, 235n3 International Life Saving Federation World Drowning research needs, 234 Report, 59 results of study, 230–31, 230f, 231t International Road Assessment Program, 41 Smokeless Village (SV) campaign, 224, 235 International Social Security Association (ISSA), 97 solid fuels, use of, 32, 223 International Society for Burn Injuries (ISBI), 58 stove types, 226, 228t, 231t International Standards Organization, 144 time savings and related job-seeking behavior from interpersonal violence, 71–96 adoption of cleaner technologies, 234 bullying, 72 treatment-seeking behaviors and associated costs, burden of, 71–72 228, 229t economic burden, 76–81, 207 Ujjwalla program, 225, 235 community violence, 72 Indonesia consequences of, 71, 74–76 climate change-related costs in, 161 costs and cost-effectiveness of interventions, 87, diarrheal disease burden in, 164 88–89t, 200, 201–2, 205 water, sanitation, and hygiene (WASH) services in costs framework, 76–81, 77–80t benefit-cost analysis for wastewater cultural and social norms permitting, 81t treatment, 190 DALYs cost-effectiveness, 191 among youth ages 10 to 24 years, 74 Total Sanitation and Sanitation Marketing LMICs (all ages and genders), 28t, 29 (TSSM) pilot, 185 dating violence, prevention of, 83 indoor air pollution. See household air pollution definitions of, 71–72 industrialization, effects of, 109 evidence on health and genetic effects of, 76 infanticide, 29 family or partner violence, 72 infectious diseases linked to workplace exposures, 29. future violence, increased risk of, 75 See also occupational injuries implementation of prevention strategies, 87–90 informal sector, occupational injuries in, 30, 106, 116t prevention delivery, 90 injuries and injury prevention, 3–4. See also prevention support, 90 interventions; specific type of injury and prevention synthesis and translation, 87–90 environmental risk intergenerational effects of, 75 deaths possible to avert through interventions, interventions, 82–87, 84–85t 20, 20t change of cultural and social norms, 85t, 86 economic evaluation of injury prevention, 10–15, essential interventions to address, 16t, 19 200–201, 201t gender equality promotion, 84t, 86 270 Index life skills in children and adolescents, support for educational programs to target risks, 199 development of, 83, 84t high-income countries already using, 2 parental relationships, household air pollution, 143–48 support for development of, 83, 84t assessment of improved biomass stove reducing access to lethal means, 84t, 86 interventions, 144–45b social determinants of, policies to address, 87 challenges with, 147–48 street outreach workers to mediate conflicts, 86 cleaner cooking methods, 145–46, 150n8 victim identification, care, and support community level initiatives, 146–47, 148 programs, 85t, 87 compliance and stacking, 148 intimate partner violence (IPV) costs and cost- cost and subsidy, 148 effectiveness of interventions, 88t cost-effectiveness of, 149 LMICs vs. HICs, 3, 4t, 76 deaths possible to avert through, 20 mental health problems from, 75 essential interventions to address, 17t microfinance program to reduce, 86, 201–2 health sector development of, 147 mortality trends in, 9, 13f, 72–74 impact on outdoor pollution, 147, 235n3 LMICs (all ages and genders), 27t, 29 new paradigms, 145–47 by type and income level, 72, 72t old paradigms, 143–44 noncommunicable diseases, increased risk of, 76 India, household air pollution case study rape or sexual assault, 72 clean fuel initiatives, 223, 225, 235 reproductive problems, increased risk of, 75 distribution of biomass stoves, 223, 224 risk and protective factors for, 81–82, 81t PAHAL subsidy scheme, 225 sexually transmitted infections (STIs) linked to, 75 by stove type, 225, 228t, 231t special populations, increased risk of, 75 interpersonal violence, 82–87, 84–85t violence against children and youth, 74. See also change of cultural and social norms, 84–85t, 86 child abuse essential interventions to address, 16t, 19 violence against elderly persons, 74 gender equality promotion, 84t, 86 violence against women, 74 life skills in children and adolescents, support for interventions to promote gender equality, 84t, 86 development of, 83, 84t Violence Prevention: The Evidence (series), 87 parental relationships, youth violence, 74 support for development of, 83, 84t costs and cost-effectiveness of interventions, reducing access to lethal means, 84t, 86 88–89t, 88t social determinants of, policies to address, 87 essential interventions to address, 16t street outreach workers to mediate conflicts, 86 interventions, 199–200 victim identification, care, and support burns and heat exposure, 61t, 63, 67 programs, 85t, 87 essential interventions to address, 16t intersectoral benefits of, 19, 199 child abuse, 83 laws and regulations to target risks, 199 climate change, 157–58, 159t occupational injuries, 112–19 costs of, 157, 208 agriculture, 117–18 defined, 157 capacity building and retention, 112–13 essential interventions to address, 18t, 19 construction worksite safety, 117 most effective measures, 156 diffusion to SMEs, 119 need to develop, 167 ergonomics, 15, 115 reductions in greenhouse gas emissions, essential interventions to address, 16t, 19 153–54, 166 fatigue risk management, 118 cost-effective for injuries and environmental health care sector, 117 health conditions, 1, 2, 200–211. See also injury care and return to work, 112 cost-effectiveness of interventions machine safety, 118 deaths possible to avert through, 20 migrant workers’ safety training, 117 drowning, 61t, 66–67 participatory approaches, 115–16, 116t essential interventions to address, 16t planning, monitoring, and evaluation data, 113 evidence for prevention, 62 primary prevention, 112 prevention challenges in LMICs, 62–63 protective covering and equipment, 107, 112 Index 271 psychosocial risk management, 118–19 Jamison, D. T., 224 regulation and enforcement, 113–14 Jan, S., 201 responsible parties for, 199 Jeuland, M., 205, 224 safety climate and safety culture, 114–15 Johns Hopkins International Injury safety incentive programs, 115 Research Unit, 49 slip, trip, and fall prevention, 116–17 worker training, 114 K workplace violence, 118 Kenya packages of interventions, development of, 3b benefit-cost analysis of clean energy in, 14, 225 poisoning, 61t, 63–64, 67 child abuse in, 74 essential interventions to address, 16t, 18 climate change-related costs in, 162 road traffic injuries, 40–45 cost-effectiveness of road traffic injury interventions economic analysis of, 44–45, 208 s in, 202 essential interventions to address, 15t, 18–19 diarrheal disease burden in, 164 implementation of, 45, 48 physical violence in the workplace in, 105t infrastructure construction for safety, 41–42 violence against women in, 74 promotion of, 47 kerosene proven and promising interventions, 42–43t, lamps, burns from, 60 42–44 stoves, 149–50n3, 225 safety legislation and regulation, 40–41, 47–48 burns from, 59 seatbelts and child restraints, 40–41, 45, Korea, Republic of, costs of road traffic injuries in, 38 46b, 49t Kovats, S., 163 speed limits, 40, 42 Krug, A., 64 vehicle safety, 41 subsidies. See subsidies L water, sanitation, and hygiene (WASH), 181–84 Lambert, D. M., 108 approaches to improve effectiveness of service landmine clearance, cost-effectiveness of, 201, 205 delivery, 185–86 Larsen, B., 205 behavioral change and promotion of, 184 Latin America and the Caribbean. See also specific community-driven development (CDD) countries programs, 186 climate change-related costs in, 162 Community-Led Total Sanitation (CLTS), 184 costs of interpersonal violence in, 76, 77–78t deaths possible to avert through interventions, diarrheal disease burden in, 178t 20, 20t labor productivity, climate change’s impact demand-based approaches, 184 on, 163 effectiveness of service delivery models, 184–86 migrant workers in, 107 essential interventions to address, 17t, 19 occupational injuries in, 97, 106 nutrition interventions promoting hygiene, 186 physical violence in the workplace in, 104t public regulation, 208 trade unions’ role, 104 results-based approaches, 185, 193n11 violence against women in, 74, 75 results for, 183, 183t, 208 Latin New Car Assessment Program (NCAP), 41 safety-net programs, 186 laws and regulations to target risks, 199 School-Led Total Sanitation (SLTS), 184 motorcycle helmet use. See Vietnam helmet supply-side approaches, 184–85 regulation case study ISBI (International Society for Burn Injuries), 58 occupational health, 109–10, 113–14 ISSA (International Social Security Association), 97 road safety, 40–41, 47–48 water, sanitation, and hygiene (WASH), 208 J Laxminarayan, R., 224 Jabil (manufacturer), 110 lead exposure, interventions to address, 18t Jamaica Linnan, M., 63 costs of interpersonal violence in, 76, 78t liquefied petroleum gas (LPG) costs of road traffic injuries in, 38 beneficial use of, 224, 225, 228t, 231, 231t, 233, 235, violence against women in, 74 235n2 272 Index burns, risk factor for, 60 Morocco Locke, R. M., 110 intimate partner violence, costs of, 80t Lu, Z., 240 violence against women, costs of, 81 lung cancer, 139 mortality trends age-standardized data, 6, 8–9f M better medical treatment as well as better economic malaria, 153, 154, 155, 161, 162, 165 conditions as factors in cross-county Malawi comparisons, 20 climate change-related costs in, 162 burns and heat exposure, 9, 12f, 55 violence against women in, 74 income and regions, 57t, 58 Malaysia LMICs (all ages and genders), 27t, 29, 58 costs of road traffic injuries in, 38 males and age groups, 57t, 58 electronics industry in, 110 climate change, 165 motorcycles and electric bikes in, 41 heat-related mortality, 163 Mali’s Community-Led Total Sanitation (CLTS), 184 collective violence and conflicts Malla, M. B., 225 by income levels, 72t manufacturing sector and occupational injuries, LMICs (all ages and genders), 27t 102–3, 105t by region, 73f Markandya, A., 161, 162 drowning, 9, 12f, 56–58, 56f Marucci-Wellman, H., 113 age group, gender, and country type, 58t Mehta, S., 224 income and regions, 56t, 57t menstrual hygiene management, 180 LMICs (all ages and genders), 27t, 29 mental health problems, 75, 177 males and age groups, 57t, 58 Mexico falls, 9, 12f, 56, 56f clean energy in, 14, 148 income and regions, 57t climate change-related costs in, 161 LMICs (all ages and genders), 27t, 28 groundwater overdraft in, 180 males and age groups, 57t interpersonal violence in household air pollution, 31t, 32, 133 costs of, 78t India, coal-fired power plants case study, 239, 241, physical violence in the workplace, 105t 242m, 242t, 246n9 interventions in averted deaths, 243 costs of air pollution control, 205 India, household air pollution case study, 223, 235 retrofitting vehicles for air pollution control, averted mortality costs, 230, 231t, 232f 14, 205 interpersonal violence, 9, 13f, 72–74 labor productivity, climate change’s impact on, 163 LMICs (all ages and genders), 27t, 29 occupational risk in by type and income level, 72, 72t electronics industry, 110 mortality transition over time in causes of death, 6, 7f informal sector, 106 occupational injuries, 9, 12f, 29–30 workers’ rights in maquiladoras (Mexican age-standardized mortality, 8–9f manufacturing plants operating in free-trade deaths possible to avert through interventions, zone), 114 20, 20t microfinance used as intervention LMICs, by all ages and genders, 30t household sanitation, 185 LMICs, by type of risk, 104f interpersonal violence, 86, 201–2 overlap among causes, 20 Middle East. See also specific countries poisoning, 9, 13f, 58 climate change-related costs in, 161, 162 income and regions, 57t occupational injuries in, 97–98 LMICs (all ages and genders), 27t, 28 physical violence in the workplace in, 104t males and age groups, 57t, 58 migrant workers, 98–99, 107–9 road traffic injuries, 9, 12f, 26–27t, 26–28, 35, 213 Millennium Development Goals (MDGs), 5, 31–32, by gender and country’s income level, 36t, 37 171, 172, 174, 175, 186, 191, 199 water, sanitation, and hygiene (WASH), 31–32, 31t Mischke, C., 113 motorcycle helmet use Modi, Narendra, 224 cost-effectiveness of, 11 Index 273 legislation mandating, 40 nontransport unintentional injuries, 55–70. See also Vietnam helmet regulation, 213–22. See also unintentional injuries (nontransport) Vietnam helmet regulation case study Nonyelum, S. C., 64 Mozambique North Africa. See also specific countries burns from lamp use in, 60 climate change-related costs in, 161, 162 landmine clearance, cost-effectiveness of, 201 occupational injuries in, 98 physical violence in the workplace in, 105t physical violence in the workplace in, 104t water, sanitation, and hygiene (WASH) services women as workers in, 101 in, 176f IRC WASHCost project, 187 O occupational injuries, 4, 97–132 N acute and chronic medical diseases, 119–21 National Institute for Occupational Safety and age of workers, 101 Health, 117 agriculture, forestry, and fishing sector, 102 natural disasters biologic hazards, 120 cholera and, 178 burden of, 33, 103–8, 110 costs arising in health sector from, 160–61 chemical dumping and, 121 deaths and DALYs caused by, 56f, 57t, 58 child labor, 30, 108, 116 flood-related drowning, 59, 177 common risks, 110–12 waterborne diseases resulting from, 177 in comparative risk assessment framework, 21n4 Nepal construction sector, 102 benefit-cost analysis of clean energy in, 14, 225 cost-effectiveness of interventions, 119 burns from lamp use in, 60 lack of data, 208 cataracts, cooking with solid fuels as costs of, 97, 207 cause of, 139 DALYs in LMICs (all ages and genders), 29–30, 30t cost-effectiveness of road traffic injury interventions electronics industry, 110–11 in, 202 employment sectors, 101–3, 105–6t menstrual hygiene management in, 180 ergonomic problems, 4, 29–30, 30t Netherlands, Sustainable Safety Model of, 45 European Survey of Enterprises on New and New Zealand Emerging Risks, 118 agricultural safety in, 118 European Working Conditions Survey, 99, 100 costs exposure to chemicals and toxics, 4, 29, 30t, 119–21 of homicide in, 76 essential interventions to address, 17t of interpersonal violence in, 77t prevention, 120–21 of road traffic injuries in, 38 types of hazards, 120 NGOs regulating working conditions, 109–10, 114 exposure to physical workplace hazards, 99, Nicaragua 104t, 120 intimate partner violence, costs of, 79t falls as, 59 violence against women in, 74 gender distribution of workers, 30, 101 costs of, 81 global supply chain, 108–12, 114 Niger, diarrheal disease burden in, 164 health care and social services, 103, 105t Nigeria health problems, 101 diarrheal disease burden in, 164 home-based work, 98, 99t occupational injuries in, 105t, 107 individual factors, 100–101 migrant workers’ risk, 107 in informal sector, 30, 106, 116t physical violence in the workplace in, 105t interventions, 112–19 safety warnings on medications and consumer agriculture, 117–18 products in, 64 capacity building and retention, 112–13 noise exposure, 4, 30, 30t, 121 construction worksite safety, 117 noncommunicable diseases, increased risk related to diffusion to SMEs, 119 interpersonal violence, 76 ergonomics, 15, 115 nongovernmental organizations (NGOs) regulating essential interventions to address, 16t, 19 working conditions, 109–10, 114 fatigue risk management, 118 274 Index health care sector, 117 diarrheal disease burden in, 164 injury care and return to work, 112 poisoning interventions, economic analysis of, 66 machine safety, 118 road safety interventions, economic analysis of, 44 migrant workers’ safety training, 117 safety warnings on medications and consumer participatory approaches, 115–16, 116t products in, 64 planning, monitoring, and evaluation data, 113 Pandey, K., 161 primary prevention, 112 Pant, K. P., 225 protective covering and equipment, 107, 112 Paraffin Safety Association of Southern Africa, 63–64 psychosocial risk management, 118–19 Paraguay regulation and enforcement, 113–14 clean fuel initiatives in, 149 responsible parties for, 199 climate change-related costs in, 162 safety climate and safety culture, 114–15 Paraguay–Argentina migration corridor, 107 safety incentive programs, 115 violence against women in, 74 slip, trip, and fall prevention, 116–17 Parry, M., 161 worker training, 114 particulate matter exposure workplace violence, 118 generally. See air pollution LMICs vs. HICs, 4, 5t in households. See household air pollution manufacturing sector, 102–3, 105t in workplace, 4 migrant workers, 98–99, 107–9 Pattanayak, S. K., 205, 224 mortality trends in, 9, 12f, 29–30 pedestrian injuries. See road traffic injuries age-standardized mortality, 8–9f Peru deaths possible to avert through interventions, costs of interpersonal violence in, 76, 77t 20, 20t physical violence in the workplace in, 105t LMICs, by all ages and genders, 30t Water and Sanitation Program’s Global Scaling Up LMICs, by type of risk, 104f Handwashing Projects, 184 NGOs regulating working conditions, 109–10, 114 pesticides, exposure to, 118, 120 noise exposure, 4, 30, 30t, 121 Philippines outsourcing of hazardous jobs, 4, 97, 98, 99t, 110 health care sector’s occupational risks in, 106t psychosocial exposures, 99–100 water, sanitation, and hygiene (WASH) services in risk management of, 118–19 cost-effectiveness, 191 recession of 2007–09, effect of, 97 water pollution, 181 regulation of occupational health, 109–10 Pierfederici, R., 161 repetitive motion, 99, 110 poisoning small and medium enterprises (SMEs), 98, 106, 113 behavioral strategies for prevention of, 63–64 temporary vs. permanent workers, 98 burden of injuries, 56f, 66 textile, clothing, and footwear industry, 111–12, 111t child poisoning cases, 60 trade unions’ role, 104 child safety caps on medications, 64 wholesale and retail trade, 103 community-based educational interventions, 63 work–life conflict and adverse outcomes, 100 costs and cost-effectiveness of interventions, work organization and shifts, 100, 118 65t, 66 OECD countries. See high-income countries DALYs trends in, 57t, 58 open defecation, incidence of, 31, 174, 185 LMICs (all ages and genders), 28, 28t organic solvents, exposure to, 120 environmental strategies for prevention of, 64 Ortiz, R. A., 162 gender differences, 60 outdoor worker productivity, impact of climate change, interventions, 61t, 63–64, 67 162–63 essential interventions to address, 16t, 18 outsourcing of hazardous jobs, 4, 97, 98, 99t, 110 LMICs vs. HICs, 4t Ozdemir, R., 63 mortality trends in, 9, 13f, 58 income and regions, 57t P LMICs (all ages and genders), 27t, 28 Pakistan males and age groups, 57t, 58 children in poor households, increased health risk pesticide poisoning of agricultural workers, 118 of, 190 risk factors for, 60 Index 275 Pope, C. A., III, 241 promotion of, 47 power outages and climate change, 163 proven and promising interventions, 42–43t, ProcureITfair (NGO coalition), 110 42–44 Prüss-Ustün, A., 178 safety legislation and regulation, 40–41, 47–48 public health services. See also laws and regulations to seatbelts and child restraints, 40–41, 45, 46b, 49t target risks speed limits, 40, 42 climate change’s impact on, 161, 164 vehicle safety, 41 mass media communication by, 19 Johns Hopkins International Injury Research Unit violence as priority for, 71, 90 training program for prevention and control, 49 Latin New Car Assessment Program (NCAP), 41 R LMICs vs. HICs, 3, 4t, 36 Ramakrishnan, S., 162 mortality trends in, 9, 12f, 26–27t, 26–28, 35, 213 rape or sexual assault, 72, 180. See also interpersonal by gender and country’s income level, 36t, 37 violence operationalizing action for road safety, 47–49 Reddy, M. S., 240 postcrash risk factors including trauma and regulation and enforcement. See laws and regulations to emergency care services, 39–40 target risks interventions targeting, 40, 45 Rehfuess, E., 224 precrash risk factors of, 39 Renner, J., 202 interventions targeting, 40 Ribeiro, J., 202 research needs, 49 risk factors. See also specific type of injury resource mobilization, 47 for injuries and environmental health conditions, 1 risk factors for, 38–40 predictable nature of, 2 road safety policies and integrated approaches, 45 road traffic injuries (RTIs), 35–54 Russian Federation case study of seatbelt use, 46b Bloomberg Philanthropies Global Road Safety safe systems approach, 45 Program, 45, 46b societal burden of, 38 capacity development, 48–49 Sustainable Safety Model of the Netherlands, 45 clinical research agenda, development of, 47 Swedish Vision Zero, 45 cost-effectiveness of interventions, 11, 200, 202 targeting high-risk individuals and groups, 47 costs of RTI as potentially impoverishing, 213 technical assistance, 49 costs of RTI prevention programs, 205 traffic calming measures, 42 crash risk factors of, 39 trends in burden of, 32 interventions targeting, 40 vulnerable road users, risk level of, 36–37, 41–42 DALYs in LMICs, 35 World Report on Road Traffic Injury Prevention, 42 by age and gender, 26–27t, 37 young people as most affected age group, 37 all ages and genders, 28t Romis, M., 110 by gender and country’s income level, 36t RTIs. See road traffic injuries driving under the influence, 40, 45 Russian Federation economic burden of, 37–38 climate change-related labor productivity future research needs, 49 improvements in, 162 Global Road Safety Partnership, 49 road traffic accidents, mortality and DALYs trends Haddon matrix on multifactorial nature of causes in, 37 and risk factors of, 38–39, 39t seatbelt use, case study of, 46b health burden of, 35–37 Rwanda helmet regulation. See motorcycle helmet use; climate change-related costs in, 162 Vietnam helmet regulation case study violence against women in, 74 International Road Assessment Program, 41 intersectoral partnerships to address, 47, 49t S interventions, 40–45 Saint Lucia, climate change-related costs in, 162 economic analysis of, 44–45, 208 salmonellosis, 163 essential interventions to address, 15t, 18–19 sanitation, 174, 175f. See also water, sanitation, and implementation of, 45, 48 hygiene infrastructure construction for safety, 41–42 Community Approach to Total Sanitation 276 Index (CATS), 185 labor productivity, climate change’s impact on, 163 Community-Led Total Sanitation (CLTS), 184 migrant workers in, 107, 109 environmental consequences for poorly managed poisoning intervention programs in, 63, 64, 66 human excreta, 180–81 PREPARE program for youth in, 83 on-plot sanitation, benefits of, 179–80 road traffic injuries in, 38 technologies to reduce transmission of South America. See Latin America and the Caribbean pathogens, 182 South Asia. See also specific countries Total Sanitation and Sanitation Marketing burns in, 29 (TSSM), 185 child mortality due to diarrheal illnesses in, 31 schistosomiasis, 178 household air pollution in, 143 schools. See education occupational injuries in, 97, 106 Schwebel, D. C., 63 open defecation, incidence of, 174 scrubbers. See India, coal-fired power plants case study road traffic mortality rate in, 35 self-harm and suicide textile, clothing, and footwear industry in, 111 assembly line work in electronics manufacturing universal access to WASH services in, 186 and, 110 water insecurity in, 180 DALYs in LMICs (all ages and genders), 28t women as workers in, 101 firearms associated with, 74 South-East Asia. See also specific countries LMICs vs. HICs, 3, 4t burns in, 29 mortality trends in mortality rates, 58 by age and gender, 73t child mortality due to diarrheal illnesses in, 31, 178t by income level, 72t climate change-related costs in, 161, 162 LMICs (all ages and genders), 27t falls, incidence in, 28 by region, 73f occupational injuries in, 97, 106 Senegal road traffic injuries in costs of sewer-based systems in, 207 legislation and enforcement strategies, 44 sanitation costs in, 188 motorcycles and electric bikes, 41 Water and Sanitation Program’s Global Scaling Up unintentional injuries (nontransport) in, 57t Handwashing Projects, 184, 193n10 water, sanitation, and hygiene (WASH) services in sexually transmitted infections (STIs), 75 access to on-plot sanitation, 180 Shahpar, C., 224 cost-effectiveness, 191 Shrestha, R., 203 hygiene, 175 small and medium enterprises (SMEs), occupational Stevens, G., 202 injuries in, 98, 106, 113 Stevenson, M., 202 Smith, K. R., 139, 142, 155 Stockholm Environment Institute, 162 smoke detectors, use of, 63 stove types. See household air pollution; India, snakebites, mortality trends in, 9, 12f household air pollution case study social norms. See cultural and social norms stunting, 177 solid cookfuels. See household air pollution; India, Sub-Saharan Africa. See also specific countries household air pollution case study burns in, 29 South Africa child labor in, 108 benefit-cost analysis for small-scale water schemes child mortality in in, 190 diarrheal illnesses and, 31 firearms licensing, effect of, 86 WASH initiatives reducing, 190 informal sector’s occupational risk in, 106 diarrheal disease in, 31, 177 interpersonal violence in household air pollution in, 133, 143 Intervention with Microfinance and Gender malaria in, 153 Equity program, 86, 201–2 occupational injuries in, 97–98, 106 Parent Centre to prevent violence against women health care sector, 106t and children, 90 informal sector, 106 physical violence in the workplace in, 105t physical violence in the workplace in, 105t Soul City initiative to raise awareness of road traffic injuries in laws, 86 cost-effectiveness of interventions, 202 Index 277 legislation and enforcement strategies, 44 Thailand mortality rate, 35 costs Stepping Stones program in, 86 of interpersonal violence in, 76, 78t trade unions’ role, 104 of road traffic injury prevention, 205 violence against women in, 74 occupational injuries in, 105t water, sanitation, and hygiene (WASH) services informal sector’s risk, 106 in, 175–76 migrant workers’ risk, 109 access to water, 31 participatory discussions with health care child mortality reduced by, 190 providers to prevent, 116 financial and economic consequences, 181 road traffic injuries in hygiene, 175 alcohol-impaired driving, 44 open defecation, incidence of, 174 motorcycles and electric bikes, 41 universal access, 186 safety interventions, economic analysis of, 201 water supply, 173 trade unions’ role in occupational protections, 104 subsidies Traerup, S. L. M., 162 household air pollution interventions training programs. See education Give It Up subsidy program (India), 146, 150n8, trauma care 224, 225, 230, 233, 235 impoverishment of injured worker for, 207 household air pollution interventions, 148 as postcrash risk factor, 39–40 sanitation subsidies and financing targeted to conditional-cash transfer (CCT) participants, U 185–86, 193n11 Uganda Sudan diseases in benefit-cost analysis of clean energy in, 14, 225 diarrheal disease burden in, 164 climate change-related costs in, 162 home-based education for HIV/AIDS treatment, diarrheal disease burden in, 164 cost-effectiveness of, 203 suicide. See self-harm and suicide health care sector’s occupational risks, 105t surveillance, effectiveness of, 19, 32 interpersonal violence in Survey of Occupational Injuries and Illnesses (U.S. costs of intimate partner violence, 80t Bureau of Labor Statistics), 102 Raising Voices program in, to prevent violence Sustainable Development Goals (SDGs), 3b, 171, 172, against women and children, 90 192, 199 violence against women, 74 interpersonal violence, elimination of, 71 road traffic injuries in Swart, L., 63 costs of, 38 Sweden safety interventions, economic analysis of, 44, agricultural safety in, 118 200, 205 Swedish Vision Zero, 45 undernutrition swimming lessons, cost-effectiveness of, 11 climate change and reduced food availability, 155, 165 T water, sanitation, and hygiene (WASH) and, 179 Tanzania UNICEF child abuse in, 74 data collection on water, sanitation, and hygiene climate change-related costs in, 162 (WASH) services in schools, 176 cost-effectiveness of road traffic injury interventions Multiple Indicator Cluster Surveys and s in, 202 Demographic and Health Surveys, 74 physical violence in the workplace in, 105t unintentional injuries (nontransport), 55–70. See Total Sanitation and Sanitation Marketing (TSSM) also burns and heat exposure; drowning; falls; pilot in, 185 poisoning Water and Sanitation Program’s Global Scaling Up burden of injuries, 56, 56f, 66 Handwashing Projects, 184 costs and cost-effectiveness of interventions, Tediosi, F., 224 64–66, 65t textile, clothing, and footwear industry, occupational decreasing mortality rates of, 55 injuries in, 111–12, 111t exposure to forces of nature, 55 278 Index HICs vs. LMICs, 55 repetitive motion, 99 interventions, 60–64, 61t Pennsylvania’s Evidence-Based Prevention and mortality trends in Intervention Support Center (EPISCenter) at income and regions, 57t University of Pennsylvania to support violence males and age groups, 57t prevention programs, 90 United Kingdom Safe Dates program in, 83 interpersonal violence in urban areas costs of, 77t access to water in, 173–74 Wales program to reduce violence-related sanitation in, 174 injuries, 82 U.S. Environmental Protection Agency, 208 occupational injuries in exposure to chemicals/toxics and physical V hazards, 121 vector-borne disease, 155, 162, 165 Health and Safety Executive’s rools for safety in Venezuela construction industry, 117 Colombia–Venezuela migration corridor, 107 manufacturing sector, 103 costs of interpersonal violence in, 77t road traffic injuries, costs of, 38 Venkataraman, C., 240 United Nations Verguet, S., 224 Children’s Fund, 185 Vietnam. See also Vietnam helmet regulation data collection on water, sanitation, and hygiene case study (WASH) services in institutions, 176 costs Framework Convention on Climate Change, 166 of fall injuries in, 64 Human Right to Safe Drinking Water and Sanitation of road traffic injuries in, 38, 64, 213 (HRTWS), 171 occupational injuries in, 105t Road Safety Collaboration, 26, 35, 40, 45, 49 Worker Improvement in Neighborhood United Nations Development Programme Development program, 116 (UNDP), 162 water, sanitation, and hygiene (WASH) services in United States cost-effectiveness, 191 air pollution-related mortality in, 241 Water and Sanitation Program’s Global Scaling burns, risk factors in, 59 Up Handwashing Projects, 184 Clean Air Act Amendments (1990), 239 Vietnam helmet regulation case study, 213–22 climate change-related costs in, 162 cost consequences and affordability, 215–16 coal-fired power plants in, 239, 245 cost-effectiveness of, 219 costs decrease in crash deaths, 218–19 of child abuse in, 77–78t, 81 design of study and input parameters, 214, 215t of interpersonal violence in, 76, 77t enforcement and fines for failure to wear a of intimate partner violence in, 79t, 81 helmet, 214 of road traffic injuries in, 38 extended cost-effectiveness analysis (ECEA) of, 14b, of youth violence in, 80t 213–14 firearms use in financial risk protection, calculation of, 216, 218f associated with homicides in, 82 government’s cost of implementing, 216 associated with suicides in, 74 health consequences of, 215 global supply chains of, 108 injury prevention shown to be poverty prevention as labor productivity, climate change’s impact on, 163 well, 218 lung cancer studies in, 139 legislation mandating use of motorcycle occupational injuries in helmets, 213 agricultural safety, 118 limitations in study, 219 burden of, 97 major assumptions in study, 214–15 decline in, 97 as model for other LMICs, 213 electronics industry, 110 National Traffic Safety Committee data used in exposure to chemicals/toxics and physical study, 214 hazards, 121 sensitivity analysis, 216 health care sector, 117 setting of study, 214 Index 279 study results, 216–18 safety-net programs, 186 variables in study, 214 School-Led Total Sanitation (SLTS), 184 wealthy accruing larger share of benefit, 218 supply-side approaches, 184–85 violence. See collective violence; interpersonal violence; Joint Monitoring Programme’s (JMP) focus, 172, women, violence against 174f, 175, 190 vitamin D supplementation, 64 LMICs vs. HICs, 4–5, 5t menstrual hygiene management, 180 W microfinance linked to household sanitation, 185 water, sanitation, and hygiene (WASH), 4–5, 171–98 Millennium Development Goal (MDG) targets, 5, arsenic in groundwater, 177 31–32, 171, 172, 174, 175, 186, 191, 199 benefit-cost analysis of, 14, 189–90, 190t, 203, 204t morbidity and mortality attributable to lack of, benefits of services, 188, 189t 31–32, 31t children in poor households, increased health risk of on-plot sanitation, 179–80 as justification for targeting, 190 reporting system for institutions, 175–76 climate change and, 192 research needs, 192 cost-effective analysis of, 190–91, 191f, 200, 203 risks linked to income, 6–7, 9f costs of universal access and improved services, scope of services, 171, 172t 186–88, 187f, 188t, 205, 206–7t social welfare consequences of inadequate WASH, DALYs in LMICs from, 31–32, 190 179–80 definitions, 172–73 status of, 172–76 dehydration, 177 sustainability of, 191 diarrheal disease and, 177–78 Sustainable Development Goals (SDGs), diseases transmitted through inadequate 171, 172, 192 services, 177–78 undernutrition and, 179 distribution of services, 175–77 universal access and improved services, 186–88, effectiveness and costs of technologies and practices, 187f, 191–92 181–84, 182f, 208 vulnerability of children and, 177 environmental consequences of inadequate WASH, willingness-to-pay (WTP) studies, 188, 208 180–81 water pollution, consequences of, 180–81 financial and economic consequences of inadequate water supply, 173–74. See also water, sanitation, and WASH, 181 hygiene global monitoring of, 172, 173t supply-side approaches to rural areas, 185 Global Public-Private Partnership for Handwashing treatment of, 182 (PPPHW), 184 virtual water trade, 180, 193n6 health consequences of inadequate WASH, 177–79 Watkiss, P., 163 helminth infections, incidence of, 31, 178–79, 183 weather variables. See climate change; natural interventions, 181–84 disasters approaches to improve effectiveness of service Western Asia. See also specific countries delivery, 185–86 informal sector’s occupational risk in, 106 behavioral change and promotion of, 184 women as workers in, 101 community-driven development (CDD) Western Pacific. See also specific countries programs, 186 benefit-cost analysis of clean energy in, 14, 225 Community-Led Total Sanitation (CLTS), 184 diarrheal disease burden in, 164 deaths possible to avert through interventions, road traffic deaths as leading cause of mortality for 20, 20t people 15–49 years, 213 demand-based approaches, 184 unintentional injuries (nontransport) in, 57t effectiveness of service delivery models, willingness-to-pay approach, 37, 38 184–86 India, coal-fired power plants case study, 245, essential interventions to address, 17t, 19 247n26 nutrition interventions promoting hygiene, 186 water, sanitation, and hygiene, 188, 208 public regulation, 208 Wilson, A., 202 results-based approaches, 185, 193n11 women. See also gender results for, 183, 183t, 208 access to water and, 31, 173, 177, 179–80 280 Index burns, incidence of, 29 on climate-change related child mortality, 178 menstrual hygiene management, 180 on climate-resilient system, 157 women, violence against, 74, 75. See also interpersonal Global Health Estimates database, 26, 35, 58 violence Global Status Report on Road Safety, 20, 48, acid attacks on women, 29 49–50n1 costs of, 81, 207 on household air pollution, 136, 144 interventions to promote gender equality, 84t, 86 Indoor Air Quality Guidelines, 133, 150n4 rape or sexual assault, 72, 180 on interpersonal violence, 71–72 Working Group on Child Drowning in LMICs, 62 on kerosene use as household fuel, 150n3 work–life conflict and adverse outcomes, 100 on occupational health workplaces integration with primary health care, 104 safety. See occupational injuries regulations for, 109 water, sanitation, and hygiene (WASH) services on pesticide poisoning, 118 in, 180 on road traffic injuries responsible for life years World Bank lost, 37 on diarrheal disease and malaria in climate-change Strengthening Road Safety Legislation, 47 scenario, 161 WHO-CHOICE methodology to determine Global Partnership on Output-Based Aid cost-effectiveness of interventions, (GPOBA), 185 14, 44, 65, 202, 203 on household air pollution, 136 World Mental Health Survey, 75 International Benchmarking Network for Water and World Report on Child Injury Prevention, 62–63 Sanitation Utilities (IBNET), 181 World Report on Road Traffic Injury Prevention, 42 Program-for-Results Financing (PforR), 185, 193n11 Y on WASH initiatives youth child mortality reduction, 190 road traffic injuries, young people as most affected costs of universal access to WASH services, age group, 37 186–87 violence. See interpersonal violence output-based aid, 185, 193n11 Water and Sanitation Program, 185 Z World Development Report (1993), 2 Zambia, violence against women in, 74 World Health Assembly Zimbabwe resolution to improve trauma and emergency care child abuse in, 74 services, 40 occupational injuries in violence as public health priority for, 71 health care sector, 106t World Health Organization (WHO) migrant workers, 107 on alcohol and health, 83 miners’ risk, 105t on burns data and surveillance, 58 violence against women in, 74 Index 281 ECO-AUDIT Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, we leverage electronic publishing options and print- on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. We follow the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on Forest Stewardship Council (FSC)–certified paper, with nearly all containing 50–100 percent recycled con- tent. The recycled fiber in our book paper is either unbleached or bleached using totally chlorine-free (TCF), processed chlorine–free (PCF), or enhanced elemen- tal chlorine–free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://www.worldbank.org/corporateresponsibility.