86304 Transport for Health The Global Burden of Disease from Motorized Road Transport Foreword by World Bank Group President Jim Yong Kim Global Road Safety Facility INSTITUTE FOR HEALTH METRICS AND EVALUATION The World Bank Group UNIVERSITY OF WASHINGTON Transport for Health The Global Burden of Disease from Motorized Road Transport Global Road Safety Facility INSTITUTE FOR HEALTH METRICS AND EVALUATION The World Bank Group UNIVERSITY OF WASHINGTON 4 | Transport for Health 5 | Transport for Health Lead authors: Institutional affiliation: ABOUT THE GLOBAL ROAD SAFETY FACILITY AT Kavi Bhalla Johns Hopkins Bloomberg School of Public Health THE WORLD BANK GROUP Marc Shotten The World Bank The Global Road Safety Facility (GRSF), a global partnership program administered by the World Bank, was established to help address the growing crisis of road traffic Aaron Cohen Health Effects Institute deaths and injuries in low- and middle-income countries. GRSF provides funding, Michael Brauer University of British Columbia knowledge, and technical assistance that catalyze further investments through Saeid Shahraz Schneider Institute for Health Policy World Bank projects addressing road safety. GRSF also partners and collaborates Richard Burnett Health Canada with other multilateral organizations, the private sector and NGOs, and country- based agencies. To express interest in collaborating with the GRSF, or to receive Katherine Leach-Kemon Institute for Health Metrics and Evaluation copies of this publication, please contact the GRSF at www.worldbank.org/grsf. Greg Freedman Institute for Health Metrics and Evaluation Christopher J.L. Murray Institute for Health Metrics and Evaluation ABOUT IHME Contributing authors: Institutional affiliation: The Institute for Health Metrics and Evaluation (IHME) is an independent global Rita Van Dingenen European Commission, Joint Research Centre health research center at the University of Washington that provides rigorous and Frank Dentener European Commission, Joint Research Centre comparable measurement of the world’s most important health problems and Theo Vos Institute for Health Metrics and Evaluation evaluates the strategies used to address them. IHME makes this information freely Mohsen Naghavi Institute for Health Metrics and Evaluation available so that policymakers have the evidence they need to make informed decisions about how to allocate resources to best improve population health. Jerry Abraham Harvard School of Public Health David Bartels Massachusetts General Hospital To express interest in collaborating, participating in GBD training workshops, or Pon-Hsiu Weh School of Medicine, University of Texas receiving updates of GBD or copies of this publication, please contact IHME at: Health Science Center, San Antonio Institute for Health Metrics and Evaluation 2301 Fifth Ave., Suite 600 Seattle, WA 98121 USA Telephone: +1-206-897-2800 Fax: +1-206-897-2899 Email: comms@healthmetricsandevaluation.org www.healthmetricsandevaluation.org 7 | Transport for Health ACKNOWLEDGMENTS The Global Burden of Diseases, Injuries, and Risk Factors Study 2010 (GBD 2010) was implemented as a collaboration among seven institutions: the Institute for Health Metrics and Evaluation (IHME) as the coordinating center, the University of Queensland School of Population Health, Harvard School of Public Health, the Johns Hopkins Bloomberg School of Public Health, the University of Tokyo, Imperial College London, and the World Health Organization. This summary draws on seven GBD 2010 papers published in The Lancet (2012 Dec 13; 380). GBD 2010 had 488 co-authors from 303 institutions in 50 countries. The World Bank and IHME oversaw the production of this publication. We would like to thank the World Bank peer reviewers for their comments, particularly Anne-Maryse Pierre-Louis of the Human Development Network, Veronica Raffo of the Latin America Transport Unit, and Andreas Kopp of the Transport Anchor. We are grateful to the report’s writer, Kavi Bhalla; to William Heisel, Stephen Lim, and Rhonda Stewart at IHME for content guidance; to Daniel Greenbaum and Robert O’Keefe at the Health Effects Institute for valuable comments; to Ryan Barber, Stan Biryukov, Megan Coggeshall, Daniel Dicker, and Diego Gonzalez-Medina for data analysis; to Brittany Wurtz for program coordination and fact checking; to Katherine Leach-Kemon for publication management; to Patricia Kiyono for produc- tion oversight; to Brian Childress for editorial support; and to Ann Kumasaka for design. This report would not have been possible without the ongoing contributions of Global Burden of Disease collaborators around the world. In particular, members of the Global Burden of Disease Injury Expert Group and Air Pollution Expert Group made substantial contributions to the findings presented in this report. Finally, we would like to extend our gratitude to the Global Road Safety Facility at the World Bank for co-financing this report and for supporting the work of the Injury Expert Group, and to the Bill & Melinda Gates Foundation for generously funding IHME and for its consistent support of Global Burden of Disease research. 8 | Transport for Health 9 | Transport for Health List of figures List of Tables PAge table PAge Figure 24 1. Percentage of global health loss that can be attributed to motorized 22 1. Leading causes of death worldwide, associated DALYs, and burden road transport compared with other leading risk factors, 2010 attributable to motorized road transport, 2010 25 2. Global shifts in healthy years lost due to road injuries and ambient 35 2. Underreporting in official statistics of road traffic deaths in China and air pollution from all sources, 1990 to 2010 India, 2010 26 3a. Overall ambient air pollution levels (PM2.5), 1990 37 3. Countries with the highest death rates due to road injuries among those with at least three epidemiological measurements of road injury 26 3b. Change in ambient air pollution levels (PM2.5), 1990 to 2010 mortality, 2010 27 4. Change in motor vehicle ownership per capita, 1990 to 2010 40 4. Leading causes of death globally by age groups for males and 29 5. Trends in road injury death rates, 1980 to 2010 females, 2010 30 6. Percentage change in global road deaths since 2007 GBD 2010 countries by region 54 A1. 31 7. Death rates from injuries and air pollution due to motorized road transport, 2010 32 8. Ranking of health loss (DALYs) attributable to injuries and air pollution due to road transport compared with leading risk factors in global regions, 2010 33 9. Ranking of the top 25 global causes of deaths and their ranking in different regions, 2010 34 10. Underreporting of deaths from road injuries in official government statistics, 2010 36 11. National road injury death rates estimated by the GBD 2010 study compared with official national statistics from high-income countries that report to the International Road Traffic and Accident Database, 2010 38 12. Rate of healthy years lost to injuries and air pollution from motorized road transport, 2010 39 13. Global decline in age-specific death rate among males and females, 1980 to 2010 43 14. Deaths in road crashes by type of road user and region, 2010 44 15. Correlation between regional pedestrian injury death rates and total road injury death rates 45 16. Rate of nonfatal road injuries warranting health care, 2010 46 17. Cases and burden of nonfatal road injuries warranting hospital admission, 2010 59 A1. Country-level differences in estimated fraction of PM2.5 attributable to motor vehicles 10 | Transport for Health 11 | Transport for Health foreword foreword Transport for Health focuses timely attention on the growing burden that motorized This report summarizes the findings of a long and meticulous journey of data road transport imposes on global health development. By quantifying the burden of gathering and analysis to quantify the health losses from road deaths and injuries disease attributable to both road injury and air pollution from vehicles, the authors worldwide, as part of the path-finding Global Burden of Disease (GBD) study. It is have found that motorized road transport deaths exceed those from diseases such as important, first, to acknowledge the profound contribution made by the lead authors HIV, tuberculosis, or malaria. That is a powerful wake-up call. and global team of injury prevention professionals to estimate the disease burden of road trauma, before absorbing their findings and recommendations. Without their Mobility solutions need to address the whole gamut of human needs: transport that dedication and tenacity, the way forward would be less certain. The first GBD study, is clean and affordable must also be safe for people who want access to jobs, health, published nearly two decades ago, signaled an emerging road safety crisis in devel- and education. Road injuries now rank as the world’s eighth-leading cause of death oping regions of the world. It triggered a remarkable program of global advocacy and the number-one killer of young people ages 15 to 24. While the disease burden that culminated in the United Nations Decade of Action for Road Safety and Global attributed to ambient air pollution has declined among richer regions such as Western Plan to bring road safety outcomes under control in these regions by 2020. However, Europe and North America, over the last 20 years we have seen a sharp rise in South limited investment has been mobilized so far to implement the UN initiative. The Asia and East Asia. second GBD study, and related analyses presented in this report, confirm the impor- tance of road safety as a global development priority and the urgency with which it These alarming findings underscore the urgent need to spread improvements in must be addressed. transport pollution and safety across world regions. Building the institutional capacity to address this challenge, by mobilizing expertise across health and transport sectors, The report’s findings highlight the growth in road deaths and injuries globally, and is also crucial. It is a matter of life, death, and equity: approximately 90% of all road their substantial impacts on maternal and child health, despite sustained reductions crashes now happen in low- and middle-income countries; yet they own only half of over the last three to four decades in high-income countries. Combined with the the world’s motor vehicles. More than half of global deaths are among pedestrians deaths arising from vehicle pollution, the road transport death toll exceeds that of, and operators of motorized two-wheeled vehicles. Rates are higher in the world’s for example, HIV/AIDS, tuberculosis, malaria, or diabetes. This statistic further rein- poorest regions. These losses are tragic and needless. Families often lose their forces the call for global action. Without these GBD estimates we would not have breadwinners or have to pay for expensive medical treatment. Many are plunged into a clear picture of the true situation because official country data in the developing poverty as a result. world vastly understate the scale of road transport health losses. Road crashes cost an estimated 1% to 5% of GDP in developing countries, under- The report’s findings also underscore the policy complexities of managing the mining efforts to reduce poverty and boost shared prosperity. In the coming years, the adverse health impacts of increasing road vehicle travel. These include premature World Bank Group, our partners through the Global Road Safety Facility, the interna- death and disability arising from road crashes, air pollution, diminished physical tional donor community, and governments worldwide need to scale up efforts to save activity, and greenhouse gas emissions. While estimates of road crash injuries millions of lives and avoid serious injuries, as mandated by the United Nations Decade have been improved, and conservative estimates of the disease burden attribut- of Action for Road Safety 2011-2020. able to road vehicle pollution have been made, the adverse health impacts of reduced physical activity and greenhouse gas emissions resulting from increased The work conducted by the Institute for Health Metrics and Evaluation to update the road vehicle use are yet to be reliably estimated. Nevertheless, it is clear from the Global Burden of Disease (GBD) dataset is particularly valuable. These findings will evidence presented that the potential is substantial for safer active transport facili- help us sensitize policymakers to the staggering cost of current and future health ties to reduce the growing burden of non-communicable diseases and climate trends, and to mobilize appropriate responses to transport and health challenges in an change impacts. increasingly urbanized and motorized world. The issue remains of how to best use the report’s findings. A strong case is made Facts in hand, I am convinced we can achieve safer, cleaner, and more affordable for improved data collection and more robust accounting of adverse health transport solutions that benefit the poor, create resilient economies, and save millions impacts within a sustainable development framework. Multisectoral collaboration, of lives. strengthened institutional management capacity, and the development of more Jim Yong Kim comprehensive transport evaluation tools will be integral to this desired shift in President The World Bank Group 12 | Transport for Health 13 | Transport for Health practice. Above all, the report’s findings make it clear that the economic benefits of reducing the health losses accruing to road transport are too huge to ignore. In the Preface case of road safety and air quality improvements, they have proved to outweigh their costs – as demonstrated by the experience of high-income countries over the last four decades – and the business case for integrated action is compelling. Two decades ago, the first iteration of the Global Burden of Disease study brought attention to the growing toll of premature death and disability from road injury. Tony Bliss Since that time, deaths from road injuries increased to become the eighth-leading cause of death worldwide in 2010. Road injuries are a universal threat to population Global Road Safety Advisor health across rich and poor countries alike and disproportionately impact the most Monash University Accident Research Centre economically productive age groups in society. Principal Advisor The increasing use of vehicles over time has led to more air pollution, which also Commission for Global Road Safety negatively impacts health. This report marks the first effort to quantify both the burden of disease attributable to road injury and the burden linked to air pollu- tion from vehicles. While the consequences of road injury tend to be immediate and severe, air pollution from vehicles is more insidious and can lead to ischemic heart disease, stroke, chronic lung diseases, and lower respiratory infections. To strengthen our understanding of this health risk, we need to improve the data available on vehicle pollution and the methods for estimating its contribution to disease burden. If this challenge can be met, it will allow researchers and policy- makers to monitor progress in reducing the burden from vehicle emissions. Ministry of health officials are typically viewed as the chief stewards of countries’ population health, but reducing the disease burden from motor vehicles requires action from multiple sectors. As economies grow and demand for cars and roads increases, the transport sector plays a vital role in designing, building, and main- taining an infrastructure and regulatory system that will encourage economic growth while minimizing health loss. Policymakers can improve health by imple- menting measures shown to effectively reduce disease burden from transport, such as vehicle safety and emissions regulations, seatbelt and helmet requirements, and speeding and drunk-driving laws. These regulations are only as effective as they are enforced legally, requiring investment to ensure compliance with these laws. The success we see in high-income countries, where premature death and disability from road injuries have dropped, shows us it is possible to reverse this growing problem. The United States reduced its burden of road injuries by 16% between 1990 and 2010, despite a significant increase in population and vehicles on the road. Many Western European and high-income countries in the Asia Pacific region reduced their burdens even more dramatically. Japan reduced its disease burden from road injuries by 42% between 1990 and 2010, and Sweden lowered its burden by 30%. Case studies of interventions, policies, regulations, and institutional capacity to deliver them in these high-achieving countries could help elucidate key lessons that other nations can follow to reduce the burden of road injuries. Annual updates of the Global Burden of Disease will allow decision-makers to continually monitor the impact of road injuries in their countries as they implement 14 | Transport for Health 15 | Transport for Health new policies and design programs to address this important health problem. By collaborating across sectors, replicating effective policies carried out in other Executive Summary countries, and making an ongoing commitment to improve data and research, we can get closer to realizing a world where everyone enjoys the benefits of safe transport and lives a long and healthy life. Rapid improvements in road transport have helped many nations make progress toward their development goals. Transport is often one of the most highly funded Christopher J.L. Murray sectors in development bank lending portfolios due to ongoing demand from borrowers as well as its role in stimulating economic growth and competitiveness. Director and Professor of Global Health At the same time, the global development community is increasingly concerned Institute for Health Metrics and Evaluation about the social costs of growth in road transport, particularly the impacts on University of Washington human health due to the rise in road traffic injuries and impacts on non-communi- cable diseases via emissions and decreased physical activity. This report quantifies, for the first time, the global health loss from injuries and air pollution that can be attributed to motorized road transport. It combines estimates of the global burden of road injuries based on a large pool of new data from the most information-poor regions with estimates of the health effects of pollution from vehicles. The results of this analysis show the following: • Motorized road transport imposes a large burden on population health, resulting in more than 1.5 million deaths and 79.6 million healthy years of life lost annu- ally. Deaths from road transport exceed those from HIV, tuberculosis, or malaria. Injuries and pollution from vehicles contribute to six of the top 10 causes of death globally. • Road injuries have a substantial impact on maternal and child health. Health loss attributable to motorized road transport exceeds that from key risk factors affecting children, including childhood underweight and suboptimal breast- feeding. Road injuries rank among the top 10 causes of death after the first year of life through age 59. In addition, road injuries are a top-10 cause of death among women of childbearing age and are the fourth-leading cause among women aged 15 to 29 years. • The burden due to motorized road transport is growing. Over the last two decades, deaths due to road crashes grew by 46%. Deaths attributable to air pollu- tion, to which motor vehicles are an important contributor, grew by 11%. • Health loss resulting from the combined effects of road injuries and pollution from transport is substantial in all regions. While the deaths from road transport are dominated by road injuries in poorer regions, such as sub-Saharan Africa, health loss due to pollution from vehicles tends to be highest among richer regions, such as Western Europe. • Injuries are responsible for most of the burden of motorized road transport, accounting for 95% of the healthy life years lost. Road crashes result in 1.3 million deaths annually and 78.2 million nonfatal injuries warranting medical care. • Pedestrians alone account for 35% of road injury deaths globally and over 50% in East and Central sub-Saharan Africa. 16 | Transport for Health 17 | Transport for Health • Pollution from vehicles is the cause of 184,000 deaths globally, including 91,000 deaths from ischemic heart disease, 59,000 deaths from stroke, and 34,000 deaths INTRODUCTION from lower respiratory infections, chronic obstructive pulmonary disease, and lung cancer. • Official government statistics substantially underreport road injuries. Estimates Motorized road transport poses a growing threat based on Global Burden of Disease 2010 data suggest, for example, that road to population health injury deaths are more than twice the official statistics in India, four times those in International and national development agencies have long viewed building roads China, and more than six times the official numbers in parts of Africa. as a key strategy for driving economic growth and improving the health and well-being of people. Providing reliable transport infrastructure can stimulate Key conclusions and recommendations economic development in several ways. For instance, foreign direct investment is attracted to regions that provide high-quality road infrastructure to facilitate efficient This report reaffirms the need for safe and clean transport for achieving global logistics.1 Within a country, road infrastructure connects remote areas with centers health and development goals. It calls for a multisectoral collaboration that includes of trade and connects centers of industry to global markets, spurring the growth of the transport, health, and urban sectors, among others, to help achieve beneficial trade and reducing costs by improving access to goods and services.1 and sustainable development. In particular: Since 2000, the Millennium Development Goals (MDGs) have been the central focus • Road injuries are a major contributor to the Global Burden of Disease. Thus, of global development. While the MDGs did not explicitly address the transport rapidly scaling up road safety programs alongside the expansion of transport is sector, roads have been viewed as crucial for successfully achieving several of the vital for saving lives while promoting development. Mitigating the health risks MDGs. In 2005, the African Union and the United Nations (UN) Economic Commis- requires a long-term investment strategy to build the capacity of national insti- sion for Africa wrote a report2 calling for widespread infrastructure improvements, tutions so they can actively manage safety and mobility performance through stating, “The significance of transport services to each of the MDGs means that targeted interventions. This is necessary given the multisectoral complexity of effective pursuit of the latter requires priority attention to those transport services, road safety and demands a systematic approach rather than isolated efforts with which are relevant to each.” Roads bring people closer to health care facilities and specific interventions. educational opportunities. In particular, rural connectivity helps reduce maternal • While malnutrition, diarrhea, and many infectious diseases occur in settings of mortality through better access to maternal care.3 Rural roads have impact on extreme poverty, the health burden associated with road transport spreads with increasing the enrollment of girls in school.3 In addition to being an end in itself, economic growth and rapid motorization. This need not be the case, provided education of girls is important to population health as it helps reduce fertility rates countries, aid agencies, and donors develop comprehensive and country-specific and improve maternal and child health, among other mechanisms.4 Similarly, roads policy frameworks for investing in the health and well-being of populations. It facilitate access to food markets and can promote better nutrition.3 took developed countries 70 years to reverse negative health trends from road transport, but developing countries can accelerate this process through strategic Motorized road transport has grown briskly in recent decades, especially in regions investments and collaboration across sectors. with the most rapidly growing economies. In the last two decades, China has built a highway system that, by Chinese government estimates, rivals that of the • Due to limitations of available data and methods, we have likely underestimated United States, and it plans to further develop the network substantially over the the effects of pollution from vehicles. Additional research is needed to obtain next decade.5 In India, rapid expansion of the highway infrastructure is currently more detailed geographic information on human exposure to air pollution in underway because insufficient road transport is viewed as a key impediment to rapidly motorizing regions and to better understand the health effects of exposure industrial growth and has been viewed as a reason for the country’s failure to to traffic-related pollutants. In addition, a comprehensive accounting of the burden achieve the full benefits of economic reforms.6 In sub-Saharan Africa, where most of road transport requires research to quantify the loss of physical activity due to people do not have access to all-weather roads, road transport is seen as a key solu- motorization, which is not possible with currently available data and methods. tion to providing basic services, reducing poverty, and driving economic growth.3 • There is an urgent need for better tracking of the health impacts of road transport. Statistical systems need to be expanded and improved to collect key indicators to However, motorized road transport is also closely linked to several threats to human monitor and evaluate these effects. The absence of reliable accounting of health health and well-being that have not been previously assessed in a comprehensive impacts not only endangers effective multisectoral action, it can waste government manner. These harms include the direct health effects of road injuries and vehicular resources or lead to development aid funding being targeted at ineffective solutions. emissions, the indirect health effects of sedentary lifestyles resulting from frequent use of motorized transport, and the threat of catastrophic environmental damage 18 | Transport for Health 19 | Transport for Health through climate change, to which vehicular emissions are a key contributor. Each of unless these activities are safe.10 In fact, a growing body of literature shows that these issues is increasingly receiving attention. Our knowledge of global road safety such programs are most successful when they employ an integrated approach that has substantially improved in recent years due to several epidemiological studies includes providing safe infrastructure such as sidewalks and bike lanes, supportive such as the Global Burden of Diseases, Injuries, and Risk Factors Study 2010 (GBD land use planning, and advocacy and education.11 2010), which is the basis of this publication. While the Global Burden of Disease due to ambient air pollution has also been characterized, the contribution of emissions Providing infrastructure that makes active modes of travel safe, secure, and conve- from vehicles has never been estimated. In this report, we estimate these for the nient is particularly important, making the transport sector the key enabler for the first time. Although studies to estimate the global burden of physical inactivity have multisectoral goal of healthy development. Major international institutions, such become increasingly sophisticated, the contribution of motorization to physical inac- as development banks, are attempting to reframe their development agenda to tivity has not been measured and represents an important gap in our assessment of be consistent with these principles of sustainable transport and development.12 the health impacts of motorized road transport. Finally, although action to address Commitments for sustainable transport growth, as seen at the 2012 Rio+20 climate change is urgently needed, the connections between transport, climate, and Summit,12 although small, are important steps toward this broader vision. health have been extremely difficult to quantify.7 Purpose of this report Despite these numerous limitations, the systematic and comprehensive assessment of the health effects of motorized road transport presented in this publication can This report is the first attempt to assess the combined direct global health loss that help guide decision-making in the transport sector for a safer and more prosperous can be attributed to motorized road transport. Specifically, we quantify the health future. impacts of motorized road transport from 1) injuries due to road traffic crashes over the last two decades, and 2) air pollution generated by motorized road transport (“pollution from vehicles”). Need for a multisectoral approach Since the identification of road injuries as one of the top 10 causes of death and We did not estimate indirect health impacts, such as physical inactivity, and distal disability by the original GBD study in 1993, substantive global efforts have been impacts, such as climate change, because we lacked sufficient epidemiological data undertaken to establish road safety as a development priority. In May 2011, the UN to construct estimates of these effects at a global scale. Finally, we discuss the impli- Decade of Action for Road Safety 2011-2020 was launched in more than 100 coun- cations of our findings for transport and health policy. tries with the goal of preventing 5 million road traffic deaths and 50 million serious injuries. The launch was a culmination of substantial efforts by international and national agencies. These included the release of the 2004 World Report on Road Traffic Injury Prevention by the World Health Organization (WHO) and the World Bank, multiple UN and World Health Assembly resolutions calling on governments to improve road safety, and the first Global Ministerial Conference on Road Safety in 2009.8 Numerous co-sponsoring country governments, key UN regional agencies, and multilateral development banks have endorsed the call for the Decade of Action and its goals. At the same time, non-communicable diseases, which are linked closely to motor- ized transport through the health effects of air pollution and physical inactivity, have emerged as a major health problem. In September 2011, the High-level Meeting of the UN General Assembly adopted the Political Declaration on the Prevention and Control of Non-communicable Diseases.9 It is now considered likely that non- communicable diseases will figure prominently in the post-MDG era beginning in 2015. A multisectoral approach for addressing the health impacts of transport is key for maximizing public health improvements. For example, promoting active transport, such as walking and biking, has emerged as a guiding vision among public health professionals. However, people tend not to walk, bike, or take public transport 20 | Transport for Health 21 | Transport for Health across the globe for the year 2005 and extrapolated it to 1990 and 2010 based on methods observed trends. Second, we estimated the contribution of road transport to overall PM2.5 for 2010 in all countries using a global air quality source-receptor model that links emissions of pollutants within a given source region with downwind impacts. Our estimates of the health losses due to motorized road transport are based We estimated the country-specific burden of disease attributable to total PM2.5 expo- on GBD 2010, which is a systematic, scientific effort to quantify the compara- sure using nonlinear functions for the leading global causes of death and disability tive magnitude of health loss due to 291 diseases and injuries, 1,160 sequelae associated with outdoor air pollution: ischemic heart disease, stroke, lung cancer, (direct consequences of disease and injury, also known as complications), and 67 chronic obstructive pulmonary disease, and acute lower respiratory tract infections risk factors for 20 age groups and both sexes in 1990, 2005, and 2010. GBD 2010 in children. These risk functions were used in GBD 2010 to generate country-level produced estimates for 187 countries and 21 regions. GBD 2010 generated estimates estimates of the burden of disease attributable to exposure to ambient PM2.5. We of the burden of road injuries as well as the burden that can be attributed to outdoor then estimated the country-specific burden of disease attributable to PM2.5 from road air pollution from all sources. For the first time, we estimated for this report the transport by multiplying the country-specific total PM2.5-attributable burden by the amount of premature death and disability that is attributable to air pollution from region-specific (country-specific for some large countries) proportion of ambient motor vehicles (“pollution from vehicles”) and combined that with the burden of PM2.5 from road transport. road injuries to construct estimates of the total health loss due to motorized road We report burden of disease using the following standard metrics of population transport. This chapter provides a synopsis of methods. For more details, see health loss: Annex 1 and associated GBD 2010 publications.13 • Years of life lost (YLLs): This is the number of years of life lost due to premature death. It is calculated by multiplying the number of deaths at each age by a Estimating the burden of road injuries standard life expectancy at that age. As part of GBD 2010, we accessed and assessed all empirical measurements of • Years lived with disability (YLDs): Years of life lived with any short-term or population health that could inform estimates of the incidence of fatal and nonfatal long-term health loss, adjusted for severity. road injuries. These included data from vital registration systems, verbal autopsy studies, mortuary/burial registers, household surveys, hospital databases, and • Disability-adjusted life years (DALYs): This is the sum of YLLs and YLDs. prospective studies of disability outcomes following injuries. Prior to analysis, these DALYs are also defined as years of healthy life lost. data sources were systematically cleaned and standardized. Next, we estimated road injury mortality in all countries using Cause of Death Ensemble Modeling (CODEm), which is an analytical tool used in GBD 2010 that tests a wide variety of possible statistical models of causes of death and creates a combined “ensemble” model that provides the best predictive performance. We assessed the burden of nonfatal road injuries by first calculating the incidence of nonfatal crashes using population-based data (e.g., household surveys), then estimating the complica- tions from crashes using hospital data. Next, we estimated the duration of disability based on prospective follow-up studies, and finally, estimated health loss by applying disability weights (a number on a scale from 0 to 1 that represents the severity of health loss associated with a health state). Estimating the burden of vehicular emissions We estimated human exposure to annual average levels of fine particulate air pollution (PM2.5) from road transport in two steps. First, we developed a database of geo-referenced, annual average PM2.5 measurements from surface monitors in 2005, combined with estimates of PM2.5 derived from satellite-based observations and estimates of PM2.5 from a global atmospheric chemical transport model (TM5). Thus, we estimated PM2.5 levels at a grid resolution of 10 x 10 square kilometers 22 | Transport for Health 23 | Transport for Health Global Health Loss Due to Injuries and Injuries resulting from road crashes account for 95% of the combined burden of ill Pollution from Road Transport health from motorized road transport. Road injuries killed 1.33 million people glob- ally in 2010 and were the eighth-leading cause of death, accounting for 2.5% of all global deaths. The road injury death toll exceeded that from diseases such as tuber- Motorized road transport and health loss culosis and malaria that receive substantial attention in the global health research and development community. They were the 10th-leading cause of healthy life years Injuries and air pollution generated by motorized road transport were associated lost, contributing 3.0% of the total global health burden. They were also the eighth- with six of the top 10 causes of death and five of the top 10 causes of premature leading cause of premature mortality. death and disability, also known as disability-adjusted life years (DALYs), in 2010 (Table 1). In fact, the top three causes of death, premature mortality (YLLs), and In addition to injuries, pollution from vehicles causes a broad range of acute and premature death and disability are diseases that are linked to air pollution, which is chronic health effects, ranging from minor physiologic disturbances to death from closely associated with motorized road transport. Overall, injuries and air pollution respiratory and cardiovascular diseases. In 2010, we estimate that exposure to from road transport caused 1.5 million deaths globally, representing 2.9% of deaths pollution from vehicles, in terms of particulate matter pollution (PM2.5) derived from from all causes. Together, they were the sixth-leading cause of death in 2010, with a vehicular emissions, resulted in 184,000 deaths globally. This includes 91,000 deaths death toll exceeding those from HIV/AIDS, tuberculosis, malaria, and diabetes. They from ischemic heart disease, 59,000 deaths from stroke, and an additional 34,000 were responsible for 79.6 million healthy life years lost, or DALYs, which is 3.2% of deaths due to lower respiratory infections, chronic obstructive pulmonary disease the total global burden of disease and injuries. (COPD), and lung cancer combined. As explained in Annex 1, we expect that these results underestimate the health loss attributable to pollution from vehicles. While this report was being prepared, the International Council for Clean Transpor- tation (ICCT) completed an analysis of the mortality attributable to ambient PM2.5 Table 1: Leading causes of death worldwide, associated DALYs, and burden attributable to from motor vehicles, using similar methodology but some different input data- motorized road transport, 2010 sets.14 Overall, the results of this analysis were similar to ours. The ICCT estimated 230,000 deaths per year in 2005, compared to the 184,000 deaths per year in 2010 Burden attributable to that we estimate. As in our analysis, the ICCT found that the greatest disease burden Global burden of disease motorized road transport attributable to air pollution from motor vehicles was observed in East Asia, followed Rank Cause Deaths DALYs Deaths DALYs by Western Europe, South Asia, and North America. In both our work and the ICCT analysis, mortality rates attributable to ambient PM2.5 from motor vehicles were 1 Ischemic heart disease 7,029,270 129,795,464 90,639 1,909,563 highest in Western, Central, and Eastern Europe, high-income Asia Pacific countries, 2 Stroke 5,874,181 102,238,999 58,827 1,148,699 and North America. 3 COPD 2,899,941 76,778,819 17,266 346,376 Figure 1 compares the combined burden of injuries and air pollution from motor- 4 Lower respiratory infections 2,814,379 115,227,062 5,670 489,540 ized road transport with other leading risk factors. In 2010, road transport ranked 5 Lung cancer 1,527,102 32,405,411 11,395 232,646 eighth, with a burden comparable to alcohol use, which is also a key contributor to 6 HIV/AIDS 1,465,369 81,549,177 – – the burden of road injuries. It ranked ahead of important global health risks such as 7 Diarrheal diseases 1,445,798 89,523,909 – – childhood malnutrition and risks faced in the workplace. 8 Road injury 1,328,536 75,487,102 1,328,536 75,487,104 9 Diabetes mellitus 1,281,345 46,857,136 – – 10 Tuberculosis 1,195,990 49,399,351 – – All other causes 24,207,527 1,682,995,639 – – Total 52,769,676 2,482,258,070 1,512,333 79,613,928 Note: In the “burden attributable to motorized road transport” column, emissions from road transport contribute to deaths and DALYs from ischemic heart disease, stroke, COPD, lower respiratory infections, and lung cancer. Road transport accidents contribute to deaths and DALYs from road injury. 30 | Transport for Health 31 | Transport for Health There is an urgent need to make all relevant data available and comparable to Comparisons of burden of disease from road reduce the uncertainty in monitoring health impacts of transport. In particular, transport across regions WHO and the World Bank have in-country networks with access to rich data on Figure 7 shows death rates from motorized road transport across different regions in health outcomes and transport covariates. Pooling these together can create an 2010. The figure shows considerable variation in the relative contribution of pollution unprecedented dataset to assess global injury metrics. GBD 2010 has already made and injuries by region. The general pattern suggests that in the poorest regions of the substantial advances in developing the tools to combine data from a wide range of world, deaths from road transport are dominated by road injuries. For example, inju- sources into coherent global health metrics. Further efforts to make transport and ries accounted for approximately 99% of total deaths attributable to road transport in related health data available could narrow the existing gaps in our knowledge about sub-Saharan Africa. In contrast, in Western Europe, pollution from vehicles contrib- global road safety. uted nearly half the burden (44%). These variations are caused by a wide range of factors. These include the relative success or failure of different regions in reducing road injuries and controlling pollution from vehicles. In addition, they reflect differ- Figure 6: Percentage change in global road deaths since 2007 ences in ages across regions because road injuries tend to affect young adults, while air pollution has a greater impact on young children and older people. 45% The importance of a risk factor for population health depends on its ranking relative to other risk factors for premature death and disability. Figure 8 ranks the leading risk factors according to their contribution to disease burden in each region. Color- 35% coding is used to indicate how high a risk factor ranks in a region. The leading risk % change in road injury deaths since 2007 25% Figure 7: Death rates from injuries and air pollution due to motorized road transport, 2010 15% Global High-income Asia Pacific Western Europe 5% Australasia High-income North America Central Europe WHO Global Status Report GBD 2010 Southern Latin America -5% Eastern Europe East Asia Tropical Latin America -15% Central Latin America Southeast Asia Central Asia -25% Andean Latin America North Africa and Middle East Note: Vertical lines represent the uncertainty interval surrounding the estimate. Caribbean South Asia Oceania Southern sub-Saharan Africa Eastern sub-Saharan Africa Central sub-Saharan Africa Western sub-Saharan Africa 0 10 20 30 40 Death rate (per 100,000) Road injuries Air pollution 40 | Transport for Health 41 | Transport for Health Table 4: Leading causes of death globally by age groups for males and females, 2010 Males Under 1 1-4 years 5-14 years 15-29 years 30-44 years 45-59 years 60-74 years 75+ years Cause Cause Rank Cause Cause Cause Cause Cause Cause 1 Preterm birth Malaria Road injury Road injury HIV/AIDS Ischemic heart disease Ischemic heart disease Ischemic heart disease complications 2 Lower respiratory Lower respiratory HIV/AIDS Interpersonal violence Road injury Stroke Stroke Stroke infections infections 3 Neonatal encephalopathy Diarrheal diseases Diarrheal diseases Self-harm Ischemic heart disease Cirrhosis COPD COPD 4 Neonatal sepsis Protein-energy Lower respiratory HIV/AIDS Tuberculosis Lung cancer Lung cancer Lower respiratory malnutrition infections infections 5 Diarrheal diseases HIV/AIDS Malaria Tuberculosis Self-harm Tuberculosis Lower respiratory Lung cancer infections 6 Congenital anomalies Drowning Drowning Drowning Interpersonal violence Road injury Diabetes Diabetes 7 Malaria Meningitis Typhoid fevers Malaria Cirrhosis HIV/AIDS Tuberculosis Hypertensive heart disease 8 Meningitis Road injury Meningitis Lower respiratory Stroke Liver cancer Cirrhosis Prostate cancer infections 9 Protein-energy malnutrition Measles Congenital anomalies Mechanical forces Lower respiratory infections COPD Stomach cancer Other cardio & circulatory 10 Syphilis Fire Forces of nature Diarrheal diseases Liver cancer Self-harm Liver cancer Chronic kidney disease Females Under 1 1-4 years 5-14 years 15-29 years 30-44 years 45-59 years 60-74 years 75+ years Cause Cause Rank Cause Cause Cause Cause Cause Cause 1 Preterm birth Malaria Diarrheal diseases HIV/AIDS HIV/AIDS Ischemic heart disease Ischemic heart disease Ischemic heart disease complications 2 Lower respiratory Diarrheal diseases HIV/AIDS Maternal disorders Maternal disorders Stroke Stroke Stroke infections 3 Neonatal encephalopathy Lower respiratory Malaria Self-harm Tuberculosis Breast cancer COPD COPD infections 4 Neonatal sepsis Protein-energy Lower respiratory Road injury Ischemic heart disease Diabetes Diabetes Lower respiratory malnutrition infections infections 5 Diarrheal diseases HIV/AIDS Road injury Tuberculosis Self-harm HIV/AIDS Lower respiratory Diabetes infections 6 Congenital anomalies Meningitis Meningitis Malaria Stroke COPD Lung cancer Hypertensive heart disease 7 Malaria Measles Drowning Fire Road injury Lung cancer Breast cancer Alzheimer’s disease 8 Protein-energy malnutrition Congenital anomalies Typhoid fevers Diarrheal diseases Lower respiratory infections Cirrhosis Hypertensive heart disease Other cardio & circulatory 9 Meningitis Drowning Congenital anomalies Lower respiratory Breast cancer Tuberculosis Diarrheal diseases Lung cancer infections 10 Syphilis Road injury Fire Interpersonal violence Diarrheal diseases Cervical cancer Cirrhosis Chronic kidney disease 42 | Transport for Health 43 | Transport for Health Road injuries are an important threat to maternal Figure 14: Deaths in road crashes by type of road user and region, 2010 and child health Remarkably, road injuries rank among the top 10 causes of death for children after 100 the first year of life, ranking eighth among boys 1 to 4 years old and 10th among girls 1 to 4 years old, globally (Table 4). Road injuries are the leading cause of death 80 among 1- to 4-year-olds in the high-income countries of North America and the second-leading cause of death in Western Europe, Australasia, and the high-income 60 % deaths countries of Asia Pacific. In fact, road injuries are a top-five cause of death for 1- to 4-year-olds in 11 of 21 global regions. 40 MDG 5 focuses on reducing maternal mortality by giving special priority to preg- nancy and childbirth. Although the cause of death ranking of road injuries is lower 20 among women than men, road injuries were among the top 10 causes of death 0 for women in every age group between 5 and 44 in 2010 (Table 4). Road injuries Global High-income Asia Pacific Western Europe Australasia High-income North America Central Europe Southern Latin America Eastern Europe East Asia Tropical Latin America Central Latin America Southeast Asia Central Asia Andean Latin America North Africa and Middle East Caribbean South Asia Oceania Southern sub-Saharan Africa Eastern sub-Saharan Africa Central sub-Saharan Africa Western sub-Saharan Africa were the fourth- and seventh-leading cause of death for women in the age groups 15 to 29 and 30 to 44, respectively. In the maternal age range of 15 to 49, we have estimated 3.5 million deaths from all causes in 2010. While maternal conditions accounted for 7.3% of these deaths, road injuries accounted for 4.0%. Road injuries rank prominently as a threat to the health of young women in most regions. For instance, among women 15 to 19 years old, road injuries are the leading cause of death in nine regions, are among the top three causes in 17 Pedestrian injury by road vehicle Pedal cycle vehicle regions, and are among the top 10 causes in all global regions with the exception of Motorized vehicle with 2 wheels Motorized vehicle with 3+ wheels Oceania. In Western sub-Saharan Africa, road injuries were the third-leading cause of death; they ranked among the top 10 causes in other regions of sub-Saharan Other road vehicle Africa. Regional variation in death rates among different types of road users Figure 15 illustrates that regional pedestrian death rates are strongly associated Figure 14 illustrates that at the global level, occupants of motor vehicles with three with overall road injury death rates and differ greatly across regions. Rates of death or more wheels are at the highest risk, accounting for 36% of all fatalities. Pedes- due to pedestrian injury vary by almost an order of magnitude. They are highest in trians account for 35% of global road deaths, and riders of motorized two-wheelers the three sub-Saharan Africa regions and lowest in Western Europe and Austral- make up another 16%. However, there is dramatic variation in death patterns among asia. Total road injury death rates in these regions mirror trends in death rates from road users across regions. In general, car occupants make up higher proportions pedestrian injury. of total road injury deaths in high-income regions, while pedestrians account for higher proportions in poorer regions. In two of the four sub-Saharan African regions In addition to being a key part of reducing road injuries, ensuring the safety of (Eastern and Central sub-Saharan Africa), pedestrian deaths make up over half of pedestrians is essential for reducing emissions from vehicles and increasing all fatalities. In contrast, in North America, vehicle occupants account for almost physical activity. Promoting active transport by protecting vulnerable road users can three-fourths (73%) of all fatalities. However, this general pattern has several excep- reduce the burden of non-communicable diseases, including ischemic heart disease, tions. There are many middle-income regions where car occupants make up a high stroke, lower respiratory infections, COPD, and lung cancer. Research shows that proportion of total road traffic deaths. Overall, in nine of 21 global regions, more the provision of safety infrastructure for walking and biking is among the most than half of all road deaths are vehicle occupants; six of those regions are low- and important ways to encourage these active modes of transport.11 Such infrastructure middle-income regions, including North Africa and the Middle East, Central Asia, includes traffic calming measures to reduce vehicle speeds, such as the use of speed Southern Latin America, and the Caribbean. These results suggest that vastly bumps, curb extensions, chicanes, and roundabouts, and the provision of separated different road safety strategies will be needed to maximize lives saved in different sidewalks and bicycle lanes to reduce exposure to motor vehicles.18 settings across the globe. 48 | Transport for Health 49 | Transport for Health and are therefore exposed to higher levels of air pollution. Residing near heavily protocol as the reporting standard in the World Bank-financed Ibero-American Road trafficked roads is associated with poor health outcomes including asthma in Safety Observatory in Latin America, utilizing twinning partnerships between high- children. and low-income countries’ agencies to help develop statistical capacity. There are three additional reasons why we believe our estimates of the burden On the other hand, in countries with the highest underreporting, the existing of disease from vehicle pollution are too low. First, it is unlikely that the data we national health surveillance infrastructure is too weak to reliably track road injury use to estimate air pollution exposure globally capture emissions from all types of mortality. Although developing such should be an ongoing priority, it is a slow vehicles. Second, our estimates of the burden of disease from vehicle pollution do process that will take many years. Immediate attention therefore should be given not include ozone, which is a pollutant formed in the atmosphere that comes from to using all existing data sources to construct statistical estimates of the national transport and other sources. Ozone pollution is linked to death and disability from burden of road injuries for guiding safety programs. GBD 2010 takes such an chronic lung diseases such as COPD. Finally, because the relationship between air approach with an emphasis on developing global and regional models. Such pollution and mortality is nonlinear, other approaches to the statistical analysis used work needs to be extended further at the national level to better utilize the known to estimate the burden of disease from vehicle pollution may produce somewhat strengths and biases of local data sources. In addition to generating immediate larger estimates. More detailed information about these limitations can be found in evidence to guide policy, this approach also helps identify potential country data Annex 1 and the Web appendix. systems, such as mortuary surveillance, national household surveys, and hospital registries, that can be strengthened and expanded to build reliable information infra- Currently, these limitations constrain our ability to quantify premature death and structure. disability from vehicle pollution on a global scale. Improved data and disease burden estimates are possible, however, and will be required to guide governments As acknowledged earlier in the report, the expansion of the road and transit network as they design and implement transportation policies designed to reduce the public has long been viewed as a key strategy for driving economic growth and improving health burden of road transport. Development of better data warrants financial the health and well-being of people. However, unmanaged growth without the requi- support from all concerned with this complex and growing problem. site capacity and oversight from country agency and regulatory bodies can result, in the case of developing countries, in decades of motorized road transit systems that inflict large amounts of harm on their populations, without government capacity to Recommendations target interventions correctly. While it took high-income countries decades to reduce 1) Rapidly scale up road safety programs and crash reporting capacity to save lives their road injury death toll, low- and middle-income countries can greatly reduce and promote economic development this timeframe by rapidly investing in a long-term strategy to actively manage safety Road injuries are a major contributor to the Global Burden of Disease and are vastly and mobility performance, the topic of the next section. underreported. Governments in many low- and middle-income countries report a substantially lower road injury death toll than our estimates. In the poorest coun- 2) Promote strong institutional development for multisectoral collaboration in the tries of sub-Saharan Africa, which have the highest road injury death rates, official emerging sustainable development era of safe and clean mobility government statistics often report less than one-fifth of road injury deaths. Even in Global health is undergoing a rapid transition away from mostly infectious diseases the rapidly developing economies of Asia, such as China and India, official statistics that affect children to non-communicable diseases and injuries that affect adults. often account for less than half of all road injury deaths. At the global level, the sum This requires adjustment on multiple fronts on the part of different actors ranging of countries’ official death counts (641,000 deaths) published by the WHO in their from country governments to the private sector to the NGO community. Devel- 2013 Global Status Report on Road Safety is less than half of our global estimate opment agencies will need to create new comprehensive policy frameworks to (1.3 million). Unless accounting of road injuries in official statistics reported to WHO implement cross-sectoral change in a logical and sequenced manner in client coun- is improved, it is likely that road safety will continue to be neglected in national tries and in their own internal global practice areas. This reflects the complexity of health and development priorities. an ever-changing, multidimensional environment that demands a clear accounting of emerging health threats to the planet. A substantial international effort should address how road injury data are collected and the health burden is estimated. Some low- and middle-income countries already For example, the 2004 World Report on Road Traffic Injury Prevention, authored have a relatively strong information infrastructure, such as high-quality national jointly by the WHO and the World Bank, focused on creating empowered lead agen- vital registration systems, and underreporting is relatively low. In these countries, cies with statutory responsibility to reduce road injuries. This went hand-in-hand strengthening existing systems for recording and reporting road injury statistics will with recommendations to ensure adequate resources for these agencies to manage enhance the quality of disease burden estimates. Such efforts should include stan- road safety in a multidisciplinary manner across relevant sectors of government. dardization of definitions and methodologies, such as the use of the OECD/IRTAD 50 | Transport for Health 51 | Transport for Health Unfortunately, attempts by many actors in the internal community to address road of economic growth and rapid motorization. An appropriate proportion of this safety in low- and middle-income countries still tend to take an “intervention-first” economic gain needs to be assigned to managing the negative impacts of the trans- approach, focusing on individual risk factors such as wearing helmets, using seat- port sector. belts, preventing drunk driving, and social marketing campaigns. Addressing the enormous and growing health losses from road transport will While these interventions are needed, sustainability of road safety programs require large investments in building and managing transport systems that are requires a government commitment to systematically invest in building transporta- safe, clean, and affordable. Although this report does not aim to estimate monetary tion systems that promote safe mobility in a holistic manner. Establishing a losses, calculations based on the 2001 WHO Report of the Commission on Macro lead road safety agency, building reliable data systems, and other system-wide Economics and Health have shown that such losses are substantial – equivalent investments that encompass vehicle quality, enforcement, safe infrastructure, and to 1% to 3% of gross domestic product per annum and calculated in 2014 for road road users in the pre-crash, in-crash, and post-crash stages are key. 21 crashes as high as 4.6% for India and 10.1% for Uganda – potentially exceeding the amount of international development assistance flowing into these countries.21 In Therefore, the overarching new science of delivering effective transfer of road safety comparison, current investments in road safety remain miniscule.8 Clearly, there is a knowledge means taking weak existing management capacity within a complex strong financial case for increasing investments in safe and clean road transport. system and creating the ability to shift rapidly to a safe system approach focused on getting results.21 An example is the case of Argentina, which has made road safety a Five decades of experience from high-income regions suggests that growth in trans- national priority through investment in its National Road Safety Agency, which itself port systems can be managed to reduce injuries and air pollution with appropriate is the recipient of a standalone project loan for road safety from the World Bank, one investments. In most high-income countries, road safety has steadily improved of the first of its kind. Cooperation across different ministries, in the transport and since the early 1970s despite increasing vehicle ownership rates and continued health sectors in particular, is central to the lending package, which will undertake expansion of highway infrastructure. As noted elsewhere, the policy history of systematic, measurable, and accountable investment through targeted programs. these countries suggests that they established national road safety agencies with Such an example can help provide lessons for other low- and middle-income coun- legislative powers and a mandate to manage safety in the transport system. These tries striving to replicate an “agency-first” model as advocated in the World Report agencies instituted a long series of interventions that targeted highway infrastruc- and the 2013 World Bank Safe System Projects Guidelines. ture (e.g., by requiring median barriers, guard rails, traffic calming designs), vehicle safety (e.g., by requiring airbags, seatbelts, child seats, crashworthiness standards, Lowering motor vehicle pollution is also important to reducing non-communicable crash avoidance technologies), and road users (e.g., through stronger enforcement diseases. Further, encouraging walking, biking, and active lifestyles has emerged of and social marketing campaigns for seat-belt use, helmet use, and preven- as a key strategy because of their cardiovascular health benefits. However, as our tion of drunk driving). Developing countries can utilize new tools and World Bank report indicates, active modes of transport make people highly vulnerable to road Guidelines to prioritize short-, medium-, and long-term road safety investment and injuries. Thus, partnerships between the transport, health, and urban planning sequence it in a way that makes sense for sustaining gains.21 sectors are necessary for developing solutions for healthy mobility. Increasingly, one of the key drivers of 21st-century competitiveness will be how countries – Similarly, as awareness of the health effects of air pollution grew, high-income coun- particularly urban centers – design their land use patterns to improve the health of tries developed comprehensive policies to reduce motor vehicle emissions. This their populations. This represents a fundamental switch toward linking transit to included a range of regulatory strategies, including emissions control technologies, health outcomes. This change is currently being discussed in the context of a fuel-composition modifications mandated to meet various air quality objectives, and post-2015 Sustainable Development Goal world, and is particularly important to vehicle inspection programs. For developing countries, reduction technologies such protect vulnerable generations in low- and middle-income countries. as three-way catalysts and particle traps, the elimination of leaded fuel, and agree- ment on a globally sustainable level of fuel sulfur content will become critical in the 3) Commit the resources needed to realize the health and economic gains from a coming years. safe and clean transport system Developing transport solutions that deliver health will require financial commit- 4) Systematically account for the health impact of road projects ments to support a wide range of strategies. Although the health impacts of road Successfully addressing concerns that are posed by rapid growth in the road sector transport compete for attention with other pressing global health concerns, these in low- and middle-income countries requires improved accounting of health occur in a development context that is different from that of many diseases and impacts. At present, more effort is invested in estimating the economic rate of return illnesses. As GBD 2010 illustrates, malnutrition, diarrhea, and many infectious of road projects than in estimating either the social benefits (access to health care, diseases occur in settings of extreme poverty where financial resources are severely education, markets) or social harms (burden of disease, environmental costs) of limited. However, the disease burden associated with road transport is an outcome transport. For example, to obtain a loan to finance road projects from infrastructure 52 | Transport for Health 53 | Transport for Health lending agencies, such as multilateral development banks, governments must prove Annex 1: the viability of these loans through careful accounting of the economic returns on GBD 2010 Methods investment. To accomplish this, transport planners work in conjunction with trans- port economists to develop models that can project demand for transport, estimate savings in travel time, monetize savings, identify revenue streams, and estimate What is the Global Burden of Disease 2010 (GBD 2010) rates of return for loans. These analyses are successfully done in complex transpor- Study? tation markets that are shaped by individual behavioral decisions about interrelated In 1991, the World Bank commissioned the first Global Burden of Disease study choices of residential location, transport modes, trip destinations, and vehicle char- to develop a comprehensive and comparable assessment of the burden of 107 acteristics, among others. diseases and injuries and 10 selected risk factors for the world and eight major To fully realize the societal benefit of transport infrastructure, it is important that regions. The findings represented a major improvement in global knowledge of comparable efforts are made to quantify the full costs and benefits of road proj- population health metrics and proved to be influential in shaping the global health ects. We need analytical models of crash causation that can be used to estimate the priorities of international health and development agencies. The study also stimu- injuries and deaths that will be caused or avoided by proposed road projects. Impor- lated numerous national burden of disease analyses that have informed debates on tantly, such analysis will require empirical measurement of the risks associated with health policy over the last two decades. different types of road infrastructure. Our existing knowledge of causal relationships GBD 2010, the most recent iteration of the study, is a comprehensive update of the orig- between road injuries and environmental factors is remarkably weak, especially in inal study and presents estimates for 291 diseases and injuries, 67 risk factors, and 1,160 low- and middle-income countries where most new road building is currently occur- sequelae (nonfatal health consequences) disaggregated by sex and 20 age groups for ring. 21 regions (Table A1) covering the entire globe. The study is a collaboration of hundreds Similarly, analytical models are needed to characterize health impacts of changes in of researchers around the world, led by the Institute for Health Metrics and Evaluation vehicular emissions that accompany road projects. Such analysis requires building at the University of Washington and a consortium of several other institutions, including spatially refined emission inventories that rely on local driving patterns, fuel types, Harvard University, Imperial College London, Johns Hopkins University, University of and vehicle types. In addition, these models need to account for noncombustion Queensland, University of Tokyo, and the World Health Organization. sources of particulate matter, such as resuspended road dust, tire wear, and brake Diseases and injuries result in either premature death or life lived with ill health. GBD wear. These models will need to be validated against actual measurements through aims to quantify the gap between the ideal of a population that lives a full life in full field valuation and verification of particulate matter emissions. health and reality. GBD uses the following concepts to measure this health burden: Finally, analytical work is needed to characterize the impact of individual transport • Years of life lost (YLLs) are the number of years of life lost due to premature death. projects and the broader transport sector on human physical activity and green- They are calculated by multiplying the number of deaths at each age by a stan- house gas emissions. In 2010, physical inactivity was the 11th-largest risk factor for dard life expectancy at that age. years of healthy life lost, accounting for 2.8% of the total disease burden. We were unable to quantify the contribution of motorized transport to physical inactivity due • Years of life lived with disability (YLDs) are the number of years of life lived with to lack of information. Similarly, vehicular emissions are an important contributor short-term or long-term health loss weighted by the severity of the disabling to anthropogenic climate change. However, although the impact of climate change sequelae of diseases and injuries. on human health is likely to be large, it acts through complex causal pathways. Our • Disability-adjusted life years (DALYs) are the main summary measure of popu- ability to model the diverse effects from greenhouse gases across populations is lation health used in GBD to quantify health loss. DALYs provide a metric that stymied by the lack of systematic studies in this area. allows comparison of health loss across different diseases and injuries. They are calculated as the sum of YLLs and YLDs; thus they are a measure of the number of Improving human health and well-being is ultimately a key goal of all development years of healthy life that are lost due to death and nonfatal illness or impairment. projects. Health impacts should not be viewed as a potential externality but rather be part of the holistic development objective of transport projects, which aim to How did we construct estimates of the burden of support, among other things, healthy, productive lives. This is what we mean with road transport? the title of this report, Transport for Health. This report brings together two streams of work undertaken within GBD 2010: first, a comprehensive effort to improve the evidence base of the estimates of the burden of road injuries using new data sources and improved methods; and second, advances in GBD 2010 in estimating the burden of disease that can be attributed to 54 | Transport for Health 55 | Transport for Health Table A1: GBD 2010 countries by region long-term exposure to air pollution, which we have further partitioned to estimate the contribution from air pollution caused by motorized road transport. Andean Latin America Bolivia, Ecuador, Peru Estimating the global burden of road injuries Australasia Australia, New Zealand The guiding principle of the burden of disease approach is that estimates of popula- Caribbean Antigua and Barbuda, Bahamas, Barbados, Belize, Cuba, Dominica, tion health metrics (such as incidence and prevalence) should be generated after Dominican Republic, Grenada, Guyana, Haiti, Jamaica, Saint Lucia, careful analysis of all available data sources and correction for bias. A substantial Saint Vincent and the Grenadines, Suriname, Trinidad and Tobago project-wide effort was made to incorporate data from vital registration and sample Central Asia Armenia, Azerbaijan, Georgia, Kazakhstan, Kyrgyzstan, Mongolia, Tajikistan, registration systems, demographic surveillance systems, and many others. This Turkmenistan, Uzbekistan broad search was coupled with a targeted effort to improve data on road injuries from the most information-poor settings. As a result, a wealth of data from regions Central Europe Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Hungary, such as sub-Saharan Africa was used for the first time in epidemiological research. Macedonia, Montenegro, Poland, Romania, Serbia, Slovakia, Slovenia Key data sources for injuries included the following: Central Latin America Colombia, Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Venezuela • Vital registration statistics: These are tabulations from national vital registration systems, which usually record causes of death listed on death certificates. Central sub-Saharan Africa Angola, Central African Republic, Congo, Democratic Republic of the Congo, Equatorial Guinea, Gabon • Verbal autopsy: This is a method of determining cause of death in which a trained interviewer uses a structured questionnaire to collect information about symp- East Asia China, North Korea, Taiwan toms that preceded an individual’s death. Such surveillance is commonly done in Eastern Europe Belarus, Estonia, Latvia, Lithuania, Moldova, Russia, Ukraine regions that do not have reliable vital registration systems. Eastern sub-Saharan Africa Burundi, Comoros, Djibouti, Eritrea, Ethiopia, Kenya, Madagascar, Malawi, Mauritius, • Mortuary/burial registers: Medico-legal records from mortuaries and burial permit Mozambique, Rwanda, Seychelles, Somalia, Sudan, Tanzania, Uganda, Zambia offices were another important source of data for information-poor regions. High-income Asia Pacific Brunei, Japan, Singapore, South Korea • Household surveys: These were a critical source for estimating the incidence of High-income North America Canada, United States nonfatal injuries. North Africa and Middle East Algeria, Bahrain, Egypt, Iran, Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco, • Hospital databases: Large hospital registries were used as a valuable source of Palestine, Oman, Qatar, Saudi Arabia, Syria, Tunisia, Turkey, United Arab Emirates, information about the sequelae resulting from injuries. Yemen • Prospective studies of disability outcomes: The results from follow-up studies of Oceania Fiji, Kiribati, Marshall Islands, Micronesia, Papua New Guinea, Samoa, patients after an injury were used to estimate the duration of disability and the Solomon Islands, Tonga, Vanuatu probability that an injury results in permanent disability. South Asia Afghanistan, Bangladesh, Bhutan, India, Nepal, Pakistan Prior to analysis, these data sources were subjected to systematic harmoniza- Southeast Asia Cambodia, Indonesia, Laos, Malaysia, Maldives, Myanmar, Philippines, Sri Lanka, tion and data cleaning. This includes adjusting for completeness of mortality data Thailand, Timor-Leste, Vietnam sources, mapping across different coding schemes, and reattribution of poorly Southern Latin America Argentina, Chile, Uruguay specified causes. Southern sub-Saharan Africa Botswana, Lesotho, Namibia, South Africa, Swaziland, Zimbabwe We estimated mortality from road crashes in 40 age-sex groups for all countries Tropical Latin America Brazil, Paraguay from 1980 to 2010 using Cause of Death Ensemble Modeling (CODEm), which involves developing a large range of plausible statistical models between the cause Western Europe Andorra, Austria, Belgium, Cyprus, Denmark, Finland, France, Germany, Greece, and known covariates, testing all possible permutations of covariates, and gener- Iceland, Ireland, Israel, Italy, Luxembourg, Malta, Netherlands, Norway, Portugal, ating ensembles of the component models. The performance of all component Spain, Sweden, Switzerland, United Kingdom models and ensembles is evaluated based on their out-of-sample predictive validity Western sub-Saharan Africa Benin, Burkina Faso, Cameroon, Cape Verde, Chad, Côte d’Ivoire, Gambia, Ghana, and the best-performing model or ensemble is chosen. Guinea, Guinea-Bissau, Liberia, Mali, Mauritania, Niger, Nigeria, São Tomé and Príncipe, Senegal, Sierra Leone, Togo We estimated the burden of nonfatal outcomes of injuries by first constructing estimates of the incidence of the external causes of injuries using household survey data, hospital data, and the injury mortality estimates. We used hospital databases 56 | Transport for Health 57 | Transport for Health to estimate the incidence of the health outcomes (e.g., fractures, dislocations) that To estimate the risk of mortality across the full global range of estimated ambient result from road injuries. We estimated the long-term disability from these health concentrations of PM2.5, exposure-response functions were developed that inte- outcomes using data collected from studies that have followed patients after they grated epidemiologic evidence for the hazardous effects of particulate matter at sustained a road injury. Finally, we computed YLDs by applying disability weights. different concentrations from different sources and environments. Study-level These methods rely on many assumptions and will likely undergo substantial refine- estimates of the relative risk of mortality associated with any or all of ambient PM2.5, ments in the years to come. However, they are the only known attempt at large-scale secondhand smoke, household air pollution, and active smoking were compiled coupling of empirical data to construct global estimates of the burden of nonfatal for the following outcomes: ischemic heart disease, stroke, lung cancer, chronic road injuries. obstructive pulmonary disease, and acute lower respiratory tract infection in children. Several nonlinear functions with up to three parameters for fitting the inte- More details about GBD 2010 data sources and methods can be obtained from the grated exposure-response relationship were evaluated and assessed with respect to following four publications: goodness of fit. These integrated exposure-response curves were used to generate • Lozano RL, et al. Global and regional mortality from 235 causes of death for GBD 2010 estimates of the burden of disease attributable to exposure to ambient 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden PM2.5. For more details about methods and findings, see Burnett et al.22 of Disease Study 2010. The Lancet. 2012 Dec 13;380:2095–2128.  To estimate the motor-vehicle contribution to PM2.5, we used the global air quality • Vos T, et al. Years lived with disability (YLDs) for 1,160 sequelae of 289 diseases source-receptor model, TM5-FASST, developed by the Joint Research Centre of the and injuries, 1990-2010: a systematic analysis for the Global Burden of Disease European Commission.23 Briefly, this model links emissions of pollutants within a Study 2010. The Lancet. 2012 Dec 13;380:2163–2196. given source region with downwind impacts, using knowledge of meteorology and • Murray CJL, et al. Disability-adjusted life years (DALYs) for 291 diseases and atmospheric chemistry. TM5-FASST reproduces, for country and regional aver- injuries in 21 regions, 1990-2010: a systematic analysis for the Global Burden of ages, concentration levels of air pollutants simulated by the full TM5 Chemical Disease Study 2010. The Lancet. 2012 Dec 13;380:2197–2223. Transport Model, which has been fully evaluated in a large number of international comparisons (e.g., the Task Force on Hemispheric Transport of Air Pollution24) and • Salomon J, et al. Common values in assessing health outcomes from disease and is described in more detail in Brauer et al. and references cited therein. The emis- injury: disability weights measurement study for the Global Burden of Disease sions information used to assess the contribution of various source sectors to Study 2010. The Lancet. 2012 Dec 13;380:2129–2143. ambient PM2.5 concentrations are those used in the Global Energy Assessment25 and described in detail by Rao et al.26 Using TM5-FASST, PM2.5 concentrations attributable Estimating the burden of air pollution from motorized road transport to road transport were calculated for each of 56 global regions along with estimates A database of geo-referenced, annual averaged fine particulate matter (PM2.5) of PM2.5 levels attributable to all anthropogenic emissions sources, which allowed measurements from surface monitors in 2005 was assembled from national, us to estimate the proportional contribution of PM2.5 from road transport to the regional, and local air-quality monitoring reports and published literature. As total PM2.5 levels for 1990 and 2010 in each of the 21 GBD 2010 regions. The concen- surface monitor-based measurements of PM2.5 do not cover all populations, these trations attributable to road transport were adjusted at the country level to also data were combined with two other estimates of PM2.5: 1) estimates of PM2.5 derived consider the contributions from “natural” sources such as mineral dust and sea salt from satellite-based observations of aerosol optical depth, a proxy measure for by comparing the total anthropogenic source concentrations with the concentrations PM2.5, combined with information from a global atmospheric chemistry transport of PM2.5 from all sources, as described in Brauer et al. We then estimated the country- model (GEOS Chem); and 2) estimates of PM2.5 from a separate global atmospheric specific burden of disease attributable to PM2.5 from road transport by multiplying chemistry transport model (TM5). Both of these estimates were linked to the geo- the country-specific total PM2.5-attributable burden by the region-specific (country- referenced surface monitor-based estimates of PM2.5 in the database described specific in some large countries) proportion of ambient PM2.5 from road transport.15 above, and a regression model was used to relate the average of the satellite-based and chemical transport model estimates of PM2.5 with the surface monitoring data. We have, in all likelihood, underestimated the burden of disease attributable to Together, these estimates of PM2.5 allowed for the quantification of PM2.5 levels at vehicular emissions. This is the result of the limitations of the available data and a grid resolution of 10 x 10 square kilometers across the globe for the year 2005. methods, which are described below. Information on trends in air pollution concentrations and emissions were used to estimate PM2.5 levels for 1990 and 2010. Population estimates for each grid cell from Geographic misalignment of exposure data the Gridded Population of the World were used to estimate population-weighted We applied a regional fraction to both urban and rural areas, which likely underesti- average exposure for each GBD region. For more information, see Brauer et al.15 mates contribution in urban areas, where pollution is higher. For example, the traffic fractions applied to China and India are approximately 2% and 6%, respectively, but in Delhi and Beijing it is estimated to be as high as 20%.27 Similarly, in the US, we have estimated the fraction of ambient PM2.5 attributable to road transport as 15%, 58 | Transport for Health 59 | Transport for Health while a population-weighted estimate of this contribution derived from state-level Figure A1: Country-level differences in estimated fraction of PM2.5 attributable to motor vehicles emissions information across the US estimates this fraction as 26%.28 While this report was being prepared, the International Council for Clean Transporta- 90 tion (ICCT) completed a similar analysis of the mortality attributable to ambient PM2.5 80 from motor vehicles.14 Overall, results were similar, with the ICCT estimating 230,000 deaths per year in 2005, compared to the 180,000 deaths per year in 2010 that we 70 estimate. The same general regional patterns in mortality attributable to PM2.5 from motor vehicles were observed in the two analyses. In addition to the differ- 60 ence in years (2005 and 2010), the ICCT analysis uses a different chemical transport Frequency 50 model, the Model for Ozone and Related Tracers, version 4 (MOZART-4). As the ICCT analysis did not use age-stratified exposure-response functions and utilized some- 40 what different underlying mortality data, the two attributable mortality estimates are not directly comparable. Accordingly, while we cannot fully attribute the 28% lower 30 mortality in our estimate to the spatial misalignment in estimated motor vehicle 20 contributions, this comparison provides a rough estimate of the magnitude of this error and supports our suggestion that the estimates provided here are low. 10 Both our analysis and that conducted by ICCT used the same database of ambient 0 -14.4 -12.2 -9.9 -7.7 -5.5 -3.2 -1.0 1.3 3.5 5.7 8.0 10.2 12.5 >12.6 PM2.5 concentrations, however, and therefore allow us to directly compare the % Difference in motor vehicle fraction (MOZART 4 - TM5 FASST) fraction of ambient PM2.5 attributable to motor vehicles. Consequently, it is clear that the differences in these two estimates are due to differences in the underlying emissions databases and chemical transport models and to the different spatial resolution of the estimated contribution of motor vehicles to ambient PM2.5. While Limitations of the health data our analysis was conducted at the country level, the ICCT analysis was at 0.4° x 0.5°, Although there is considerable evidence regarding the adverse health effects of approximately 40 x 50 km at the equator. residential proximity to road traffic, GBD 2010 did not quantify the burden of disease We also compared the country-level contributions of motor vehicles to ambient that might be attributable to it. 29,30 As a result, our estimates do not include some PM2.5 between the two approaches (Figure A1). adverse effects of traffic-related air pollution, such as increased asthma incidence and severity in children. We have also not included the health impact of motor ICCT estimates were systematically higher than those used in our analysis, although vehicle contributions to ozone via emission of precursor compounds. GBD 2010 absolute differences were rather small. There was reasonable agreement between estimated that ozone exposure contributes to the incidence of chronic obstructive the two estimates overall (r=0.78), and generally rather small differences (mean pulmonary disease and globally was responsible for 152,000 deaths in 2010. Esti- absolute difference of 3.0%; SD of differences of 3.6%) in the estimated contribu- mates in the US suggest that mortality attributable to ozone via precursors emitted tions for specific countries, although large discrepancies were observed in some from road transport is roughly 10% to 15% of that attributable to PM2.5 from road locations. For example, ICCT estimated that contributions were more than 10% transport.27,30 Further, we do not assess impacts of mobile source “air toxics”31 such higher in Belgium, Canada, Denmark, Germany, Japan, Luxembourg, Malaysia, as benzene, 1,3-butadiene, formaldehyde, and acetaldehyde, although these are Mexico, the Netherlands, Singapore, and Venezuela, while TM5-FASST estimated expected to be much smaller than those attributable to the motor vehicle contribu- that contributions were more than 10% higher in Egypt and Malta. tions to PM2.5.32 In China (4.9% versus 2.2%) and India (5.6% versus 6.2%), differences between the Limitations of the statistical model ICCT study and our analysis were quite small, while in the US (23.5% versus 14.9%), The estimates reflect the impact of removing road transport given current levels differences were larger. of PM2.5 concentrations; however, due to nonlinearities in the integrated exposure- response curves, these numbers are smaller than the burden of road transport Limitations of the emissions data absent any other source of PM2.5. These issues are discussed in more detail in the The vehicular contribution to ambient PM2.5 is emissions-based, although emissions Web appendix. are processed through a chemical transport model. These estimates therefore only include those transportation sources that are included in currently available emis- sions inventories. 60 | Transport for Health 61 | Transport for Health Annex 2: Country Estimates 1990 2010 2010 Road injury deaths Road injury deaths Nonfatal road injuries Motor vehicle air pollution Total burden (air pollution + road injuries) Official country GBD 2010 Uncertainty Motorcycle Vehicle Injuries warranting Deaths Rate statistics road deaths range Rate Pedestrian Bicyclist rider occupant Other admission Total nonfatal injuries Deaths Uncertainty range Cause of death YLL YLD DALY Country count per 100,000 count count 95% CI per 100,000 % % % % % count count count 95% CI rank rank rank rank Afghanistan 4,590 34 1,501 10,213 (5,054 - 15,093) 32 14 6 12 64 4 36,483 345,765 1,388 (1,130 - 1,705) 10 8 5 8 Albania 332 10 352 395 (259 - 533) 12 16 7 13 51 13 4,149 33,823 136 (112 - 161) 12 11 5 9 Algeria 3,765 15 N.A. 4,283 (3,570 - 5,371) 12 6 9 8 74 2 51,649 432,149 417 (366 - 474) 9 7 7 6 Andorra 5 10 3 6 (3 - 8) 7 19 11 14 55 1 63 519 6 (3 - 10) 10 10 9 9 Angola 6,563 63 4,042 9,408 (2,450 - 31,110) 49 69 2 4 24 1 17,608 184,103 11 (8 - 13) 5 4 5 4 Antigua and Barbuda 3 5 N.A. 5 (4 - 7) 6 14 8 8 69 1 104 837 0 (0 - 1) 11 10 8 10 Argentina 3,389 10 5,094 6,067 (4,484 - 7,015) 15 15 7 14 63 1 39,338 341,421 278 (170 - 436) 9 9 10 8 Armenia 676 19 285 474 (352 - 776) 15 34 4 3 56 3 4,230 35,180 162 (134 - 192) 12 11 9 11 Australia 2,836 17 1,363 2,024 (1,629 - 2,590) 9 18 3 12 67 1 23,097 189,314 18 (11 - 28) 10 9 7 9 Austria 1,411 18 552 723 (622 - 991) 9 18 10 15 56 1 6,272 52,525 393 (311 - 489) 10 10 10 9 Azerbaijan 1,143 16 1,202 882 (585 - 1,510) 10 27 4 3 62 4 13,728 111,534 340 (297 - 383) 12 12 7 12 Bahamas 48 19 N.A. 57 (43 - 72) 17 8 5 12 73 2 397 3,375 1 (0 - 2) 8 5 7 6 Bahrain 128 26 73 256 (185 - 327) 20 3 5 5 87 1 2,162 18,211 10 (7 - 13) 6 3 8 4 Bangladesh 3,432 3 2,872 6,113 (4,148 - 10,330) 4 34 14 11 29 11 298,166 2,304,607 2,667 (2,129 - 3,177) 12 11 6 10 Barbados 30 11 19 31 (23 - 38) 11 14 6 7 72 1 332 2,709 2 (1 - 3) 10 9 7 8 Belarus 2,332 23 1,190 2,117 (1,637 - 2,687) 22 37 6 6 47 3 14,393 123,448 665 (537 - 806) 11 10 11 10 Belgium 1,921 19 840 1,345 (1,139 - 1,720) 12 15 16 18 49 2 7,362 64,902 928 (752 - 1,120) 10 10 9 9 Belize 19 10 41 59 (42 - 71) 19 10 11 11 64 2 337 2,911 1 (0 - 1) 8 5 8 5 Benin 982 21 816 1,726 (1,245 - 2,155) 19 36 5 16 39 4 10,624 91,918 22 (18 - 27) 6 6 8 7 Bhutan 75 14 79 87 (53 - 147) 12 39 10 11 36 3 1,482 11,794 12 (9 - 16) 10 6 7 7 Bolivia 1,476 22 1,681 1,989 (1,310 - 2,571) 20 45 5 4 45 2 11,183 97,324 23 (17 - 30) 8 6 9 7 Bosnia and Herzegovina 45 1 336 65 (34 - 132) 2 29 11 7 49 4 4,893 37,699 187 (156 - 221) 13 13 7 13 Botswana 155 11 385 283 (191 - 484) 14 36 7 14 38 5 1,962 18,242 0 (0 - 0) 8 4 9 4 Brazil 31,443 21 36,499 43,985 (35,301 - 52,857) 23 34 4 23 38 1 166,013 1,538,102 618 (426 - 852) 9 6 12 7 Brunei 40 16 46 50 (37 - 58) 12 17 13 12 58 0 356 3,076 1 (0 - 2) 8 7 9 7 Bulgaria 1,219 14 775 913 (739 - 1,092) 12 19 6 7 64 4 7,818 65,719 844 (726 - 978) 12 11 6 11 Burkina Faso 2,844 30 966 5,585 (4,271 - 7,113) 34 34 3 14 33 15 14,308 144,032 54 (41 - 70) 8 5 8 4 Burundi 2,097 37 357 2,534 (812 - 5,044) 30 35 16 12 25 12 9,588 89,842 17 (12 - 24) 8 7 9 6 Cambodia 875 9 1,816 2,394 (1,414 - 3,298) 17 10 6 23 52 10 21,924 183,274 129 (110 - 150) 10 7 7 8 Cameroon 4,051 33 1,353 6,951 (4,682 - 9,920) 35 41 4 13 37 4 20,934 205,855 52 (44 - 62) 5 4 8 4 Canada 4,191 15 2,227 2,962 (2,559 - 3,909) 9 17 5 10 68 1 33,251 275,144 607 (474 - 758) 10 9 10 10 Cape Verde 45 13 63 80 (36 - 177) 16 44 7 10 35 4 646 5,490 1 (1 - 1) 9 5 9 8 Central African Republic 916 31 145 1,911 (899 - 3,835) 43 47 4 15 26 9 4,357 44,557 12 (10 - 15) 5 6 5 6 Chad 954 16 3,226 2,765 (2,144 - 3,536) 24 38 4 16 36 7 11,096 102,518 23 (17 - 31) 10 9 4 9 62 | Transport for Health 63 | Transport for Health 1990 2010 2010 Road injury deaths Road injury deaths Nonfatal road injuries Motor vehicle air pollution Total burden (air pollution + road injuries) Official country GBD 2010 Uncertainty Motorcycle Vehicle Injuries warranting Deaths Rate statistics road deaths range Rate Pedestrian Bicyclist rider occupant Other admission Total nonfatal injuries Deaths Uncertainty range Cause of death YLL YLD DALY Country count per 100,000 count count 95% CI per 100,000 % % % % % count count count 95% CI rank rank rank rank Chile 1,587 12 2,071 2,204 (1,573 - 2,572) 13 47 7 7 37 2 17,104 144,068 220 (159 - 287) 9 8 9 7 China 155,521 14 70,134 282,576 (205,235 - 414,850) 21 37 3 17 26 16 1,903,239 16,300,000 27,379 (23,028 - 31,278) 10 6 6 7 Colombia 6,260 19 5,502 7,503 (5,997 - 9,241) 16 41 7 24 27 1 30,559 281,963 105 (73 - 145) 9 6 11 6 Comoros 143 33 14 213 (122 - 411) 29 49 18 5 24 4 851 7,871 0 (0 - 0) 5 6 8 6 Congo 1,005 42 269 1,916 (633 - 5,519) 47 65 2 5 27 1 3,236 35,686 15 (12 - 18) 6 4 10 4 Costa Rica 429 14 700 753 (625 - 913) 16 36 9 17 36 2 4,876 41,535 24 (16 - 34) 9 5 12 5 Côte d’Ivoire 3,383 27 699 6,536 (4,232 - 8,893) 33 37 4 17 39 4 19,363 188,795 68 (57 - 84) 8 6 10 6 Croatia 1,019 23 426 537 (443 - 669) 12 20 10 15 53 2 3,799 32,727 267 (207 - 335) 10 10 6 10 Cuba 2,247 21 809 1,162 (995 - 1,578) 10 32 17 12 35 5 13,751 112,686 83 (53 - 116) 12 9 7 8 Cyprus 167 26 60 111 (93 - 140) 15 17 4 21 56 2 1,076 9,006 26 (20 - 33) 9 9 7 8 Czech Republic 1,532 15 802 988 (795 - 1,229) 9 21 12 10 56 1 13,665 110,901 663 (552 - 776) 10 10 5 10 Democratic Republic of the Congo 6,497 18 332 7,733 (5,107 - 11,060) 12 35 5 23 28 8 82,668 678,838 136 (110 - 170) 15 12 7 12 Denmark 763 15 255 476 (394 - 603) 9 18 13 13 55 1 4,028 33,958 183 (135 - 236) 10 10 9 9 Djibouti 303 54 N.A. 345 (167 - 723) 39 65 8 3 22 2 993 9,634 1 (1 - 2) 4 2 11 2 Dominica 10 15 8 9 (7 - 11) 13 11 5 11 72 1 87 708 0 (0 - 0) 10 8 8 7 Dominican Republic 1,185 16 2,470 2,231 (1,730 - 2,581) 22 4 1 6 89 1 9,040 84,034 55 (41 - 72) 8 5 6 6 Ecuador 2,366 23 3,222 3,498 (2,798 - 4,157) 24 56 5 7 31 1 13,732 126,386 23 (16 - 32) 7 4 8 3 Egypt 7,025 12 9,602 11,708 (9,030 - 13,959) 14 30 3 2 64 1 116,416 964,142 11,315 (9,876 - 12,579) 8 7 7 7 El Salvador 1,375 26 1,017 1,589 (1,333 - 2,116) 26 10 14 14 58 3 5,344 50,112 15 (7 - 26) 8 5 12 6 Equatorial Guinea 178 47 53 524 (109 - 1,855) 75 68 2 3 26 1 560 6,796 1 (1 - 2) 4 3 7 3 Eritrea 682 22 N.A. 1,202 (898 - 1,673) 23 41 14 7 31 8 5,966 54,463 9 (7 - 11) 7 6 8 7 Estonia 383 25 78 126 (100 - 182) 9 21 9 4 57 8 1,417 11,820 26 (14 - 41) 12 11 11 10 Ethiopia 15,103 31 2,506 21,520 (16,689 - 27,821) 26 47 11 4 30 9 65,191 642,113 118 (96 - 143) 6 4 7 4 Federated States of Micronesia 11 11 2 14 (9 - 23) 12 19 8 7 58 7 163 1,340 0 (0 - 0) 10 10 10 10 Fiji 45 6 52 63 (53 - 78) 7 17 9 6 61 7 1,344 10,737 0 (0 - 0) 13 12 10 11 Finland 660 13 272 387 (326 - 529) 7 12 12 11 63 2 5,092 41,515 44 (25 - 69) 12 9 8 10 France 10,009 18 3,992 5,523 (4,699 - 7,626) 9 11 8 21 58 1 46,255 388,852 3,529 (2,808 - 4,280) 10 8 8 8 Gabon 586 63 327 1,267 (340 - 3,485) 84 68 2 5 25 1 1,172 14,868 2 (2 - 3) 4 2 9 2 Gambia 223 23 N.A. 387 (283 - 519) 22 37 4 15 38 6 2,016 17,906 4 (3 - 5) 7 5 7 6 Georgia 1,206 22 685 515 (378 - 795) 12 9 4 4 81 2 5,320 43,693 228 (181 - 279) 12 11 8 12 Germany 11,771 15 3,648 5,469 (4,689 - 7,584) 7 17 12 15 55 1 61,846 507,966 7,359 (6,118 - 8,729) 9 11 7 11 Ghana 2,053 14 1,986 4,844 (3,267 - 6,097) 20 38 5 7 46 4 32,905 281,393 72 (61 - 84) 7 5 8 6 Greece 2,179 21 1,451 1,773 (1,498 - 2,242) 16 15 8 25 41 11 12,006 103,222 742 (559 - 980) 11 9 7 9 Grenada 8 8 N.A. 13 (9 - 16) 12 8 7 7 76 2 126 1,042 1 (0 - 1) 11 8 8 9 Guatemala 590 7 958 944 (722 - 1,200) 7 37 5 8 45 4 17,381 137,969 44 (36 - 54) 14 13 9 12 Guinea 1,019 18 503 1,869 (1,409 - 2,305) 19 33 5 17 39 6 12,295 105,831 33 (26 - 41) 9 8 8 9 Guinea- Bissau 309 30 134 443 (288 - 600) 29 30 4 22 38 6 186 15,686 4 (3 - 6) 8 6 9 8 64 | Transport for Health 65 | Transport for Health 1990 2010 2010 Road injury deaths Road injury deaths Nonfatal road injuries Motor vehicle air pollution Total burden (air pollution + road injuries) Official country GBD 2010 Uncertainty Motorcycle Vehicle Injuries warranting Deaths Rate statistics road deaths range Rate Pedestrian Bicyclist rider occupant Other admission Total nonfatal injuries Deaths Uncertainty range Cause of death YLL YLD DALY Country count per 100,000 count count 95% CI per 100,000 % % % % % count count count 95% CI rank rank rank rank Guyana 68 9 112 127 (72 - 171) 17 10 10 11 66 3 846 7,203 4 (2 - 6) 9 8 7 9 Haiti 1,168 16 N.A. 1,395 (988 - 1,745) 14 7 9 10 65 8 11,270 94,986 62 (46 - 86) 14 13 7 12 Honduras 654 13 1,217 1,231 (982 - 1,542) 16 36 7 11 43 4 6,920 60,980 27 (18 - 37) 11 8 11 8 Hungary 2,194 21 740 1,246 (998 - 1,579) 12 28 18 9 44 1 13,285 109,725 816 (688 - 957) 12 11 7 11 Iceland 27 10 8 16 (12 - 19) 5 7 8 10 69 7 253 2,035 2 (1 - 3) 9 8 7 9 India 145,378 17 130,037 273,835 (176,843 - 440,771) 22 44 11 17 21 8 2,197,047 18,500,000 38,804 (32,697 - 44,928) 10 9 8 9 Indonesia 43,407 24 31,234 65,335 (53,625 - 80,627) 27 12 6 19 57 6 360,187 3,170,472 1,374 (1,167 - 1,606) 7 5 6 6 Iran 15,399 28 23,249 27,486 (19,719 - 34,419) 37 28 5 12 48 7 173,153 1,413,027 2,602 (2,265 - 2,951) 8 3 6 4 Iraq 1,638 9 5,708 2,593 (2,027 - 3,652) 8 27 7 9 54 3 43,833 349,096 931 (803 - 1,055) 10 10 7 10 Ireland 461 13 212 292 (238 - 412) 6 19 6 9 64 1 3,449 28,547 112 (75 - 154) 11 9 9 9 Israel 573 12 352 729 (578 - 858) 10 38 4 7 51 1 5,362 45,641 304 (248 - 376) 9 8 8 7 Italy 11,212 20 4,237 6,832 (5,829 - 9,084) 11 27 11 16 46 1 40,682 356,709 5,895 (4,731 - 7,104) 10 10 8 8 Jamaica 42 2 319 85 (36 - 122) 3 16 6 16 59 2 3,366 26,529 17 (12 - 22) 13 13 7 9 Japan 14,299 12 5,772 10,017 (8,284 - 14,084) 8 37 18 14 31 0 118,924 974,382 8,280 (6,524 - 10,200) 10 10 9 10 Jordan 543 16 670 728 (593 - 882) 12 24 6 5 63 2 8,715 71,435 185 (156 - 213) 9 7 8 7 Kazakhstan 3,768 23 3,379 3,965 (3,167 - 5,133) 25 26 4 4 64 2 19,011 170,027 283 (243 - 322) 10 10 8 10 Kenya 3,648 16 2,966 7,820 (5,183 - 13,628) 19 51 12 4 29 3 48,022 427,257 15 (10 - 21) 7 4 7 5 Kiribati 10 14 6 15 (11 - 19) 16 14 8 7 62 9 135 1,130 0 (0 - 0) 9 9 10 9 Kuwait 311 15 374 493 (415 - 595) 18 12 12 12 61 2 4,276 36,137 40 (34 - 45) 9 4 7 6 Kyrgyzstan 1,045 24 850 1,161 (900 - 1,394) 22 26 6 7 57 4 7,089 61,481 102 (87 - 118) 11 10 8 7 Laos 555 13 767 1,068 (670 - 1,539) 17 11 4 25 53 7 7,668 65,649 68 (55 - 82) 11 9 6 9 Latvia 882 33 218 344 (269 - 511) 15 37 8 5 48 2 4,830 39,225 84 (57 - 116) 11 10 11 9 Lebanon 492 17 533 516 (369 - 715) 12 17 6 4 71 1 6,348 51,495 245 (203 - 291) 9 9 8 8 Lesotho 76 5 362 232 (106 - 405) 11 26 8 18 37 11 2,362 19,535 7 (5 - 11) 15 14 10 16 Liberia 428 20 78 561 (199 - 983) 14 23 5 29 35 9 5,091 42,338 8 (6 - 12) 10 11 9 12 Libya 811 19 N.A. 1,322 (985 - 1,775) 21 17 5 5 72 1 9,136 76,161 83 (72 - 97) 9 6 9 7 Lithuania 1,085 29 299 613 (510 - 857) 18 33 10 6 49 2 4,914 41,634 130 (98 - 168) 10 10 11 9 Luxembourg 73 19 32 48 (36 - 60) 10 9 3 10 70 8 372 3,153 30 (19 - 43) 10 9 9 9 Macedonia 144 8 452 133 (96 - 157) 6 6 3 4 83 4 2,720 21,354 128 (104 - 153) 12 12 5 11 Madagascar 2,891 26 422 3,405 (2,631 - 4,846) 16 48 11 5 30 6 25,756 217,814 14 (9 - 20) 8 7 8 9 Malawi 2,722 29 976 4,867 (3,293 - 6,560) 32 43 12 7 32 6 16,259 154,318 3 (2 - 5) 7 6 11 6 Malaysia 2,638 15 6,872 4,106 (3,124 - 4,968) 14 4 4 21 69 1 52,427 422,519 405 (339 - 475) 9 8 9 7 Maldives 30 14 6 29 (22 - 38) 9 9 13 17 55 6 535 4,267 2 (0 - 4) 7 5 5 7 Mali 1,813 21 739 3,133 (2,379 - 3,924) 20 35 4 17 38 5 14,787 131,881 33 (26 - 44) 8 5 7 6 Malta 15 4 15 16 (12 - 19) 4 14 8 18 53 7 330 2,615 39 (22 - 58) 10 10 9 9 Marshall Islands 6 12 4 7 (5 - 10) 11 18 8 7 61 6 91 742 0 (0 - 0) 9 9 10 8 Mauritania 514 26 163 1,016 (743 - 1,383) 29 35 5 15 40 5 3,090 30,463 5 (4 - 6) 6 5 8 6 Mauritius 107 10 158 123 (79 - 151) 9 8 10 13 63 6 1,841 14,819 1 (0 - 3) 10 9 9 9 Mexico 15,954 19 17,301 20,096 (16,217 - 24,578) 18 41 4 7 46 2 55,622 558,214 2,179 (1,892 - 2,478) 9 7 11 7 Moldova 1,069 24 452 534 (447 - 745) 15 29 7 10 45 9 5,660 46,667 261 (215 - 309) 12 10 10 10 Mongolia 462 21 477 661 (456 - 908) 24 21 4 5 66 4 3,668 32,991 26 (23 - 30) 10 8 9 8 66 | Transport for Health 67 | Transport for Health 1990 2010 2010 Road injury deaths Road injury deaths Nonfatal road injuries Motor vehicle air pollution Total burden (air pollution + road injuries) Official country GBD 2010 Uncertainty Motorcycle Vehicle Injuries warranting Deaths Rate statistics road deaths range Rate Pedestrian Bicyclist rider occupant Other admission Total nonfatal injuries Deaths Uncertainty range Cause of death YLL YLD DALY Country count per 100,000 count count 95% CI per 100,000 % % % % % count count count 95% CI rank rank rank rank Montenegro 70 12 95 82 (66 - 96) 13 48 7 11 33 2 768 6,389 31 (26 - 36) 10 10 6 10 Morocco 2,210 9 3,778 2,857 (2,421 - 3,872) 9 7 7 6 77 2 42,311 339,187 454 (394 - 522) 9 8 9 9 Mozambique 2,264 17 2,549 7,154 (5,493 - 11,166) 31 41 11 8 30 10 26,996 238,575 3 (2 - 5) 6 4 8 4 Myanmar 4,528 12 2,464 9,277 (5,037 - 13,985) 19 9 6 23 53 9 56,886 490,076 548 (422 - 703) 9 9 8 6 Namibia 111 8 292 222 (157 - 385) 10 32 9 11 42 6 2,107 18,961 0 (0 - 1) 12 11 9 12 Nepal 2,598 14 1,689 3,293 (2,493 - 4,197) 11 28 13 15 26 19 57,934 461,572 675 (570 - 800) 11 9 8 11 Netherlands 1,780 12 640 1,068 (898 - 1,493) 6 12 28 14 46 1 12,579 102,940 1,092 (865 - 1,323) 10 10 8 10 New Zealand 743 22 375 454 (390 - 562) 10 11 3 9 76 1 4,654 38,567 3 (2 - 5) 11 9 8 8 Nicaragua 512 12 742 639 (525 - 798) 11 22 10 13 50 4 6,810 56,293 12 (7 - 18) 9 9 11 9 Niger 1,496 19 703 2,078 (1,412 - 2,821) 13 30 5 19 39 7 18,891 156,679 46 (35 - 61) 14 13 5 13 Nigeria 32,606 33 5,279 74,548 (55,477 - 91,154) 47 41 3 18 30 8 154,369 1,608,482 297 (241 - 362) 3 2 8 2 North Korea 3,518 17 N.A. 3,728 (1,602 - 6,331) 15 15 6 12 26 42 40,189 329,849 725 (593 - 867) 11 9 9 10 Norway 514 12 208 279 (235 - 378) 6 13 6 13 67 1 3,629 29,595 15 (9 - 25) 11 10 10 11 Oman 857 46 820 1,090 (856 - 1,331) 40 68 2 2 28 0 3,717 36,799 41 (36 - 46) 3 1 7 1 Pakistan 8,867 8 5,192 16,573 (12,746 - 22,510) 10 39 13 13 28 7 331,613 2,651,023 4,496 (3,855 - 5,238) 12 10 7 12 Palestine 286 14 N.A. 440 (321 - 560) 11 15 6 5 73 2 5,630 46,594 132 (108 - 155) 8 6 7 6 Panama 439 18 422 591 (501 - 774) 17 42 6 4 46 2 3,343 29,132 8 (5 - 13) 10 6 12 7 Papua New Guinea 475 11 269 871 (545 - 1,197) 13 23 7 10 52 7 9,780 79,442 1 (0 - 1) 12 11 8 13 Paraguay 500 12 1,206 1,247 (823 - 1,469) 19 41 3 28 27 2 6,362 55,947 8 (3 - 15) 9 7 10 8 Peru 2,682 12 2,514 3,973 (3,103 - 4,649) 14 41 4 3 51 1 35,026 291,197 69 (50 - 90) 9 4 7 5 Philippines 4,317 7 6,739 8,396 (6,464 - 10,535) 9 21 5 19 50 5 110,309 900,551 554 (416 - 714) 10 12 9 14 Poland 7,513 20 3,907 5,681 (4,590 - 7,152) 15 34 10 5 50 0 46,151 391,422 1,814 (1,537 - 2,117) 11 11 6 11 Portugal 2,853 29 937 1,327 (1,097 - 1,940) 12 25 9 12 49 5 4,914 46,631 593 (443 - 768) 11 9 9 8 Qatar 80 17 228 306 (212 - 380) 17 7 12 14 64 3 3,564 29,191 7 (6 - 8) 3 1 8 2 Romania 4,159 18 2,377 2,906 (2,389 - 3,687) 14 22 9 8 56 5 12,279 112,734 1,581 (1,353 - 1,831) 12 11 7 11 Russia 40,747 28 26,567 33,379 (27,469 - 40,921) 24 42 3 5 48 2 179,432 1,569,191 6,572 (5,439 - 7,787) 10 10 10 10 Rwanda 2,885 41 438 2,492 (1,431 - 5,488) 23 53 13 5 23 6 12,724 111,807 16 (12 - 20) 7 6 7 6 Saint Lucia 26 19 14 25 (20 - 33) 14 10 6 6 75 2 211 1,791 1 (0 - 1) 10 8 8 8 Saint Vincent and the Grenadines 8 7 5 11 (8 - 13) 10 18 10 9 61 2 135 1,095 1 (0 - 1) 11 9 8 9 Samoa 16 10 55 15 (11 - 21) 8 18 9 5 60 7 272 2,153 0 (0 - 0) 11 10 8 11 São Tomé and Príncipe 12 10 33 15 (10 - 24) 9 38 4 14 40 4 220 1,775 0 (0 - 0) 10 7 10 11 Saudi Arabia 5,757 36 6,596 9,128 (7,304 - 10,400) 34 5 1 1 93 0 42,258 342,789 333 (285 - 376) 5 2 8 4 Senegal 392 5 277 645 (307 - 1,406) 5 48 5 16 26 5 14,528 114,598 12 (10 - 16) 15 16 7 15 Serbia 1,176 12 660 988 (769 - 1,141) 10 24 11 12 52 1 12,272 100,090 564 (475 - 654) 10 10 6 10 Seychelles 8 12 13 12 (8 - 18) 15 16 6 17 61 1 123 1,008 0 (0 - 0) 9 9 8 9 Sierra Leone 951 24 357 1,095 (627 - 1,505) 19 29 5 24 36 6 7,063 61,334 14 (11 - 18) 10 9 10 9 Singapore 256 9 193 164 (123 - 212) 4 23 9 40 28 1 3,574 28,787 44 (23 - 64) 11 9 7 9 Slovakia 971 18 515 618 (527 - 801) 11 29 11 7 51 1 5,796 48,554 356 (298 - 415) 10 10 6 10 68 | Transport for Health 69 | Transport for Health 1990 2010 2010 Road injury deaths Road injury deaths Nonfatal road injuries Motor vehicle air pollution Total burden (air pollution + road injuries) Official country GBD 2010 Uncertainty Motorcycle Vehicle Injuries warranting Deaths Rate statistics road deaths range Rate Pedestrian Bicyclist rider occupant Other admission Total nonfatal injuries Deaths Uncertainty range Cause of death YLL YLD DALY Country count per 100,000 count count 95% CI per 100,000 % % % % % count count count 95% CI rank rank rank rank Slovenia 487 24 138 220 (183 - 297) 11 16 9 16 58 1 3,023 24,560 107 (85 - 132) 10 10 5 9 Solomon Islands 32 10 12 62 (44 - 84) 11 14 8 9 59 10 765 6,272 0 (0 - 0) 10 8 9 10 Somalia 1,898 28 N.A. 2,083 (1,509 - 3,255) 22 43 10 7 29 12 10,805 95,810 14 (11 - 18) 8 9 7 9 South Africa 2,597 7 14,804 4,479 (3,339 - 5,571) 9 50 5 7 36 1 51,312 422,129 231 (180 - 291) 10 10 9 11 South Korea 12,262 28 5,505 7,839 (6,365 - 9,651) 16 43 7 20 30 0 42,262 374,837 2,126 (1,673 - 2,586) 9 9 9 9 Spain 7,949 20 2,478 3,950 (3,403 - 5,439) 9 23 4 14 58 1 9,698 102,160 1,848 (1,487 - 2,253) 10 10 9 9 Sri Lanka 1,294 7 2,483 2,650 (1,832 - 4,226) 13 14 8 13 60 5 27,914 228,517 217 (168 - 271) 10 9 8 10 Sudan 5,511 21 3,582 10,278 (7,877 - 13,730) 24 65 7 3 22 3 27,318 273,830 43 (34 - 54) 5 4 9 5 Suriname 63 15 87 80 (55 - 97) 15 10 7 18 64 1 615 5,117 1 (1 - 2) 9 7 7 6 Swaziland 53 6 216 218 (127 - 346) 18 33 7 14 40 5 906 9,055 2 (1 - 3) 10 10 9 10 Sweden 1,011 12 266 512 (413 - 750) 6 14 9 13 63 1 7,134 57,704 159 (110 - 221) 11 9 8 10 Switzerland 1,120 17 327 594 (488 - 793) 8 28 12 11 47 2 2,691 24,535 444 (344 - 551) 10 10 7 9 Syria 825 7 2,118 1,100 (768 - 1,660) 5 18 6 5 68 3 24,325 194,113 970 (857 - 1,111) 9 9 7 10 Taiwan 5,330 26 N.A. 4,156 (3,234 - 5,562) 18 14 11 34 20 22 38,323 320,120 444 (363 - 537) 8 7 7 7 Tajikistan 803 15 442 619 (468 - 873) 9 20 5 4 66 5 9,690 78,354 97 (82 - 110) 15 12 6 14 Tanzania 4,857 19 3,582 9,404 (6,482 - 14,042) 21 53 7 5 29 6 51,035 464,028 12 (8 - 18) 7 5 8 6 Thailand 12,337 22 13,365 19,867 (14,779 - 24,943) 29 18 3 35 43 0 56,372 542,010 1,521 (1,276 - 1,828) 11 5 8 5 Timor-Leste 65 9 99 90 (49 - 134) 8 10 6 21 56 8 1,726 13,741 1 (1 - 2) 10 9 8 10 Togo 835 23 742 1,401 (966 - 1,733) 23 28 5 20 38 9 6,958 62,337 16 (13 - 20) 6 6 9 7 Tonga 10 11 6 12 (8 - 16) 11 20 8 4 59 8 148 1,207 0 (0 - 0) 11 11 9 10 Trinidad and Tobago 158 13 200 230 (170 - 295) 17 21 6 6 65 1 1,571 13,337 3 (1 - 5) 9 8 7 7 Tunisia 2,038 25 1,208 2,719 (1,880 - 3,317) 26 30 3 14 52 1 12,578 113,041 214 (181 - 255) 9 5 8 6 Turkey 8,022 15 5,253 5,810 (4,839 - 8,418) 8 19 3 8 68 3 63,339 520,623 2,402 (2,141 - 2,701) 9 9 8 10 Turkmenistan 704 19 N.A. 704 (487 - 1,118) 14 23 4 4 67 3 7,197 59,607 66 (55 - 75) 10 11 8 10 Uganda 3,185 18 2,954 7,365 (5,368 - 10,509) 22 54 10 7 24 5 37,368 332,414 30 (23 - 40) 5 4 9 5 Ukraine 12,059 23 6,116 8,007 (6,323 - 9,784) 18 37 7 5 47 5 49,729 431,242 4,272 (3,589 - 5,058) 11 10 9 10 United Arab Emirates 527 29 826 1,838 (1,128 - 2,654) 25 12 5 7 75 1 11,495 103,262 53 (45 - 62) 4 2 6 3 United Kingdom 5,526 10 1,905 3,710 (3,169 - 4,822) 6 24 5 16 54 1 45,987 376,369 3,384 (2,730 - 4,053) 12 10 9 11 United States 49,643 20 32,885 44,001 (36,199 - 53,473) 14 14 2 10 73 0 247,223 2,195,212 15,374 (12,643 - 18,263) 10 9 11 9 Uruguay 420 14 556 428 (332 - 514) 13 10 11 18 57 3 2,974 25,493 52 (26 - 84) 10 9 11 8 Uzbekistan 3,566 17 2,731 4,683 (3,598 - 6,555) 17 50 3 1 43 2 39,414 334,218 513 (448 - 582) 11 11 8 9 Vanuatu 19 13 4 35 (23 - 54) 14 16 8 7 61 8 332 2,776 0 (0 - 0) 9 8 9 8 Venezuela 4,696 24 7,714 7,616 (6,017 - 10,598) 26 33 8 12 45 2 25,149 240,924 150 (118 - 186) 8 5 11 5 Vietnam 9,146 14 11,859 16,371 (12,460 - 19,166) 19 13 7 58 15 7 249,726 2,034,092 607 (485 - 728) 8 5 6 6 Yemen 1,860 16 3,843 3,520 (2,003 - 5,220) 15 16 5 8 68 3 32,778 272,129 581 (486 - 710) 11 8 6 9 Zambia 2,276 29 1,348 2,798 (2,077 - 3,955) 21 53 8 5 30 4 14,883 131,637 4 (3 - 5) 9 6 9 6 Zimbabwe 1,453 14 1,777 3,527 (1,375 - 5,853) 28 11 10 31 11 38 10,477 96,577 2 (1 - 4) 10 8 9 8 Note: Official country statistics are the country-reported data presented in the 2013 WHO Global Status Report on Road Safety after adjustment to 30-day definition. Countries for which official data were not available are marked N.A. 70 | Transport for Health 71 | Transport for Health 13. Available at: www.healthmetricsandevaluation.org/gbd. References 14. Chambliss S, Silva R, West JJ, Zeinali M, Minjares R. A method for estimating source-attributable health impacts: global mortality from surface transportation in 2005. Manuscript in preparation. 1. Kessides I, Noll R, Benjamin N. Regionalizing infrastructure reform in developing 15. Brauer M, et al. Exposure assessment for estimation of the Global Burden of countries. World Economics. 2010; 11(3):79–108. Disease attributable to outdoor air pollution. Environmental Science and 2. African Union and UN Economic Commission for Africa. Transport and the Technology. 2012;46(2):652–660. Millennium Development Goals in Africa. Addis Ababa, Ethiopia: UN Economic 16. Fann N, et al. The recent and future health burden of air pollution apportioned Commission for Africa, 2005. across US sectors. Environmental Science and Technology. 2013;47(8):3580– 3. World Bank. Safe, Clean, and Affordable… Transport for Development: the 3589. World Bank Group’s Transport Business Strategy for 2008-2012. Washington, 17. Wegman F, Lynam D, Nilsson G. SUNflower: a comparative study of the devel- DC: World Bank Group, 2008. opments of road safety in Sweden, the United Kingdom, and the Netherlands. 4. Gakidou E, et al. Increased educational attainment and its effect on child Leidschendam, Netherlands: SWOV Institute for Road Safety Research, 2002. mortality in 175 countries between 1970 and 2009: a systematic analysis. 18. Pollack KM, Kercher C, Frattaroli S, Peek-Asa C, Sleet D, Rivara FP. Toward The Lancet. 2010;376:959–974. environments and policies that promote injury-free active living—it wouldn’t 5. Yan C. China real-time report: road-building rage to leave U.S. in dust. hurt. Health & Place. 2012;18(1):106–114. doi:10.1016/j.healthplace.2011.07.010. Wall Street Journal China. January 18, 2011. 19. Polinder S, Haagsma JA, Lyons RA, Gabbe BJ, Ameratunga S, Cryer C, et al. 6. Ghani E, Goswami AG, Kerr W. Highway to success in India: the impact of the Measuring the population burden of fatal and nonfatal injury. Epidemiologic Golden Quadrilateral Project for the location and performance of manufacturing. Reviews. 2012;34(1):17–31. doi:10.1093/epirev/mxr022. World Bank Policy Research Working Paper no. 6320. Washington, DC: World 20. Ezzati M, Lopez AD, Rodgers A, Murray CJ. Comparative quantification of health Bank, 2013. risks. Global and regional burden of disease attributable to selected major risk 7. Costello A, Abbas M, Allen A, Ball S, Bell S, Bellamy R, et al. Managing the factors. Geneva: World Health Organization, 2012. health effects of climate change. The Lancet. 2009;373(9676):1693–1733. 21. Bliss T, Breen J. Road Safety Management Capacity Reviews and Safe System 8. Watkins K. Safe and Sustainable Roads: the Case for a Sustainable Development Projects Guidelines. Global Road Safety Facility, World Bank, 2013; Bliss T, Goal. London, UK: Make Roads Safe: The Campaign for Global Road Safety, Raffo V. Improving Global Road Safety: Towards Equitable and Sustainable 2012; iRAP 2013. The global cost of road crashes. Available at http://www.irap. Development. IUHPE and World Bank, 2013. net/about-irap-3/research-and-technical-papers; World Health Organization. 22. Burnett RT, Pope III CA, Ezzati M, Olives C, Lim SS, Mehta S, et al. An integrated Macroeconomics and Health: Investing in Health for Economic Development. risk function for estimating the Global Burden of Disease attributable to ambi- Geneva: World Health Organization; 2001. ent fine particulate matter exposure. Environmental Health Perspectives. 2014. 9. United Nations. Political Declaration of the High-Level Meeting of the General http://dx.doi.org/10.1289/ehp.1307049 (accessed 25 Feb 2014). Assembly on the Prevention and Control of Non-Communicable Diseases, vol. 23. United Nations Environmental Program (UNEP). Near-term Climate Protection 117, Sixty-Sixth Session. New York, USA: United Nations, 2011. and Clean Air Benefits: Actions for Controlling Short-Lived Climate Forcers. 10. Delinger A, Stanton C. Barriers to children walking and biking to school. Nairobi, Kenya: UNEP, 2011. Available at: http://www.unep.org/publications/eb- Morbidity and Mortality Weekly Report. 2002;51:701–704. ooks/slcf/. 11. Pucher J, Dill J, Handy S. Infrastructure, programs, and policies to increase 24. Task Force on Hemispheric Transport of Air Pollution. Assessment Reports. bicycling: an international review. Preventive Medicine. 2010;50:S106–S125. Available at: http://www.htap.org/publications/assessment_reports.htm. doi:10.1016/j.ypmed.2009.07.028. 25. International Institute for Applied Systems Analysis (IIASA). Global Energy 12. Commitment to Sustainable Transport: Joint Statement to the Rio+20 United Assessment (GEA): Toward a Sustainable Future. Laxenburg, Austria: IIASA, Nations Conference on Sustainable Development by the African Development 2012. Available at: http://www.iiasa.ac.at/web/home/research/Flagship-Projects/ Bank, Asian Development Bank, CAF-Development Bank of Latin America, Global-Energy-Assessment/Chapters_Home.en.html. European Bank for Reconstruction and Development, European Investment Bank, Inter-American Development Bank, Islamic Development Bank, and World Bank. June 2012. 72 | Transport for Health 26. Rao S, Chirkov V, Dentener F, et al. Environmental modeling and methods for estimation of the global health impacts of air pollution. Environmental Modeling & Assessment, 2012;17(6):613-622. 27. Health Effects Institute (HEI), International Scientific Oversight Committee. Outdoor Air Pollution and Health in the Developing Countries of Asia: A Comprehensive Review. Special Report 18. Boston, USA: HEI, 2010. Available at: http://pubs.healtheffects.org/view.php?id=349. 28. Caiazzo F, Ashok A, Waitz I, Yim SHL, Barrett SRH. Air pollution and early deaths in the United States; part I: quantifying the impact of major sectors in 2005. Atmospheric Environment. 2013 Nov;79:198–208. 29. Health Effects Institute (HEI). Traffic-Related Air Pollution: A Critical Review of the Literature on Emissions, Exposure, and Health Effects. Special Report 17. Boston, USA: HEI, 2010. Available at: http://pubs.healtheffects.org/getfile. php?u=553. 30. Brauer M, Reynolds C, Hystad P. Traffic-related air pollution and health in Canada. Canadian Medical Association Journal. 2013 Oct;185(18):1557–1558. doi:10.1503/cmaj.121568. 31. Health Effects Institute (HEI). Mobile-Source Air Toxics: A Critical Review of the Literature on Exposure and Health Effects. Special Report 16. Boston, USA: HEI, 2007. Available at: http://pubs.healtheffects.org/view.php?id=282. 32. Technology Transfer Network. 1996 National-Scale Air Toxics Assessment. Available at: http://www.epa.gov/ttn/atw/nata/risksum.html.