Levels & Trends in 2017 Report 2014 Child Estimates Developed by the Mortality UN Inter-agency Group for Child Mortality Estimation United Nations This report was prepared at UNICEF headquarters by Lucia Hug, David Sharrow, and Danzhen You on behalf of the United Nations Inter-agency Group for Child Mortality Estimation (UN IGME). Organizations and individuals involved in generating country-speci fic estimates of child mortality United Nations Children’s Fund Lucia Hug, David Sharrow, Yuhan Sun, Ana Marcusanu, Danzhen You World Health Organization Colin Mathers, Daniel Hogan, Jessica Ho, Wahyu Retno Mahanani World Bank Group Emi Suzuki United Nations, Department of Economic and Social Affairs, Population Division Patrick Gerland, Francois Pelletier, Lina Bassarsky, Helena Cruz Castanheira, Danan Gu, Nan Li, Cheryl Sawyer, Thomas Spooren- berg, Guangyu Zhang United Nations Economic Commission for Latin America and the Caribbean, Population Division Guiomar Bay Special thanks to the Technical Advisory Group of the UN IGME for providing technical guidance on methods for child mortality estimation Robert Black (Chair), Johns Hopkins University Bruno Masquelier, University of Louvain Leontine Alkema, University of Massachusetts, Amherst Kenneth Hill, Stanton-Hill Research Simon Cousens, London School of Hygiene and Tropical Medicine Jon Pedersen, Fafo Trevor Croft, The Demographic and Health Surveys (DHS) Program, ICF Neff Walker, Johns Hopkins University Michel Guillot, University of Pennsylvania Special thanks to the United States Agency for International Development (USAID) and the Bill and Melinda Gates Foundation for supporting UNICEF’s child mortality estimation work. Thanks also go to the Joint United Nations Programme on HIV/AIDS for sharing estimates of AIDS mortality. Further thanks go to Fengqing Chao from the National University of Singapore for assistance in preparing the UN IGME estimates as well as Jing Liu from Fafo for preparing the underlying data. Special thanks to Khin Wityee Oo and Anna Mukerjee from UNICEF for proofreading. And special thanks to colleagues in the field of fices of UNICEF and WHO for supporting the country consultation process. Thanks also go to Laurence Christian Chandy (Director, Division of Data, Research and Policy), Hongwei Gao (Deputy Director, Policy, Strategy and Network, Division of Data, Research and Policy), Mark Hereward (Associate Director, Data and Analytics, Division of Data, Research and Policy), Priscilla Idele, Attila Hancioglu, Rada Noeva, Claes Johansson, Claudia Cappa, Anshana Arora, Sebastian Bania, Ivana Bjelic, Yadigar Coskun, Emily Garin, Anna Grojec, Ahmed Hanafy, Karoline Hassfurter, Shane Khan, Bo Pedersen, Upasana Young and Turgay Unalan from UNICEF, Theresa Diaz and Mohamed Mahmoud Ali from WHO, Mary Mahy and Juliana Daher from the Joint United Nations Programme on HIV/AIDS, William Weiss from USAID and Kate Somers from the Bill and Melinda Gates Foundation for their support. Natalie Leston edited the report. Sinae Lee laid out the report. Copyright © 2017 by the United Nations Children’s Fund The United Nations Inter-agency Group for Child Mortality Estimation (UN IGME) constitutes representatives of the United Nations Children’s Fund, the World Health Organization, the World Bank Group and the United Nations Population Division. Differences between the estimates presented in this report and those in forthcoming publications by UN IGME members may arise because of differences in reporting periods or in the availability of data during the production process of each publication and other evidence. UN IGME estimates were reviewed by countries through a country consultation process but are not necessarily the of ficial statistics of United Nations Member States, which may use a single source of data or alternative rigorous methods. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of UNICEF, the World Health Organization, the World Bank Group or the United Nations Population Division concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement. United Nations Children’s Fund World Health Organization 3 UN Plaza, New York, New York, 10017 USA Avenue Appia 20, 1211 Geneva 27, Switzerland World Bank Group United Nations Population Division 1818 H Street, NW, Washington, DC, 20433 USA 2 UN Plaza, New York, New York, 10017 USA CHILD SURVIVAL: KEY FACTS AND FIGURES • The world has made substantial progress in fifth birthday, while in the world’s high-income reducing child mortality in the past several countries the ratio is 1 in 189. Among newborns decades. The total number of under-five deaths in sub-Saharan Africa, about 1 child in 36 dies in dropped to 5.6 (5.4, 6.0)1 million in 2016 from the first month, while in the world’s high-income 12.6 (12.4, 12.8) million in 1990 – 15,000 every countries the ratio is 1 in 333. day compared with 35,000 in 1990. • Many lives can be saved if the gaps across • Globally, the under-five mortality rate dropped to countries are closed. If all countries had reached 41 (39, 44) deaths per 1,000 live births in 2016 an under-five mortality rate at or below the from 93 (92, 95) in 1990 – a 56 (53, 58) per cent average rate of high-income countries – 5.3 decline. deaths per 1,000 live births – 87 per cent of under-five deaths could have been averted, and • Globally, 2.6 (2.5, 2.8) million newborns died almost 5 million children’s lives could have been in 2016 – or 7,000 every day. Neonatal deaths saved in 2016. accounted for 46 per cent of all under-five deaths, increasing from 41 per cent in 2000. • If current trends continue with more than 50 countries falling short of the Sustainable • The largest number of newborn deaths occurred Development Goal (SDG) target on child survival, in Southern Asia (39 per cent), followed by sub- some 60 million children under age 5 will die Saharan Africa (38 per cent). Five countries between 2017 and 2030 – and half of them will be accounted for half of all newborn deaths: India, newborns. Pakistan, Nigeria, the Democratic Republic of the Congo and Ethiopia. • If every country achieves the SDG target on child survival by 2030, an additional 10 million lives of • The neonatal mortality rate fell by 49 per cent children under age 5 will be saved throughout the from 37 (36, 38) deaths per 1,000 live births in period 2017–2030 – about half of them will be 1990 to 19 (18, 20) in 2016. newborns. • Children face the highest risk of dying in their • Most under-five deaths are caused by diseases first month of life, at a rate of 19 deaths per that are readily preventable or treatable with 1,000 live births. By comparison, the probability proven, cost-effective interventions. Infectious of dying after the first month but before reaching diseases and neonatal complications are age 1 is 12 and after age 1 but before turning 5 responsible for the vast majority of under-five is 11. deaths globally. • Progress is slower in reducing neonatal mortality • The probability of dying among children aged 5–14 rates than in reducing mortality rates in children was 7.5 (7.2, 8.3) deaths per 1,000 children aged 5 aged 1–59 months. While neonatal mortality in 2016 – substantially lower than among younger declined by 49 per cent, the mortality in children children. Still 1 (0.9, 1.1) million children aged 5–14 aged 1–59 months declined by 62 per cent from died in 2016. This is equivalent to 3,000 children 1990 to 2016. in this age group dying every day. Among children aged 5–14, communicable diseases are a less • Disparities in child survival exist across prominent cause of death than among younger regions and countries: in sub-Saharan Africa, children, while other causes including injuries and approximately 1 child in 13 dies before his or her non-communicable diseases become important. 1 Introduction Every year, millions of children under 5 years of With the end of the era of the Millennium age die, mostly from preventable causes such as Development Goals, the international community pneumonia, diarrhoea and malaria. In almost agreed on a new framework – the SDGs. The half of the cases, malnutrition plays a role, while SDG target for child mortality represents a renewed commitment to the world’s children: By unsafe water, sanitation and hygiene are also 2030, end preventable deaths of newborns and significant contributing factors. For this reason, children under 5 years of age, with all countries child mortality is a key indicator not only for aiming to reduce neonatal mortality to at least as child health and well-being, but for overall low as 12 deaths per 1,000 live births and under- progress towards the Sustainable Development five mortality to at least as low as 25 deaths per Goals (SDGs). 1,000 live births. 2 The world made substantial progress in reducing could have been prevented in 2016. Reducing child mortality in the past few decades. Globally, inequities and reaching the most vulnerable the under-five mortality rate dropped from 93 newborns and children as well as their mothers deaths per 1,000 live births in 1990 to 41 in 2016. are important priorities to achieve the SDG Progress in reducing child mortality has been targets on ending preventable child deaths. accelerated in the 2000–2016 period compared with the 1990s – globally, the annual rate of While the mortality risk for children aged 5–14 is reduction in the under-five mortality rate has about one fifth of the risk of dying for children increased from 1.9 per cent in 1990–2000 to 4.0 under age 5, still about 1 million children aged per cent in 2000–2016. The remarkable progress 5–14 died in 2016. Public health interventions in improving child survival since 2000 has saved need to address the particular health risks for the lives of 50 million children under age 5 – this age group, which differ from the primary children who would have died had under-five risks among younger children. Special attention mortality remained at the same level as in 2000 needs to be paid to sub-Saharan Africa where in each country. the probability that a child aged 5 dies before reaching his or her fifteenth birthday (19 deaths Despite the substantial progress in reducing per 1,000 children aged 5) is 17 times higher child mortality, child survival remains an urgent than the average in high-income countries (1.1 concern. In 2016, 5.6 million children died deaths per 1,000 children aged 5). before their fifth birthday – among them 2.6 million (46 per cent) died in the first month of Evidence-based estimation of child mortality life. It is unacceptable that 15,000 children die is a cornerstone for tracking progress towards every day, mostly from preventable causes and child survival goals and identifying priority treatable diseases, even though the knowledge areas to accelerate progress towards eliminating and technologies for life-saving interventions are preventable child deaths. Reliable estimates available. are crucial for planning national and global Inequities in child mortality across and within health strategies, policies and interventions countries remain large. At the country level, on child health and well-being. In the context the under-five mortality rate ranged from a of monitoring child survival, the United high of 133 deaths per 1,000 live births to a low Nations Inter-agency Group for Child Mortality of 2 deaths per 1,000 live births in 2016. Many Estimation (UN IGME) updates child mortality countries still have very high rates – particularly estimates annually. This report presents the in sub-Saharan Africa, home to all six countries group’s latest estimates of under-five, infant with an under-five mortality rate above 100 and neonatal mortality up to the year 2016, deaths per 1,000 live births. Hypothetically, if all and assesses progress at the country, regional countries had reached an under-five mortality and global levels. The report also presents, rate at or below the average rate of high-income for the first time, the mortality estimates for countries – 5.3 deaths per 1,000 live births – the children aged 5–14 generated by UN IGME. In toll of under-five deaths in 2016 would have been addition, the report provides an overview on 0.7 million. In other words, almost 5 million the estimation methods used for child mortality deaths (87 per cent of the total under-five deaths) indicators. 3 Levels and Trends in Child Mortality Mortality among children under Despite substantial progress, improving child age 5 survival remains a matter of urgent concern. In 2016, an estimated 5.6 (5.4, 6.0) million children Under-five mortality died before reaching their fifth birthday (Table The world has made substantial progress in 2), mostly from preventable diseases. This child survival since 1990. The global under-five translates to 15,000 under-five deaths per day, an mortality rate declined by 56 per cent (53, 58), intolerably high number of largely preventable from 93 (92, 95) deaths per 1,000 live births in child deaths. 1990 to 41 (39, 44) in 2016 (Table 1 and Figure 1). The majority of the regions in the world and The burden of under-five deaths remains 142 out of 195 countries at least halved their unevenly distributed. About 80 per cent of under-five mortality rate. Among all countries, under-five deaths occur in two regions, sub- more than a third (67) cut their under-five Saharan Africa and Southern Asia. Six countries mortality by two thirds – 28 of them are low- account for half of the global under-five deaths, or lower-middle-income countries, indicating namely, India, Nigeria, Pakistan, the Democratic that improving child survival is possible even in Republic of the Congo, Ethiopia and China. resource-constrained settings. India and Nigeria alone account for almost a TABLE Levels and trends in the under-five mortality rate, by Sustainable Development Goal region, 1 1990-2016 Annual rate of Under-five mortality rate (deaths per 1,000 live births) reduction (per cent) Decline (per cent) 1990- 1990- 2000- Region 1990 1995 2000 2005 2010 2015 2016 1990-2016 2016 2000 2016 Northern America and Europe 14 12 10 8 7 6 6 59 3.5 3.8 3.3 Northern America 11 9 8 8 7 7 6 41 2.0 2.8 1.5 Europe 15 13 10 8 7 6 5 65 4.0 3.9 4.1 Latin America and the Caribbean 55 44 33 26 25 18 18 68 4.4 5.0 4.0 Central Asia and Southern Asia 124 108 91 75 60 48 46 63 3.8 3.1 4.3 Central Asia 73 74 64 49 37 28 26 64 3.9 1.2 5.6 Southern Asia 126 109 92 76 61 49 47 63 3.8 3.2 4.2 Eastern Asia and South-Eastern Asia 57 50 40 29 22 17 16 72 4.9 3.6 5.7 Eastern Asia 51 45 35 23 15 10 10 81 6.4 3.9 8.0 South-Eastern Asia 72 59 49 40 33 28 27 63 3.8 3.9 3.8 Western Asia and Northern Africa 75 62 51 41 33 29 28 62 3.7 3.9 3.6 Western Asia 66 54 43 34 27 25 24 63 3.8 4.2 3.6 Northern Africa 84 71 60 49 40 34 33 61 3.6 3.4 3.8 Sub-Saharan Africa 183 175 157 128 102 82 79 57 3.2 1.5 4.3 Oceania 35 33 33 31 27 24 23 35 1.6 0.6 2.3 Oceania excluding Australia and New Zealand 74 69 66 63 57 50 49 34 1.6 1.1 1.9 Australia and New Zealand 10 7 6 6 5 4 4 58 3.4 4.1 2.9 Least developed countries 176 160 139 111 89 71 68 61 3.6 2.4 4.4 Landlocked developing countries 167 158 141 111 85 66 63 62 3.7 1.7 5.0 Small island developing States 79 70 62 56 79 43 42 47 2.4 2.4 2.4 World 93 87 78 64 52 42 41 56 3.2 1.9 4.0 Note: All calculations are based on unrounded numbers. 4 TABLE Levels and trends in the number of deaths of children under age 5, by Sustainable Development Goal region, 2 1990-2016 Share of global under-five deaths Under-five deaths (thousands) (per cent) Decline (per cent) Region 1990 1995 2000 2005 2010 2015 2016 1990-2016 1990 2016 Northern America and Europe 191 144 112 97 84 72 71 63 1.5 1.3 Northern America 47 40 35 35 32 28 28 41 0.4 0.5 Europe 144 104 77 62 52 44 43 70 1.1 0.8 Latin America and the Caribbean 652 513 387 293 270 194 187 71 5.2 3.3 Central Asia and Southern Asia 4,950 4,322 3,645 2,997 2,394 1,859 1,775 64 39.3 31.5 Central Asia 113 106 78 61 54 44 41 63 0.9 0.7 Southern Asia 4,836 4,217 3,566 2,936 2,339 1,815 1,734 64 38.4 30.7 Eastern Asia and South-Eastern Asia 2,312 1,688 1,203 881 675 522 495 79 18.3 8.8 Eastern Asia 1,446 1,001 646 413 286 197 180 88 11.5 3.2 South-Eastern Asia 866 687 558 468 390 326 314 64 6.9 5.6 Western Asia and Northern Africa 689 568 463 392 354 330 323 53 5.5 5.7 Western Asia 302 254 207 168 146 137 135 55 2.4 2.4 Northern Africa 388 314 256 223 208 193 188 52 3.1 3.3 Sub-Saharan Africa 3,787 4,040 4,040 3,667 3,220 2,838 2,777 27 30.1 49.2 Oceania 18 18 18 18 17 15 15 17 0.1 0.3 Oceania excluding Australia and New Zealand 15 15 16 16 15 14 13 10 0.1 0.2 Australia and New Zealand 3 2 2 2 2 2 1 50 0.0 0.0 Least developed countries 3,669 3,639 3,437 2,966 2,544 2,154 2,101 43 29.1 37.2 Landlocked developing countries 1,763 1,789 1,708 1,450 1,204 1,001 972 45 14.0 17.2 Small island developing States 94 84 74 66 96 52 51 46 0.7 0.9 World 12,598 11,293 9,868 8,344 7,014 5,831 5,642 55 100.0 100.0 Note: All calculations are based on unrounded numbers. FIGURE Under-five mortality declined in all regions third (32 per cent) of the global under-five 1 between 1990 and 2016 deaths. Under-five mortality rate by Sustainable Development Goal region, Huge disparities in under-five mortality exist 1990 and 2016 across regions and countries. Sub-Saharan 1990 2016 SDG target for 2030 Africa remains the region with the highest 200 under-five mortality rate in the world. In 2016, 183 the region had an average under-five mortality 176 167 rate of 79 deaths per 1,000 live births. This Deaths per 1,000 live births 150 translates to 1 child in 13 dying before his 124 or her fifth birthday – 15 times higher than 100 the average ratio of 1 in 189 in high-income 93 countries, or 20 times higher than the ratio 79 79 74 75 68 of 1 in 250 in the region of Australia and New 63 57 55 49 Zealand. At the country level, the under-five 46 50 42 41 28 mortality rates in 2016 ranged from 2 deaths 18 16 14 10 per 1,000 live births to 133 (Map 1). The risk of 6 4 0 dying for a child born in the highest-mortality Central Asia and Southern Asia World Australia and New Zealand Western Asia and Northern Africa Small island developing States Landlocked developing countries Northern America and Europe Least developed countries Eastern Asia and South-Eastern Asia Latin America and the Caribbean Sub-Saharan Africa Oceania* country is about 60 times higher than in the lowest-mortality country. All six countries with mortality rates above 100 deaths per 1,000 live births are in sub-Saharan Africa. Note: Oceania* refers to Oceania excluding Australia and New Zealand. 5 MAP Children in sub-Saharan Africa and Southern Asia face a higher risk of dying before 1 their fifth birthday Under-five mortality rate (deaths per 1,000 live births) by country, 2016 Deaths per 1,000 live births >100 75 to 100 50 to 75 25 to 50 ≤ 25 No data Note: The classification is based on unrounded numbers. This map does not reflect a position by UN IGME agencies on the legal status of any country or territory or the delimitation of any frontiers. Children in fragile context have about twice under-five mortality across states varied from the risk of dying under age 5 than children in 13 deaths per 1,000 live births to 62 based on non-fragile context. Among the 10 countries the Sample Vital Registration data in 2015.4 with the highest under-five mortality rates, 7 are The latest mortality estimates by wealth quintile classified as fragile countries.2 Moreover, fragile generated by UN IGME reveal that in 99 low- and states accounted for 22 per cent of the under-five middle-income countries, 5 under-five mortality deaths among low- and middle-income countries among children born in the poorest households in 2016, yet they only shared about 12 per cent of is on average twice that of children born in the the under-five population. wealthiest households.6 The burden of under- five deaths is also disproportionally concentrated The number of countries with signi ficant among poorer households, with the two poorest gender-based gaps in child mortality has fallen. quintiles accounting for about half of the under- In some countries, the risk of dying before age 5 five deaths but only for 40 per cent of the births. for girls is significantly higher than what would be expected based on global patterns. These Eliminating the gaps between the poorest and countries are primarily located in Southern Asia richest households and between countries and Western Asia. The number of countries would save millions of lives. In 2016 alone, showing these gender disparities fell by almost some 2 million7 lives would have been saved had half between 1990 and 2016, from 19 to 11. under-five mortality in the poorest households been as low as it is in the wealthiest households. Inequity persists within countries Closing the gap between countries would have geographically or by social-economic status. produced even starker results: if all countries For example, in Chad, under-five mortality had reached an under-five mortality rate at or across regions ranged from 67 deaths per 1,000 below the average rate of high-income countries live births to 230 based on the Demographic – 5.3 deaths per 1,000 live births – 87 per cent and Health Survey (DHS) 2014–2015.3 In India, of under-five deaths could have been prevented, 6 and the lives of almost 5 million children could target, and 13 countries in the region will not have been saved in 2016. reach the target until after 2050. Accelerated progress will be needed in more Achieving the SDG target on time would mean than a quarter of all countries, to achieve averting 10 million under-five deaths compared SDG targets in child survival. Among all 195 with a business-as-usual scenario. If current countries analysed, 116 already met the SDG trends continue, over 60 million children under target on under-five mortality and 27 countries 5 years of age will die between 2017 and 2030, are expected to meet the target by 2030 if current about half of them newborns. More than half of trends continue, while 52 countries need to these deaths will occur in sub-Saharan Africa accelerate progress. These countries can be found and about 30 per cent in Southern Asia. Meeting in most regions of the world, but the majority are the SDG target would reduce the number of in sub-Saharan Africa. If current trends continue, under-five deaths by 10 million between 2017 and more than three quarters of all countries in sub- 2030. Urgent efforts are needed in the countries Saharan Africa will miss the under-five mortality that are falling behind. 7 Neonatal mortality 1990 to 2016 was slower than the decline in mortality The fi rst 28 days of life – the neonatal period – among children aged 1–59 months: 49 per cent, are the most vulnerable time for a child’s survival. compared with 62 per cent, a pattern consistent Children face the highest risk of dying in their across all SDG regions (Figure 2). The relative first month of life, at a global rate of 19 deaths per decline in neonatal mortality was slower in sub- 1,000 live births (Table 3). By way of comparison, Saharan Africa than in the other regions. Despite the probability of dying after the first month but the modest decline in the neonatal mortality rate before reaching age 1 is 12 and after age 1 but in sub-Saharan Africa of 40 per cent, the number before turning 5 is 11. Globally, 2.6 (2.5, 2.8) of neonatal deaths remained almost the same from million children died in the first month of life in 1990 to 2016 due to an increasing number of births. 2016 (Table 4) – approximately 7,000 newborn deaths every day – most of which occurred in the Marked disparities in neonatal mortality exist first week, with about 1 million dying on the first across regions and countries. Among the SDG day and close to 1 million dying within the next six regions, neonatal mortality was highest in sub- days. Saharan Africa and Southern Asia, which each reported 28 deaths per 1,000 live births (Table 3). A Neonatal mortality declined globally and in all child in sub-Saharan Africa or in Southern Asia is regions but more slowly than mortality among nine times more likely to die in the first month than children aged 1–59 months. The global neonatal a child in a high-income country. Across countries, mortality rate fell from 37 (36, 38) deaths per 1,000 neonatal mortality rates ranged from 46 deaths per live births in 1990 to 19 (18, 20) in 2016. However, 1,000 live births in Pakistan to 1 each in Iceland and the decline in the neonatal mortality rate from Japan (Figure 3). FIGURE Progress in reducing neonatal mortality is slower FIGURE Huge disparities in the level of neonatal 2 than in mortality among children aged 1–59 months 3 mortality persist across countries and regions Percentage decline in neonatal mortality and in mortality among children Neonatal mortality rates, by country and Sustainable Development aged 1-59 months, by Sustainable Development Goal region, 1990-2016 Goal region, 2016 100 Neonatal mortality Mortality among children aged 1-59 months 40 75 73 73 75 71 70 69 67 67 63 63 62 59 56 56 Deaths per 1,000 live births 30 52 Per cent 50 50 49 49 50 47 42 40 29 20 22 25 SDG target for 2030 10 0 World Central Asia and Southern Asia Latin America and the Caribbean Least developed countries Sub-Saharan Africa Landlocked developing countries Eastern Asia and South-Eastern Asia Western Asia and Northern Africa Northern America and Europe Australia and New Zealand Oceania* Small island developing States 0 Central Sub- Western Eastern Latin Oceania* Northern Australia Asia and Saharan Asia and Asia and America America and New Southern Africa Northern South- and the and Europe Zealand Asia Africa Eastern Caribbean Asia Note: Oceania* refers to Oceania excluding Australia and New Zealand. All calculations are Note: Each dot represents a country. Oceania* refers to Oceania excluding Australia based on unrounded numbers. and New Zealand. 8 TABLE Levels and trends in the neonatal mortality rate, by Sustainable Development Goal region, 1990-2016 3 Annual rate of Neonatal mortality rate (deaths per 1,000 live births) reduction (per cent) Decline (per cent) 1990- 1990- 2000- Region 1990 1995 2000 2005 2010 2015 2016 1990-2016 2016 2000 2016 Northern America and Europe 7 6 5 4 4 3 3 56 3.2 3.4 3.0 Northern America 6 5 5 5 4 4 4 35 1.7 2.4 1.2 Europe 8 7 6 4 3 3 3 64 4.0 3.7 4.1 Latin America and the Caribbean 23 19 16 13 11 9 9 59 3.5 3.8 3.3 Central Asia and Southern Asia 56 51 45 39 33 28 27 52 2.8 2.2 3.2 Central Asia 28 29 27 22 18 14 13 53 2.9 0.6 4.3 Southern Asia 57 52 46 39 34 29 28 52 2.8 2.2 3.1 Eastern Asia and South-Eastern Asia 28 25 20 15 11 9 8 71 4.7 3.3 5.6 Eastern Asia 28 25 19 13 8 5 5 82 6.7 3.6 8.6 South-Eastern Asia 28 24 21 18 16 14 14 51 2.8 2.8 2.8 Western Asia and Northern Africa 31 27 23 20 17 15 15 50 2.7 2.7 2.7 Western Asia 28 24 20 17 14 13 13 54 3.0 3.1 2.9 Northern Africa 33 29 26 23 20 18 17 48 2.5 2.3 2.6 Sub-Saharan Africa 46 45 41 37 32 28 28 40 2.0 1.1 2.5 Oceania 14 13 14 13 12 11 10 24 1.0 0.0 1.7 Oceania excluding Australia and New Zealand 27 26 26 25 23 21 21 22 1.0 0.5 1.3 Australia and New Zealand 5 4 4 3 3 2 2 49 2.6 2.5 2.6 Least developed countries 52 47 42 37 31 27 26 50 2.6 2.1 2.9 Landlocked developing countries 48 45 42 36 31 26 26 47 2.4 1.3 3.1 Small island developing States 27 25 24 23 22 19 19 29 1.3 1.3 1.3 World 37 34 31 26 22 19 19 49 2.6 1.8 3.1 Note: All calculations are based on unrounded numbers. TABLE Levels and trends in the number of neonatal deaths, by Sustainable Development Goal region, 1990-2016 4 Neonatal deaths as a share of Number of neonatal deaths (thousands) Decline under-five deaths (per cent ) (per cent) Region 1990 1995 2000 2005 2010 2015 2016 1990-2016 1990 2000 2016 Northern America and Europe 98 75 60 53 45 40 39 60 51 54 55 Northern America 24 21 20 20 18 16 16 34 52 55 57 Europe 74 54 41 33 27 24 23 69 51 53 54 Latin America and the Caribbean 270 227 181 141 119 101 98 64 41 47 52 Central Asia and Southern Asia 2,277 2,076 1,824 1,561 1,317 1,083 1,044 54 46 50 59 Central Asia 45 40 32 29 28 23 21 52 39 41 51 Southern Asia 2,232 2,036 1,792 1,532 1,289 1,061 1,023 54 46 50 59 Eastern Asia and South-Eastern Asia 1,112 807 600 457 341 263 250 78 48 50 51 Eastern Asia 778 526 358 242 153 100 92 88 54 55 51 South-Eastern Asia 334 282 242 215 188 163 158 53 39 43 50 Western Asia and Northern Africa 285 245 215 196 187 176 173 40 41 46 54 Western Asia 131 115 100 87 78 74 72 45 43 48 54 Northern Africa 154 130 115 109 109 103 100 35 40 45 53 Sub-Saharan Africa 1,008 1,079 1,117 1,092 1,056 1,010 1,003 1 27 28 36 Oceania 7 7 7 8 7 7 7 4 39 42 45 Oceania excluding Australia and New Zealand 6 6 6 7 6 6 6 -5 37 40 44 Australia and New Zealand 1 1 1 1 1 1 1 39 48 55 59 Least developed countries 1,138 1,120 1,088 1,008 920 845 834 27 31 32 40 Landlocked developing countries 530 536 529 493 454 411 404 24 30 31 42 Small island developing States 33 30 29 28 27 24 23 28 35 39 46 World 5,058 4,517 4,005 3,507 3,073 2,681 2,614 48 40 41 46 Note: All calculations are based on unrounded numbers. 9 The burden of neonatal deaths is also unevenly Southern Asia, where the proportion of neonatal distributed across regions and countries. deaths is among the highest (59 per cent) despite Two regions account for almost 80 per cent of a relatively high under-five mortality rate. Many the newborn deaths in 2016 – Southern Asia countries in this region have higher-than- accounted for 39 per cent of all such deaths and expected neonatal mortality rates, given the level sub-Saharan Africa accounted for 38 per cent of under-five mortality. To save newborns in these (Table 4). At the country level, half of all neonatal countries, it is critical to understand the causes deaths are concentrated in five countries, namely, of higher-than-expected neonatal mortality rates India (24 per cent), Pakistan (10 per cent), Nigeria and the bottlenecks to prevent newborn deaths. (9 per cent), the Democratic Republic of the Congo (4 per cent) and Ethiopia (3 per cent). Many countries will lag even further behind in India and Pakistan alone accounted for about a achieving the SDG target on neonatal mortality third of all newborn deaths. than on under-five mortality if current trends continue. On current trends, more than 60 Globally, 46 per cent of under-five deaths occur countries will miss the target for neonatal during the neonatal period. Despite falling mortality by 2030, while 52 countries will miss rates of neonatal mortality, its importance in the the target for under-five mortality. About half burden of under-five deaths is increasing. Due to of these countries would not even reach the the slower decline of neonatal mortality relative neonatal mortality target by 2050. These 60+ to mortality in children aged 1–59 months, the countries carried about 80 per cent of the burden share of neonatal deaths among under-five deaths of neonatal deaths in 2016. increased from 40 per cent in 1990 to 46 per cent in 2016. This trend is expected to continue as Accelerating progress to achieve the SDG target the under-five mortality rate continues to decline on neonatal mortality would save the lives of 5 (Figure 3). million newborns from 2017 to 2030 in the 60+ countries that will miss the target for neonatal Lower under-five mortality is associated with mortality by 2030 if current trends continue. a higher concentration of under-five deaths Based on current trends, 30 million newborns occurring during the neonatal period. The share would die between 2017 and 2030. Eighty per of neonatal deaths among under-five deaths is cent of these deaths would occur in Southern still relatively low in sub-Saharan Africa (36 per Asia and sub-Saharan Africa. About one in six cent), which remains the region with the highest of these deaths (5 million) could be averted if under-five mortality rates. In the regions Australia countries at risk of missing the SDG target with and New Zealand and Northern America and low rates of progress and high neonatal mortality Europe, where under-five mortality rates are low, rates accelerate progress. Many of the countries more than half of all under-five deaths occur with low rates of progress are concentrated in during the neonatal period. The only exception is sub-Saharan Africa and Southern Asia. 10 The remarkable progress in improving child FIGURE survival over the past few decades, particularly in Infectious diseases and neonatal complications 4 some low- and lower-middle-income countries, are the leading causes of death among children provides a clear message: with the right under age 5 commitments, concerted efforts and political will, bold and ambitious goals are within reach. A. Global distribution of deaths among children under age 5, Despite the substantial progress, the unfinished by cause, 2016 business of child survival looms large. If current Deaths among children Neonatal deaths (46%) trends continue without acceleration, some 60 aged 1– 59 months (54%) Pneumonia, 3% million children under 5 years of age will die from 2017 to 2030, and about half of them will be Pneumonia, 13% newborns. Preterm birth complications, 16% Ending newborn and child deaths from Other, 12% preventable infectious diseases is critical. Despite Intrapartum-related events, 11% Congenital, 4% strong advances in fighting childhood illnesses, Intrapartum-related events, 1% infectious diseases – which are most often Preterm birth complications, 2% Sepsis or menigitis,7% diseases of the poor and thus are a marker of Meningitis, 2% AIDS, 1% equity – remain highly prevalent, particularly Malaria, 5% Other, 3% Diarrhoea, in sub-Saharan Africa and Southern Asia. 8% Injury, 1% Injury, 6% Congenital, 5% Pneumonia, diarrhoea and malaria remain Tetanus, 1% Measles, 1% among the leading causes of death among Diarrhoea, 0.3% children under age 5 – accounting for almost Nearly half of all deaths in children under age 5 are attributable to undernutrition a third of global under-five deaths, and about 40 per cent of under-five deaths in sub-Saharan B. Global distribution of deaths among newborns, by cause, 2016 Africa.8 The main killers of children under age 5 Pneumonia Diarrhoea in 2016 included preterm birth complications (18 6% 1% per cent), pneumonia (16 per cent), intrapartum- Intrapartum-related Sepsis or Congenital Preterm birth complications events menigitis abnormalities Other related events (12 per cent), diarrhoea (8 per 35% 24% 24% 15% 11% 7% cent), neonatal sepsis (7 per cent) and malaria (5 Tetanus per cent) (Figure 4). 1% Note: Estimates are rounded and therefore may not sum up to 100%. Source: WHO and Maternal and Child Epidemiology Estimation Group (MCEE) provisional Accelerating the reduction in child mortality estimates 2017 is possible by expanding effective preventive and curative interventions that target the main causes of child deaths and the most vulnerable and the first week of life, as well as care for small newborns and children. With an increasing and sick newborns. share of under-five deaths occurring during the neonatal period, accelerated change for Despite the substantial progress in reducing child child survival, health and development requires deaths, children from poorer areas or households greater focus on a healthy start to life. Children remain disproportionately vulnerable. It is critical that die in the first 28 days of life suffer from to address these inequities to further accelerate diseases and conditions that are associated with the pace of progress to fulfil the promise to quality of care around the time of childbirth and children. Without intensified efforts to reduce are readily preventable or treatable with proven, newborn and child mortality, particularly in cost-effective interventions. Further reductions in neonatal deaths in particular depend on the highest-mortality areas and in contexts of building stronger health services, ensuring that persistent inequities, the SDG targets will be every birth is attended by skilled personnel and unattainable. Countries and the international making hospital care available in an emergency. community must take immediate action to further Cost-effective interventions for newborn health accelerate progress to end preventable newborn cover the antenatal period, the time around birth and child deaths. 11 Mortality among children aged FIGURE Mortality among children aged 5–14 declined 5–14 5 in all regions between 1990 and 2016 Mortality among children aged 5–14 is low, Probability of dying among children aged 5–14 by Sustainable Development Goal region, 1990 and 2016 but 1 million children in this age group still died in 2016. The probability of dying 50 1990 2016 among children aged 5–14 was 7.5 (7.2, 8.3) 42 deaths per 1,000 children aged 5 in 2016 – 40 40 39 substantially lower than the probability of Deaths per 1,000 children aged 5 dying for children under age 5 (41 deaths per 1,000 live births). Still, 1 (0.9, 1.1) million 30 children aged 5–14 died in 2016. This is equivalent to 3,000 children aged 5–14 dying 19 20 19 every day. 15 15 15 14 13 The world has halved the mortality rate 11 10 9 8 among children aged 5–14 since 1990. From 8 7 7 6 5 1990 to 2016, the mortality rate in older 4 3 3 2 1 1 children declined by 51 (46, 54) per cent and 0 Oceania* Western Asia and Northern Africa Australia and New Zealand World Small island developing States Sub-Saharan Africa Northern America and Europe Central Asia and Southern Asia Landlocked developing countries Latin America and the Caribbean Least developed countries Eastern Asia and South-Eastern Asia the number of deaths dropped by 44 per cent from 1.7 (1.7, 1.8) million to 1 million. Most of the regions reduced the probability of dying among children aged 5–14 by at least half (Table 5 and Figure 5). Note: Oceania* refers to Oceania excluding Australia and New Zealand. TABLE Levels and trends in mortality among children aged 5–14 (probability of dying, deaths per 1,000 children 5 aged 5) and the number of deaths, by Sustainable Development Goal region, 1990-2016 Probability of dying among children aged Annual rate Number of deaths among children 5–14 (deaths per 1,000 children aged 5) Decline of reduction aged 5–14 (thousands) (per cent) (per cent ) Region 1990 2000 2010 2015 2016 1990-2016 1990-2016 1990 2000 2010 2015 2016 Northern America and Europe 3 2 2 1 1 57 3.2 42 31 18 16 15 Northern America 2 2 1 1 1 46 2.4 9 8 6 6 6 Europe 3 3 2 1 1 64 3.9 32 23 12 10 10 Latin America and the Caribbean 6 4 4 3 3 50 2.7 65 49 43 34 33 Central Asia and Southern Asia 19 13 9 7 7 66 4.1 611 470 327 258 245 Central Asia 8 5 4 3 3 56 3.2 9 7 4 4 4 Southern Asia 20 14 9 7 7 66 4.2 602 463 323 254 241 Eastern Asia and South-Eastern Asia 9 6 5 4 4 60 3.5 308 226 131 110 107 Eastern Asia 6 5 3 3 3 59 3.5 149 116 59 50 48 South-Eastern Asia 15 10 7 5 5 65 4.0 159 111 72 60 59 Western Asia and Northern Africa 11 8 5 5 5 59 3.4 83 65 47 45 43 Western Asia 9 6 4 4 4 60 3.5 34 27 21 20 18 Northern Africa 13 9 7 6 5 58 3.4 50 38 27 25 25 Sub-Saharan Africa 42 33 23 19 19 55 3.1 604 596 536 516 513 Oceania 6 5 5 4 4 38 1.9 3 3 2 2 2 Oceania excluding Australia and New Zealand 13 11 9 8 8 43 2.1 2 2 2 2 2 Australia and New Zealand 2 1 1 1 1 55 3.1 1 0 0 0 0 Least developed countries 40 28 19 16 15 61 3.7 581 512 433 393 388 Landlocked developing countries 39 28 19 16 15 60 3.6 284 264 210 195 193 Small island developing States 14 10 11 7 7 48 2.5 13 12 13 8 8 World 15 12 9 8 8 51 2.7 1,716 1,442 1,105 981 959 Note: All calculations are based on unrounded numbers. 12 MAP Countries with the highest mortality in children aged 5–14 are concentrated in parts of sub-Saharan Africa 2 Probability of dying among children aged 5–14 (deaths per 1,000 children aged 5) by country, 2016 Deaths per 1,000 children aged 5 >30 20 to 30 10 to 20 5 to 10 ≤5 No data Note: The classification is based on unrounded numbers. This map does not reflect a position by UN IGME agencies on the legal status of any country or territory or the delimitation of any frontiers. Large disparities exist in the survival sub-Saharan Africa, followed by Southern Asia chances of children aged 5–14 across regions with about 25 per cent. Half (52 per cent) of and countries. In sub-Saharan Africa, the all deaths between age 5–14 occurred in seven probability of dying among children aged countries (India, Nigeria, the Democratic 5–14 was 19 deaths per 1,000 children aged 5, Republic of the Congo, Pakistan, Ethiopia, followed by Oceania – excluding Australia and China and the Niger). New Zealand – with 8 and Southern Asia with 7. The average risk for a child in sub-Saharan Injuries become more prominent as a cause of Africa to die between age 5 and age 14 is 17 death as children get older. Among children times higher than the average for children in aged 5–9 years and younger adolescents aged high-income countries (1.1 deaths per 1,000 10–14 years, communicable diseases are a less children aged 5) and 14 times higher than in prominent cause of death than among younger Northern America and Europe. The highest children, while other causes become important. probability of dying in this age group was found For instance, injuries account for more in the Niger, with 40 deaths per 1,000 children than a quarter of the deaths among this age aged 5 versus 0.5 per 1,000 in both Denmark group, non-communicable diseases for about and Luxembourg. The top 26 countries with the another quarter and infectious diseases and highest mortality rates are all in sub-Saharan other communicable diseases, perinatal and Africa, with 15 of them having mortality rates nutritional causes for about half of the deaths.9 above 20 (Map 2). More than half (53 per cent) Drowning and road injuries alone account for of deaths to children aged 5–14 occurred in 10 per cent of all deaths in this age group. 13 Country consultation In accordance with the decision by the Statistical and mortality among children aged 5–14 for its Commission and the United Nations Economic country. The objective was to identify relevant and Social Council resolution 2006/6, UN IGME data that were not included in the UN IGME data- child mortality estimates, which are used for the base, and to allow countries to review and provide compilation of global indicators for SDG monitor- feedback on estimates. In 2017, 108 of 195 coun- ing, are produced in consultation with countries.10 tries sent responses, and 72 of those provided UNICEF and the World Health Organization comments or additional data. After the consulta- (WHO) undertook joint country consultations in tions, the UN IGME draft estimates for mortality 2017. The country consultation process gave each in children under age 5 were revised for 70 coun- country’s Ministry of Health and National Sta- tries using new data and the estimates for mortal- tistics Office the opportunity to review all data ity in children under age 5–14 were revised for inputs, the estimation methodology and the draft 75 countries due to new data. All countries were estimates for mortality in children under age 5 informed about changes in their estimates. 14 Estimating Child Mortality The United Nations Inter-agency Group for Child censuses, household surveys and sample Mortality Estimation (UN IGME), which includes registration systems. members from UNICEF, WHO, the World Bank Group and United Nations Population Division, 2. Assess data quality, recalculate data inputs was established in 2004 to advance the work on and make adjustments, if needed, by applying monitoring progress towards the achievement of standard methods. child survival goals. 3. Fit a statistical model to these data to UN IGME’s Technical Advisory Group, generate a smooth trend curve that averages comprising leading academic scholars and over possibly disparate estimates from the independent experts in demography and different data sources for a country. biostatistics, provides guidance on estimation methods, technical issues and strategies for data 4. Extrapolate the model to a target year – in analysis and data quality assessment. this case, 2016. UN IGME updates its neonatal, infant and To increase the transparency of the estimation under-five mortality estimates annually after process, UN IGME has developed a child mortality web portal, CME Info (). It includes all available data data quality. These estimates are widely used and shows estimates for each country as well as in UNICEF’s flagship publications, the United which data are currently officially used by UN Nations Secretary-General’s SDG report, and IGME. Once the new estimates are finalized, publications by other United Nations agencies, CME Info will be updated to reflect all available governments and donors. data and the new estimates. In 2017, UN IGME for the first time generated UN IGME estimates are based on national country-specific trend estimates of the mortality available data from censuses, surveys or vital in children aged 5–14, that is, the probability that registration systems. UN IGME does not use any a child aged 5 dies before reaching his or her covariates to derive its estimates. It only applies fifteenth birthday. These estimates are presented a curve fitting method to good-quality empirical in this report. data to derive trend estimates after data quality assessment. Countries often use a single source In this chapter, we summarize the methods UN for their official estimates or apply different IGME uses to generate estimates of mortality methods from UN IGME methods to derive among children under age 5 and children aged official estimates. The differences between UN 5–14. IGME estimates and national official estimates are usually not large if the empirical data are Overview of good quality. UN IGME aims to minimize UN IGME follows the following broad strategy to the errors for each estimate, harmonize trends arrive at annual estimates of child mortality: over time, and produce up-to-date and properly assessed estimates of child mortality. In the 1. Compile and assess the quality of all available absence of error-free data, there will always be nationally representative data relevant to the uncertainty around data and estimates, both estimation of child mortality, including data nationally and internationally. To allow for from vital registration systems, population added comparability, UN IGME generates such 15 estimates with uncertainty bounds. Applying the country-specific charts). In this round a consistent methodology also allows for of estimation, a substantial amount of newly comparisons between countries, despite the available data has been added to the underlying varied number and types of data sources. UN database for under-five, infant and neonatal IGME applies a common methodology across mortality. Data from 64 new surveys or censuses countries and uses original empirical data were added for 40 countries and data from vital from each country but does not report figures registration systems or sample vital registration produced by individual countries using other systems were updated for 131 countries. In total, methods, which would not be comparable to more than 6,600 country-year data points for other country estimates. 400 series were added or updated. The database, as of September 2017, contains 18,000 country- Data sources year data points from more than 1,500 series Nationally representative estimates of under-five across 195 countries from 1990 (or earlier, up to mortality can be derived from several different 1940) to 2017. The increased empirical data have sources, including civil registration and sample substantially changed the estimates generated surveys. Demographic surveillance sites and by UN IGME for some countries from previous hospital data are excluded, as they are rarely editions partly because the fitted trend line is representative. The preferred source of data is based on the entire time series of data available a civil registration system which records births for each country. The estimates presented in this and deaths on a continuous basis. If registration report may differ from and are not necessarily is complete and the system functions efficiently, comparable with previous sets of UN IGME the resulting estimates will be accurate and estimates or the most recent underlying country timely. However, in the developing world most data. For mortality among children aged 5–14 countries do not have well-functioning vital years, data were calculated from censuses registration systems, and household surveys, such and surveys, or vital registration records of as the UNICEF-supported Multiple Indicator population and deaths in the age group. The Cluster Surveys (MICS), the Demographic and database for mortality among children aged 5–14 Health Surveys (DHS) supported by the United contains more than 5,500 data points. States Agency for International Development (USAID), and periodic population censuses have Whatever the method used to derive the become the primary source of data on mortality estimates, data quality is critical. UN IGME among children under age 5 and among children assesses data quality and does not include data aged 5–14. These surveys ask women about the sources with substantial non-sampling errors survival of their children, and it is these reports or omissions as underlying empirical data in its (or microdata upon availability) that provide the statistical model to derive UN IGME estimates. basis of child mortality estimates for a majority of developing countries. Data from civil registration systems Civil registration data are the preferred data The first step in the process of arriving at source for child mortality estimation. The estimates of levels and recent trends of the under- calculation of the under-five mortality rates five mortality rate or infant mortality rate is to (U5MR), infant mortality rates (IMR), neonatal compile all newly available data, and add the data mortality rates (NMR) and mortality rates among to the CME database. Newly available data will children aged 5–14 from civil registration data include newly released vital statistics from a civil is derived from a standard period abridged life registration system, results from recent censuses table. For civil registration data (with available and household surveys and, occasionally, results data on the number of deaths and mid-year from some older census or survey not previously populations), initially annual observations were available. constructed for all observation years in a country. For country-years in which the coefficient of The full set of empirical data used in this variation exceeded 10 per cent, deaths and analysis is publicly available from the UN IGME mid-year populations were pooled over longer web portal CME Info ( under “underlying data,” as well as on combining those with adjacent previous years, 16 to reduce spurious fluctuations in countries on the coefficients of variation (a measure of where small numbers of births and deaths were sampling uncertainty) of the estimates.13 observed. In general, summary birth history data, collected The coefficient of variation is defined to by censuses and many household surveys, use the be the stochastic standard error of the 5q0 age of the woman as an indicator of exposure (5q0=U5MR/1,000) or 1q0 (1q0=IMR/1,000) time and exposure time period of the children, observation divided by the value of the 5q0 or and use models to estimate mortality indicators 1q0 observation. The stochastic standard error for periods in the past for women aged 25–29 of the observation is calculated using a Poisson through 45–49. This method is well known, approximation using live birth numbers from the but has several shortcomings. Starting with the World Population Prospects, given by sqrt(5q0/ 2014 round of estimation, UN IGME changed lb) (or similarly sqrt(1q0/lb), where 5q0 is the the method of estimation for summary birth under-five mortality rate (per 1 live birth) and histories to one based on classification of women lb is the number of live births in the year of the by the time that has passed since their first birth. observation.11 After this recalculation of the The new method has several advantages over the civil registration data, the standard errors are previous one. First, estimates based on time since set to a minimum of 2.5 per cent for input into first birth generally have lower sampling errors, the model. A similar approach was used for and second, it avoids the problematic assumption neonatal mortality and mortality among children that the estimates derived for each age group aged 5–14. In previous revisions, UN IGME adequately represent the mortality of the whole adjusted vital registration (VR) data for deficient population. As a result, the new method has less completeness in the reporting of early infant susceptibility to the selection effect of young deaths in several European countries. For more women who give birth early, since all women details on the past adjustment, see the Notes who give birth necessarily must have a first birth. section.12 Third, the method tends to show less fluctuation across time, particularly in countries with Survey data relatively low fertility and mortality. UN IGME The majority of survey data comes in one of two considers the improvements in the estimates forms: the full birth history, whereby women of based on time since first birth worthwhile when reproductive ages from 15 to 49 are asked for the compared with the estimates derived from the date of birth of each of their children, whether classification by age of the mother; hence, in the child is still alive, and, if not, the age at cases where the microdata are available, UN death; and the summary birth history, whereby IGME has reanalysed the data using the new women are asked only about the number of their method.14 children ever born and the number that have died (or, equivalently, the number still alive). Moreover, following advice from the Technical Advisory Group (TAG) of UN IGME, child Full birth history data, collected by all DHS mortality estimates from summary birth histories and, increasingly, also MICS surveys, allow the were not included if estimates from full birth calculation of child mortality indicators for histories in the same survey were available.15 specific time periods in the past. This allows DHS and MICS to publish child mortality estimates Adjustment for missing mothers in high-HIV for five 5-year periods before the survey, that settings is, 0 to 4, 5 to 9, 10 to 14, etc. UN IGME has In populations severely affected by HIV and re-calculated estimates for calendar year periods, AIDS, HIV-positive (HIV+) children will be more using single calendar years for periods shortly likely to die than other children, and will also be before the survey, and then gradually increasing less likely to be reported since their mothers will the number of years for periods further in the have been more likely to die also. Child mortality past, whenever microdata from the survey are estimates will thus be biased downward. The available. The cut-off points for a given survey for magnitude of the bias will depend on the extent shifting from estimates for single calendar years to which the elevated under-five mortality of to two years, or two years to three, etc., are based HIV+ children is not reported because of the 17 deaths of their mothers. The TAG of UN IGME and to extrapolate that trend to a defined developed a method to adjust HIV/AIDS-related time point – in this case, 2016. This method is mortality for each survey data observation described in the following section. from full birth histories during HIV and AIDS epidemics (1980–present), by adopting a set of Estimation of under-five mortality rates simplified but reasonable assumptions about the Under-five mortality rate (U5MR) estimates distribution of births to HIV+ women, primarily were produced using the Bayesian B-splines bias- relating to the duration of their infection, adjusted model, referred to as the B3 model. vertical transmission rates, and survival times This model was developed, validated and used of both mothers and children from the time of to produce previous rounds of UN IGME child the birth.16 This method was applied to all direct mortality estimates published in September estimates from full birth histories. 2013,17 September 201418 and September 2015.19 Systematic and random measurement error In the B3 model, log (U5MR) is estimated Data from these different sources require with a flexible splines regression model. The different calculation methods and may suffer spline regression model is fitted to all U5MR from different errors – for example, random observations in the country (i.e., country-year errors in sample surveys or systematic errors due data points). An observed value for U5MR to misreporting. Thus, different surveys often is considered to be the true value for U5MR yield widely different estimates of U5MR for a multiplied by an error factor – i.e., observed given time period, as illustrated in Figure 6. U5MR = true U5MR * error, or, on the log- To reconcile these differences and take better scale, log(observed u5mr) = log(true U5MR) account of the systematic biases associated + log(error), where error refers to the relative with the various types of data inputs, TAG has difference between an observation and the truth. developed a new estimation method to fit a While estimating the true U5MR, properties of smoothed trend curve to a set of observations the errors that provide information about the FIGURE Empirical data of under-five mortality rate in Nigeria and Papua New Guinea 6 Nigeria Papua New Guinea 350 Under-five mortality rate (deaths per 1,000 live births) Under-five mortality rate (deaths per 1,000 live births) 300 200 250 150 200 150 100 100 50 50 0 0 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 Year Year Note: All data available for the country are shown as coloured points, with observations from the same data series joined by lines, and each colour identifying different data sources. Grey bands in the left plot represent the standard errors of the observations where available or applicable. Series considered but not included in the statistical model due to substantial non-sampling errors or omission appear with dashed lines. 18 quality of the observation – or in other words, FIGURE Empirical under-five mortality data and the extent of error that we expect – are taken 7 estimates from the B3 model for Senegal into account. These properties include: the standard error of the observation; its source type Senegal (e.g., DHS versus census) and if the observation 500 is part of a data series from a specific survey Under-five mortality rate (deaths per 1,000 live births) (and how far the data series is from other series with overlapping observation periods). These 400 properties are summarized in the so-called data model. When estimating the U5MR, the data model adjusts for the errors in the observations, 300 including the average systematic biases associated with different types of data sources, using 200 information on data quality for different source types from every country. 100 Figure 7 displays the plot of the U5MR over time for Senegal, used here for illustrative purposes. 0 Compared with the previously applied Loess 1950 1960 1970 1980 1990 2000 2010 estimation approach, the B3 model better Year accounts for data errors, including biases and Note: The B3 estimates are in red. Ninety per cent uncertainty intervals for the sampling and non-sampling errors in the data. U5MR are given by the pink bands. All data available for the country are shown It can better capture short-term fluctuations in as coloured points, with observations from the same data series joined by lines. Solid points and lines represent data series/observations that were included the under-five mortality rate and its annual rate for curve-fitting. Grey bands in the left plot represent the standard errors of the of reduction, and thus is better able to account observations where available or applicable. for evidence of acceleration in the decline of under-five mortality from new surveys. Validation where r is the ratio of the IMR to the median exercises show that the B3 model also performs B3 estimates of U5MR in the corresponding better in short-term projections. country-year. This is to restrict the IMR to be lower than the U5MR. For the remaining The B3 method was developed and implemented countries, the IMR is derived from the U5MR, for UN IGME by Leontine Alkema from the through the use of model life tables that contain University of Massachusetts, Amherst, and Jin Rou New from the National University of known regularities in age patterns of child Singapore, with guidance and review by the mortality.21 The advantage of this approach is TAG of UN IGME. A more complete technical that it avoids potential problems with the under- description of the B3 model is available reporting of neonatal deaths in some countries elsewhere.20 and ensures that the internal relationships of the three indicators are consistent with established Estimation of infant mortality rates norms. For Sahelian countries (Burkina Faso, In general, the B3 model described above is Chad, the Gambia, Mali, Mauritania, the Niger applied to the U5MR for all countries (except and Senegal), the relationship from model life for the Democratic People’s Republic of Korea, tables does not apply between infant and child where a non-standard method was employed). For countries with high-quality VR data (covering mortality, thus a logit transform of the ratio a sufficient period of time and deemed to have of IMR/U5MR is used to estimate IMR from high levels of completeness and coverage), the U5MR using data from full birth histories and B3 model is also used to estimate IMR, but is a multilevel regression with country-specific fitted to the logit transform of r – i.e., log(r/1-r), intercept. 19 Adjustment for rapidly changing under- uses the available data by sex to estimate a five and infant mortality driven by HIV time trend in the sex ratio (male/female ratio) and AIDS of U5MR instead. Bayesian methods for UN To capture the extraordinarily rapid changes IGME estimation of sex ratios with a focus on in child mortality driven by HIV and AIDS over the estimation and identification of countries the epidemic period in some countries, the with outlying levels or trends were used. A more regression models were fitted to data points for complete technical description of the new model the U5MR from all other causes than HIV and is available elsewhere.11 AIDS, and then estimates of HIV and AIDS under-five mortality from the Joint United Estimates of neonatal mortality Nations Programme on HIV/AIDS (UNAIDS) The neonatal mortality rate (NMR) is defined as were added to estimates from the regression the probability of dying before 28 days per 1,000 model. This method was used for 17 countries live births. In 2015, the UN IGME method for where the HIV prevalence rate exceeded 5 per estimating NMR was updated. The new Bayesian cent at any point in time since 1980. Steps were as methodology is similar to that used to estimate follows: U5MR and estimates by sex. It has the advantage that, compared with the previous model, it 1. Compile and assess the quality of all newly can capture data-driven trends in NMR within available, nationally representative data countries and over time for all countries. A more relevant to the estimation of child mortality. complete technical description of the new model is available elsewhere.24 2. Adjust survey data to account for possible biases in data collection related to the HIV For neonatal mortality in HIV-affected and and AIDS epidemic. crisis-affected populations, the ratio is estimated initially for non-AIDS and non-crisis mortality. 3. Use UNAIDS estimates of AIDS child After estimation, crisis neonatal deaths are mortality22 to adjust the data points from added back on to the neonatal deaths to compute 1980 onward to exclude AIDS deaths. the total estimated neonatal mortality rate. No AIDS deaths are added back to the NMR, thereby 4. Fit the standard B3 model to the observations assuming that HIV/AIDS-related deaths only to AIDS-free data points. affect child mortality after the first month of life. 5. Extrapolate the model to the target year – in Estimation of mortality in children aged this case, 2016. 5–14 For the first time this year, UN IGME produces 6. Add back estimates of deaths due to AIDS country-specific trend estimates of the mortality (from UNAIDS). in children aged 5–14 – that is, the probability that a child aged 5 dies before reaching his or 7. For the epidemic period, a non-AIDS curve her fifteenth birthday (10q5). The methods of IMR is derived from U5MR using model used are similar to those that are used to life tables and then the UNAIDS estimates estimate under-five mortality rates (U5MR). of AIDS deaths for children under age 1 are In 39 countries, there were not enough data added to generate the final IMR estimates. inputs to estimate the probability 10q5 from vital registration, surveys or censuses. For these Estimates of under-five and infant cases, the probability 10q5 was modelled based mortality by sex on an expected relation between mortality in In 2012, UN IGME produced estimates of U5MR the age groups 0–4 and 5–14, as observed in for males and females separately for the first countries with sufficient data series. A linear time.23 In many countries, fewer sources have regression was used to regress log (10q5) against provided data by sex than have provided data log (U5MR), with region-specific dummies, and for both sexes combined. For this reason, rather the coefficients of this regression were used than estimate U5MR trends by sex directly to predict the probability 10q5 between 1990 from reported mortality levels by sex, UN IGME and 2016 for countries with insufficient data 20 sources based on the estimates of the under-five different crises for mortality among children mortality rate. The advantage of this approach aged 5–14. Because the background mortality is that no model life tables are used (since such rates were relatively low in the age group 5–14, life tables are based on the historical experience crisis deaths represented a larger share of deaths, of countries with high-quality vital registration and thus more crises met these criteria than for data and do not always adequately reflect age under-five mortality. Crisis deaths were included patterns of mortality in low- and middle-income in the estimates by first excluding data points countries). from crisis years, fitting the B3 model to the remaining data, and then adding the crisis- In populations severely affected by HIV and specific mortality rate to the fitted B3 curve. AIDS, HIV-positive (HIV+) children will be more Crisis death estimates are uncertain but presently likely to die than other children, and will also no uncertainty around crisis deaths is included be less likely to be reported since their mothers in the uncertainty intervals of the estimates. will have been also more likely to die. However, Instead, we assume that the relative uncertainty no adjustment was included for HIV-related in the adjusted estimates is equal to the relative biases in the age group 5–14, since no method uncertainty in the non-adjusted estimates; this currently exists to estimate the magnitude of assumption will be revisited in the near future. this bias in the probability 10q5. This bias should be less severe when estimating mortality in the UN IGME has assessed recent humanitarian age group 5–14, as compared with the under- crises, namely, in the Syrian Arab Republic and five mortality rate, because in the absence of Yemen. Based on the scarce currently available treatment, the majority of children infected data and the difficulties to estimate a broader through their mothers, will die before reaching impact of these crises on health systems, UN age 5. IGME decided to hold the estimates constant from the start of each of these crises while Estimating child mortality due to increasing the uncertainty over the crisis time, conflict and natural disasters where applicable direct crisis deaths have been Estimated deaths for major crises were derived added to the constant trend estimate. UN from various data sources from 1990 to the IGME will review new data, if available, in the present. Estimated deaths from natural disasters next estimation round and revise estimates were obtained from the CRED International accordingly. Disaster Database,25 with under-five proportions and for children aged 5–14 estimated as described elsewhere,26 and conflict deaths were Estimation of uncertainty intervals taken from Uppsala Conflict Data Program/ Given the inherent uncertainty in child mortality Peace Research Institute Oslo datasets as well estimates, 90 per cent uncertainty intervals as reports prepared by the United Nations and (UIs) are used by UN IGME instead of the more other organizations. Estimated child deaths due conventional 95 per cent ones. While reporting to major crises were included if they met the intervals that are based on higher levels of following criteria: (1) the crisis was isolated to uncertainty (i.e., 95 per cent instead of 90 per a few years; (2) under-five crisis deaths or crisis cent) would have the advantage that the chance deaths among children aged 5–14 were > 10 per of not having included the true value in the cent of non-crisis deaths in the age group; (3) interval is smaller, the disadvantage of choosing crisis U5MR > 0.2 per 1,000 or crisis 10q5 was > higher uncertainty levels is that intervals lose 0.2 per 1,000; and (4) the number of under-five their utility to present meaningful summaries crisis deaths or among children 5–14 years old of a range of likely outcomes if the indicator of was > 10 deaths or (5) in the event that high- interest is highly uncertain. Given this trade- quality vital registration data were available and off and the substantial uncertainty associated should not be smoothed by the B3 model. with child mortality estimates, UN IGME chose to report 90 per cent UIs, or, in other words, These criteria resulted in 16 different crises intervals for which there is a 90 per cent chance being explicitly incorporated into the UN that they contain the true value, to encourage IGME estimates for under-five mortality and 38 wider use and interpretation of the UIs. 21 Extrapolation to common reference year Nations and/or regional peacekeeping operation or c) presence of a United Nations and/ or regional peace-building and political mission. The detailed classi fication can be found If the underlying empirical data refer to an at , accessed on 6 September 2017. earlier reference period than the end year of the 3. Institut National de la Statistique des Études Économiques et Démographiques - period the estimates are reported, UN IGME INSEED/Tchad, Ministère de la Santé Publique - MSP/Tchad, and ICF International, Enquête Démographique et de Santé et à Indicateurs Multiples au Tchad (EDS-MICS) extrapolates the estimates to the common end 2014–2015, 2016, available from . year, in this round to 2016. UN IGME does not 4. Of fice of the Registrar General & Census Commissioner, India Ministry of Home Affairs Government of India New Delhi: Sample Registration System Statistical Report 2015, India: use any covariates to derive the estimates, but Sample Registration System 2015, New Delhi, 2016, available from . 5. Among the 99 countries with estimates of under- five mortality by wealth quintile. trend to extrapolate to the target year. The 6. This group of countries accounts for 93 per cent of the under- five deaths and 71 per cent average extrapolation period in the 2017 round of of the under- five population lives in these countries. The average under- five mortality rate for this group of countries in 2016 was 53 deaths per 1,000 live births. estimation was 4.5 years for under-five mortality, 7. Among the 99 countries with estimates of under- five mortality by wealth quintile. with half of the countries having data points 8. WHO and Maternal and Child Epidemiology Estimation Group (MCEE) provisional esti- mates 2017, forthcoming. within the past 3.5 years. For about 70 countries, 9. World Health Organization, Global Health Estimates 2015, WHO, 2016. the latest available child mortality estimate was 10. Economic and Social Council: Statistical Commission Report on the forty-eighth ses- sion (7–10 March 2017) E/2017/24-E/CN.3/2017/35, United Nations, New York, 2017. more than 5 years old. 11. Alkema, L., and J. R. New, ‘Global Estimation of Child Mortality Using a Bayesian B-Spline Bias-Reduction Method’, Annals of Applied Statistics vol. 8, no. 4, 2014, pp. 2122–2149. Calculating number of deaths for 12. There were concerns about the completeness of early infant mortality data from civil children under age 5 registration. A European report on perinatal indicators, for example, noted a wide varia- tion in how European countries defi ne infant mortality, due to differences in birth and A birth-week cohort method is used to calculate death registration practices (that is, differences in the cut-off points for acceptable weight the absolute number of deaths among neonates, or estimated gestation period to be registered as a birth and subsequent death). This discrepancy can lead to under-reporting of infant deaths by some countries, particularly infants and children under age 5. First, each when compared with countries that use a broader defi nition for live birth. UN IGME pre- viously carried out an analysis of the ratio of early neonatal (under seven days) deaths to annual birth cohort is divided into 52 equal total neonatal deaths, which showed that several countries, many in Eastern Europe, had birth-week cohorts. Then each birth-week cohort signi ficantly lower values than what would be expected, suggesting an undercounting of early infant deaths. The results of this analysis were used as an upwards adjustment of 10 is exposed throughout the first five years of per cent or 20 per cent to under- five mortality rates across all years for several countries in previous UN IGME reports. This year, this assessment was revisited using the latest data, life to the appropriate calendar year- and age- and the clear signal of underreporting is no longer apparent across countries. Therefore, UN IGME has removed these adjustment factors in the estimates for this publication. specific mortality rates depending on cohort age. Going forward, UN IGME will assemble fi ner age-speci fic child mortality data, and attempt For example, the twentieth birth week cohort to determine the current level of underreporting bias in different countries, and how that bias has changed over time. This analysis could lead to a different adjustment approach in of the year 2000 will be exposed to the infant future estimates. mortality rates in both 2000 and 2001. All deaths 13. Pedersen, J., and J. Liu, ‘Child Mortality Estimation: Appropriate time periods for child mortality estimates from full birth histories’, Plos Medicine, vol. 9, no. 8, 2012. from birth-week cohorts occurring as a result 14. Verhulst, A., ‘Child Mortality Estimation: An assessment of summary birth history of exposure to the mortality rate for a given methods using microsimulation’, Demographic Research, vol. 34, article 39, available from . calendar year are allocated to that year and are 15. Silva, R., ‘Child Mortality Estimation: Consistency of under- five mortality rate estimates using full birth histories and summary birth histories’, PLoS Medicine, vol. 9, no. 8, 2012. summed by age group at death to get the total 16. Walker, N., K. Hill, and F. M. Zhao, ‘Child Mortality Estimation: Methods used to adjust number of deaths for a given year and age group. for bias due to AIDS in estimating trends in under- five mortality’, PLoS Medicine, vol. 9, no. 8, 2012. Continuing with the above example, deaths 17. United Nations Inter-agency Group for Child Mortality Estimation (UN IGME), ‘Levels from the twentieth birth-week cohort of the year & Trends in Child Mortality. Report 2013’, United Nations Children’s Fund, New York, 2013, available from . 2000 would contribute to infant deaths in year 18. United Nations Inter-agency Group for Child Mortality Estimation (UN IGME), ‘Levels 2000 and 2001. Any deaths occurring among the & Trends in Child Mortality. Report 2014’, United Nations Children’s Fund, New York, 2014, available from. twentieth birth-week cohort of year 2000 after 19. United Nations Inter-agency Group for Child Mortality Estimation (UN IGME), ‘Levels the twentieth week in 2001 would contribute & Trends in Child Mortality. Report 2015’, United Nations Children’s Fund, New York, 2015, available from . to under-five deaths for year 2001 and so forth. 20. Alkema, L., and J. R. New, ‘Global Estimation of Child Mortality Using a Bayesian Under-five deaths in each calendar year are B-Spline Bias-Reduction Method’, Annals of Applied Statistics, vol. 8, 2014, pp. 2122–2149. 21. Guillot, M., et al., ‘Child Mortality Estimation: A global overview of infant and child calculated by summing up all the deaths under mortality age patterns in light of new empirical data’, PLoS Medicine, vol. 9, no. 8, 2012. age 5 across all age group cohorts in that year. 22. UNAIDS 2017 estimates, July 2017. 23. Sawyer, C.C., ‘Child Mortality Estimation: Estimating sex differences in childhood mor- The annual number of live births estimates in tality since the 1970s’, PLoS Medicine, vol. 9, no. 8, 2012. each country used to calculate the annual under- 24. Alexander, M., and L. Alkema, ‘Global Estimation of Neonatal Mortality Using a Bayesian Hierarchical Splines Regression Model’, 2016, available at . Prospects: the 2017 Revision.27 25. CRED, ‘EM-DAT: The CRED International Disaster Database’, Université Catholique de Louvain, Belgium, available from . 26. World Health Organization, ‘WHO Methods and Data Sources for Global Causes of Notes Death 2000–2015’, Global Health Estimates Technical Paper WHO/HIS/IER/GHE/2016.3, 1. Values in parentheses indicate 90 per cent uncertainty intervals for the estimates. WHO, Geneva, 2016, available from . 2. Fragile states refer to the World Bank Group ‘Harmonized List of Fragile Situations FY18’. ‘Fragile situations’ have: either a) a harmonized average Country Policy and Insti- 27. United Nations Department of Economic and Social Affairs Population Division, tutional Assessment (CPIA) country rating of 3.2 or less, or b) the presence of a United ‘World Population Prospects: The 2017 revision’, United Nations, New York, 2017. 22 23 STATISTICAL TABLE Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Under-five mortality rate (U5MR) with 90 per cent uncertainty interval Number of under-five deaths with 90 per cent (deaths per 1,000 live births) uncertainty interval (thousands) a Annual rate of reduction (ARR) 1990 2016 1990 2016 (per cent) 1990-2016 Under- Under- Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper Country U5MR U5MR ARR five five bound bound bound bound bound bound bound bound bound bound deaths deaths Afghanistan 177 162 194 70 57 85 3.6 2.8 4.5 108 99 118 80 64 96 Albania 40 35 45 14 7 25 4.2 1.8 6.6 3 3 4 0 0 1 Algeria 49 46 52 25 23 27 2.6 2.1 2.9 41 38 44 24 22 26 Andorra 9 5 15 3 2 5 4.5 1.3 7.6 0 0 0 0 0 0 Angola 221 198 249 83 41 147 3.8 1.5 6.5 136 121 152 96 48 170 Antigua and Barbuda 26 19 36 9 6 11 4.3 2.6 6.1 0 0 0 0 0 0 Argentina 29 28 29 11 11 12 3.7 3.5 3.9 20 20 21 8 8 9 Armenia 50 45 55 13 10 18 5.0 3.9 6.1 4 3 4 1 0 1 Australia 9 9 9 4 4 4 3.5 3.3 3.7 2 2 2 1 1 1 Austria 10 9 10 4 3 4 3.8 3.5 4.3 1 1 1 0 0 0 Azerbaijan 95 86 104 31 19 52 4.3 2.2 6.3 19 17 21 5 3 9 Bahamas 24 22 25 11 9 13 3.1 2.3 3.9 0 0 0 0 0 0 Bahrain 23 22 24 8 7 9 4.3 3.7 4.8 0 0 0 0 0 0 Bangladesh 144 140 148 34 31 38 5.5 5.1 5.9 532 517 548 106 96 117 Barbados 18 17 19 12 10 16 1.4 0.5 2.3 0 0 0 0 0 0 Belarus 15 15 16 4 4 4 5.2 5.0 5.5 2 2 2 0 0 0 Belgium 10 10 10 4 4 4 3.6 3.2 4.0 1 1 1 1 0 1 Belize 39 35 44 15 14 16 3.7 3.1 4.3 0 0 0 0 0 0 Benin 178 167 191 98 74 132 2.3 1.2 3.4 39 37 42 38 29 52 Bhutan 128 114 146 32 23 45 5.3 3.9 6.7 3 2 3 0 0 1 Bolivia (Plurinational State of) 124 117 131 37 26 52 4.7 3.4 6.0 29 28 31 9 7 13 Bosnia and Herzegovina 18 18 19 6 6 7 4.3 3.8 4.6 1 1 1 0 0 0 Botswana 54 47 62 41 18 77 1.1 -1.5 4.3 2 2 3 2 1 4 Brazil 64 60 69 15 13 18 5.6 4.8 6.3 240 224 256 45 37 54 Brunei Darussalam 13 13 14 10 9 11 1.1 0.6 1.7 0 0 0 0 0 0 Bulgaria 18 18 19 8 7 8 3.4 3.1 3.7 2 2 2 1 0 1 Burkina Faso 199 186 212 85 66 109 3.3 2.3 4.3 79 74 84 60 46 77 Burundi 170 155 187 72 57 91 3.3 2.3 4.3 45 40 49 31 24 39 Cabo Verde 63 60 65 21 18 25 4.1 3.5 4.7 1 1 1 0 0 0 Cambodia 116 108 125 31 19 49 5.1 3.3 7.0 44 41 47 11 7 18 Cameroon 143 133 154 80 62 103 2.3 1.3 3.2 72 67 77 66 52 86 Canada 8 8 8 5 4 6 2.0 1.4 2.5 3 3 3 2 2 2 Central African Republic 174 157 193 124 80 192 1.3 -0.5 3.0 20 18 23 20 13 31 Chad 211 196 226 127 105 150 1.9 1.3 2.7 60 56 64 77 64 91 Chile 19 19 20 8 8 9 3.2 2.9 3.5 6 5 6 2 2 2 China 54 50 59 10 9 11 6.5 5.9 7.1 1,402 1,292 1,525 168 147 194 Colombia 35 33 38 15 12 20 3.2 2.3 4.2 32 30 34 11 9 15 Comoros 126 112 140 73 39 144 2.1 -0.6 4.5 2 2 2 2 1 4 Congo 91 80 103 54 38 74 2.0 0.7 3.4 8 7 9 9 7 13 Cook Islands 24 22 27 8 5 12 4.4 2.7 6.2 0 0 0 0 0 0 Costa Rica 17 16 17 9 7 10 2.5 1.8 3.2 1 1 1 1 1 1 Côte d’Ivoire 151 141 162 92 69 122 1.9 0.8 3.0 77 72 82 78 58 103 Croatia 13 13 13 5 4 5 3.9 3.5 4.2 1 1 1 0 0 0 Cuba 13 13 14 6 5 6 3.4 3.2 3.6 2 2 2 1 1 1 Cyprus 11 11 12 3 2 3 5.6 4.7 6.5 0 0 0 0 0 0 Czechia 12 12 12 3 3 4 5.1 4.8 5.3 2 2 2 0 0 0 Democratic People’s Republic of Korea 43 34 56 20 16 26 3.0 ─ ─ 18 14 23 7 5 9 Democratic Republic of the Congo 184 167 203 94 67 129 2.6 1.3 3.9 280 254 308 304 218 416 Denmark 9 9 9 4 4 5 2.8 2.2 3.3 1 1 1 0 0 0 Djibouti 118 101 136 64 43 95 2.3 0.7 3.9 3 2 3 1 1 2 Dominica 17 16 19 34 27 43 -2.6 -3.6 -1.7 0 0 0 0 0 0 Dominican Republic 60 56 64 31 23 41 2.6 1.5 3.6 13 12 14 7 5 9 24 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Sex-specific under-five Probability of mortality rate Infant Neonatal dying among Number of (deaths per 1,000 live births) mortality rate Number of mortality rate Number of children deaths among (deaths per infant deaths (deaths per neonatal deaths aged 5–14 children 1,000 live (thousands) a 1,000 live (thousands)a (deaths per aged 5–14 1990 2016 births) births) 1,000 children (thousands)a aged 5) Country Male Female Male Female 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 Afghanistan 183 172 74 66 120 53 74 60 75 40 48 46 21 10 7 10 Albania 44 36 15 12 35 12 3 0 13 6 1 0 6 2 0 0 Algeria 53 45 27 24 41 22 34 20 23 16 19 15 9 4 7 3 Andorra 9 8 3 3 7 2 0 0 4 1 0 0 2 1 0 0 Angola 231 211 88 76 131 55 82 65 54 29 35 35 46 15 17 13 Antigua and Barbuda 29 23 9 8 25 5 0 0 17 4 0 0 5 2 0 0 Argentina 32 26 12 10 26 10 18 7 15 6 11 5 3 2 2 2 Armenia 54 45 15 12 42 12 3 0 23 7 2 0 4 2 0 0 Australia 10 8 4 3 8 3 2 1 5 2 1 1 2 1 0 0 Austria 11 8 4 3 8 3 1 0 5 2 0 0 2 1 0 0 Azerbaijan 103 86 34 28 75 27 15 5 32 18 7 3 5 3 1 0 Bahamas 25 22 11 10 20 9 0 0 14 6 0 0 4 3 0 0 Bahrain 24 22 8 7 20 7 0 0 15 3 0 0 4 2 0 0 Bangladesh 147 140 37 32 100 28 368 87 64 20 241 62 25 5 75 16 Barbados 20 16 13 11 16 11 0 0 12 8 0 0 3 2 0 0 Belarus 17 13 4 3 12 3 2 0 9 2 1 0 4 2 1 0 Belgium 11 9 4 4 8 3 1 0 5 2 1 0 2 1 0 0 Belize 43 35 16 13 32 13 0 0 20 10 0 0 5 3 0 0 Benin 185 171 102 93 107 63 24 25 46 31 11 13 46 22 7 7 Bhutan 134 121 36 29 90 27 2 0 43 18 1 0 18 7 0 0 Bolivia (Plurinational State of) 130 117 40 33 85 30 20 7 42 19 10 5 14 6 2 1 Bosnia and Herzegovina 20 16 7 5 16 5 1 0 11 5 1 0 3 1 0 0 Botswana 58 50 44 37 42 33 2 2 26 26 1 1 10 8 0 0 Brazil 69 59 16 14 53 14 198 40 26 8 96 23 5 3 17 8 Brunei Darussalam 14 12 11 9 10 9 0 0 6 4 0 0 4 2 0 0 Bulgaria 21 16 8 7 15 7 2 0 8 4 1 0 4 2 0 0 Burkina Faso 206 191 89 80 99 53 40 38 46 26 19 19 41 26 11 14 Burundi 180 160 77 66 103 48 27 21 41 24 11 11 57 20 9 6 Cabo Verde 67 58 23 19 48 18 1 0 20 10 0 0 6 2 0 0 Cambodia 124 108 34 27 85 26 32 10 40 16 15 6 36 5 9 2 Cameroon 151 135 85 74 89 53 45 44 42 24 22 20 37 30 13 19 Canada 9 7 5 5 7 4 3 2 4 3 2 1 2 1 1 0 Central African Republic 181 166 130 117 114 89 14 14 52 42 6 7 36 22 3 3 Chad 219 202 133 121 111 75 33 46 52 35 16 22 55 26 10 11 Chile 21 17 9 8 16 7 5 2 9 5 2 1 3 2 1 0 China 56 52 11 9 42 9 1,090 144 30 5 759 86 7 3 138 45 Colombia 39 31 17 14 29 13 26 10 18 9 16 6 5 3 4 2 Comoros 133 118 78 68 88 55 2 1 51 33 1 1 17 6 0 0 Congo 96 85 58 49 59 39 5 7 28 21 3 4 37 11 3 2 Cook Islands 27 22 9 7 21 7 0 0 13 4 0 0 5 2 0 0 Costa Rica 19 15 10 8 14 8 1 1 9 6 1 0 3 2 0 0 Côte d’Ivoire 163 139 101 82 104 66 54 57 50 37 26 32 31 28 11 18 Croatia 14 11 5 4 11 4 1 0 8 3 0 0 3 1 0 0 Cuba 15 12 6 5 11 4 2 1 7 2 1 0 4 2 1 0 Cyprus 12 10 3 2 10 2 0 0 6 1 0 0 2 1 0 0 Czechia 14 10 4 3 10 3 1 0 7 2 1 0 2 1 0 0 Democratic People’s Republic of Korea 47 39 22 18 33 15 14 5 21 11 9 4 8 4 3 1 Democratic Republic of the Congo 192 175 101 87 118 72 184 235 41 29 67 96 44 28 44 64 Denmark 10 8 5 4 7 4 0 0 4 3 0 0 2 1 0 0 Djibouti 126 109 70 58 91 54 2 1 49 33 1 1 17 2 0 0 Dominica 18 16 36 31 14 31 0 0 11 24 0 0 3 2 0 0 Dominican Republic 64 55 34 28 46 26 10 5 25 21 5 4 8 3 1 1 25 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Under-five mortality rate (U5MR) with 90 per cent uncertainty interval Number of under-five deaths with 90 per cent (deaths per 1,000 live births) uncertainty interval (thousands)a Annual rate of reduction (ARR) 1990 2016 1990 2016 (per cent) 1990-2016 Under- Under- Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper Country U5MR U5MR ARR five five bound bound bound bound bound bound bound bound bound bound deaths deaths Ecuador 57 51 63 21 13 35 3.8 1.9 5.7 17 16 19 7 4 11 Egypt 86 82 90 23 18 29 5.1 4.1 6.0 165 157 172 57 45 74 El Salvador 60 55 65 15 10 22 5.3 3.8 6.8 10 9 11 2 1 3 Equatorial Guinea 191 165 221 91 61 132 2.8 1.3 4.5 3 3 4 4 2 5 Eritrea 151 138 166 45 29 70 4.7 2.9 6.5 19 17 21 7 5 11 Estonia 18 17 18 3 3 3 7.0 6.3 7.5 0 0 0 0 0 0 Ethiopia 203 189 218 58 47 73 4.8 3.9 5.7 441 411 473 187 149 233 Fiji 28 24 34 22 18 26 1.0 0.0 1.9 1 1 1 0 0 0 Finland 7 7 7 2 2 3 4.1 3.8 4.6 0 0 0 0 0 0 France 9 9 9 4 4 4 3.2 2.9 3.6 7 7 7 3 3 3 Gabon 92 80 107 47 32 68 2.6 1.1 4.2 3 3 4 3 2 4 Gambia 168 150 188 65 41 103 3.6 1.9 5.4 7 6 8 5 3 8 Georgia 47 42 53 11 10 12 5.7 5.0 6.3 4 4 5 1 1 1 Germany 9 8 9 4 4 4 3.1 2.9 3.3 7 7 7 3 3 3 Ghana 127 120 134 59 45 77 3.0 1.9 4.0 70 66 74 51 39 66 Greece 11 10 11 4 3 4 3.9 3.3 4.6 1 1 1 0 0 0 Grenada 22 21 24 16 14 19 1.3 0.6 2.0 0 0 0 0 0 0 Guatemala 82 77 87 29 24 35 4.1 3.3 4.8 29 27 31 12 10 15 Guinea 235 220 251 89 72 111 3.7 2.9 4.6 63 59 67 39 31 48 Guinea-Bissau 219 196 244 88 61 123 3.5 2.2 5.0 10 9 11 6 4 8 Guyana 60 55 66 32 22 48 2.4 0.8 3.9 1 1 1 1 0 1 Haiti 145 136 155 67 52 88 3.0 1.9 3.9 38 35 40 17 14 23 Honduras 58 54 63 19 14 25 4.4 3.2 5.5 11 10 12 4 3 5 Hungary 17 17 18 5 4 6 4.6 4.0 5.3 2 2 2 0 0 1 Iceland 6 6 7 2 2 3 4.3 3.2 5.3 0 0 0 0 0 0 India 126 122 130 43 39 47 4.1 3.7 4.6 3,396 3,287 3,511 1,081 975 1,188 Indonesia 84 81 88 26 21 33 4.5 3.6 5.4 395 378 413 131 104 165 Iran (Islamic Republic of) 57 52 61 15 11 21 5.1 3.7 6.4 107 98 115 20 15 29 Iraq 54 50 59 31 23 42 2.1 1.0 3.3 35 32 38 38 28 51 Ireland 9 9 10 4 3 4 3.6 3.0 4.2 0 0 1 0 0 0 Israel 12 11 12 4 3 4 4.5 4.1 4.9 1 1 1 1 1 1 Italy 10 10 10 3 3 4 4.1 3.8 4.4 5 5 6 2 2 2 Jamaica 30 26 36 15 10 25 2.6 0.8 4.4 2 2 2 1 0 1 Japan 6 6 6 3 3 3 3.3 3.2 3.5 8 8 9 3 3 3 Jordan 37 34 39 18 13 24 2.8 1.5 4.1 4 4 5 4 3 6 Kazakhstan 52 48 57 11 11 12 5.9 5.5 6.3 20 19 22 4 4 5 Kenya 98 92 104 49 41 60 2.7 1.9 3.4 95 90 101 74 61 90 Kiribati 96 83 111 54 35 84 2.2 0.4 4.0 0 0 0 0 0 0 Kuwait 18 17 18 8 8 9 2.8 2.5 3.2 1 1 1 1 0 1 Kyrgyzstan 65 58 73 21 20 22 4.4 3.9 4.8 9 8 10 3 3 3 Lao People’s Democratic Republic 162 146 178 64 46 88 3.6 2.3 4.9 28 26 31 10 7 14 Latvia 17 16 18 5 4 5 5.0 4.5 5.5 1 1 1 0 0 0 Lebanon 33 29 37 8 4 14 5.4 3.2 7.8 2 2 2 1 0 1 Lesotho 91 83 100 94 72 121 -0.1 -1.1 1.0 5 5 6 6 4 7 Liberia 258 237 280 67 51 93 5.2 4.0 6.3 25 23 27 10 8 14 Libya 42 36 49 13 9 19 4.5 3.0 6.1 5 5 6 2 1 2 Lithuania 15 14 16 5 5 6 4.0 3.5 4.4 1 1 1 0 0 0 Luxembourg 9 8 10 2 2 3 5.0 4.0 6.0 0 0 0 0 0 0 Madagascar 160 149 171 46 32 66 4.7 3.4 6.2 78 73 83 37 26 53 Malawi 232 220 246 55 43 71 5.5 4.6 6.5 98 93 104 36 28 46 Malaysia 17 16 17 8 8 9 2.7 2.5 2.9 8 8 9 4 4 5 Maldives 94 86 103 9 7 11 9.2 8.3 10.2 1 1 1 0 0 0 26 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Sex-specific under-five Probability of mortality rate Infant Neonatal dying among Number of (deaths per 1,000 live births) mortality rate Number of mortality rate Number of children deaths among (deaths per infant deaths (deaths per neonatal deaths aged 5–14 children 1,000 live (thousands) a 1,000 live (thousands)a (deaths per aged 5–14 1990 2016 births) births) 1,000 children (thousands)a aged 5) Country Male Female Male Female 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 Ecuador 62 51 23 18 44 18 13 6 24 11 8 4 8 3 2 1 Egypt 86 86 24 22 63 19 120 49 33 13 64 32 11 5 16 9 El Salvador 64 55 17 13 46 13 8 2 23 8 4 1 7 4 1 0 Equatorial Guinea 200 180 97 84 129 66 2 3 51 32 1 1 40 17 0 0 Eritrea 162 139 49 39 93 33 12 5 34 18 4 3 48 13 4 2 Estonia 20 15 3 3 14 2 0 0 10 1 0 0 5 1 0 0 Ethiopia 215 191 64 53 121 41 268 132 60 28 137 90 81 17 117 48 Fiji 31 26 24 20 24 19 1 0 13 9 0 0 10 4 0 0 Finland 7 6 3 2 6 2 0 0 4 1 0 0 2 1 0 0 France 10 8 4 4 7 3 6 2 4 2 3 2 2 1 2 1 Gabon 99 86 52 43 60 34 2 2 32 22 1 1 22 16 1 1 Gambia 175 160 70 61 82 42 3 3 50 28 2 2 35 13 1 1 Georgia 53 42 12 9 40 10 4 1 25 7 2 0 7 3 1 0 Germany 10 7 4 4 7 3 6 2 3 2 3 2 2 1 2 1 Ghana 135 119 64 53 80 41 44 36 42 27 24 24 29 13 12 9 Greece 11 10 4 4 9 3 1 0 7 2 1 0 2 1 0 0 Grenada 24 21 17 15 18 14 0 0 12 8 0 0 4 3 0 0 Guatemala 87 76 31 26 60 24 21 10 29 14 10 6 12 4 3 1 Guinea 243 227 94 84 139 58 38 26 63 25 18 11 52 22 9 8 Guinea-Bissau 234 203 96 80 130 58 6 4 64 38 3 3 34 18 1 1 Guyana 67 53 37 28 46 27 1 0 31 20 1 0 11 6 0 0 Haiti 153 136 73 61 100 51 26 13 39 25 10 6 31 15 6 4 Honduras 63 53 21 17 45 16 9 3 22 10 4 2 9 4 1 1 Hungary 19 15 6 5 15 4 2 0 11 3 1 0 3 1 0 0 Iceland 7 6 2 2 5 2 0 0 3 1 0 0 2 1 0 0 India 122 130 42 44 88 35 2,385 867 57 25 1,570 640 21 6 449 160 Indonesia 91 78 29 23 62 22 288 110 30 14 142 68 15 5 69 25 Iran (Islamic Republic of) 57 56 16 15 44 13 82 17 25 10 47 13 9 3 15 3 Iraq 58 50 34 28 42 26 28 32 26 18 18 22 8 6 4 5 Ireland 10 8 4 3 8 3 0 0 5 2 0 0 2 1 0 0 Israel 12 11 4 3 10 3 1 0 6 2 1 0 2 1 0 0 Italy 11 9 4 3 8 3 5 1 7 2 4 1 2 1 1 0 Jamaica 34 26 17 13 25 13 2 1 20 11 1 1 5 3 0 0 Japan 7 6 3 3 5 2 6 2 3 1 3 1 2 1 3 1 Jordan 38 35 19 17 30 15 4 4 20 11 3 3 5 4 1 1 Kazakhstan 59 45 13 10 44 10 17 4 22 6 8 2 6 3 2 1 Kenya 103 92 53 45 63 36 62 53 28 23 28 34 18 11 13 14 Kiribati 102 89 59 49 69 42 0 0 35 23 0 0 17 9 0 0 Kuwait 19 16 9 8 15 7 1 0 10 4 0 0 4 2 0 0 Kyrgyzstan 71 59 24 19 54 19 7 3 24 12 3 2 9 3 1 0 Lao People’s Democratic Republic 172 151 70 58 111 49 20 8 54 29 10 5 27 10 3 2 Latvia 19 15 5 4 13 4 0 0 8 2 0 0 6 2 0 0 Lebanon 34 31 8 8 27 7 2 1 21 5 1 0 7 1 0 0 Lesotho 98 84 101 86 73 72 4 4 40 39 2 2 14 14 1 1 Liberia 270 245 72 62 172 51 16 8 58 23 5 4 34 17 2 2 Libya 46 38 14 12 36 11 5 1 21 7 3 1 9 5 1 1 Lithuania 17 13 6 5 12 4 1 0 8 3 0 0 4 2 0 0 Luxembourg 10 8 3 2 7 2 0 0 4 2 0 0 2 1 0 0 Madagascar 167 152 51 42 97 34 49 28 40 19 20 15 44 13 15 9 Malawi 242 222 60 50 137 39 59 25 51 23 23 15 42 14 11 7 Malaysia 18 15 9 8 14 7 7 4 8 4 4 2 5 3 2 1 Maldives 100 88 9 8 68 7 1 0 42 5 0 0 13 3 0 0 27 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Under-five mortality rate (U5MR) with 90 per cent uncertainty interval Number of under-five deaths with 90 per cent (deaths per 1,000 live births) uncertainty interval (thousands)a Annual rate of reduction (ARR) 1990 2016 1990 2016 (per cent) 1990-2016 Under- Under- Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper Country U5MR U5MR ARR five five bound bound bound bound bound bound bound bound bound bound deaths deaths Mali 254 238 271 111 69 176 3.2 1.4 5.0 101 95 108 82 51 131 Malta 11 10 12 7 6 8 1.9 1.2 2.7 0 0 0 0 0 0 Marshall Islands 51 43 60 35 25 51 1.4 -0.2 2.9 0 0 0 0 0 0 Mauritania 117 105 129 81 45 147 1.4 -0.9 3.7 9 8 10 12 6 21 Mauritius 23 22 24 14 12 15 2.0 1.5 2.4 1 0 1 0 0 0 Mexico 46 42 50 15 14 15 4.4 4.0 4.7 110 102 120 34 32 36 Micronesia (Federated States of) 55 44 69 33 15 76 1.9 -1.1 4.9 0 0 0 0 0 0 Monaco 8 7 9 3 3 4 3.1 2.1 4.2 0 0 0 0 0 0 Mongolia 109 100 118 18 12 26 6.9 5.5 8.5 8 7 8 1 1 2 Montenegro 17 16 18 4 3 5 5.7 5.0 6.3 0 0 0 0 0 0 Morocco 80 75 85 27 20 37 4.1 2.9 5.4 58 54 62 19 14 26 Mozambique 248 229 268 71 53 97 4.8 3.6 5.9 149 138 161 78 58 106 Myanmar 116 106 126 51 39 64 3.2 2.2 4.2 128 118 140 48 37 60 Namibia 71 65 78 45 31 66 1.8 0.3 3.2 4 3 4 3 2 5 Nauru 58 36 95 35 21 56 2.0 -0.8 4.9 0 0 0 0 0 0 Nepal 141 132 150 35 29 42 5.4 4.6 6.1 99 93 105 20 16 24 Netherlands 8 8 9 4 4 4 3.0 2.8 3.2 2 2 2 1 1 1 New Zealand 11 11 12 5 5 6 2.8 2.3 3.3 1 1 1 0 0 0 Nicaragua 68 63 73 20 13 31 4.7 2.9 6.5 10 9 11 2 2 4 Niger 329 308 350 91 65 129 4.9 3.6 6.2 137 129 146 86 62 121 Nigeria 213 199 227 104 77 140 2.7 1.6 3.9 862 807 918 733 544 980 Niue 14 10 20 22 10 51 -1.8 -5.3 1.6 0 0 0 0 0 0 Norway 9 8 9 3 2 3 4.6 4.1 5.1 0 0 1 0 0 0 Oman 39 34 45 11 10 11 5.0 4.4 5.5 3 2 3 1 1 1 Pakistan 139 134 144 79 61 102 2.2 1.2 3.2 584 563 605 424 329 550 Palau 36 31 42 16 9 31 3.2 0.5 5.7 0 0 0 0 0 0 Panama 31 27 35 16 10 27 2.4 0.4 4.5 2 2 2 1 1 2 Papua New Guinea 88 80 97 54 32 94 1.9 -0.3 3.8 13 12 14 12 7 21 Paraguay 47 42 52 20 12 32 3.3 1.4 5.2 6 6 7 3 2 4 Peru 80 76 85 15 12 20 6.4 5.3 7.4 52 50 55 9 7 12 Philippines 58 54 62 27 20 37 2.9 1.8 4.1 116 109 124 64 48 87 Poland 17 17 18 5 5 5 5.0 4.9 5.2 10 10 10 2 2 2 Portugal 15 14 15 4 3 4 5.5 4.8 6.3 2 2 2 0 0 0 Qatar 21 19 22 9 8 9 3.4 3.0 3.9 0 0 0 0 0 0 Republic of Korea 16 15 17 3 3 4 5.9 5.6 6.2 10 10 11 2 1 2 Republic of Moldova 33 28 38 16 11 23 2.8 1.2 4.3 3 2 3 1 0 1 Romania 31 31 32 9 8 10 4.8 4.4 5.2 10 10 11 2 2 2 Russian Federation 22 21 22 8 6 9 4.0 3.2 4.8 45 45 46 14 12 17 Rwanda 151 142 160 39 25 60 5.2 3.5 7.0 48 45 51 14 9 22 Saint Kitts and Nevis 32 29 35 9 6 14 4.7 3.2 6.2 0 0 0 0 0 0 Saint Lucia 21 20 23 13 11 16 1.8 1.0 2.7 0 0 0 0 0 0 Saint Vincent and the Grenadines 24 23 26 17 14 20 1.5 0.7 2.3 0 0 0 0 0 0 Samoa 31 27 35 17 12 24 2.2 0.9 3.8 0 0 0 0 0 0 San Marino 11 9 14 3 1 6 5.2 2.2 8.2 0 0 0 0 0 0 Sao Tome and Principe 105 92 119 34 23 50 4.4 2.8 6.0 0 0 1 0 0 0 Saudi Arabia 45 36 55 13 7 25 4.8 2.0 7.2 25 21 31 8 5 16 Senegal 140 134 148 47 37 60 4.2 3.3 5.1 44 42 46 25 20 32 Serbia 28 28 29 6 5 7 6.1 5.6 6.5 4 4 4 1 0 1 Seychelles 17 15 18 14 11 18 0.6 -0.4 1.6 0 0 0 0 0 0 Sierra Leone 262 239 286 114 88 141 3.2 2.4 4.2 50 45 54 29 23 36 Singapore 8 7 8 3 3 3 3.9 3.5 4.3 0 0 0 0 0 0 Slovakia 15 15 15 6 6 6 3.5 3.3 3.7 1 1 1 0 0 0 28 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Sex-specific under-five Probability of mortality rate Infant Neonatal dying among Number of (deaths per 1,000 live births) mortality rate Number of mortality rate Number of children deaths among (deaths per infant deaths (deaths per neonatal deaths aged 5–14 children 1,000 live (thousands) a 1,000 live (thousands)a (deaths per aged 5–14 1990 2016 births) births) 1,000 children (thousands)a aged 5) Country Male Female Male Female 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 Mali 263 245 115 105 130 68 53 51 73 36 30 27 47 24 12 13 Malta 12 10 7 6 10 6 0 0 8 5 0 0 2 1 0 0 Marshall Islands 55 46 39 31 40 29 0 0 20 16 0 0 9 6 0 0 Mauritania 125 108 88 74 71 54 6 8 45 34 4 5 22 9 1 1 Mauritius 26 20 15 12 20 12 0 0 15 8 0 0 4 2 0 0 Mexico 49 42 16 13 37 13 88 29 22 8 54 18 6 3 12 6 Micronesia (Federated States of) 59 50 37 30 43 28 0 0 25 17 0 0 10 6 0 0 Monaco 9 7 4 3 6 3 0 0 4 2 0 0 2 1 0 0 Mongolia 124 93 21 14 77 15 5 1 30 10 2 1 19 4 1 0 Montenegro 18 16 4 4 15 4 0 0 11 2 0 0 2 1 0 0 Morocco 84 75 30 24 63 23 46 16 36 18 26 13 10 3 7 2 Mozambique 257 238 76 67 165 53 99 59 61 27 38 30 68 15 28 13 Myanmar 123 108 55 46 82 40 90 38 48 25 53 23 20 8 20 8 Namibia 76 66 49 41 48 32 3 2 26 18 1 1 17 11 1 1 Nauru 61 51 38 31 45 29 0 0 29 22 0 0 11 6 0 0 Nepal 141 141 37 32 98 28 69 16 59 21 42 12 29 5 15 3 Netherlands 9 7 4 3 7 3 1 1 5 3 1 0 2 1 0 0 New Zealand 13 10 6 5 9 5 1 0 4 3 0 0 3 1 0 0 Nicaragua 73 62 22 17 51 17 8 2 24 9 4 1 8 4 1 0 Niger 333 325 95 87 133 51 57 50 55 26 24 26 71 40 18 26 Nigeria 223 202 110 98 126 67 516 476 52 34 218 247 42 21 117 107 Niue 15 12 25 20 12 19 0 0 7 12 0 0 3 4 0 0 Norway 10 8 3 2 7 2 0 0 4 2 0 0 2 1 0 0 Oman 43 36 12 10 32 9 2 1 17 5 1 0 6 2 0 0 Pakistan 141 136 82 75 106 64 450 346 64 46 278 248 14 11 39 48 Palau 40 32 18 14 31 14 0 0 19 8 0 0 7 3 0 0 Panama 34 28 18 15 26 14 2 1 18 10 1 1 6 3 0 0 Papua New Guinea 94 82 59 50 64 42 9 9 31 24 5 5 15 9 2 2 Paraguay 50 43 22 18 37 17 5 2 22 11 3 2 7 4 1 0 Peru 84 76 17 14 57 12 37 7 28 8 18 5 11 4 6 2 Philippines 64 52 30 24 41 22 82 51 20 13 40 30 14 7 23 14 Poland 20 15 5 4 15 4 9 1 11 3 6 1 3 1 2 0 Portugal 16 13 4 3 12 3 1 0 7 2 1 0 4 1 1 0 Qatar 23 19 9 8 18 7 0 0 11 4 0 0 4 2 0 0 Republic of Korea 17 14 4 3 14 3 9 1 8 2 5 1 5 1 4 0 Republic of Moldova 37 29 18 14 27 14 2 1 19 12 2 0 5 2 0 0 Romania 35 28 10 8 25 8 8 1 15 4 5 1 5 2 2 0 Russian Federation 25 18 9 7 18 7 38 12 11 3 22 6 5 2 11 4 Rwanda 159 142 42 35 93 29 29 11 41 17 13 6 67 12 16 4 Saint Kitts and Nevis 35 28 10 8 26 8 0 0 19 6 0 0 5 2 0 0 Saint Lucia 24 19 15 12 18 12 0 0 13 9 0 0 4 2 0 0 Saint Vincent and the Grenadines 27 22 18 15 20 15 0 0 13 10 0 0 4 4 0 0 Samoa 34 28 19 16 26 15 0 0 16 9 0 0 6 4 0 0 San Marino 12 10 3 3 10 3 0 0 7 1 0 0 2 1 0 0 Sao Tome and Principe 111 98 37 30 67 26 0 0 26 15 0 0 20 9 0 0 Saudi Arabia 47 42 14 12 36 11 21 7 22 7 13 4 7 2 3 1 Senegal 147 134 51 43 72 34 23 18 40 21 13 11 37 16 9 7 Serbia 30 26 6 5 24 5 3 0 17 4 2 0 3 1 0 0 Seychelles 18 15 16 13 14 12 0 0 12 9 0 0 5 4 0 0 Sierra Leone 274 250 120 106 156 83 30 21 54 33 11 9 55 21 7 4 Singapore 8 7 3 3 6 2 0 0 4 1 0 0 2 1 0 0 Slovakia 17 13 6 5 13 5 1 0 9 3 1 0 3 1 0 0 29 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Under-five mortality rate (U5MR) with 90 per cent uncertainty interval Number of under-five deaths with 90 per cent (deaths per 1,000 live births) uncertainty interval (thousands)a Annual rate of reduction (ARR) 1990 2016 1990 2016 (per cent) 1990-2016 Under- Under- Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper Country U5MR U5MR ARR five five bound bound bound bound bound bound bound bound bound bound deaths deaths Slovenia 10 10 11 2 2 3 5.8 5.3 6.3 0 0 0 0 0 0 Solomon Islands 38 33 44 26 20 34 1.5 0.3 2.7 0 0 1 0 0 1 Somalia 181 151 221 133 73 243 1.2 -0.9 3.2 61 51 74 79 44 144 South Africa 57 51 65 43 37 50 1.1 0.4 1.9 63 56 71 51 43 59 South Sudan 256 213 301 91 56 144 4.0 2.0 6.0 67 56 78 38 24 61 Spain 9 9 9 3 3 4 3.9 3.4 4.4 4 4 4 1 1 2 Sri Lanka 21 21 21 9 8 11 3.1 2.6 3.6 8 7 8 3 3 3 State of Palestine 45 41 48 19 14 28 3.2 1.8 4.5 4 4 4 3 2 4 Sudan 131 122 142 65 53 80 2.7 1.9 3.6 106 99 115 83 67 102 Suriname 46 38 57 20 10 41 3.2 0.5 6.1 1 0 1 0 0 0 Swaziland 66 58 75 70 48 102 -0.3 -1.7 1.2 2 2 3 3 2 4 Sweden 7 7 7 3 3 3 3.4 3.1 3.6 1 1 1 0 0 0 Switzerland 8 8 8 4 4 4 2.7 2.4 3.0 1 1 1 0 0 0 Syrian Arab Republic 37 34 40 18 14 25 2.9 1.4 3.9 16 15 18 7 6 10 Tajikistan 107 96 120 43 26 74 3.5 1.4 5.4 22 20 25 11 7 18 Thailand 38 35 40 12 7 20 4.3 2.4 6.3 41 39 44 9 5 15 The former Yugoslav Republic of Macedonia 37 36 38 12 9 19 4.3 2.5 5.5 1 1 1 0 0 0 Timor-Leste 175 159 193 50 33 75 4.8 3.2 6.4 5 5 6 2 1 3 Togo 145 134 156 76 61 94 2.5 1.7 3.3 22 21 24 19 15 24 Tonga 22 18 26 16 11 26 1.1 -0.8 3.0 0 0 0 0 0 0 Trinidad and Tobago 30 25 36 19 9 43 1.8 -1.5 4.6 1 1 1 0 0 1 Tunisia 57 50 65 14 10 19 5.5 4.1 7.0 12 11 14 3 2 4 Turkey 74 70 79 13 12 13 6.8 6.5 7.1 104 97 111 16 16 17 Turkmenistan 86 75 100 51 22 109 2.0 -0.9 5.2 11 10 13 7 3 15 Tuvalu 57 48 68 25 15 43 3.1 0.9 5.3 0 0 0 0 0 0 Uganda 175 165 186 53 45 62 4.6 4.0 5.3 143 135 152 90 76 106 Ukraine 19 18 22 9 9 10 2.9 2.5 3.5 13 12 15 4 4 5 United Arab Emirates 17 14 19 8 7 8 3.0 2.2 3.7 1 1 1 1 1 1 United Kingdom 9 9 10 4 4 5 3.0 2.7 3.2 7 7 7 3 3 4 United Republic of Tanzania 179 169 189 57 46 71 4.4 3.6 5.2 192 181 203 117 96 146 United States 11 11 11 7 6 7 2.1 1.9 2.4 44 43 45 26 24 28 Uruguay 23 23 24 9 9 10 3.6 3.3 3.9 1 1 1 0 0 0 Uzbekistan 72 63 83 24 20 29 4.2 3.7 4.7 51 44 58 16 13 19 Vanuatu 36 30 42 28 18 42 1.0 -0.8 2.7 0 0 0 0 0 0 Venezuela (Bolivarian Republic of) 30 29 31 16 15 18 2.3 1.8 2.8 17 16 17 10 9 11 Viet Nam 51 47 56 22 19 26 3.3 2.6 3.9 99 91 108 34 29 41 Yemen 126 118 134 55 40 76 3.2 1.9 4.4 76 72 81 48 34 66 Zambia 182 171 194 63 46 85 4.1 2.9 5.3 63 59 67 39 28 52 Zimbabwe 75 69 82 56 44 72 1.1 0.2 2.1 28 26 30 30 23 38 Estimates of mortality among children under age 5 and children aged 5–14 by Sustainable Development Goal regionb Northern America and Europe 14 14 14 6 6 6 3.5 3.2 3.6 191 189 193 71 68 74 Northern America 11 11 11 6 6 7 2.0 1.9 2.3 47 46 48 28 26 30 Europe 15 15 16 5 5 6 4.0 3.8 4.3 144 142 146 43 40 46 Latin America and the Caribbean 55 54 57 18 17 19 4.4 4.1 4.7 652 633 672 187 178 202 Central Asia and Southern Asia 124 121 127 46 42 51 3.8 3.4 4.1 4,950 4,837 5,072 1,775 1,630 1,945 Central Asia 73 68 78 26 22 34 3.9 2.9 4.6 113 106 121 41 35 54 Southern Asia 126 123 129 47 43 52 3.8 3.4 4.1 4,836 4,724 4,958 1,734 1,586 1,902 Eastern Asia and South-Eastern Asia 57 55 60 16 15 18 4.9 4.4 5.2 2,312 2,200 2,439 495 455 548 Eastern Asia 51 47 55 10 8 11 6.4 5.8 7.0 1,446 1,336 1,570 180 159 206 South-Eastern Asia 72 70 75 27 24 31 3.8 3.3 4.3 866 842 891 314 281 360 Western Asia and Northern Africa 75 73 77 28 26 32 3.7 3.3 4.0 689 674 706 323 297 361 Western Asia 66 63 68 24 21 29 3.8 3.2 4.4 302 292 313 135 118 161 Northern Africa 84 81 87 33 29 37 3.6 3.1 4.1 388 376 400 188 167 215 Sub-Saharan Africa 183 179 187 79 73 89 3.2 2.8 3.5 3,787 3,714 3,869 2,777 2,570 3,113 Oceania 35 33 38 23 16 37 1.6 -0.2 3.1 18 17 19 15 10 24 Oceania excluding Australia and New Zealand 74 68 81 49 31 80 1.6 -0.3 3.3 15 14 16 13 9 22 Australia and New Zealand 10 9 10 4 4 4 3.4 3.2 3.6 3 3 3 1 1 2 Least developing countries 176 173 179 68 65 75 3.6 3.3 3.9 3,669 3,615 3,731 2,101 1,990 2,310 Landlocked developing countries 167 164 171 63 59 70 3.7 3.4 4.0 1,763 1,725 1,805 972 913 1,070 Small island developing States 79 76 82 42 37 51 2.4 1.7 2.9 94 91 97 51 45 62 World 93 92 95 41 39 44 3.2 2.9 3.4 12,598 12,426 12,801 5,642 5,409 6,043 30 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Sex-specific under-five Probability of mortality rate Infant Neonatal dying among Number of (deaths per 1,000 live births) mortality rate Number of mortality rate Number of children deaths among (deaths per infant deaths (deaths per neonatal deaths aged 5–14 children 1,000 live (thousands)a 1,000 live (thousands)a (deaths per aged 5–14 1990 2016 births) births) 1,000 children (thousands)a aged 5) Country Male Female Male Female 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 Slovenia 12 9 3 2 9 2 0 0 6 1 0 0 2 1 0 0 Solomon Islands 41 35 28 23 31 22 0 0 15 10 0 0 7 5 0 0 Somalia 189 173 139 126 109 83 37 50 45 39 16 24 38 24 8 10 South Africa 63 52 48 39 45 34 49 40 20 12 22 15 11 5 10 6 South Sudan 266 246 96 85 152 59 39 26 67 38 18 17 54 17 9 6 Spain 10 8 4 3 7 3 3 1 5 2 2 1 2 1 1 0 Sri Lanka 23 19 10 9 18 8 6 3 13 5 5 2 6 2 2 1 State of Palestine 47 42 21 18 36 17 3 3 22 11 2 2 7 3 0 0 Sudan 139 123 70 60 82 45 68 58 43 29 36 38 30 9 18 10 Suriname 51 41 22 18 40 18 0 0 22 11 0 0 7 3 0 0 Swaziland 71 60 76 65 50 52 2 2 22 21 1 1 19 11 0 0 Sweden 8 6 3 3 6 2 1 0 4 2 0 0 1 1 0 0 Switzerland 9 7 4 4 7 4 1 0 4 3 0 0 2 1 0 0 Syrian Arab Republic 40 34 19 16 30 14 13 6 17 9 7 4 7 3 3 2 Tajikistan 116 98 48 38 84 37 18 9 32 20 7 5 18 3 2 1 Thailand 43 33 14 11 31 11 33 8 21 7 23 5 8 3 10 3 The former Yugoslav Republic of Macedonia 39 35 13 11 34 11 1 0 17 8 1 0 3 1 0 0 Timor-Leste 182 167 54 46 132 42 4 2 57 22 2 1 30 8 1 0 Togo 154 135 82 70 89 51 14 13 43 26 7 7 39 23 4 5 Tonga 20 24 15 18 19 14 0 0 10 7 0 0 5 3 0 0 Trinidad and Tobago 33 27 20 17 26 17 1 0 20 13 0 0 4 3 0 0 Tunisia 60 53 15 12 44 12 9 2 27 8 6 2 7 3 2 1 Turkey 77 71 13 12 56 11 77 14 33 7 46 8 9 2 12 3 Turkmenistan 100 72 60 42 70 43 9 6 29 22 4 3 7 4 1 0 Tuvalu 61 53 28 23 44 21 0 0 30 17 0 0 10 5 0 0 Uganda 187 163 58 48 104 38 88 65 39 21 34 37 33 16 17 20 Ukraine 22 17 10 8 17 8 11 4 12 5 8 3 5 2 3 1 United Arab Emirates 19 15 9 7 14 7 1 1 8 4 0 0 6 1 0 0 United Kingdom 10 8 5 4 8 4 6 3 5 3 4 2 2 1 1 1 United Republic of Tanzania 185 172 60 53 108 40 118 84 41 22 47 46 31 12 23 20 United States 13 10 7 6 9 6 37 23 6 4 23 15 2 1 9 5 Uruguay 26 21 10 8 21 8 1 0 12 5 1 0 3 2 0 0 Uzbekistan 80 64 27 21 59 21 42 14 31 14 22 9 6 3 3 2 Vanuatu 38 33 30 25 29 23 0 0 15 12 0 0 7 5 0 0 Venezuela (Bolivarian Republic of) 33 27 18 15 25 14 14 8 13 10 7 6 5 3 2 2 Viet Nam 58 43 25 18 37 17 71 27 23 12 46 18 13 3 22 4 Yemen 131 120 59 51 88 43 55 37 44 27 28 23 21 6 8 4 Zambia 191 173 68 58 110 44 39 27 37 23 14 15 30 14 7 7 Zimbabwe 81 69 62 51 50 40 19 21 24 23 9 12 13 15 4 6 Estimates of mortality among children under age 5 and children aged 5–14 by Sustainable Development Goal regionb (continued) Northern America and Europe 16 12 6 5 12 5 159 60 7 3 98 39 3 1 42 15 Northern America 12 10 7 6 9 6 40 24 6 4 24 16 2 1 9 6 Europe 17 13 6 5 13 5 120 36 8 3 74 23 3 1 32 10 Latin America and the Caribbean 60 51 19 16 44 15 517 159 23 9 270 98 6 3 65 33 Central Asia and Southern Asia 122 126 46 46 88 37 3,530 1,433 56 27 2,277 1,044 19 7 611 245 Central Asia 81 64 30 23 60 23 93 36 28 13 45 21 8 3 9 4 Southern Asia 124 128 47 47 89 38 3,437 1,397 57 28 2,232 1,023 20 7 602 241 Eastern Asia and South-Eastern Asia 60 54 18 15 44 14 1,753 410 28 8 1,112 250 9 4 308 107 Eastern Asia 53 49 10 9 40 8 1,124 153 28 5 778 92 6 3 149 48 South-Eastern Asia 79 66 30 24 53 22 629 257 28 14 334 158 15 5 159 59 Western Asia and Northern Africa 78 71 31 26 55 23 513 258 31 15 285 173 11 5 83 43 Western Asia 69 62 26 22 50 20 230 111 28 13 131 72 9 4 34 18 Northern Africa 87 81 35 30 61 26 282 147 33 17 154 100 13 5 50 25 Sub-Saharan Africa 192 173 84 73 109 54 2,301 1,910 46 28 1,008 1,003 42 19 604 513 Oceania 38 32 25 21 27 18 14 12 14 10 7 7 6 4 3 2 Oceania excluding Australia and New Zealand 79 69 53 44 55 38 11 11 27 21 6 6 13 8 2 2 Australia and New Zealand 11 8 4 4 8 3 2 1 5 2 1 1 2 1 1 0 Least developed countries 183 168 73 63 109 48 2,306 1,498 52 26 1,138 834 40 15 581 388 Landlocked developing countries 176 159 68 59 101 43 1,083 674 48 26 530 404 39 15 284 193 Small island developing States 84 73 46 38 56 32 67 39 27 19 33 23 14 7 13 8 World 96 91 43 39 65 31 8,787 4,242 37 19 5,058 2,614 15 8 1,716 959 31 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Estimates of mortality among children under age 5 and children aged 5–14 by UNICEF regionb Under-five mortality rate (U5MR) with 90 per cent uncertainty interval Number of under-five deaths with 90 per cent (deaths per 1,000 live births) uncertainty interval (thousands) Annual rate of reduction (ARR) 1990 2016 1990 2016 (per cent) 1990–2016 Under- Under- Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper Region U5MR bound bound U5MR bound bound ARR bound bound five bound bound five bound bound deaths deaths Sub-Saharan Africa 181 177 185 78 73 88 3.2 2.8 3.5 3,893 3,820 3,977 2,860 2,655 3,198 West and Central Africa 199 193 205 95 83 110 2.9 2.3 3.3 2,042 1,981 2,109 1,756 1,547 2,042 Eastern and Southern Africa 164 160 169 61 57 69 3.8 3.3 4.1 1,851 1,807 1,900 1,104 1,026 1,241 Middle East and North Africa 66 64 67 24 22 27 3.9 3.3 4.2 558 544 573 237 216 270 South Asia 129 126 133 48 44 53 3.8 3.4 4.2 4,730 4,618 4,851 1,713 1,566 1,881 East Asia and Pacific 57 54 60 16 15 18 4.8 4.4 5.2 2,329 2,218 2,457 510 471 564 Latin America and Caribbean 55 54 57 18 17 19 4.4 4.1 4.7 652 633 672 187 178 202 North America 11 11 11 6 6 7 2.0 1.9 2.3 47 46 48 28 26 30 Europe and Central Asia 31 30 32 10 9 11 4.5 4.0 4.7 388 379 400 107 100 120 Eastern Europe and Central Asia 47 45 48 14 13 17 4.5 4.0 4.8 331 321 342 88 81 101 Western Europe 11 10 11 4 4 4 3.8 3.8 3.9 58 57 58 19 19 20 World 93 92 95 41 39 44 3.2 2.9 3.4 12,598 12,426 12,801 5,642 5,409 6,043 Estimates of mortality among children under age 5 and children aged 5–14 by World Health Organization regionb Under-five mortality rate (U5MR) with 90 per cent uncertainty interval Number of under-five deaths with 90 per cent (deaths per 1,000 live births) uncertainty interval (thousands) Annual rate of reduction (ARR) 1990 2016 1990 2016 (per cent) 1990–2016 Under- Under- Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper Region U5MR bound bound U5MR bound bound ARR bound bound five bound bound five bound bound deaths deaths Africa 178 174 182 77 71 86 3.2 2.8 3.5 3,764 3,692 3,844 2,720 2,512 3,043 Americas 44 42 45 14 14 15 4.3 4.0 4.5 699 681 719 215 205 230 Eastern Mediterranean 102 100 105 52 46 61 2.6 2.0 3.1 1,374 1,345 1,407 877 776 1,032 Europe 31 30 32 10 9 11 4.5 4.0 4.8 390 380 401 108 100 121 South-East Asia 119 116 122 39 36 42 4.3 4.0 4.6 4,626 4,513 4,746 1,407 1,297 1,521 Western Pacific 52 49 56 13 12 15 5.4 4.8 5.8 1,742 1,632 1,867 313 285 351 World 93 92 95 41 39 44 3.2 2.9 3.4 12,598 12,426 12,801 5,642 5,409 6,043 32 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Estimates of mortality among children under age 5 and children aged 5–14 by UNICEF regionb (continued) Sex-specific under-five Probability of mortality rate Infant Neonatal dying among Number of (deaths per 1,000 live births) mortality rate Number of mortality rate Number of children deaths among (deaths per infant deaths (deaths per neonatal deaths aged 5–14 children 1,000 live (thousands) 1,000 live (thousands) (deaths per aged 5–14 1990 2016 births) births) 1,000 children (thousands) aged 5) Region Male Female Male Female 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 Sub-Saharan Africa 190 171 84 73 108 53 2,369 1,968 46 28 1,044 1,041 42 18 621 522 West and Central Africa 208 189 101 89 116 63 1,211 1,184 49 31 531 589 42 23 293 321 Eastern and Southern Africa 173 155 66 56 101 43 1,158 784 44 25 513 452 41 14 328 201 Middle East and North Africa 68 63 26 22 50 20 427 198 28 14 239 135 10 4 67 33 South Asia 127 132 48 48 92 39 3,355 1,380 59 28 2,185 1,010 21 7 587 238 East Asia and Pacific 60 54 18 15 43 14 1,766 422 28 8 1,119 257 9 4 311 109 Latin America and Caribbean 60 51 19 16 44 15 517 159 23 9 270 98 6 3 65 33 North America 12 10 7 6 9 6 40 24 6 4 24 16 2 1 9 6 Europe and Central Asia 34 28 11 9 25 8 312 92 14 5 174 57 4 2 55 18 Eastern Europe and Central Asia 51 42 16 13 38 13 264 76 21 7 144 45 6 2 42 13 Western Europe 12 9 4 4 9 3 48 16 6 2 30 12 2 1 13 4 World 96 91 43 39 65 31 8,787 4,242 37 19 5,058 2,614 15 8 1,716 959 Estimates of mortality among children under age 5 and children aged 5–14 by World Health Organization regionb (continued) Sex-specific under-five Probability of mortality rate Infant Neonatal dying among Number of (deaths per 1,000 live births) mortality rate Number of mortality rate Number of children deaths among (deaths per infant deaths (deaths per neonatal deaths aged 5–14 children 1,000 live (thousands) 1,000 live (thousands) (deaths per aged 5–14 1990 2016 births) births) 1,000 children (thousands) aged 5) Region Male Female Male Female 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 Africa 186 168 82 71 107 52 2,296 1,879 46 27 1,010 993 41 18 602 505 Americas 47 40 16 13 35 12 556 183 18 8 295 114 5 3 75 38 Eastern Mediterranean 105 100 55 49 76 41 1,019 690 43 28 597 475 13 8 131 108 Europe 34 28 11 9 25 8 313 93 14 5 175 57 4 2 55 18 South-East Asia 117 120 39 39 84 32 3,261 1,136 53 23 2,088 817 20 6 643 217 Western Pacific 55 49 14 12 40 11 1,337 260 27 7 891 156 8 3 209 72 World 96 91 43 39 65 31 8,787 4,242 37 19 5,058 2,614 15 8 1,716 959 33 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Estimates of mortality among children under age 5 and children aged 5–14 by World Bank regionb Under-five mortality rate (U5MR) with 90 per cent uncertainty interval Number of under-five deaths with 90 per cent (deaths per 1,000 live births) uncertainty interval (thousands) Annual rate of reduction (ARR) 1990 2016 1990 2016 (per cent) 1990–2016 Under- Under- Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper Region U5MR U5MR ARR five five bound bound bound bound bound bound bound bound bound bound deaths deaths Low income 188 185 192 73 68 81 3.6 3.2 3.9 2,571 2,522 2,628 1,671 1,565 1,852 Middle income 91 89 93 38 36 41 3.4 3.0 3.6 9,861 9,693 10,050 3,903 3,673 4,237 Lower middle income 121 118 123 51 47 56 3.3 3.0 3.6 7,400 7,271 7,539 3,386 3,154 3,711 Upper middle income 52 49 54 14 13 15 5.0 4.7 5.3 2,462 2,351 2,589 518 493 557 Low and middle income 102 101 104 44 42 48 3.2 2.9 3.4 12,432 12,259 12,635 5,575 5,341 5,975 East Asia and Pacific 60 57 63 17 16 19 4.8 4.4 5.2 2,307 2,196 2,435 504 465 558 Europe and Central Asia 47 45 48 14 13 17 4.5 4.0 4.8 331 321 342 88 81 101 Latin America and the Caribbean 57 55 58 18 17 19 4.4 4.1 4.7 644 625 664 184 175 199 Middle East and North Africa 69 67 71 26 23 29 3.8 3.3 4.2 529 516 543 227 206 259 South Asia 129 126 133 48 44 53 3.8 3.4 4.2 4,730 4,618 4,851 1,713 1,566 1,881 Sub-Saharan Africa 181 177 185 78 73 88 3.2 2.8 3.5 3,891 3,817 3,974 2,859 2,654 3,196 High income 13 12 13 5 5 6 3.3 2.8 3.6 166 161 172 67 63 75 World 93 92 95 41 39 44 3.2 2.9 3.4 12,598 12,426 12,801 5,642 5,409 6,043 Estimates of mortality among children under age 5 and children aged 5–14 by United Nations Population Division regionb Under-five mortality rate (U5MR) with 90 per cent uncertainty interval Number of under-five deaths with 90 per cent (deaths per 1,000 live births) uncertainty interval (thousands) Annual rate of reduction (ARR) 1990 2016 1990 2016 (per cent) 1990–2016 Under- Under- Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper Region U5MR U5MR ARR five five bound bound bound bound bound bound bound bound bound bound deaths deaths More developed regions 13 13 13 6 5 6 3.4 3.2 3.6 202 200 205 75 72 78 Less developed regions 103 102 105 45 43 48 3.2 3.0 3.4 12,396 12,224 12,599 5,567 5,334 5,968 Least developed countries 176 173 179 68 65 75 3.6 3.3 3.9 3,669 3,615 3,731 2,101 1,990 2,310 Excluding least developed countries 88 86 89 37 34 40 3.3 3.0 3.6 8,727 8,560 8,915 3,466 3,235 3,780 Excluding China 116 115 118 50 48 54 3.3 3.0 3.4 10,995 10,859 11,149 5,399 5,165 5,797 Sub-Saharan Africa 183 179 187 79 73 89 3.2 2.8 3.5 3,787 3,714 3,869 2,777 2,570 3,113 Africa 165 162 169 72 67 81 3.2 2.7 3.4 4,175 4,100 4,258 2,965 2,760 3,307 Asia 89 88 91 32 30 35 3.9 3.6 4.2 7,563 7,405 7,742 2,404 2,259 2,590 Europe 15 15 16 5 5 6 4.0 3.8 4.3 144 142 146 43 40 46 Latin America and the Caribbean 55 54 57 18 17 19 4.4 4.1 4.7 652 633 672 187 178 202 Northern America 11 11 11 6 6 7 2.0 1.9 2.3 47 46 48 28 26 30 Oceania 35 33 38 23 16 37 1.6 -0.2 3.1 18 17 19 15 10 24 World 93 92 95 41 39 44 3.2 2.9 3.4 12,598 12,426 12,801 5,642 5,409 6,043 Definitions Under-five mortality rate: Probability of dying between birth and exactly 5 years of age, expressed per 1,000 live births. Infant mortality rate: Probability of dying between birth and exactly one year of age, expressed per 1,000 live births. Neonatal mortality rate: Probability of dying in the first 28 days of life, expressed per 1,000 live births. Probability of dying among children aged 5 –14: Probability of dying at age 5–14 years expressed per 1,000 children aged 5. Note: Upper and lower bounds refer to the 90 per cent uncertainty intervals for the estimates. The estimates generated by the United Nations Inter-agency Group for Child Mortality Estimation are not necessarily the official statistics of United Nations Member States, which may use alternative rigorous methods. a. Numbers of deaths are rounded to thousands. A zero indicates that the number of deaths is below 500. Unrounded numbers of deaths are available for download on childmortality.org. b. The sum of the number of deaths by region may differ from the world total because of rounding. 34 STATISTICAL TABLE (CONTINUED) Country, regional and global estimates of mortality among children under age 5 and children aged 5–14 Estimates of mortality among children under age 5 and children aged 5–14 by World Bank regionb (continued) Sex-specific under-five Probability of mortality rate Infant Neonatal dying among Number of (deaths per 1,000 live births) mortality rate Number of mortality rate Number of children deaths among (deaths per infant deaths (deaths per neonatal deaths aged 5–14 children 1,000 live (thousands) 1,000 live (thousands) (deaths per aged 5–14 1990 2016 births) births) 1,000 children (thousands) aged 5) Region Male Female Male Female 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 Low income 196 180 78 68 112 51 1,565 1,176 50 27 713 632 46 18 427 332 Middle income 93 89 39 36 65 29 7,084 3,009 39 19 4,261 1,944 14 6 1,253 613 Lower middle income 122 119 53 49 84 38 5,148 2,570 48 25 3,038 1,679 20 8 984 510 Upper middle income 55 49 15 13 41 12 1,936 439 26 7 1,223 265 7 3 269 103 Low and middle income 105 99 46 42 71 33 8,649 4,185 40 20 4,973 2,576 17 8 1,681 945 East Asia and Pacific 63 56 19 16 46 14 1,749 417 29 9 1,109 254 10 4 303 108 Europe and Central Asia 51 42 16 13 38 13 264 76 21 7 144 45 6 2 42 13 Latin America and the Caribbean 61 52 19 16 45 15 510 156 23 9 266 96 6 3 64 32 Middle East and North Africa 71 67 28 24 53 22 403 190 29 15 225 129 10 4 63 32 South Asia 127 132 48 48 92 39 3,355 1,380 59 28 2,185 1,010 21 7 587 238 Sub-Saharan Africa 190 171 84 73 108 53 2,367 1,967 46 28 1,043 1,040 42 18 621 522 High income 14 11 6 5 10 5 138 57 6 3 84 38 3 1 35 14 World 96 91 43 39 65 31 8,787 4,242 37 19 5,058 2,614 15 8 1,716 959 Estimates of mortality among children under age 5 and children aged 5–14 by United Nations Population Division regionb (continued) Sex-specific under-five Probability of mortality rate Infant Neonatal dying among Number of (deaths per 1,000 live births) mortality rate Number of mortality rate Number of children deaths among (deaths per infant deaths (deaths per neonatal deaths aged 5–14 children 1,000 live (thousands) 1,000 live (thousands) (deaths per aged 5–14 1990 2016 births) births) 1,000 children (thousands) aged 5) Region Male Female Male Female 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 1990 2016 More developed regions 15 12 6 5 11 5 168 64 7 3 103 41 3 1 45 17 Less developed regions 106 101 47 42 71 33 8,619 4,179 40 20 4,955 2,573 17 8 1,670 942 Least developed countries 183 168 73 63 109 48 2,306 1,498 52 26 1,138 834 40 15 581 388 Excluding least developed countries 89 86 38 35 63 28 6,313 2,681 38 18 3,817 1,738 13 6 1,089 554 Excluding China 119 113 52 47 79 37 7,529 4,035 43 23 4,196 2,487 20 9 1,533 897 Sub-Saharan Africa 192 173 84 73 109 54 2,301 1,910 46 28 1,008 1,003 42 19 604 513 Africa 173 157 77 67 101 50 2,584 2,057 44 26 1,162 1,103 36 17 653 538 Asia 90 88 33 31 65 26 5,513 1,954 41 18 3,520 1,367 14 5 952 370 Europe 17 13 6 5 13 5 120 36 8 3 74 23 3 1 32 10 Latin America and the Caribbean 60 51 19 16 44 15 517 159 23 9 270 98 6 3 65 33 Northern America 12 10 7 6 9 6 40 24 6 4 24 16 2 1 9 6 Oceania 38 32 25 21 27 18 14 12 14 10 7 7 6 4 3 2 World 96 91 43 39 65 31 8,787 4,242 37 19 5,058 2,614 15 8 1,716 959 35 Regional Classifications The regional classi fications referred to in the report are Sustainable Development Goal regions (see below). Aggregates presented in the statistical table for member organizations of the United Nations Inter-agency Group for Child Mortality may differ, and regional classi fications with the same name from different member organizations (e.g. “Sub-Saharan Africa”) may include different countries. Whether a country belongs to the group of Least developed countries (LDC), Landlocked developing countries (LLDC) and/or Small island developing States (SIDS) is indicated in the brackets after the country name. Sub-Saharan Africa South-Eastern Asia Angola (LDC), Benin (LDC), Botswana (LLDC), Burkina Faso (LDC, Brunei Darussalam, Cambodia (LDC), Indonesia, Lao People’s Demo- LLDC), Burundi (LDC, LLDC), Cabo Verde (SIDS), Cameroon, Central cratic Republic (LDC, LLDC), Malaysia, Myanmar (LDC), Philippines, African Republic (LDC, LLDC), Chad (LDC, LLDC), Comoros (LDC, Singapore (SIDS), Thailand, Timor-Leste (LDC, SIDS), Viet Nam SIDS), Congo, Côte d’Ivoire, Democratic Republic of the Congo (LDC), Djibouti (LDC), Equatorial Guinea (LDC), Eritrea (LDC), Ethiopia (LDC, LLDC), Gabon, Gambia (LDC), Ghana, Guinea (LDC), Guinea- Bissau (LDC, SIDS), Kenya, Lesotho (LDC, LLDC), Liberia (LDC), Mad- Latin America and the Caribbean agascar (LDC), Malawi (LDC, LLDC), Mali (LDC, LLDC), Mauritania Antigua and Barbuda (SIDS), Argentina, Bahamas (SIDS), Barbados (LDC), Mauritius (SIDS), Mozambique (LDC), Namibia, Niger (LDC, (SIDS), Belize (SIDS), Bolivia (Plurinational State of) (LLDC), Brazil, LLDC), Nigeria, Rwanda (LDC, LLDC), Sao Tome and Principe (SIDS), Chile, Colombia, Costa Rica, Cuba (SIDS), Dominica (SIDS), Domini- Senegal (LDC), Seychelles (SIDS), Sierra Leone (LDC), Somalia (LDC), can Republic (SIDS), Ecuador, El Salvador, Grenada (SIDS), Guatemala, South Africa, South Sudan (LDC, LLDC), Swaziland (LLDC), Togo Guyana (SIDS), Haiti (LDC, SIDS), Honduras, Jamaica (SIDS), Mexico, (LDC), Uganda (LDC, LLDC), United Republic of Tanzania (LDC), Nicaragua, Panama, Paraguay (LLDC), Peru, Saint Kitts and Nevis Zambia (LDC, LLDC), Zimbabwe (LLDC) (SIDS), Saint Lucia (SIDS), Saint Vincent and the Grenadines (SIDS), Suriname (SIDS), Trinidad and Tobago (SIDS), Uruguay, Venezuela (Bolivarian Republic of) Northern Africa and Western Asia Northern Africa Algeria, Egypt, Libya, Morocco, Sudan (LDC), Tunisia Oceania Australia and New Zealand Australia, New Zealand Western Asia Armenia (LLDC), Azerbaijan, Bahrain, Cyprus, Georgia, Iraq, Israel, Jordan, Kuwait, Lebanon, Oman, Qatar, Saudi Arabia, State of Palestine, Syrian Arab Republic, Turkey, United Arab Emirates, Yemen (LDC) Oceania (excluding Australia and New Zealand) Cook Islands (SIDS), Fiji (SIDS), Kiribati (LDC, SIDS), Marshall Islands (SIDS), Micronesia (Federated States of) (SIDS), Nauru (SIDS), Niue (SIDS), Palau (SIDS), Papua New Guinea (SIDS), Samoa (SIDS), Solo- Central and Southern Asia mon Islands (LDC, SIDS), Tonga (SIDS), Tuvalu (LDC, SIDS), Vanuatu (LDC, SIDS) Central Asia Kazakhstan (LLDC), Kyrgyzstan (LLDC), Tajikistan (LLDC), Turkmeni- stan (LLDC), Uzbekistan (LLDC) Europe and Northern America Northern America Canada, United States of America Southern Asia Afghanistan (LDC, LLDC), Bangladesh (LDC), Bhutan (LLDC), India, Iran (Islamic Republic of), Maldives (SIDS), Nepal (LDC, LLDC), Paki- stan, Sri Lanka Europe Albania, Andorra, Austria, Belarus, Belgium, Bosnia and Herzegovina, Bulgaria, Croatia, Czechia, Denmark, Estonia, Finland, France, Ger- many, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Lux- Eastern and South-Eastern Asia embourg, Malta, Monaco, Montenegro, Netherlands, Norway, Poland, Portugal, Republic of Moldova (LLDC), Romania, Russian Federation, Eastern Asia San Marino, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, The China, Democratic People’s Republic of Korea, Japan, Mongolia former Yugoslav Republic of Macedonia (LLDC), Ukraine, United King- (LLDC), Republic of Korea dom of Great Britain and Northern Ireland 36 Cover photo: © UNICEF/UN025707/Bongyereirwe Photo on page 2: © UNICEF/UN059878/Romeo Photo on page 7: © UNICEF/UN072236/Phelps Photo on page 10: © UNICEF/UN065191/Phelps Photo on page 14: © UNICEF/UN046741/Haque Photo on page 23: © UNICEF/UN025694/Bongyereirwe United Nations The United Nations Inter-agency Group for Child Mortality Estimation (UN IGME) was formed in 2004 to share data on child mortality, harmonize estimates within the UN system, improve methods for child mortality estimation, report on progress towards child survival goals and enhance country capacity to produce timely and properly assessed estimates of child mortality. The UN IGME includes the United Nations Children’s Fund, the World Health Organization, the World Bank Group and the United Nations Population Division of the Department of Economic and Social Affairs as full members. UN IGME’s independent Technical Advisory Group, comprising eminent scholars and independent experts in demography and biostatistics, provides technical guidance on estimation methods, technical issues and strategies for data analysis and data quality assessment. UN IGME updates its child mortality estimates annually after reviewing newly available data and assessing data quality. This report contains the latest UN IGME estimates of child mortality at the country, regional and global levels. Country-specific estimates and the data used to derive them are available at . Suggested citation: United Nations Inter-agency Group for Child Mortality Estimation (UN IGME), ‘Levels & Trends in Child Mortality: Report 2017, Estimates Developed by the UN Inter-agency Group for Child Mortality Estimation’, United Nations Children’s Fund, New York, 2017.