Global Causes of Death WHO

8/13/2019 Global Causes of Death WHO

http://slidepdf.com/reader/full/global-causes-of-death-who 1/72

WHO methods and data sources

for global causes of death 2000‐2011

Department of Health Statistics an d Information Systems WHO, Geneva

Ju ne 20 13

Global Health Estimates Technical Paper WHO/HIS/HSI/GHE/2013.3



i

Acknowledgments

This Technical Report was written by Colin Mathers, Gretchen Stevens and Doris Ma Fat with inputs and

assistance from Wahyu Retno Mahanani, Jessica Ho and Li Liu. Estimates of regional deaths by cause for

years 2000‐2011 were primarily prepared by Colin Mathers, Gretchen Stevens, Jessica Ho, Doris Ma Fat and Wahyu Retno Mahanani, of the Mortality and Burden of Disease Unit in the WHO Department of

Health Statistics and Information Systems, in the Health Systems and Innovation Cluster of the World

Health Organization (WHO), Geneva, drawing heavily on advice and inputs from other WHO

Departments, collaborating United Nations (UN) Agencies, and WHO expert advisory groups and

academic collaborators.

Many of the inputs to these estimates result from collaborations with Interagency Groups, expert

advisory groups and academic groups. The most important of these include the Interagency Group on

Child Mortality Estimation (UN‐IGME), the UN Population Division, the Child Health Epidemiology

Reference Group (CHERG), the Maternal Mortality Expert and Interagency Group (MMEIG), the

International Agency for Research on Cancer, WHO QUIVER, and the Global Burden of Disease 2010

Study Collaborating Group. While it is not possible to name all those who provided advice, assistance or data, both inside and outside WHO, we would particularly like to note the assistance and inputs

provided by Kirill Andreev, Diego Bassani, Bob Black, Ties Boerma, Phillipe Boucher, Freddie Bray, Tony

Burton, Harry Campbell, Doris Chou, Richard Cibulskis, Simon Cousens, Jacques Ferlay, Marta Gacic‐

Dobo, Richard Garfield, Alison Gemmill, Patrick Gerland, Peter Ghys, Philippe Glaziou, Danan Gu, Ken

Hill, Kacem Iaych, Mie Inoue, Robert Jakob, Dean Jamison, Prabhat Jha, Hope Johnson, Joy Lawn, Nan Li,

Li Liu, Rafael Lozano, Chris Murray, Lori Newman, Mikkel Oestergaard, Max Parkin, Margie Peden,

Francois Pelletier, Juergen Rehm, Igor Rudan, Lale Say, Emily Simons, Charalampos Sismanidis, Thomas

Spoorenberg, Karen Stanecki, Peter Strebel, Emi Suzuki, Tamitza Toroyan, Theo Vos, Tessa Wardlaw,

Richard White, John Wilmoth and Danzhen You.

Estimates and analysis are available at: http://www.who.int/gho/mortality_burden_disease/en/index.html

For further information about the estimates and methods, please contact [email protected]

In this series 1. WHO methods and data sources for life tables 1990‐2011 (Global Health Estimates Technical Paper

WHO/HIS/HSI/GHE/2013.1)

2. CHERG‐WHO methods and data sources for child causes of death 2000‐2011 (Global Health Estimates

Technical Paper WHO/HIS/HSI/GHE/2013.2) 3. WHO methods and data sources for global causes of death 2000‐2011 (Global Health Estimates

Technical Paper WHO/HIS/HSI/GHE/2013.3)



ii

Table of Contents

Acknowledgments .......................................................................................................................................... i

Table of Contents .......................................................................................................................................... ii

1 Introduction ……………………………………………………………………………………………………………………………………….1

2 Population and all‐cause mortality estimates for years 2000‐2011 ........................................................ 3

2.1 All‐cause mortality and population estimates ................................................................................. 3

2.2 Estimation of neonatal, infant and under‐5 mortality rates............................................................ 3

2.3 All‐cause mortality computed from civil registration data .............................................................. 4

2.4 All‐cause mortality projected from civil registration data ............................................................... 4

2.5 Countries with other information on levels of adult mortality ....................................................... 5

2.6 Mortality shocks – epidemics, conflicts and disasters ..................................................................... 6

3 Countries with useable death registration data ...................................................................................... 7 3.1 Data and estimates .......................................................................................................................... 7

3.2 Inclusion criteria for countries with high quality death registration data ....................................... 7

3.3 Redistribution of unknown sex/age and ‘garbage’ codes and adjustment for incomplete death

registration ..................................................................................................................................... 12

3.4 Mapping to GHE cause lists............................................................................................................ 12

3.5 Interpolation and extrapolation for missing country‐years........................................................... 14

3.6 Adjustment of specific causes ........................................................................................................ 14

3.7 Other national‐level information on causes of death ................................................................... 14

4 Child mortality by cause ........................................................................................................................ 19

4.1 Causes of under 5 death in countries with good death registration data ..................................... 19

4.2 Causes of neonatal death (deaths at less than 28 days of age) ..................................................... 19

4.3 Causes of child death at ages 1‐59 months –low mortality countries ........................................... 20

4.4 Causes of child death at ages 1‐59 months –high mortality countries ......................................... 20

4.5 Causes of child death for China and India ..................................................................................... 21

4.6 Inclusion of WHO‐CHERG estimates in Global Health Estimates 2000‐2011 ................................ 21

5 Methods for specific causes with additional information ..................................................................... 22

5.1 Tuberculosis ................................................................................................................................... 22

5.2 HIV/AIDS and sexually transmitted diseases ................................................................................. 22

5.3 Malaria ........................................................................................................................................... 22

5.4 Whooping cough ............................................................................................................................ 23

5.5 Measles .......................................................................................................................................... 23

5.6 Schistosomiasis .............................................................................................................................. 24



iii

5.7 Maternal causes of death .............................................................................................................. 24

5.8 Cancers ........................................................................................................................................... 24

5.9 Alcohol use and drug use disorders ............................................................................................... 25

5.10 Epilepsy .......................................................................................................................................... 25

5.11 Road injuries .................................................................................................................................. 25

5.11.1 Countries with death registration data............................................................................. 26

5.11.2 Countries with other sources of information on causes of death .................................... 26

5.11.3 Countries with populations less than 150 000 ................................................................. 26

5.11.4 Countries without eligible death registration data .......................................................... 26

5.12 Conflict and natural disasters ........................................................................................................ 28

6 Other causes of death for countries without useable data ................................................................... 30

7 Uncertainty of estimates ....................................................................................................................... 33

References…………………………………………………………………………………………………………………………………………….37

Annex Table A GHE cause categories and ICD‐10 codes ........................................................................... 43

Annex Table B First‐level categories for analysis of child causes of death ............................................... 48

Annex Table C Re‐assignment of ICD‐10 codes for certain neonatal deaths. .......................................... 49

Annex Table D Country groupings used for regional tabulations ............................................................. 51

D.1 WHO Regions and Member States ................................................................................................ 51

D.2 Countries grouped by WHO Region and average income per capita* .......................................... 52

D.3 World Bank income grouping* ...................................................................................................... 53

D.4 World Bank Regions ....................................................................................................................... 54

D.5 Millennium Development Goal (MDG) Regions ............................................................................ 55

Annex Table E Mapping of India MDS categories to GHE causes ............................................................. 56

Annex Table F Methods used for estimation of child and adult mortality levels, and causes of death, by

country, 2000‐2011 ........................................................................................................... 58

Annex Table G Methods used to estimate road traffic deaths for 182 participating countries ............... 64



World Health Organization Page 1

1 Introduction

Global, regional, and country statistics on population and health indicators are important for assessing

development and health progress and for guiding resource allocation. The demand is growing for timely

data to monitor progress in health outcomes such as child mortality, maternal mortality, life expectancy

and age‐ and cause‐specific mortality rates. Much of the current focus is on monitoring progress

towards the targets of the (health‐related) Millennium Development Goals (MDGs), including time series and country‐level estimates that are regularly updated. But increasingly, the demand is for

comprehensive estimates across the full spectrum, including noncommunicable diseases (NCDs) and

injuries.

WHO has previously published comprehensive estimates of deaths by region, cause, age and sex for

years 2000 and 2002 (1), 2001 (2), 2004 (3) and 2008 (4). Beginning with the 2004 estimates, WHO has

also released summary estimates of causes of death for its Member States (5). These successive single

year estimates did not form a time series, as each revision involved revisions to data and methods for a

range of inputs. To address the increasing demand for time series, for country‐level estimates, and for

comprehensive estimates across NCD and injury causes, as well as the more traditional priorities in

infectious and parasitic diseases, updated Global Health Estimates (GHE) are being released,

commencing with regional‐level estimates of deaths by cause, age and sex for years 2000‐2011 (6).

This technical paper documents the data sources and methods used for preparation of these regional‐

level cause of death estimates for years 2000‐2011. Annex Table A lists the cause of death categories

and their definitions in terms of the International Classification of Diseases, Tenth Revision (ICD‐10) (7 ).

These estimates are available for years 2000 and 2011 for selected regional groupings of countries (6),

defined in Annex Tables D, at http://www.who.int/healthinfo/global_health_estimates/en/.

Comprehensive estimates of mortality, causes of death, DALYs for diseases, injuries and risk factors

were released in December 2012 (8‐10) by the Institute of Health Metrics and Evaluation (IHME) as part

of the Global Burden of Disease 2010 study (GBD 2010). WHO was a collaborator in the study from 2007

to 2011, but did not endorse the final results, as it was unable to obtain full access to the results prior to

publication or to evaluate them. In some areas, the results of the GBD 2010 differ substantially from

existing analyses done by WHO and other United Nations agencies at global, regional and country levels.

In many other areas, the GBD 2010 results are updates that are broadly similar to previous WHO

analyses. Further work with IHME and expert groups is needed to examine the reasons for current

differences.

One of the six core functions of WHO is monitoring of the health situation, trends and determinants in

the world. Over the years it has cooperated closely with other UN partner agencies like UNICEF, UNAIDS,

UNFPA and the UN Population Division to collect and compile global health statistics. There are a

number of established UN multi‐agency expert group mechanisms for cross cutting topics such as child

mortality (the UN‐IGME including UNICEF/WHO/ UNPD/World Bank and the UN‐IGME Technical

Advisory Group) and child causes of death (CHERG, WHO/UNICEF), specific diseases such as HIV/AIDS

(UNAIDS Reference Group), maternal mortality (MMEIG including WHO/UNICEF/UNFPA/World Bank),

tuberculosis (WHO STAG), malaria (Malaria Reference Group and Roll Back Malaria‐ Malaria Monitoring

and Evaluation Reference Group).

These WHO Global Health Estimates provide a comprehensive and comparable set of cause of death

estimates from year 2000 onwards, consistent with and incorporating UN agency, interagency and WHO

estimates for population, births, all‐cause deaths and specific causes of death, including:

o most recent vital registration (VR) data for all countries where the VR data quality is assessed as

useable;




o updated and additional information on levels and trends for child and adult mortality in many

countries without good death registration data

o improvements in methods used for the estimation of causes of child deaths in countries without

good death registration data.

o Updated assessments of levels and trends for specific causes of death by WHO programs and

interagency groups. These include:

Tuberculosis –WHO

HIV – UNAIDS and WHO

Malaria – WHO

Vaccine‐preventable child causes – WHO

Other major child causes – WHO and CHERG

Maternal mortality –MMEIG

Cancers – IARC

Road traffic accidents – WHO

Conflict and natural disasters – WHO and the Collaborating Center for Research on the

Epidemiology of Disasters (CRED)

o GBD 2010 study estimates for other causes in countries without useable VR data or other

nationally representative sources of information on causes of death.

Because these estimates draw on new data and on the result of the GBD 2010 study, and there have

been substantial revisions to methods for many causes, these estimates for the years 2000‐2011 are not

directly comparable with previous WHO estimates for 2008 and earlier years. These are provisional

estimates and will be further revised in the process of extending the series to 2012 for release at

country level in late 2013. WHO and collaborators will continue to include new data and improve

methods, and it is anticipated that some causes will be substantially updated in the next revision.

These Global Health Estimates represent the best estimates of WHO, based on the evidence available to

it up until May 2013, rather than the official estimates of Member States, and have not necessarily been

endorsed by Member States. They have been computed using standard categories, definitions and

methods to ensure cross‐national comparability and may not be the same as official national estimates

produced using alternate, potentially equally rigorous methods. The following sections of this document

provide explanatory notes on data sources and methods for preparing mortality estimates by cause.




2 Population and all‐cause mortality estimates for years 2000‐2011

2.1 All‐cause mortality and population estimates

Life tables have been developed for all Member States for years 1990‐ 2011 starting with a systematic

review of all available evidence from surveys, censuses, sample registration systems, population

laboratories and vital registration on levels and trends in under‐five and adult mortality rates. Annex table F summarizes the methods used for preparing life tables. Data sources are documented in more

detail in Technical Paper 2013.1 (11).

In recent years, WHO has liaised more closely with the UN Population Division (on life tables for

countries, in order to maximize the consistency of UN and WHO life tables, and to minimize differences

in the use and interpretation of available data on mortality levels. For countries where WHO previously

predicted levels of adult mortality from estimated levels of child mortality, this update has taken into

account additional country‐specific sources of information on levels of adult mortality as reflected in the

life tables prepared by the UN Population Division for its World Population Prospects (WPP).

Total deaths by age and sex were estimated for each country by applying the WHO life table death rates

to the estimated de facto resident populations prepared by the UN Population Division in its 2010

revision (12). They may thus differ slightly from official national estimates for corresponding years. All‐

cause mortality and deaths by cause will be updated in the next WHO GHE revision to take account of

revisions to population estimates included in the WPP 2012 (released mid‐June 2013) (13).

2.2 Estimation of neonatal, infant and under‐5 mortality rates

Methods for estimating time series for neonatal, infant and under‐5 mortality rates have been

developed and agreed upon within the Inter‐agency Group for Child Mortality Estimation (UN‐IGME)

which is made up of WHO, UNICEF, UN Population Division, World Bank and academic groups. UN‐IGME

annually assesses and adjusts all available surveys, censuses and vital registration data, to then estimate

the country‐specific trends in under‐five mortality per 1000 live births (U5MR) over the past few

decades in order to predict the rates for the reference years (14). All data sources and estimates are

documented on the UN‐IGME website.1 For countries with complete recording of child deaths in death

registration systems, these are used as the source of data for the estimation of trends in neonatal, infant

and child mortality. For countries with incomplete death registration, all other available census and

survey data sources, which meet quality criteria, are used. UN‐IGME methods are documented in a

series of papers published in a collection in 2012 (15).

For data from civil registration, the neonatal mortality per 1000 live births (NMR) is calculated as the

number of neonatal death divided by the live births reported from the country when available. For

household surveys, child and neonatal mortality rates are calculated from the full birth history (FBH)

data, where women are asked for the date of birth of each of their children, whether the child is still

alive, and if not the age at death FBH data, collected by all Demographic Health Surveys (DHS), allow the

calculation of child mortality indicators for specific time periods in the past; DHS publishes child

mortality estimates for five 5‐year periods before the survey, that is, 0 to 4, 5 to 9, 10 to 14 etc.

A database consisting of pairs of NMRs and U5MRs was compiled. For a given year, NMR and U5MR

were included in the database when data for both of these were available. To ensure consistency with

1 www.childmortality.org




U5MR estimates produced by UN‐IGME, U5MR and NMR data points were rescaled for all years to

match the UN‐IGME estimates.

For countries where child mortality is strongly affected by HIV, the NMR was estimated initially using

neonatal and child mortality observations for non‐AIDS deaths, calculated by subtracting from total

death rates the estimated HIV death rates in the neonatal and 1‐59 month periods respectively, and

then AIDS neonatal deaths be added back on to the non‐HIV neonatal deaths to compute the total estimated neonatal death rate.

The following statistical model was used to estimate NMR:

log(NMR/1000) = α0+ β1*log(U5MR/1000) + β2*([log(U5MR/1000)] 2)

with additional random effect intercept parameters for both country and region. For countries with

good vital registration data covering the period 1990‐2011, random effects parameters for slope or

trend parameters were also added. Based on predictive performance evaluation using ten‐fold cross‐

validation, the statistical model fitted to data point for 1990 onwards were retained and only the most

recent data point from each survey was included (16).

2.3 All‐cause mortality computed from civil registration data

For 133 Member States with vital registration and sample vital registration systems, demographic

techniques (such as Brass Growth–Balance method, Generalized Growth–Balance method or Bennett–

Horiuchi method) were first applied to assess the level of completeness of recorded mortality data in

the population above five years of age and then those mortality rates were adjusted accordingly. The

proportion of all deaths which are registered in the population covered by the vital registration system

(referred to as completeness) has been estimated by WHO and is given for the latest available years in

the annex table.

Where vital registration data for all the reference years were available, the age specific mortality rates,

adjusted for completeness if necessary were used directly to construct the life tables. Death registration data up to and including year 2011 were available for 53 Member States.

2.4 All‐cause mortality projected from civil registration data

For another 60 Member States where vital registration data for 2011 was not available, life table

parameters were projected from those for available data years from 1985 onwards. Adjusted levels of

child mortality (5q0) and adult mortality (45q15), excluding HIV/AIDS deaths where necessary, were used

to estimate levels of two life table parameters (l 5, l 60) for each available year. The life table parameter l 60

was projected forward to 2011 using a weighted regression model giving more weight to recent years

(using an exponential weighting scheme such that the weight for each year t was 25% less than the

weight for year t +1). For Member States with a total population less than 750,000 or where the root

mean square error from this regression was greater than or equal to 0.011, a shorter‐term trend was

estimated by applying a weighting factor with 50% annual exponential decay. These projected values of

l 60, together with values of l 5 based on 5q0 from UN‐IGME were then applied to a modified logit life table

model, using the most recent national data as the standard, to predict the full life tables in the reference

years (17 ). Where necessary, HIV/AIDS death rates were then added to total mortality rates.

For two small countries without available death registration data, Andorra and Monaco, life tables were

based on mortality rates from neighbouring regions of Spain and France, respectively.




2.5 Countries with other information on levels of adult mortality

For 81 Member States without useable death registration data, assessments of mortality rates for ages 5

and over were based on life table analyses of the UN Population Division (12). The sources of available

data used in the WPP are listed elsewhere (18). Annual age‐sex‐specific death rates for years 1990‐2011

were interpolated from the WPP life tables, where necessary first subtracting out conflict and disaster

deaths occurring in each specific 5‐year time period. Annual estimates for conflict and disaster deaths

were then added back as described below.

For 39 of these Member States, with high levels of HIV mortality, the UN Population Division explicitly

estimated HIV deaths in preparing life table time series. For these Member States, HIV‐free mortality

rates were computed for interpolation of annual death rates (making use of unpublished supplementary

tabulations provided by the UN Population Division for estimated HIV deaths by age and sex in these

countries). The latest estimates of annual HIV death rates prepared by UNAIDS (19) were then added

back to the annual mortality rates to compute total all‐cause death rates by year. The high‐HIV countries

for which this method was used are identified in the Annex Table F.

For six countries, additional data inputs for the most recent period were also taken into account based on provisional analyses for the WPP 2012 provided by the UN Population Division (20). Data sources for

these countries are listed in the Annex Table F, and the following notes provide an overview of the

analyses used.

Afghanistan

The 2012 revision of child mortality estimates for Afghanistan by UN‐IGME took into account data from

the 2010 Afghanistan Mortality Survey (21) and the 2011 UNICEF MICS4 survey (22).

Adjusted estimates of adult mortality (45q15) derived from

recent household deaths data from the 2010 Afghanistan Mortality Survey (AMS);

parental orphanhood from the 2010 AMS (excluding the Southern region);

siblings deaths from the 2010 AMS (excluding the Southern region) adjusted for age

misreporting and recall biases

were also considered, but the implied low level of adult mortality could not be reconciled with

intercensal survival between the 1979 Afghan census and 2003‐05 Afghan household listing, or with

population estimates from 2003‐05 Household listing and more recent surveys in 2007‐2008 and 2011,

or with intercensal estimates of the trends in fertility, and international migration based on UNHCR

statistics on the number of Afghan refugees. Additionally, they would imply that Afghan adult mortality

levels were substantially lower than those in neighboring countries.

As a result, the life tables for Afghanistan are based on provisional analyses by UN Population Division

using the West model of the Coale‐Demeny Model Life Tables with three parameters: (1) estimates of

infant mortality, (2) estimates of child mortality, and (3) adjusted estimates of adult mortality (45q15) derived from (a) recent household deaths data from the 1979 census; (b) implied relationship between

child mortality and adult mortality based on the UN South Asian and West model of the Coale‐Demeny

Model Life Tables, and (c) levels of adult mortality based on sample registration data from neighboring

countries for recent years.




China Life tables for years since 2000 have been revised to take into account a faster rate of decline for adult

mortality than previously projected in the World Population Prospects 2010 revision. Unpublished

analyses of the China 2010 census data on adult mortality by UN Population Division have adjusted for

under‐reporting of deaths resulting in estimates of adult mortality rates for 2010 quite similar to those

reported by the China Disease Surveillance Points System (23). Egypt Life tables have been based on official estimates of life expectancy available through 2012, and in turn

derived from death registration data for Egypt. The age pattern of mortality is based on official life

tables for various years from 1960 to 2010 adjusted for infant and child mortality as estimated by UN‐

IGME, and adult mortality.

Saudi Arabia The World Population Prospects 2010 revision based estimates of adult mortality for Saudi Arabia using

model life tables with estimates of child mortality as input. Estimates of adult mortality have been

provisionally updated using adjusted death rates by age and sex from the 1999 Demographic Survey,

2004 Census and 2007 Demographic Survey adjusted for infant and child mortality, and old‐age mortality. Life tables based on annual deaths from the 2000 Demographic Survey, as well as on 2005 and

2009 registered deaths were also considered.

South Sudan and Sudan The former Sudan became two countries, South Sudan and Sudan, on 9 July 2011. Previously published

WHO and UN life tables refer to the former Sudan. Life tables for the two Member States of South

Sudan and Sudan are based on provisional analyses of population and mortality rates for the territories

corresponding to the current South Sudan and Sudan over the period 1990 to 2011.

Infant and child mortality for South Sudan and Sudan are derived from UN‐IGME estimates published in

2012 (14). Life tables are based on provisional unpublished analyses of the UN Population Division,

deriving adult mortality rates from estimates of infant and child mortality by assuming that the age pattern of mortality conforms to the North model of the Coale‐Demeny Model Life Tables. The

demographic impacts of AIDS and conflict have also been factored into the mortality estimates.

2.6 Mortality shocks – epidemics, conflicts and disasters

Country‐specific estimates of deaths for organized conflicts and major natural disasters were prepared

for years 1990‐2011 using data and methods documented in Section 5.12. For country‐years where total

death rates from these conflicts and disasters exceeded 1 per 10,000 population, these deaths were

added to the life table death rates for the relevant year.

The revised WHO estimates for conflict deaths were taken into account in preparing final life tables for

Member States for years 1990‐2011 as follows. For country‐years where death rates from conflict or disasters exceeded 1 per 10,000 population, the estimated annual age‐sex‐specific conflict deaths were

added to the life table death rates for the relevant year. In cases of extended conflicts where death rates

fluctuated above and below 1 per 10,000, only the death rate in excess of 1 per 10,000 was added to

relevant years.

Measles outbreaks and epidemics were identified as described in Section 5.5 below and similarly added

to all‐cause envelopes for relevant country‐years.




3 Countries with useable death registration data

3.1 Data and estimates

Cause‐of ‐death statistics are reported to WHO on an annual basis by country, year, cause, age and sex.

Most of these statistics can be accessed in the WHO Mortality Database (24). The number of countries

reporting data using ICD‐10 has continued to increase. For these estimates, a total of 114 countries had data covering 80% or more of deaths in the country, of which 93 countries were reporting data coded to

the third or fourth character of ICD‐10 and 59 countries had data for years 2010 or 2011.

For countries with a high‐quality vital registration system including information on cause of death, we

used the vital registration data recorded in the WHO Mortality Database. We analyzed the data using

the following steps:

1) application of inclusion criteria to select countries with high‐quality vital registration data;

2) extraction of deaths by cause group, with a short or a detailed cause list used depending on

the ICD revision used in each country‐year;

3) redistribution of deaths of unknown sex/age and deaths assigned to garbage codes and

adjustment for incomplete registration of deaths in some countries; 4) interpolation/extrapolation of number of deaths for missing country‐years;

5) adjustments for certain specific causes using additional information to adjust for over‐ or

under‐reporting

6) scaling of total deaths by age and sex to previously estimated WHO all‐cause envelopes for

years 2000‐2011

Details are provided below.

3.2 Inclusion criteria for countries with high quality death registration data

We applied the following inclusion criteria to data in the WHO mortality database:

At least five years of data are available during 1998‐present;

The data are available for 5‐year age groups to ages 85 and over;

The data are for a country whose population in 2008 was greater than 500,000;

The data are for a country that is currently a WHO Member State;

The data fulfill quality criteria pertaining to garbage codes and completeness, as described

below.

For 131 Member States with vital registration systems who have provided summary data to WHO,

demographic techniques (such as Brass Growth–Balance method, Generalized Growth–Balance method

or Bennett– Horiuchi method) were first applied to assess the level of completeness of recorded

mortality data in the population above five years of age. We then calculated the proportion of deaths with underlying cause coded to a short list of so‐called “garbage” codes:

symptoms, signs and ill‐defined conditions (ICD10 codes R00‐R99),

injuries undetermined whether intentional or unintentional (ICD10 Y10‐Y34, Y87.2),

ill‐defined cancers (C76, C80, and C97), and

ill‐defined cardiovascular diseases ( I47.2, I49.0, I46, I50, I51.4, I51.5, I51.6, I51.9 and I70.9).




Table 3.1. Characteristics of useable country vital registration data

(Only countries fulfilling the first four inclusion criteria listed above are included in this table. ICD‐10

codes included in the “garbage” category are given in the text above).

Country First year

1998+

available

Last year

available

Average

usability

2000+

Range of

completeness

Range of

garbage

fraction

Notes

Albania 1998 2004 55% 67% 71% 18% 20% Excluded due to low

usability

Argentina 1998 2010 79% 100% 100% 20% 22% Excluded due to high

proportion garbage

Armenia 1998 2011 66% 66% 81% 3% 6% Excluded due to low

usability

Australia 1998 2011 95% 100% 100% 5% 6%

Austria 1998 2011 90% 100% 100% 1% 14%

Azerbaijan 1998 2007 84% 81% 96% 2% 34% Excluded due to high

proportion garbage

Belarus 1998 2009 88% 99% 100% 10% 13% Summarized cause list

used

Belgium 1998 2009 88% 100% 100% 12% 15%

Brazil 1998 2010 76% 87% 91% 12% 21%

Bulgaria 1998 2011 79% 100% 100% 16% 28% Excluded due to high

proportion garbage

Canada 1998 2009 94% 100% 100% 6% 8%

Chile 1998 2009 94% 100% 100% 6% 11%

Colombia 1998 2009 89% 93% 96% 6% 8%

Costa Rica 1998 2011 87% 90% 95% 4% 7%

Croatia 1998 2011 87% 98% 100% 8% 17%

Cuba 1998 2010 90% 96% 98% 1% 9%

Cyprus 2004 2011 73% 90% 91% 16% 24%

Czech Republic 1998 2011 88% 99% 100% 10% 15%

Denmark 1998 2011 87% 100% 100% 12% 14%

Ecuador 1998 2010 59% 72% 73% 16% 23% Excluded due to low

usability

Egypt 2000 2011 61% 99% 100% 32% 41% Excluded due to low

usability

El Salvador 1998 2009 58% 75% 75% 18% 25% Excluded due to low

usability

Estonia 1998 2011 94% 100% 100% 5% 8%

Finland 1998 2011 97% 100% 100% 2% 3%




Country First year

1998+

available

Last year

available

Average

usability

2000+

Range of

completeness

Range of

garbage

fraction

Notes

France 1998 2009 85% 100% 100% 14% 16%

Georgia 1998 2010 53% 78% 83% 7% 69% Excluded due to low

usability

Germany 1998 2011 87% 100% 100% 11% 14%

Greece 1998 2010 75% 100% 100% 24% 27% Excluded due to high

proportion garbage

Guatemala 1998 2009 73% 89% 90% 12% 22% Excluded due to high

proportion garbage

Hungary 1998 2011 94% 99% 100% 4% 7%

Iceland 1998 2009 94% 100% 100% 5% 6%

Ireland 1998 2010 94% 100% 100% 5% 8%

Israel 1998 2010 90% 100% 100% 8% 14%

Italy 1998 2010 90% 100% 100% 8% 12%

Japan 1998 2011 89% 100% 100% 9% 13%

Kazakhstan 1998 2010 83% 84% 89% 3% 11% Summarized cause list

used

Kuwait 1998 2011 87% 98% 98% 9% 14%

Kyrgyzstan 1998 2010 90% 91% 95% 3% 8%

Latvia 1998 2010 92% 99% 100% 5% 11%

Lithuania 1998 2010 94% 99% 100% 2% 6%

Mauritius 1998 2011 90% 100% 100% 8% 15%

Mexico 1998 2010 95% 100% 100% 5% 6%

Montenegro 2000 2009 70% 93% 93% 23% 28% Excluded due to low

usability

Netherlands 1998 2011 86% 100% 100% 13% 15%

New Zealand 1998 2009 97% 100% 100% 3% 4%

Norway 1998 2011 89% 100% 100% 11% 12%

Panama 1998 2009 80% 84% 91% 8% 14%

Philippines 1998 2008 83% 91% 93% 10% 13%

Poland 1999 2011 74% 100% 100% 25% 28% Excluded due to high

proportion garbage

Portugal 1998 2011 82% 100% 100% 17% 22%

Qatar 2004 2009 74% 100% 100% 22% 32% Excluded due to high

proportion garbage




Country First year

1998+

available

Last year

available

Average

usability

2000+

Range of

completeness

Range of

garbage

fraction

Notes

Republic of

Korea

1998 2011 85% 90% 100% 13% 21%

Republic of

Moldova

1998 2011 88% 89% 91% 2% 7%

Romania 1998 2011 92% 99% 100% 0% 8%

Russian

Federation

1998 2010 95% 100% 100% 4% 6% Summarized cause list

used

Serbia 1998 2011 72% 84% 89% 12% 18%

Singapore 1998 2011 74% 74% 84% 2% 4%

Slovakia 1998 2010 94% 100% 100% 4% 11%

Slovenia 1998 2010 89% 99% 100% 9% 12%

South Africa 1998 2009 68% 81% 88% 19% 32% Excluded due to low

usability

Spain 1998 2011 89% 100% 100% 9% 13%

Sri Lanka 1998 2006 55% 74% 74% 23% 32% Excluded due to low

usability

Sweden 1998 2010 89% 100% 100% 10% 12%

Switzerland 1998 2010 89% 100% 100% 10% 13%

TFYR

Macedonia

1998 2010 84% 96% 98% 9% 15%

Thailand 1998 2006 48% 78% 88% 39% 54% Excluded due to low

usability

Trinidad and

Tobago

1998 2008 95% 100% 100% 2% 5%

Ukraine 1998 2011 96% 100% 100% 3% 6% Summarized cause list

used

United

Kingdom

1998 2010 93% 100% 100% 6% 8%

United States

of America

1998 2008 93% 100% 100% 7% 10%

Uruguay 1998 2009 83% 100% 100% 16% 17%

Uzbekistan 1998 2005 83% 85% 87% 2% 6% Summarized cause list

used for some years

Venezuela

(Bolivarian

Republic of)

1998 2009 86% 93% 95% 7% 9%




A summary usability score was calculated as follows:

(Percent Usable) = Completeness (%) * (1 ‐ Proportion Garbage)

All countries with a mean percent usable below 70% during the period 2000 to latest available year

were excluded (see Table 3.1).

The quality of cause‐of ‐death coding was further investigated in the remaining countries. The proportion of deaths assigned to an expanded list of ill‐defined causes (Table 3.2) was calculated for each year in

the period 2000‐2011. For the period 2005‐2011 countries had reported an average of 5 years of data.

Data from a country were excluded if the average proportion of ill‐defined causes was above 25% for

2005‐2011 (if available) or 2000‐2004 (if more recent data were not available). Based on this analysis,

data from Argentina, Azerbaijan, Bulgaria, Greece, Guatemala, Poland, and Qatar were excluded (Table

3.1).

Table 3.2. Expanded list of garbage codes ICD‐10 code(s) Description

A40‐A41 Streptococcal and other septicaemia C76, C80, C97 Ill‐defined cancer sites

D65 Disseminated intravascular coagulation [defibrination syndrome]

E86 Volume depletion

I10 Essential (primary) hypertension

I269 Pulmonary embolism without mention of acute cor pulmonale

I46 Cardiac arrest

I472 Ventricular tachycardia

I490 Ventricular fibrillation and flutter

I50 Heart failure

I514 Myocarditis, unspecified I515 Myocardial degeneration

I516 Cardiovascular disease, unspecified

I519 Heart disease, unspecified

I709 Generalized and unspecified atherosclerosis

I99 Other and unspecified disorders of circulatory system

J81 Pulmonary oedema

J96 Respiratory failure, not elsewhere classified

K72 Hepatic failure, not elsewhere classified

N17 Acute renal failure

N18 Chronic renal failure N19 Unspecified renal failure

P285 Respiratory failure of newborn

Y10‐Y34, Y872 External cause of death not specified as accidentally or purposely inflicted




3.3 Redistribution of unknown sex/age and ‘garbage’ codes and adjustment

for incomplete death registration

First, deaths of unknown sex pro‐rata within cause‐age groups of known sexes were redistributed, and

then deaths of unknown age pro‐rata within cause‐sex groups of known ages. Deaths coded to garbage

codes were reassigned using previously published methods (25). We redistributed deaths coded to

symptoms, signs and ill‐defined conditions pro‐rata to all non‐injury causes of death, and injuries with undetermined intent pro‐rata to all injury causes of death. Cancers with unspecified site were

redistributed pro‐rata to all sites excluding liver, pancreas, ovary, and lung. Additionally, we

redistributed cancer of uterus, part unspecified (C55) pro‐rata to cervix uteri (C53) and corpus uteri

(C54). Ill‐defined cardiovascular causes were redistributed to ischaemic heart disease and other

cardiovascular causes of death. Finally, the total number of deaths was adjusted for incomplete

recording of deaths using the completeness estimates described in Section 3.2.

3.4 Mapping to GHE cause lists

Included vital registration data were coded according to ICD9, ICD10, or one of several abbreviated

cause lists derived from ICD9 or ICD10. Total deaths by cause, age and sex were mapped to the GHE cause list (Annex Table A). We used the complete cause list in Annex Table A if the data were coded

using 3‐ or 4‐digit ICD‐10 codes. In other cases, we extracted the number of deaths by cause, age and

sex, using only the broad cause categories listed in Table 3.3. This shortlist in Table 3.3 was used for all

data from the Philippines.

For Russia, Belarus and Ukraine, HIV deaths recorded in the death registration data were substantially

miscoded to tuberculosis (GHE3), lower respiratory infections (GHE39), other infectious diseases

(GHE37), lymphomas and multiple myeloma (GHE76), other malignant neoplasms (GHE78), and

endocrine, blood and immune disorders (GHE81). Deaths in these categories falling in the characteristic

HIV age pattern were recoded to HIV (GHE10), according to the age‐sex‐specific HIV mortality estimates

from UNAIDS (refer Section 5.2).

For countries with deaths data grouped by the shortlist in Table 3.3, shortlist categories were expanded

to the full cause list using the cause‐fraction distribution within each shortlist category by year, age, sex

and GBD 2010 region from the GBD 2010 study results (26).

Coding of natural causes of death for neonates varies a great deal across countries. Some countries code

these deaths to the ‘P chapter’ (conditions originating in the perinatal period) while others use a

combination of P codes and other codes as well. In some instances the age of death is not always taken

into account. Some conditions, such as septicaemia and pneumonia, have specific codes within P00–P96

which should be used for neonates (0–27 days). For countries with vital registration data, we have

recoded all the deaths aged 0–27 days from natural causes that were initially coded outside the ‘P

chapter’ to codes in the ‘P chapter’ whenever possible. In a number of countries, neonatal septicaemia

(P36) is frequently assigned to A40 and A41 (septicaemia). In this case we have recoded them back to

P36, thus identifying more deaths due to causes originating in the perinatal period.




Table 3.3. Short cause list used for vital registration data coded using ICD‐9 or ICD‐10 abbreviated cause

lists

GHE

code Shortlist cause category

1 I. Communicable, maternal, perinatal and nutritional conditions

3 A1. Tuberculosis

9 A3. HIV/AIDS

20 A. Infectious and parasitic diseases

38 B. Respiratory infections

39 B1. Lower respiratory infections

42 C. Maternal conditions

49 D. Neonatal conditions

60 II. Noncommunicable diseases

61 A. Malignant neoplasms

62 A1. Mouth and oropharynx cancers

63 A2. Oesophagus cancer

64 A3. Stomach cancer

65 A4. Colon and rectum cancers

66 A5. Liver cancer

68 A7. Trachea, bronchus and lung cancers

70 A9. Breast cancer

71 A10. Cervix uteri cancer

72 A13. Prostate cancer

80 C. Diabetes mellitus

82+94 E/F. Mental and neurological disorders

110 H. Cardiovascular diseases

117 I. Respiratory diseases

121 J. Digestive disorders

126 K. Genitourinary diseases

140 N. Congenital anomalies

151 III. Injuries

152 A. Unintentional injuries

153 A1. Road injury

160 B. Intentional injuries

161 B1. Self ‐harm

162 B2. Interpersonal violence

163 B3. Collective violence and legal intervention




3.5 Interpolation and extrapolation for missing country‐years

For many countries, data were missing for some years. In order to create a continuous time‐series of

data, we interpolated mortality rates for each country and cause, and then extrapolated up to three

years of data at the beginning and end of the data series. To interpolate, a logistic regression was fitted

for each missing country‐sex‐cause group, using death rates six years prior and six years after the

missing data year as the dependent variable and year as the independent variable. In some cases, few deaths were recorded for a specific country‐sex‐cause group and the logistic regression did not

converge. In that case, the death rate was estimated as the average rate in the three years prior and

three years following the missing data year. To extrapolate for up to three years, a logistic regression

was fitted to the first or the final six years of data (including interpolated estimates) for each country‐

sex‐cause. Again, if the logistic regression did not converge due to the small number of deaths recorded,

the death rate was estimated as the average of the first or last three years’ death rates.

3.6 Adjustment of specific causes

Estimates for HIV deaths were compared with UNAIDS/WHO estimates for 46 countries where fewer

HIV deaths were recorded in the death registration data than estimated by UNAIDS/WHO (19). UNAIDS/WHO estimates were used except in the cases of Australia, Chile, Costa Rica, France, Trinidad

and Tobago, Uruguay and USA.

Estimates for malaria deaths were compared with WHO estimates (see Section 5.3) and replaced by

WHO estimates for 63 country years where the WHO estimates were larger than those from the death

registration data. This affected malaria deaths for Brazil (12 years), Columbia (10), Venezuela (9),

Philippines (8) and Panama (3).

WHO estimates for maternal deaths include an upwards adjustment for under‐recording of maternal

deaths in death registration data (27 ). Maternal deaths were adjusted using these country‐specific

factors, and all other causes adjusted pro‐rata.

Deaths due to alcohol and drug use disorders include alcohol and drug poisoning deaths coded to the injury chapter of ICD (see Annex Table A). Further adjustments for under‐reporting in some countries

will be undertaken in the next revision of these estimates.

Where necessary, road injury deaths were adjusted upwards to take account of additional surveillance

data provided by countries (see Section 5.11).

Estimates of deaths due to conflicts (see Section 5.12) were compared with estimates from the death

registration data year by year and added “outside‐the‐envelope” for country‐years where they are not

included in death registration data.

3.7 Other national‐level information on causes of death

Cause of death estimates for a number of countries drew on non‐national death registration data or

other data sources with cause of death information as follows.

China Cause‐specific mortality data for China were available from two sources – the sample vital registration

system data for years 1987 to 2010 (28) and summary deaths tabulations from the Diseases Surveillance

Points (DSP) system for years 1995‐1998 and 2004‐2010 (29, 30). Table 3.4 summarizes the deaths and




population covered by these two systems. The sample vital registration system data for years 1987 to

2010 was provided in separate tabulations for urban and rural sampled populations, with more urban

than rural sampling. The urban and rural crude deaths rates by age, sex and cause were weighted for

each year using the UN Population Division’s estimated urban and rural population fractions, and the

resulting death rates re‐applied to the UN total estimated population by age and sex. The DSP sample

sites are considered to be nationally representative and the resulting total deaths by age, sex and cause

were not reweighted. For both sets of data, annual data were rescaled so total deaths by age and sex

matched the estimated all‐cause envelopes for China (see Section 2.5).

Table 3.4. Total deaths and population covered by the Chinese vital registration system (VR) and the

Disease Surveillance Points system (DSP)

Year

Vital registration

system

Disease Surveillance

Points

Vital registration

system

Disease Surveillance

Points

Number of deaths Population

2000 711,946 … 117,183,678 …

2001

… … … …

2002 … … … …

2003 626,392 … 102,889,945 …

2004 295,906 430,994 55,288,841 71,173,205

2005 310,826 437,490 57,272,144 71,487,277

2006 379,057 347,057 72,240,261 66,012,299

2007 475,289 401,008 79,101,646 71,476,477

2008 471,219 424,683 … 73,928,499

2009 505,021 437,550 … 75,020,489

2010 558,915 453,211 90,158,748 78,766,626

… data not available.

Both sets of data were assessed and compared for suitability in estimating 2000‐2011 cause‐specific

mortality for China at the national level. As seen in Figure 3.1, both sets of data gave quite similar cause

distributions at major cause group level by age, across the period 2000‐2010. Additionally, comparison

for more detailed major causes of death did not give any clear indication that one data set was of

systematically higher quality than the other. We therefore based the update of cause of death estimates

for China on an average of the estimates from the two systems.

For all except the leading causes of death, there are considerable fluctuations across 5‐year age groups

and year in numbers of deaths, due to stochastic variation and perhaps also variations in recording

cause of death from year to year or sample site to sample site. In order to smooth these fluctuations and to estimate underlying trends, cubic spline smoothing was used as follows. For the VR data, cubic spline

curves were fitted to age‐sex‐cause specific deaths for years 1987‐2010 using a negative binomial model

with population as offset and with knot points at years 1992, 1997, 2003, and 2007. For the DSP data,

cubic spline curves were fitted to age‐sex‐cause specific deaths for years 1995‐2010 using a negative

binomial model with population as offset and with knot points at years 2004, 2007 and 2010. Final

estimates for China were calculated as the average of the fitted spline estimates from VR and DSP for

years 2000‐2011.




Figure 3.1. Sample vital registration data (VR) and Disease Surveillance Points data (DSP), China: comparison of cause fractions for three major cause groups by age, late 1990s, 2005 and 2010

The resulting

cause

‐specific

estimates

were

further

adjusted

with

information

from

WHO

technical

programmes and UNAIDS on specific causes (see Section 5) and from the GBD 2010 for certain specific

subcause categories where deaths were either not recorded or recorded to only selected categories in

the DSP and/or VR datasets. GBD 2010 analyses were used for GHE causes 5‐9 (STDs), 20 (hepatitis C),

26 (leishmaniasis), 34‐36 (intestinal nematode infections), 115 (inflammatory heart diseases), and 119

(asthma). Additionally, DSP broad cause group totals were redistributed to detailed subcauses using

GBD 2010 cause fractional distributions for the following categories: 82+94 (mental and behavioural

disorders and neurological conditions), 134 (musculoskeletal disorders) and 147 (oral conditions). Rabies

deaths were revised using data on reported human rabies deaths from the Chinese Center for Disease

Control and Prevention (31).

For estimates of causes of death under age 5, a separate analysis was undertaken based on an analysis

of 206

Chinese

community

‐based

longitudinal

studies

that

reported

multiple

causes

of

child

death

(see

Section 4.5 below. The CHERG conducted a systematic search of publically available Chinese databases

in collaboration with researchers from Peking University. Information was obtained from the Chinese

Ministry of Health and Bureau of Statistics websites, Chinese National Knowledge Infrastructure (CNKI)

database and Chinese Health Statistics Yearbooks published between 1990‐2008. A model was

developed to assign the total number of child deaths to provinces, age groups and main causes of child

death.




India Analysis of causes of death for India was based on data over a period of 3 years (2001–2003) recorded

by the Million Death Study (32,33), a comprehensive study based on verbal autopsy that assigned causes

to all deaths in areas of India covered by the Sample Registration System. The Sample Registration

System monitors a representative sample population of 6.3 million people in over 1 million homes in

India. The 1991 census was used to randomly select 6671 areas from approximately 1 million having about 1000 inhabitants in each.

In 2001 the Indian Registrar General Surveyor introduced an enhanced form of verbal autopsy for

assessing the cause of death. Verbal autopsy is a method of ascertaining the cause of death by

interviewing a family member or caretaker of the deceased to obtain information on the clinical signs,

symptoms and general circumstances that preceded the death. Details of methods and validation have

been reported elsewhere (33). Verbal autopsy reports were independently coded to ICD‐10 categories

by at least two of a total of 130 physicians trained in ICD‐10 coding. In case of disagreement on the ICD‐

10 codes at the chapter level, reconciliation between reports was conducted, followed by a third senior

physician’s adjudication.

A total of 136,000 deaths were enumerated between January 2001 and December 2003. Verbal

autopsies could not be conducted for 12% of the deaths for reasons such as family migration or change of residence. An additional 9% of the reports could not be coded because of data quality problems,

resulting a final dataset of 122,848 coded records.

The cause‐specific proportion of deaths in each five‐year age category from 0 to 79 years and for people

aged 80 years and over was weighted by the inverse probability of a household being selected within

rural and urban subdivisions of each state to account for the sampling design. National estimates for

deaths and mortality rates were based on United Nations 2005 estimates for India, by age, sex and area.

Figure 3.2. Percentage of deaths by cause and age for India: comparison of final GHE estimates for

year 2002

with

national

‐level

results

from

the

Million

Death

Study,

2001

‐2003

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

Age (years)

Global Health Estimates: India, 2002

Suicide, homicide and conflict

Other unintentional injuries

Road injury

Other noncommunicable

Chronic respiratory diseases

Cancers

Cardiovascular diseases

Maternal, neonatal, nutritional

Other infectious diseases

Lower respiratory infections

Diarrhoeal diseases

HIV, TB and malaria

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

Age (years)

Million Death Study: India, 2001‐2003




The GHE analysis is based on the resulting national‐level cause‐specific mortality proportions derived for

GHE cause categories from the Million Death Study. The mapping of the MDS cause categories to GHE

cause categories, and the use of GBD 2010 analyses to redistribute deaths to detailed subcause

categories is summarized in Annex Table E. GHE cause categories 26 (leishmaniasis) and 124

(appendicitis) were also estimated using GBD 2010 results.

The resulting cause‐specific estimates were further adjusted with information from WHO technical programmes and UNAIDS on specific causes (see Section 5) and adjusted to match WHO estimates of

age‐sex specific all‐cause mortality for India in 2002. Cause‐specific trends for India estimated in the

GBD 2010 study (26) were used to project cause‐fractions forwards to 2011 and backwards to 2000.

Figure 3.2 provides a comparison of the final proportional distributional estimates of deaths by cause

and age for India in the year 2002 with the original distributions in the Million Death Study for 2001‐

2003.




4 Child mortality by cause

Cause‐specific estimates of deaths for children under age 5 were estimated for 17 cause categories

using methods described elsewhere by Liu et al. (34) and on the WHO website (35). These previously

published estimates for years 2000‐2010 were updated to take account of revisions in child mortality

levels (14), as well as cause‐specific estimates for HIV, tuberculosis, measles and malaria deaths (as

described in Section 5). Inputs to the multivariate cause composition models were also updated as described below.

4.1 Causes of under 5 death in countries with good death registration data

Death registration data were used directly for estimating causes of neonatal and under 5 child deaths

for countries with good quality vital registration (VR) data with population coverage of >80%. VR data

were considered as of good quality if the following criteria were met: (a) reasonable distribution of

deaths by cause were reported without excessive use of implausible codes or certain codes, and (b)

sufficient details of the coding was provided so that deaths could be grouped into appropriate

categories used in the analysis. For countries with adequate death registration, data on causes of child

deaths were extracted from the WHO mortality database, adjusted for coverage incompleteness where

needed, and grouped according to the standard International Classification of Diseases, 10th revision (ICD‐10). For earlier years when ICD‐9 codes were used, a mapping system was applied to convert them

into ICD‐10 codes (34,webappendix ). Certain neonatal codes were re‐assigned from ill‐defined codes to

more plausible codes (see Annex Table C). Annual data for years 2000 to the latest available year were

included with data closest to the estimating year used where possible. Where the latest year available

was earlier than 2011, the cause distribution for the latest available year was assumed to apply for

subsequent year(s), which was then applied to the age‐specific total number of child deaths.

4.2 Causes of neonatal death (deaths at less than 28 days of age)

The CHERG neonatal working group undertook an extensive exercise to derive mortality estimates for six

causes of neonatal death, including preterm birth, asphyxia, severe infection, diarrhoea, congenital

malformation and other causes (36). These cause categories are defined in Annex Table B. Death registration data were used directly for 61 countries considered to have reliable information. For

another 51 low mortality countries, the cause distribution was estimated using a multinomial model

applied to death registration data. For 80 high mortality countries the cause distribution was estimated

using a multinomial model applied to (largely) verbal autopsy (VA) data from research studies (34). A

total of 90 studies in 34 countries in high mortality populations met the inclusion criteria. The

multinomial model for high mortality countries was generally used for countries with average U5MR>35

for the period 2000‐2010.

A separate cause category for neonatal pneumonia is included in the model, and the neonatal sepsis

category includes a number of neonatal infections, such as meningitis and tetanus, not separately

identified. The number of tetanus deaths was also modeled separately in a single cause model using

using a logistic regression model with percent of women who were literate, percent of births with skilled attendant, and percent protected at birth by tetanus toxoid vaccine as covariates. The resulting cause‐

specific inputs were adjusted country‐by‐country to fit the estimated neonatal death envelopes for

corresponding years.

Pending further revisions of the neonatal tetanus model to estimate longer‐term trends in neonatal

tetanus deaths, estimates for 2011 and 2000 were based on projection and back‐projection of the 2008

estimates using estimates of trends in tetanus deaths from the GBD 2010 study (26).




4.3 Causes of child death at ages 1‐59 months –low mortality countries

For 51 low mortality countries without VR data or with VR data not meeting quality criteria (see Section

4.1), the cause distribution was estimated using a multinomial model applied to death registration data.

This multinomial model applied to death registration data was generally used for countries with average

U5MR<35 for the period 2000‐2010.

For the estimates for years 2000‐2011, the previous vital registration‐based multicause model (VRMCM)

model was revised to include additional death registration data and to update time series for covariates

and extend them to 2011. The choice of covariates included in the model was not revisited for this

regional‐level update. The multinomial logistic regression model was estimated using death registration

data from countries with >80% complete cause of death (CoD) certification for years 1990‐2011 to

estimate the proportion of deaths due to pneumonia, diarrhea, meningitis, injuries, perinatal, congenital

anomalies, other NCDs and other causes.

The current version of the model used death registration data for the years 1990 to 2011, including

1,123 data points, representing 63 countries. The model included the following covariates that were

determined a priori : U5MRs, GNI per capita (PPP, $international), WHO European and American regions.

Adjustments for the scaling‐up of Hib vaccine occurred within the model. The proportional distribution

of causes of death was then applied to the HIV‐free and measles‐free envelope for children 1‐59 months

of age. Jack‐knife and Monte Carlo simulation methods were used to estimate uncertainty.

4.4 Causes of child death at ages 1‐59 months –high mortality countries

For 79 high mortality countries (average U5MR>35 for the period 2000‐2010), the cause distribution was

estimated using a multinomial model applied to (largely) verbal autopsy (VA) data from research studies

(34,36,37 ). The multicause model for deaths at ages 1‐59 months was used to derive mortality estimates

for seven causes of postneonatal death, including pneumonia, diarrhea, malaria, meningitis, injuries,

congenital malformations, causes arising in the perinatal period (prematurity, birth asphyxia and

trauma, sepsis and other conditions of the newborn), and other causes, based on 113 data points from

74 studies of postneonatal deaths from 33 countries that met inclusion criteria2. Studies were

predominantly from lower income high mortality countries. Malnutrition deaths were included in the other cause of death category. Deaths due to unknown causes were excluded from the analysis. Deaths

due to measles and HIV/AIDS were estimated separately.

The resulting cause‐specific inputs were adjusted country‐by‐country to fit the estimated 1‐59 month

death envelopes (excluding HIV and measles deaths) for corresponding years and then estimates were

further adjusted for intervention coverage (pneumonia and meningitis estimates adjusted for use of Hib

vaccine; malaria estimates adjusted for insecticide treated mosquito nets (ITNs)). This method was used

for countries without useable death registration data and with U5MR>26 and gross national income per

capita less than $7,510.

2 Studies conducted in year 1980 or later, a multiple of 12 months in study duration, cause of death available for

more than a single cause, with at least 25 deaths in children <5 years of age, each death represented once, and less

than 25% of deaths due to unknown causes were included. Studies conducted in sub-groups of the study population

(e.g. intervention groups in clinical trials) and verbal autopsy studies conducted without use of a standardizedquestionnaire or the methods could not be confirmed were excluded from the analysis.




4.5 Causes of child death for China and India

In order to estimate trends in under 5 causes of death for India, the previously developed subnational

analyses were further refined and used to develop national estimates for years 2000‐2011 (38). For

neonates, a verbal autopsy multi‐cause model (VAMCM) based on 37 sub‐national Indian community‐

based VA studies was used to predict the cause distribution of deaths at state level. The resulting cause‐

specific proportions were applied to the estimated total number of neonatal deaths to obtain the estimated number of deaths by cause at state level prior to summing to obtain national estimates.

For children who died in the ages of 1‐59 months in India, the previously developed multicause model

was rerun for years 2000‐2011 using a total of 23 sub‐national community‐based VA studies plus 22 sets

of observations for the Indian states derived from the Million Death Study (39). Nine cause categories

were specified, including measles plus the eight specified in the post‐neonatal VAMCM for other

countries. State‐level measles deaths were then normalized to fit the national measles estimates

produced by the WHO IVB. State‐level AIDS and malaria estimates were provided by UNAIDS and WHO

malaria program, respectively. All cause fractions were adjusted to sum to one. The state‐level estimates

were collapsed to obtain national estimates at the end.

For China, updated IGME U5MR estimates in 2000‐2011 were applied to the VA‐based national cause‐

specific models developed by Rudan and colleagues (40) to derive cause‐fractions annually in this

period. Together with cause‐specific inputs from WHO technical programmes and UNAIDS for measles,

meningitis, malaria and AIDS, the resulting cause‐specific inputs for China were adjusted to fit the

estimated total deaths at ages 0‐1 month and 1‐59 months, respectively.

4.6 Inclusion of WHO‐CHERG estimates in Global Health Estimates 2000‐

2011

The seventeen cause categories used for the WHO‐CHERG estimates of under 5 deaths for years 2000‐

2011 (see Annex Table B) include all the major causes of neonatal, postneonatal and 1‐4 year deaths and

two residual categories containing all remaining causes of death. These residual categories (“Other

Group 1” and “Other Group 2”) and cause groups such as “Congenital malformations” and “Injuries” were expanded to the full GHE cause list (Annex Table A) for neonatal and under 5 deaths using cause

distributions derived from VR data for countries with useable VR data (see Annex Table F) and from the

GBD 2010 estimates for other countries (26).




5 Methods for specific causes with additional information

5.1 Tuberculosis

For countries with death registration data, tuberculosis mortality estimates were generally based on the

most recently available vital registration data. For other countries, total tuberculosis deaths were

derived from latest published WHO estimates (41), together with more detailed unpublished age distributions based on the VR data and notifications data.

5.2 HIV/AIDS and sexually transmitted diseases

For countries with death registration data, HIV/AIDS mortality estimates were generally based on the

most recently available vital registration data except where there was evidence of misclassification of

HIV/AIDS deaths. In such cases, a time series analysis of causes where there was likely misclassified

HIV/AIDS deaths was carried out to identify and re‐assign such deaths. For other countries, estimates

were based on UNAIDS estimated HIV/AIDS mortality (19). It was assumed based on advice from

UNAIDS that 1% of HIV deaths under age 5 occurred in the neonatal period.

5.3

Malaria

Countries outside the WHO African Region and low transmission countries in Africa3.

Estimates of the number of cases were made by adjusting the number of reported malaria cases for

completeness of reporting, the likelihood that cases are parasite‐positive and the extent of health

service use. The procedure, which is described in the World Malaria Report 2012 (42), combines data

reported by National Malaria Control Programs (reported cases, reporting completeness, likelihood that

cases are parasite positive) with those obtained from nationally representative household surveys on

health service use. If data from more than one household survey was available for a country, estimates

of health service use for intervening years were imputed by linear regression. If only one household

survey was available then health service use was assumed to remain constant over time; analysis

summarized in the World Malaria Report 2008 (43) indicated that the percentage of fever cases seeking

treatment in public sector facilities varies little over time in countries with multiple surveys. Such a procedure results in an estimate with wide uncertainty intervals around the point estimate.

The number of deaths was estimated by multiplying the estimated number of P. falciparum malaria

cases by a fixed case fatality rate for each country as described in the World Malaria Report 2012 (42).

This method is used for all countries outside the African Region and for countries within the African

Region where estimates of case incidence were derived from routine reporting systems and where

malaria causes less than 5% of all deaths in children under 5. A case fatality rate of 0·45% is applied to

the estimated number of P. falciparum cases for countries in the African Region and a case fatality rate

of 0·3% for P. falciparum cases in other Regions. In situations where the fraction of all deaths due to

malaria is small, the use of a case fatality rate in conjunction with estimates of case incidence was

considered to provide a better guide to the levels of malaria mortality than attempts to estimate the

fraction of deaths due to malaria.

Somalia, Sudan and high transmission countries in the WHO African Region.

Child malaria deaths were estimated using the VAMCM described in Section 4.4. The VAMCM derives

mortality estimates for malaria, as well as 7 other causes (pneumonia, diarrhea, congenital

3 Botswana, Cape Verde, Eritrea, Madagascar, Namibia, Swaziland, South Africa, and Zimbabwe




malformation, causes arising in the perinatal period, injury, meningitis, and other causes) using

multinomial logistic regression methods to ensure that all 8 causes are estimated simultaneously with

the total cause fraction summing to 1. Malaria deaths were retrospectively adjusted for coverage of

insecticide‐treated nets (ITNs) and use of Haemophilus influenzae type b vaccine (34). The bootstrap

method was employed to estimate uncertainty intervals by re‐sampling from the study‐level data to

estimate the distribution of the predicted percent of deaths due to each cause. The estimated malaria

mortality rate in children under 5 years for a country was used to determine malaria transmission

intensity and the corresponding malaria‐specific mortality rates in older age groups (43).

5.4 Whooping cough

An updated model of whooping cough (pertussis) mortality is being developed by the WHO Department

of Immunization, Vaccines and Biologicals (IVB). This model has not been finalized in time for the release

of these regional‐level estimates but will be used to update the GHE estimates at country level later in

2013. In the interim, whooping cough mortality estimates from the GBD 2010 (26) have been used as an

input to the WHO‐CHERG analysis of child causes of death under age five (see Section 4).

5.5 Measles

To estimate deaths attributable to measles, a new model of measles mortality developed by WHO

Department of Immunization, Vaccines and Biologicals (IVB) was used to first estimate country‐and‐

year‐specific cases using surveillance data (44). The improved statistical model firstly estimates measles

cases by country and year using surveillance data and making explicit projections about dynamic

transitions over time as well as overall patterns in incidence.

The age distribution of measles cases are then estimated using a logistic regression function fitted to

172,191 measles cases with data on age at infection from 102 countries over 2005‐2009 extracted from

WHO's monthly measles case‐based reporting system. Two explanatory variables were included in the

regression: 1) the 5 year rolling average of estimated MCV1 coverage, categorized in <60%, 60‐84%, and

85‐100%; and 2) geographic region classified in to 7 groups.

Country‐specific measles case‐fatality ratios (CFRs) for children 1‐4 years of age were taken from a comprehensive review of community‐based studies (45). This review included 102 field studies

conducted in 29 countries during the period 1974‐2007. The set of CFRs were revised for two countries

(India and Nepal) where additional studies have been published subsequent to the review (46, 47 ). The

same CFRs were used for infants and for children aged 1‐4 years. For the period 2000‐2011, we assumed

that age‐specific CFRs are not declining over time.

Age‐specific deaths are aggregated to derive measles deaths for all children below five and for ages five

and over. The new method takes into account herd immunity and produces results that are fairly

consistent with previous ones (48). Uncertainty is estimated by bootstrap sampling from the

distribution of incidence and age distribution estimates. Updated estimates of measles deaths by age

and country for years 2000‐2011 were prepared using the above methods at the end of 2012 and

summary results published in the Weekly Epidemiological Record (49).These were used for this update of GHE causes of death for years 2000‐2011.

For countries experiencing measles outbreaks, the measles deaths were split into outbreak and endemic

deaths, the latter of which were smoothed using local regression (50). For the ages of 1‐59 months, the

endemic measles deaths and AIDS deaths were added to the measles‐ and AIDS‐free all‐cause deaths for

which the VAMCM derived cause fractions were applied. The measles outbreak deaths were added back

at the end. In places where the outbreak deaths resulted in an increase in the all‐cause deaths by 10% or




more, the original survey data were screened to examine whether a real increase in child mortality was

indicated for the outbreak year. If there were survey data available for the years around the outbreak

but no evidence of an increased mortality, the measles outbreak deaths were truncated at 10% of the

all‐cause deaths. This was only necessary in few countries, almost all of which are in Africa and all

occurred in the early 2000s when more measles deaths were estimated.

5.6

Schistosomiasis

For the last WHO update of burden of disease for year 2004 (3), the incidence and prevalence of cases of

schistosomiasis infection were separately estimated by country for S. mansoni , S. haematobium and

S. japonicum plus S. mekongi. The GBD 2004 estimated that schistosomiasis was responsible for around

41 000 deaths globally (excluding attributable cancer deaths) and 36 000 in sub‐Saharan Africa, although

others have argued that the figure should be much higher (51). Van der Werf et al (52), using limited

data from Africa, estimated that schistosomiasis caused 210 000 deaths annually. For the GBD 2004

update (3), very limited available data was used to conservatively estimate annual case fatality rates for

prevalent cases at 0.01% for S. mansoni, 0.02% for S. haematobium, and 0.03% for S. japonicum and

S. mekongi . There were estimated to be 261 million prevalent cases of schistosomiasis infection in 2004.

The GBD 2010 study estimated that there 11,650 deaths due to schistosomias in 2010, of which 1,813 were in the Middle East and North Africa, and only 61 in sub‐Saharan Africa in 2010. Divided by the

numbers of prevalent cases estimated by the GBD 2010, the implied case fatality rates for the Middle

East and North Africa, and for Latin America are 0.01% and 0.02% respectively. In comparison, the

implied African case fatality rate is almost 400 times smaller. Implied case fatality rates for non‐African

regions in the GBD 2010 were generally consistent with those previously estimated by WHO for the year

2004. Revised case fatality rates of 0.0075% for S. mansoni, 0.015% for S. haematobium were applied to

the prevalence rates estimated by GBD 2010 (53) to revise the estimates of schistosomiasis deaths for

GHE. This resulted in an estimate of 17,600 deaths in sub‐Saharan Africa and 23,300 deaths globally in

2011.

5.7 Maternal causes of death

Country‐specific estimates for maternal mortality were based on the recent Interagency estimates for

years 2000‐2011 (27,54). For 62 Member States with relatively complete data from national death

registration systems, these data were used directly for estimating and projecting maternal mortality

ratios. For other Member States, a multilevel regression model was developed using available national‐

level data from surveys, censuses, surveillance systems and death registration. This regression model

included national income per capita, the general fertility rate and the presence of a skilled attendant at

birth (as a proportion of total births) as covariates to predict trends in maternal mortality.

Note that numbers of maternal deaths were adjusted upwards by a country‐specific fraction, or by 50%,

for countries with useable death registration data but without country‐specific data on misclassification

of maternal deaths, to allow for under‐identification of maternal deaths. Note also that the maternal

mortality estimates include those HIV deaths occurring in pregnant women or within 42 days of end of pregnancy which were considered to be indirect maternal deaths rather than incidental. These HIV

maternal deaths were subtracted from total HIV deaths as estimated by UNAIDS.

5.8 Cancers

Cause‐specific estimates for cancer deaths were derived from Globocan 2008 (55) for countries without

useable death registration data. Site‐specific deaths were projected back to year 2000 and forwards to

year 2011 using trend estimates from the GBD 2010 (26).




Karposi sarcoma was excluded from the Globocan estimates as this is almost entirely a manifestation of

HIV/AIDS, already included in the estimates for HIV/AIDS deaths. Deaths due to non‐melanoma skin

cancers were included in these estimates along with melanoma, unlike in Globocan 2008.

5.9 Alcohol use and drug use disorders

The injury codes for accidental poisoning by alcohol and by opioids are now used to code acute intoxication deaths from alcohol and acute overdose deaths by opioids. These deaths have been

remapped to alcohol use disorders and drug use disorders respectively (see Annex Table A). WHO

estimates of direct deaths associated with alcohol use disorders and total deaths attributable to alcohol

consumption are under revision for a forthcoming report. The interim estimates included here for

alcohol use disorders will be revised in the next revision to take these updates into account.

GBD 2010 estimates of deaths due to drug use disorders were revised to correct an extremely low

implied case fatality rate for opioid dependent drug users in South Asia and for consistency with

estimates of prevalence and mortality associated use of illicit opiate drugs reported by the UN Office on

Drugs and Crime (UNODC) (56). UNODC estimated that there were around 17 million opiate users

globally in 2010, with higher than average prevalence of opioid users in North America, Oceania, Eastern

Europe and South East Europe. These estimates were quite similar to those of the IHME‐GBD 2010, which estimated a global prevalence of 17.3 million for opioid dependence in 2010 (53). The IHME‐GBD

2010 estimated a total of 77,615 deaths for drug use disorders in 2010, of which 43,000 were for opioid

use disorders. The implied case fatality rate of opioid use was 0.25% globally, 0.23% in the Middle East

and North Africa, and just under 0.1% in East and South East Asia. In contrast, the implied case fatality

rate of 0.025% in South Asia was only 1/10th of the global average. Estimated opioid dependence deaths

were conservatively revised upwards for South Asia to give an implied case fatality rate similar to that of

the other Asian regions. The resulting GHE estimate of 91,900 deaths for all drug use disorders in 2011 is

similar to the UNODC estimate of around 100,000 total direct drug use deaths in 2010 (with an

additional 100,000 deaths from other causes, such as infectious diseases, also attributable to drug use

disorders).

5.10 Epilepsy

The Million Death Study for India (32,33) recorded relatively high proportions of epilepsy deaths,

resulting in an initial GHE estimate of 73,600 epilepsy deaths in India in 2010 compared to an estimated

21, 650 by the GBD 2010. GBD 2010 estimates of untreated idiopathic epilepsy prevalence were used to

calculate implied regional case fatality rates (CFR) and the implied Indian CFR of 0.34 was substantially

higher than those for South East Asia (0.09) or the Middle East and North Africa (0.05). Indian epilepsy

deaths were adjusted downwards to give an implied case fatality rate of 0.17 (close to the global

average), resulting in an estimated 35,480 epilepsy deaths for India in 2010.

5.11 Road injuries

For the second WHO Global status report on road safety (57 ), updated estimates of road injury deaths

were prepared for 182 Member States for the year 2010. These estimates drew on death registration

data, on reported road traffic deaths from official road traffic surveillance systems (collected in a WHO

survey of Member States for the report), and on a revised regression model for countries without

useable death registration data. The same methods were used to develop time series estimates of road

injury deaths for years 2000‐2011 for all Member States.

The methods used for four groups of countries are summarized below and the method used for each

country is documented in Annex Table G.




5.11.1 Countries with death registration data

This group includes 87 countries with death registration data meeting one of the following

completeness criteria, viz. completeness for the year estimated at 80% or more, or average

completeness for the decade including the country‐year was 80% or more.

These countries fell into three categories:

1. For countries with death registration data for the year 2010 which exceeded the number of road

traffic deaths reported in the survey – death registration data was used. There were 33

countries in this category.

2. For countries where the latest death registration data submitted to WHO was earlier than 2010,

but not earlier than 2005 – deaths for 2010 were estimated based on a projection of the most

recent death registration data using the trends obtained through the survey. There were 40

countries in this category.

3. For countries where the reported road traffic deaths for 2010 obtained through the survey

exceeded the estimate based on death registration data: The reported road traffic deaths

(adjusted to the 1 year definition) were used. There were 12 countries in this category. There

were an additional 2 countries where reported data for earlier years were projected to 2010 and used because they exceeded the death registration numbers.

5.11.2 Countries with other sources of information on causes of death

For India, Iran, Thailand and Viet Nam, data on total deaths by cause were available for a single year or

an earlier recent single year or group of years (33,58‐60). For these countries, the regression method

described below was used to project forward from the most recent year for which an estimate of total

road traffic deaths were available.

5.11.3 Countries with populations less than 150 000

For 13 small countries with populations of less than 150 000 people the deaths reported in the survey

were used directly without adjustment.

5.11.4 Countries without eligible death registration data

For 78 countries which did not fall into any of the above groups, a regression model was used to

estimate total road traffic deaths. The regression model produced estimates of total road traffic deaths

according to the accepted International Classification of Disease definition, which counts all deaths that

follow from a road traffic death, regardless of the time period in which they occur (unlike many official

road traffic surveillance data sources, where road traffic death data is based on a 30‐day definition

following a road traffic crash). Where total deaths reported by Member States surveillance systems

were greater than the deaths estimated from the regression, these were used.

Three classes of models were tested and a negative binomial counts model was chosen:

Pop X X X C N nnln....ln 2211

(1)

where N is the total road traffic deaths (for a country‐year), C is a constant term, X i are a set of

explanatory covariates, Pop is the population for the country‐year, included as an offset, and ε is the

negative binomial error term. This model was estimated using death registration data for the period

1950–2010 that were 80% or more complete for a given year or where the average completeness for




the last decade was greater or equal to 80%. It also included nationally representative verbal autopsy or

sample death registration data for India, China and Vietnam.

Three models (A, B and C) were chosen that had good in‐sample‐ and out‐of ‐sample fit, and for which all

the covariates (see Table 5.1) were statistically significant at the 95% level. The final estimates were

based on the average predictions of these three best models.

Age distributions for road injury deaths were based on regional age distributions estimated in the GBD

2010 study (26).

Table 5.1. Covariates used in the model for road injury deaths

Independent

variables

Description Source of information Included

in models

ln(GDP) WHO estimates of Gross Domestic Product

(GDP) per capita (international dollars or

purchasing power parity dollars, 2005 base)

WHO database A, B, C

ln(vehicles

per

capita) Total vehicles per 1000 persons GSRRS surveys and WHO database A, B, C

Road density Total roads (km) per 1000 hectares International Futures database (63) A, B, C

National speed limits

on rural roads

The maximum national speed limits on rural

roads (km/h) from WHO questionnaire

GSRRS survey (57 ) A, B, C

National speed limits

on urban roads

The maximum national speed limits on urban

roads (km/h) from WHO questionnaire

GSRRS survey (57 ) A, B, C

Health system access Health system access variable (principal

component score based on a set of coverage

indicators for each country)

Institute for Health Metrics and

Evaluation dataset (61)

A, B, C

Alcohol apparent

consumption

Liters of alcohol (recorded plus unrecorded)

per adult aged 15+

WHO database A, B, C

Population working

Proportion of population aged 15‐16 years World Population Prospects 2010

revision (UNDESA)

A, B, C

Percentage

motorbikes

Per cent of total vehicles that are motorbikes GSRRS survey (57 ) B

Corruption index Control of corruption index (units range from

about ‐2.5 to +2.5 with higher values

corresponding to better control of corruption

World Bank (62), International

Futures database (63)

B

National policies for walking /cycling

Existence of national policies that encourage

walking and / or cycling

GSRRS survey (57 ) C

Population Total population (used as offset in negative

binomial regression

World Population Prospects 2010

revision (12)

A, B, C




5.12 Conflict and natural disasters

Estimated deaths for major natural disasters were obtained from the CRED International Disaster

Database (64). For country‐years where disaster death rates exceeded 1 per 10,000 population, these

deaths were added to the life table death rates for the relevant year. Age‐sex distributions were based

on a number of studies of earthquake deaths (65,66) and tsunami deaths (67,68).

Country‐specific estimates of war and conflict deaths were updated for the entire period 1990‐2011

using revised methods together with information on conflict intensity, time trends, and mortality

obtained from a number of war mortality databases (described below). These estimates relate to deaths

for which the underlying cause (following ICD conventions) was an injury due to war, civil insurrection or

organized conflict, whether or not that injury occurred during the time of war or after cessation of

hostilities. The estimates include injury deaths resulting from all organized conflicts, including organized

terrorist groups, whether or not a national government was involved. They do not include deaths from

other causes (such as starvation, infectious disease epidemics, lack of medical intervention for chronic

diseases), which may be counterfactually attributable to war or civil conflict.

Methods used previously by WHO for estimation of direct conflict deaths were developed in the early

2000s and applied adjustment factors for under‐reporting to estimates of battlefield or conflict deaths

from a variety of published and unpublished conflict mortality databases (69‐72). Murray et al. (73)

summarized the issues with estimation of war deaths, and emphasized the very considerable

uncertainty in the original Global Burden of Disease estimates (74) and subsequent WHO estimates for

conflict deaths. WHO published estimates for the years 2000 through 2008 used adjustment factors

based on conflict intensity developed from an analysis of likely levels of under‐reporting (1‐4). These

adjustment factors ranged from around 3 to higher than 4 in sub‐Saharan Africa.

Obermeyer, Murray and Gakidou (75) more recently analyzed data on deaths due to conflict from post‐

conflict sibling histories collected in the 2002 to 2003 WHO World Health Survey (WHS) program. They

used data from 13 countries with more than 5 reported sibling deaths from war injuries in at least one

10‐year period to estimate total war deaths for these countries for the period 1955‐2002. The authors

then compared their estimates of war deaths to the number of war deaths estimated in the UCDP Battle

Deaths database (76) to derive an average adjustment factor of 2.96. Garfield and Blore (77 ) noted that

a very small number of war deaths for Georgia resulted in an outlier ratio of 12.0 which heavily

influenced the overall ratio of 2.96. They reanalyzed the WHS‐derived war deaths dataset excluding

Georgia, to obtain an overall revised adjustment factor of 2.21.

The revised WHO country‐specific estimates of war and conflict deaths for the period 1990‐2011 make

use of estimates of direct deaths from three datasets: Battle‐Related Deaths (version 5), Non‐State

Conflict Dataset (UCDP version 2.4), and One‐sided Violence Dataset (UCDP version 1.4) from 1989 to

2011 (78‐80). Using these three datasets, instead of focusing solely on battle‐related deaths, reduces

the likelihood that overall direct conflict deaths are underestimated. However, it likely that a degree of

undercounting still occurs in the count‐based datasets, and the adjusted ratio obtained by Garfield and

Blore (77 ) of 2.21 is applied to the annual battle death main estimates for state‐state conflicts (78). No

adjustments were applied to estimated conflict deaths (main estimates) for non‐state conflict deaths

(79), and one‐sided violence (80).

Additional information from epidemiological studies and surveys was also used for Iraq (81,82). Deaths

due to landmines and unexploded ordinance were estimated separately by country (83). Age‐sex

distributions for conflict deaths were revised based on available distributions of conflict deaths by age

and sex for specific conflicts (73,75,81‐86).

The following tables summarizes and compares various time series of conflict deaths estimates.




Table 5.2. Estimated total global injury deaths (thousands) due to conflict: comparison of various time

series and WHO estimates

Year GBD 1990 (a) WHO 2000‐2008 UCDC ‐PRIO

(h)

WHO 2013 (i) IHME ‐GBD 2010 (j)

1990 502 ‐ 95 138 63

2000 656 310 (b)

2000 230 (c)

2000 187 (d) 85 122 53

2004 182 (e) 30 95

2005 238 (f) 18 69 26

2008 182 (g) 34 84

2010 834 28 57 18

(a) Estimates and projections by Murray and Lopez (74)

(b) World Health Report 2001 (87 ) and World report on violence and health (88).

(c) World Health Report 2002 (1)

(d) Revision for Disease Control Priorities Study (2)

(e) Global burden of disease: 2004 update (3)

(f) World Health Statistics 2007 (89)

(g) WHO estimates of causes of death for year 2008 (4)

(h) Sum of main estimates of conflict deaths for state‐state, state‐nonstate and one‐sided conflicts (78‐80)

(i) Revised WHO estimates for years 1990‐2011 as documented here.

(j) IHME Global Burden of Disease Study 2010 (26).

The revised WHO estimates for total conflict deaths (in the column WHO 2013) are considerably lower

than the previous WHO estimates for years 2000‐2008 which used the earlier higher adjustment factor

for under‐reporting, which in turn are lower than the previous estimates and projections in the original

GBD study (74). The recently estimates for conflict deaths published by IHME in the GBD 2010 study,

shown in the rightmost column, are considerable lower than the revised WHO estimates. For the year

2010, the IHME estimates are also lower than the main estimate from the UCDC‐PRIO databases for the

same year. The IHME methods were based on a regression analysis of available all‐cause mortality data

for country‐years in which battle deaths were reported in various databases. Lozano et al (26) cite (90)

for more detailed documentation of their methods.

The revised WHO estimates for conflict deaths were taken into account in preparing final all‐cause mortality envelopes for Member States for years 1990‐2011 as follows. For country‐years where death

rates from conflict or disasters exceeded 1 per 10,000 population, the estimated annual age‐sex‐specific

conflict deaths were added to the life table death rates for the relevant year. In cases of extended

conflicts where death rates fluctuated above and below 1 per 10,000, only the death rate in excess of 1

per 10,000 was added to relevant years.




6 Other causes of death for countries without useable data

Previous WHO comprehensive estimates of causes of death have relied on cause‐of ‐death modelling

and available data on cause of death distributions within each analysis subregion to estimate causes of

death for countries without useable data and where WHO cause‐specific analyses were not available (2,

3). The IHME developed covariate based estimation models for a large number of single causes as

inputs to its overall estimation of numbers of deaths by country, cause, age and sex for years 1990‐2010 in the GBD 2010 study (8‐10). Results from these models are used as inputs to WHO Global Health

Estimates for causes of death not addressed by WHO and UN Interagency estimation processes and

where countries did not have useable death registration data, as described below.

Six different modelling strategies were used by IHME for causes of death depending on the availability of

data (26,webappendix ). For all major causes of death except HIV/AIDS and measles, IHME used

ensemble modelling to create a weighted average of many individual covariate‐based models (ranging

from hundreds to thousands in some cases) for each specific cause (26,91). IHME cause of death

estimation methods are thus complex and highly computer‐intensive. The overall out‐of ‐sample

predictive validity of the ensemble is usually not much different to that of the top‐ranked model, but

uncertainty ranges are generally much wider and more plausible than for single models.

IHME results for priority causes such as HIV, TB, malaria, cancers, maternal mortality, child mortality

differ to varying degrees from those of WHO and UN agency partners. In part, this reflects differences in

modelling strategies, but also the inclusion by IHME of data from verbal autopsy (VA) studies which has

been mapped to ICD categories using IHME‐developed computer algorithms. WHO aims to work with

IHME and expert groups to further improve data and methods, which requires that all input data and

detailed analysis methods and results are made available. Figure 6.1 provides a comparison of major

cause group death rates for the GBD 2010 and WHO GHE results for year 2010 for seven broad regional

groupings.

To ensure that the results of all the single‐cause models summed to the all‐cause mortality estimate for

each age‐sex‐country‐year group, IHME applied a final step called CoDCorrect to rescale the cause‐

specific estimates. This was done using repeated random draws from the uncertainty distributions of

each single cause and from the all‐cause envelope, and proportionately rescaling each single cause

estimate so they collectively summed to the envelope estimate. The overall effect is to “squeeze” or

“expand” causes with wider uncertainty ranges more than those with narrower uncertainty ranges.

GBD 2010 results, post‐CoDCorrect, were used as inputs to estimate cause fractions by country, age, sex

and year for causes of death at ages five years and above for which death registration data and/or WHO

and UN Interagency analyses (described in Section 5) were not available. For this set of causes, GBD

2010 country‐level estimates for death rates at ages 5 and over for years 1990, 1995, 2000, 2005 and

2010 were interpolated to death rates for all years in the range 2000‐2011 using cubic spline

interpolation of log(death rates. Cause fraction distributions were then computed for the set of causes

excluding WHO/Interagency cause‐specific estimates. For countries where these cause fractions were

used (see Annex Table G), they were applied to the country‐level residual mortality envelopes by age

and sex after the WHO/Interagency cause‐specific estimates were subtracted from the WHO all‐cause

envelopes.

Table 6.1 summarizes the overall percentage change in the GBD 2010 estimates for each of these

residual causes resulting from the above process. This provides a rough metric of how much

inconsistency there is between the GBD 2010 and the GHE 2010 estimates for ages five and over as a

result of differences in all‐cause envelopes and WHO/Interagency estimates for specific causes.




Figure 6.1. Comparison of GHE and IHME death rates per 100,000 population, major cause groups, 2010




Table 6.1. Ratio of GHE total deaths for residual causes to GBD 2010 total deaths for residual causes,

low ‐ and middle‐income countries without useable death registration data, by WHO Region and age

group, 2000 and 2011

Ratio 2000 Ratio 2011

5‐14 15‐49 50‐69 70+ 5‐14 15‐49 50‐69 70+

AFR 3.70 1.50 1.36 1.44 4.15 1.57 1.23 1.36

AMR 1.75 1.25 1.10 1.01 2.05 1.28 1.14 1.07

EMR 1.48 0.91 0.84 1.09 1.65 0.81 0.88 1.12

EUR 0.95 0.89 0.98 1.11 0.54 0.62 0.83 1.08

SEAR 1.29 1.09 1.04 1.11 1.05 0.83 1.02 1.10

WPR 1.38 0.78 0.77 1.00 1.21 0.58 0.68 0.95

World 2.47 1.20 1.06 1.16 2.77 1.12 1.02 1.15




7 Uncertainty of estimates

Country‐level estimates of mortality for 2004 and 2008 previously released on the WHO website

included guidance to users on the data sources and methods used for each country, in terms of four

levels of evidence. Comprehensive uncertainty ranges have not yet been addressed for the GHE cause of

death estimates although uncertainty ranges are available for many of the component analyses for

specific causes (refer to the detailed documentation of sources in Sections 4 and 5). General guidance on the quality and uncertainty of these cause of death estimates for years 2000‐2011 is provided in

terms of the quality of data inputs and methods used. These are broadly summarized for WHO Member

States in Annex Table F for general mortality and cause‐of ‐death methods.

WHO’s adoption of health estimates is affected by a number of factors, including a country consultation

process for country‐level health estimates, existing multi‐agency and expert group collaborative

mechanisms, and compliance with minimum standards around data transparency, data and methods

sharing. More detailed information on quality of data sources and methods, as well as estimated

uncertainty intervals, is provided in referenced sources for specific causes (Sections 4 and 5).

Calculated uncertainty ranges depend on the assumptions and methods used. In practice, estimating

uncertainty in a consistent way across health indicators has had limited success (i.e., estimates with

uncertainty typically reflect some, but not all, source of uncertainty). Most methods for estimation of

uncertainty rely on statistical techniques to assess variations across observations and take into account

sampling error but are less successful in dealing with unknown systematic bias in observations. In

particular, there is not yet sufficient research or consensus on the interpretation and use of verbal

autopsy studies to ensure that systematic bias in assigning underlying cause of death can be fully

addressed or resulting uncertainty fully quantified.

The type and complexity of models used for global health estimates varies widely by

research/institutional group and health estimate. More complex models are necessary to generate more

accurate uncertainty intervals. As expected, these are more difficult to transfer across research groups

and require greater researcher expertise and time and computational resources to run. Where data are

available and of high quality, estimates from different institutions are generally in agreement.

Discrepancies are more likely to arise for countries where data are poor and for conditions where data

are sparse and potentially biased. This is best addressed through improving the primary data.

Country health information systems, including vital registration, need to be strengthened as a matter of

priority, in order to provide a more solid empirical basis for monitoring health situation and trends is

essential. Such data are also crucial for Member States’ monitoring of local trends in order to respond to

the changing needs of their populations.

To improve monitoring of mortality, morbidity and risk factors the improving health information

systems should focus on strengthening:

Death registration through civil registration and vital statistics systems (CRVS), local health and

demographic studies and other sources

Cause of death data collection through vital registration and verbal autopsy in communities

Regular household health surveys that include biological and clinical data collection

Complete facility recording and reporting with regular quality control

Although the GHE estimates for years 2000‐2011 have large uncertainty ranges for some causes and

some regions, they provide useful information on broad relativities of disease burden, on the relative




importance of different causes of death, and on regional patterns and inequalities. The data gaps and

limitations in high‐mortality regions reinforces the need for caution when interpreting global

comparative cause of death assessments and the need for increased investment in population health

measurement systems. The use of verbal autopsy methods in sample registration systems, demographic

surveillance systems and household surveys provides some information on causes of death in

populations without well‐functioning death registration systems, but there remain considerable

challenges in the validation and interpretation of such data.

Figure 7.1 summarizes the proportional distributions of deaths by age, sex and cause for years 2000 and

2011. More detailed regional tabulations of deaths by cause, age and sex for years 2000 and 2011 are

available in the WHO Global Health Observatory (www.who.int/gho) and as downloadable Excel

spreadsheets at http://www.who.int/healthinfo/global_health_estimates/en/.

Figure 7.1 Percentage of deaths by cause for global age‐sex groups, 2000 and 2011.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

Age (years)

Males, 2000



Road injury



Cancers





Diarrhoeal diseases

HIV, TB and malaria

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

Age (years)

Females, 2000

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

Age (years)

Males, 2011



Road injury



Cancers





Diarrhoeal diseases

HIV, TB and malaria

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

Age (years)

Females, 2011




These estimates for years 2000‐2011 supercede and replace all previous estimates for global and

regional causes of death published by WHO. They are not directly comparable with previous WHO

estimates for 2008 and earlier years and differences should not be interpreted as trends. Figures 7.2

and 7.3 provide summary comparisons of the GHE estimates for year 2008 with the previous WHO

estimates for year 2008 published in 2011 (4,5). These figures illustrate that there has been little change in the relative ranking for the leading causes of death, although estimated numbers of deaths are

somewhat lower for most causes. This partially reflects downwards revision of all‐cause envelopes in

recent successive revisions by UN‐IGME and UN Population Division, but also reflects accelerating

declines in child mortality, and to a lesser extent, adult mortality in recent years.

These are provisional estimates and will be further revised in the process of updating to 2012 for release

at country level in late 2013. WHO and collaborators will continue to include new data and improve

methods, and it is anticipated that some causes will be further updated in the next revision.

Figure 7.2. Change in 10 leading causes of death at global level, GHE estimates for 2011 compared with

previous WHO

cause

of

death

(COD)

estimates

for

year

2008

(4,5)

COD08 (a) GHE 2011 (b)Total

deaths

(millions) Rank Rank

Total

deaths

(millions)

Ischaemic heart d i s e a s e 7.25 1 1 7.02 Ischaemic heart d i s e a s e

Cerebrovascular d i s e a s e 6.15 2 2 6.25 Cerebrovas cular d i s e a s e

Lower respiratory infections 3.46 3 3 3.20 Lower respiratory infections

COPD 3.28 4 4 2.97 COPD

Diarrhoeal d i s e a s e s 2.46 5 5 1.89 Diarrhoeal d i s e a s e s

HIV/AIDS 1.78 6 6 1.59 HIV/AIDS

Lung cancer 1.39 7 7 1.48 Lung

cancer

Diabetes mellitus 1.26 8 8 1.39 Diabetes mellitus

Road injury 1.21 9 9 1.26 Road injury

Hypertensive heart d i s e a s e 1.15 10 10 1.17 Preterm birth complications

Preterm birth compl i ca ti ons 1.00 13 11 1.06 Hypertensive heart d i s e a s e

Disease or injury Disease or injury




Figure 7.3. Comparison of death rates per 100,000 for nine major cause groups, GHE estimates for year

2008 and previous WHO COD estimates for year 2008, for world, high income countries, and low ‐ and

middle‐income countries grouped by WHO region

0

200

400

600

800

1000

1200

1400

COD08 GHE COD08 GHE COD08 GHE COD08 GHE



Road injury


Cancers




HIV, TB and malaria

0

200

400

600

800

1000

1200

1400

COD08 GHE COD08 GHE COD08 GHE COD08 GHE



Road injury


Cancers



Other infectious

diseases

HIV, TB and malaria

World Africa Americas Eastern

Mediterranean

High income Europe South East Asia Western Pacific




References

(1) World Health Organization. World health report 2002. Reducing risks, promoting healthy life.

Geneva, World Health Organization, 2002.

(2) Lopez, A.D., Mathers, C.D., Ezzati, M., Murray, C.J.L., & Jamison, D.T. Global burden of disease and

risk factors. New York, Oxford University Press, 2006.

(3) World Health Organization. The global burden of disease: 2004 update. Geneva, World Health

Organization, 2008.

(4) World Health Organization. Causes of death 2008: data sources and methods.

http://www.who.int/healthinfo/global_burden_disease/cod_2008_sources_methods.pdf .

(5) World Health Organization. Country‐level mortality estimates by cause, age, and sex for the year

2008. Geneva: WHO. Available at

http://www.who.int/healthinfo/global_burden_disease/estimates_country/en/index.html

(6) World Health Organization. Global health estimates for deaths by cause, age, and sex for years

2000‐2011. Geneva: WHO. Available at

http://www.who.int/healthinfo/global_health_estimates/en/ (7) International Classification of Diseases – 10th Revision. Geneva, World Health Organization, 1990.

(8) Murray CJ, Ezzati M, Flaxman AD, et al. GBD 2010: a multi‐investigator collaboration for global

comparative descriptive epidemiology. Lancet , 2012, 380(9859):2055‐8.

(9) Wang H, Dwyer‐Lindgren L, Lofgren KT, et al. Age‐specific and sex‐specific mortality in 187

countries, 1970–2010: a systematic analysis for the Global Burden of Disease Study 2010. The

Lancet. 2012 Dec 13; 380: 2071–2094.

(10) Institute for Health Metrics and Evaluation 2013. GBD Compare. Available at

http://viz.healthmetricsandevaluation.org/gbd‐compare/ (accessed 24 June 2013).

(11) World Health Organization. WHO methods and data sources for life tables 1990‐2011. Global

Health Estimates Technical Paper WHO/HIS/HSI/GHE/2013.1)

(12) United Nations, Department of Economic and Social Affairs, Population Division. World Population

Prospects ‐ the 2010 revision. New York, United Nations, 2011.

(13) United Nations, Department of Economic and Social Affairs, Population Division. World

Population Prospects ‐ the 2012 revision. New York, United Nations, 2013.

(14) UNICEF, WHO, The World Bank and UN Population Division. Levels and Trends of Child Mortality ‐

Report 2012, Estimates developed by the UN Inter ‐agency Group for Child Mortality Estimation.

UNICEF, New York, 2012.

(15) The PLoS Medicine Collection on Child Mortality Estimation Methods. PLoS Medicine, 2012.

Available at:

http://www.ploscollections.org/article/browseIssue.action?issue=info:doi/10.1371/issue.pcol.v07.i19.

(16) Oestergaard MZ, et al. Neonatal Mortality Levels for 193 Countries in 2009 with Trends since 1990:

A Systematic Analysis of Progress, Projections, and Priorities. PLoS Medicine, 2011, 8(8):

e1001080. doi:10.1371/journal.pmed.1001080

(17) Murray CJL, Ferguson BD, Lopez AD, Guillot M, Salomon JA, Ahmad O. Modified logit life table

system: principles, empirical validation and application. Population Studies, 2003, 57(2):1‐18.




(18) United Nations, Department of Economic and Social Affairs, Population Division. File 0‐2 Latest

data sources used to derive estimates for total population, fertility, mortality and migrations by

countries or areas in World Population Prospects – the 2010 revisions. New York, United Nations

Population Division, 2012. Available at: http://esa.un.org/wpp/Excel‐

Data/WPP2010_F02_METAINFO.xls

(19) UNAIDS. 2012 UNAIDS

Report

on

the

Global

AIDS

Epidemic. Geneva, UNAIDS, 2012.

(20) United Nations, Department of Economic and Social Affairs, Population Division. World Population

Prospects: The 2012 Revision, Provisional results (unpublished).

(21) Afghan Public Health Institute, Ministry of Public Health, Central Statistics Organization , ICF

Macro, Indian Institute of Health Management Research, and World Health Organization.

Afghanistan Mortality Survey 2010. Calverton, Maryland, USA: APHI/MoPH, CSO, ICF Macro,

IIHMR and WHO/EMRO, 2011.

(22) Central Statistics Organisation (CSO) and UNICEF. Afghanistan Multiple Indicator Cluster Survey

2010‐2011: Final Report . Kabul, Central Statistics Organisation and UNICEF, 2012.

(23) Chinese Center for Disease Control and Prevention. National Disease Surveillance System

monitoring causes

of

death

2010. Beijing, Military Medical Science Press, 2012.

(24) World Health Organization. Mortality Database. Available at:

http://www.who.int/healthinfo/mortality_data/en/index.html

(25) Mathers CD, Lopez AD, Murray CJL, Ezzati M, Jamison DT. The burden of disease and mortality by

condition: data, methods and results for 2001. Global burden of disease and risk factors. New York,

Oxford University Press, 2006. p. 45–240.

(26) Lozano R, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990

and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet , 2012,

380(9859):2095‐128.

(27) WHO, UNICEF, UNFPA, World Bank. Trends in maternal mortality: 1990 to 2010. Geneva: World Health Organization; 2012.

(28) China Ministry of Health‐Unpublished tabulations ‐‐ Vital Registration System cause‐of ‐death data

submitted annually to WHO.

(29) 全国疾病监测系统死因监测数据集 [National Disease Surveillance System monitoring causes of

death 2010]. Chinese Center for Disease Control and Prevention, Beijing, Military Medical Science

Press, 2012, ISBN 978‐7‐80245‐827‐7.

(30) 中国疾病监测报告[Cause‐of ‐death data from Chinese Disease Surveillance Points], China

Ministry of Health, 2004‐2009.

(31) Chinese Center for Disease Control and Prevention. CDC. Reported human rabies deaths 1950‐

2010. Chinese CDC: Beijing.

(32) Registrar General of India. Causes of Death in India in 2001‐2003. New Delhi, Registrar General of

India, Government of India, 2009.

(33) Jha P, Gajalakshmi V, Gupta PC, Kumar R, Mony P, Dhingra N et al. Prospective study of one

million deaths in India: rationale, design, and validation results. PLoS Med 2006 February;3(2):e18.




(34) Liu L, Johnson HL, Cousens S, Perin J, Scott S, Lawn JE, Rudan I, Campbell H, Cibulskis R, Li M,

Mathers C, Black RE, for the Child Health Epidemiology Reference Group of WHO and UNICEF.

Global, regional, and national causes of child mortality: an updated systematic analysis for 2010

with time trends since 2000. Lancet, 2012, 379:2151‐61.

(35) World Health Organization. Methodology for WHO mortality estimates. Available at:

http://www.who.int/healthinfo/statistics/mortality/en/index2.html

(36) Black RE, Cousens S, Johnson H et al. Global, Regional and National Causes of Child Mortality,

2008. Lancet, 2010, 375(9730):1969‐87.

(37) Johnson H, Liu L, Walker CF, Black RE. Estimating the distribution of causes of child deaths in high

mortality countries with incomplete death certification. Int J Epidemiol , 2010, 39(4):1103‐1114.

(38) Liu et al. National, regional and state‐level causes of child mortality in India in 2000‐2010: a

systematic sub‐national analysis. Under preparation, 2013.

(39) Bassani DG, Kumar R, Awasthi S et al. Causes of neonatal and child mortality in India: a nationally

representative mortality survey. Lancet , 2010, 376(9755):1853‐1860.

(40). Rudan I, Chan KY, Zhang JS et al. Causes of deaths in children younger than 5 years in China in

2008. Lancet , 2010, 375(9720):1083‐1089.

(41) World Health Organization. Global Tuberculosis Report 2012. Geneva, WHO, 2012.

(42) World Health Organization. World Malaria Report 2012. Geneva, WHO, 2012.

(43) World Health Organization. World Malaria Report 2008. Geneva, WHO, 2008.

(44) Chen S, Fricks J, Ferrari MJ. Tracking measles infection through non‐linear state space models.

Journal of the Royal Statistical Society Series C , 2011, 61 (1).

(45) Wolfson LJ, Grais RF, Luquero FJ, Birmingham ME, Strebel PM. Estimates of measles case fatality

ratios: a comprehensive review of community‐based studies. Int J Epidemiol , 2009, 38(1):192‐205.

(46) Joshi AB, Luman ET, Nandy R, Subedi BK, Liyanage JBL, Wierzba TF. Measles deaths in Nepal: estimating the national case–fatality ratio. Bulletin of the World Health Organization, 2009,

87(6):456–465.

(47) Sudfeld CR, Halsey NA. Measles case fatality ratio in India a review of community based studies.

Indian Pediatr , 2009, 46(11):983‐9.

(48) Wolfson LJ, Strebel PM, Gacic‐Dobo M, Hoekstra EJ, McFarland JW, Hersh BS. Has the 2005

measles mortality reduction goal been achieved? A natural history modelling study. Lancet , 2007,

369(9557):191‐200.

(49) World Health Organization. Progress in global control and regional elimination of measles, 2000‐

2011. Weekly epidemiological record, 2013, 88(3):29‐36.

(50) Cleveland WS, Loader CL. Smoothing by local regression: principles and methods. In: Haerdle W,

Schimek MG, editors. Statistical theory and computational aspects of smoothing. New York,

Springer, 1996, 10‐49.

(51) Hotez PJ, Bundy DA, Beegle K, Brooker S, Drake L, de Silva NR et al. Helminth infections: soil‐

transmitted helminth infections and schistosomiasis. In: Jamison DT, Breman JG, Measham AR,

Alleyne G, Evans D, Claeson M et al., eds. Disease control priorities in developing countries, 2nd

edit . New York, Oxford University Press, 2006: 467‐482.




(52) Van der Werf MJ, de Vlas SJ. Morbidity and infection with schistosomes or soil ‐transmitted

helminths. Rotterdam, Erasmus University, 2001.

(53) Vos T, Flaxman AD, Naghavi M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289

diseases and injuries, 1990–2010: a systematic analysis for the Global Burden of Disease Study

2010. The Lancet. 2012 Dec 13; 380: 2163–2196.

(54) Wilmoth JR, Mizoguchi N, Oestergaard MZ, Say L, Mathers CD, Zureick‐Brown S, Inoue M, Chou D.

A New Method for Deriving Global Estimates of Maternal Mortality. Statistics, Politics, and Policy.

3(2), DOI: 10.1515/2151‐7509.1038, July 2012

(55) Ferlay J, Shin H, Bray F, Foreman D, Mathers CD, Parkin DM. Estimates of worldwide burden of

cancer in 2008: Globocan 2008. International Journal of Cancer 2010;127(12):2893‐917.

(56) United Nations Office on Drugs and Crime. World Drug Report 2012. UNODC: Vienna. 2012

(57) World Health Organization 2013. Global status report on road safety 2013: supporting a decade of

action. Geneva, WHO, 2013.

(58) Khosravi A, Taylor R, Naghavi N, Lopez AD. Mortality in the Islamic Republic of Iran, 1964–2004.

Bulletin of

the

World

Health

Organization, 2007, 85:607‐14

(59) Porapakkham Y, Rao C, Pattaraarchachai J, Polprasert W, Vos T, Adair T et al. Estimated causes of

death in Thailand, 2005: implications for health policy. Population Health Metrics, 2010, 8:14.

(60) Ngo AD, Rao C, Hoa NP, Adair T, Chuc NTK. Mortality patterns in Vietnam, 2006: findings from a

national verbal autopsy survey. BMC Research Notes, 2010, 3:78.

(61) IHME. Health system access ref

(62) Kaufmann D, Kraay A, Mastruzzi M. , Massimo. Governance Matters VIII : Aggregate and

Individual Governance Indicators 1996–2008. World Bank 2009.

(63) Hughes BB, et al. The International Futures (IFs) modeling system, version 6.54. Frederick S.

Pardee Center for International Futures, Josef Korbel School of International Studies, University of Denver, www.ifs.du.edu.

(64) CRED. EM‐DAT: The CRED International Disaster Database. Belgium, Université Catholique de

Louvain, 2012.

(65) He H, Oguchi T, Zhou R, Zhang J, Qiao S. Damage and seismic intensity of the 1996 Lijiang

earthquake, Vhina: a GIS analysis. Technical report. Tokyo, Center for Spatial Information Science,

University of Tokyo, 2001. Available at: http://www.csis.u‐tokyo.ac.jp/english/dp/dp.html

(accessed 18 January 2008).

(66) Naghii MR. Public health impact and medical consequences of earthquakes. Pan American Journal

of Public Health, 2005, 18:216–221.

(67) Nishikiori N, Abe T, Costa DG, Dharmaratne SD, Kunii O, Moji K. Who died as a result of the

tsunami? Risk factors of mortality among internally displaced persons in Sri Lanka: a retrospective

cohort analysis. BMC Public Health, 2006, 6:73.

(68) Doocy S, Rofi A, Moodie C, Spring E, Bradley S, Burnham G et al. Tsunami mortality in Aceh

Province, Indonesia. Bulletin of the World Health Organization, 2007, 85:273–278.




(69) Heidelberg Institute on International Conflict Research. Conflict barometer. Department of

Political Science, University of Heidelberg, 2012. Available at:

http://www.hiik.de/en/konfliktbarometer/.

(70) Project Ploughshares. Armed conflicts report . Waterloo, Canada, Project Ploughshares, 2005.

Available at: http://www.ploughshares.ca/.

(71) Marshall MG, Gurr TR. Peace and

conflict

2005:

a global

survey

of

armed

conflicts,

self

‐

determination movements, and democracy . University of Maryland, Center for International

Development and Conflict Management, 2005.

(72) International Peace Research Institute. UCDP/PRIO Armed Conflict Dataset. Oslo, PRIO, 2009.

Available at: http://www.prio.no/CSCW/Datasets/Armed‐Conflict/ (accessed 2 November 2009).

(73) Murray CJ, King G, Lopez AD, Tomijima N, Krug EG. Armed conflict as a public health problem.

British Medical Journal , 2002, 324(7333):346‐349.

(74) Murray CJL, Lopez AD. The Global Burden of Disease: a comprehensive assessment of mortality

and disability from diseases, injuries and risk factors in 1990 and projected to 2020. Cambridge,

Harvard School of Public Health, 1996.

(75) Obermeyer Z, Murray CJL, Gakidou E. Fifty years of violent war deaths from Vietnam to Bosnia:

analysis of data from the world health survey programme. British Medical Journal , 2008,

336:1482‐6.

(76) Lacina B, Gleditsch NP. Monitoring trends in global combat: a new dataset of battle deaths. Eur J

Popul, 2005, 21:145‐166.

(77) Garfield, R, Blore J. Direct Conflict Deaths. Unpublished report prepared on behalf of the

Collective Violence Expert Group for the Global Burden of Disease Study, 2009.

(78) International Peace Research Institute 2012. UCDP/PRIO Battle‐Related Deaths Dataset v. 5‐2012b,

1989‐2011. Oslo, PRIO, 2013. Available at:

http://www.pcr.uu.se/research/ucdp/datasets/ucdp_battle‐related_deaths_dataset/ (accessed 4 February 2013).

(79) International Peace Research Institute 2012. UCDP/PRIO Non‐State Conflict Dataset v. 2.4‐2012,

1989‐2011. Oslo, UCDP, 2013. Available at:

http://www.pcr.uu.se/research/ucdp/datasets/ucdp_non‐state_conflict_dataset_/ (accessed 4

February 2013).

(80) International Peace Research Institute 2012. UCDP/PRIO One‐Sided Violence Dataset v. 1.4‐2012,

1989‐2011. Oslo, PRIO, 2013. Available at:

http://www.pcr.uu.se/research/ucdp/datasets/ucdp_one‐sided_violence_dataset/ (accessed 4

February 2013).

(81) Iraq Family Health Survey Study Group. Violence‐Related Mortality in Iraq from 2002 to 2006. N

Engl J Med , 2008, NEJMsa0707782.

(82) Iraq Body Count. Iraqi deaths from violence 2003–2011. Available at:

http://www.iraqbodycount.org/

(83) International Campaign to Ban Landmines. Landmine monitor. Available at: http://www.the‐

monitor.org/




(84) Hoeffler A. Dealing with the consequences of violent conflicts in Africa. Background Paper for the

African Development Bank, 2008. Available at: http://users.ox.ac.uk/~ball0144/consequences.pdf

(85) World Health Organization. European Programme for Intervention Epidemiology Training.

Retrospective mortality survey among the internally displaced population, Greater Darfur, Sudan,

August 2004. Geneva, World Health Organization, 2004. Available at:

http://www.who.int/disasters/repo/14652.pdf

(86) Office for the Coordination of Humanitarian Affairs ‐ Occupied Palestinian Territory. Protection of

Civilians: Casualty Database. Available at:

http://www.ochaopt.org/poc.aspx?id=1010002

(87) World Health Organization. World health report 2001. Mental Health: New Understanding, New

Hope. Geneva, World Health Organization, 2001.

(88) Krug EG, et al. World Report on violence and health. Geneva: World Health Organization, 2002.

(89) World Health Organization. World Health Statistics 2007 . Geneva: World Health Organization,

2007.

(90) Murray C, Lopez AD, Wang H. Mortality estimation

for

national

populations:

methods

and

applications. Seattle, University of Washington Press, 2012.

(91) Foreman KJ, Lozano R, Lopez AD, Murray CJL. Modeling causes of death: an integrated approach

using CODEm. Population Health Metrics. 2012; 10:1.




Annex Table A GHE cause categories and ICD‐10 codes

Code GHE cause name ICD-10 code

1 I. Communicable, maternal, perinatal andnutritional conditions

a A00-B99, G00-G04, N70-N73, J00-J22, H65-H68, O00-O99,P00-P96, E00-E02, E40-E46, E50-E64, D50-D53, D64.9, U04

2 A. Infec tious and parasi tic di seases A00-B99, G00, G03-G04, N70-N73

3 1. Tuberculosis A15-A19, B90

4 2. Sexually transmitted diseases (STDs)excluding HIV

A50-A64, N70-N73

5 a. Syphilis A50-A53

6 b. Chlamydia A55-A56

7 c. Gonorrhoea A54

8 d. Trichomoniasis A59

9 e. Other STDs A57-A58, A60-A64, N70-N73

10 3. HIV/AIDS B20-B24

11 4. Diarrhoeal diseases b A00, A01, A03, A04, A06-A09

12 5. Childhood-cluster diseases A33-A37, B0513 a. Whooping cough A37

14 b. Diphtheria A36

15 c. Measles B05

16 d. Tetanus A33-A35

17 6. Meningitis A39, G00, G03

18 7. Encephalitis b A83-A86, B94.1, G04

19 8. Hepatitis B B16-B19 (minus B17.1, B18.2)

20 9. Hepatitis C B17.1, B18.2

21 10. Parasitic and vector diseases A30, A71, A82, A90-A91, B50-B57, B65, B73, B74.0-B74.2

22 a. Malaria B50-B54, P37.3, P37.4

23 b. Trypanosomiasis B5624 c. Chagas disease B57

25 d. Schistosomiasis B65

26 e. Leishmaniasis B55

27 f. Lymphatic filariasis B74.0-B74.2

28 g. Onchocerciasis B73

29 h. Leprosy A30

30 i. Dengue A90-A91

31 j. Trachoma A71

32 k. Rabies A82

33 11. Intestinal nematode infections B76-B77, B79

34 a. Ascariasis B77

35 b. Trichuriasis B79

36 c. Hookworm disease B76

37 12. Other infectious diseases A02, A05, A20-A28, A31, A32, A38, A40-A49, A65-A70, A74- A79, A80-A81, A87-A89, A92-A99, B00-B04, B06-B15, B25-B49,B58-B60, B64, B66-B72, B74.3-B74.9, B75,B78, B80-B89, B91-B99 (minus B94.1)





38 B. Respiratory infections b

J00-J22, H65-H68,P23, U04

39 1. Lower respiratory infections J09-J22, P23, U04

40 2. Upper respiratory infections J00-J06

41 3. Otitis media H65-H68

42 C. Maternal condi tions O00-O99

43 1. Maternal haemorrhage O44-O46, O67, O72

44 2. Maternal sepsis O85-O86

45 3. Hypertensive disorders of pregnancy O10-O16

46 4. Obstructed labour O64-O66

47 5. Abortion O00-O07

48 6. Other maternal conditions O20-O43, O47-O63, O68-O71, O73-O75, O87-O99

49 D. Neonatal condi tions P00-P96 excl P37.3, P37.4

50 1. Preterm birth complications b

P05, P07, P22, P27-P28

51 2. Birth asphyxia and birth trauma b

P03, P10-P15, P20-P21, P24-P26, P29

52 3. Neonatal sepsis and infections P35-P39 (excluding P37.3, P37.4)

53 4. Other neonatal conditions P00-P02, P04, P08, P50-P96

54 E. Nutritional deficiencies E00-E02, E40-E46, E50-E64, D50-D53, D64.9

55 1. Protein-energy malnutrition E40-E46

56 2. Iodine deficiency E00-E02

47 3. Vitamin A deficiency E50

58 4. Iron-deficiency anaemia D50, D64.9

59 5. Other nutritional disorders D51-D53, E51-E64

60 II. Noncommunicable diseasesa C00-C97, D00-D48, D55-D64 (minus D 64.9), D65-D89, E03-

E07, E10-E16, E20-E34, E65-E88, F01-F99, G06-G98, H00-H61, H68-H93, I00-I99, J30-J98, K00-K92, N00-N64, N75-N98,L00-L98, M00-M99, Q00-Q99, X41-X42

b, X45

b

61 A. Malignant neopl asms C00-C97

62 1. Mouth and oropharynx cancersd

C00-C14

63 2. Oesophagus cancer d C15

64 3. Stomach cancer d C16

65 4. Colon and rectum cancersd

C18-C21

66 5. Liver cancer C22

67 6. Pancreas cancer C25

68 7. Trachea, bronchus and lung cancers C33-C34

69 8. Melanoma and other skin cancersd

C43-C44

70 9. Breast cancer d C50

71 10. Cervix uteri cancer d C53

72 11. Corpus uteri cancer

d

C54-C5573 12. Ovary cancer C56

74 13. Prostate cancer d C61

75 14. Bladder cancer d C67

76 15. Lymphomas and multiple myelomad

C81-C90, C96

77 16. Leukaemiad C91-C95

78 17. Other malignant neoplasmsd

C17, C23, C24, C26-C32, C37-C41, C45-C49, C51, C52,C57-C60, C62-C66, C68-C80, C97





79 B. Other neoplasms D00-D48

80 C. Diabetes mellitus E10-E14

81 D. Endocrine, blood , immune disorders D55-D64 (minus D64.9), D65-D89, E03-E07, E15-E34, E65-E88

82 E. Mental and behavioural disorders F04-F99, X41-X42c, X45

c

83 1. Unipolar depressive disorders F32-F33, F34.1

84 2. Bipolar affective disorder F30-F31

85 3. Schizophrenia F20-F29

86 4. Alcohol use disorders F10, X45c

87 5. Drug use disorders F11-F16, F18-F19, X41-X42c

88 6. Anxiety disorders F40-F44

89 7. Eating disorders F50

90 8. Pervasive developmental disorders F84

91 9. Childhood behavioural disorders F90-F92

92 10. Idiopathic intellectual disability F70-F79

93 11. Other mental and behavioural disorders F04-F09, F17, F34-F39 (minus F34.1), F45-F48, F51-F69, F80-

F83, F88-F89, F93-F99

94 F. Neurological condit ions F01-F03, G06 -G98

95 1. Alzheimer’s disease and other dementias F01-F03, G30-G31

96 2. Parkinson disease G20-G21

97 3. Epilepsy G40-G41

98 4. Multiple sclerosis G35

99 5. Migraine G43

100 6. Non-migraine headache G44

101 7. Other neurological conditions G06-G12, G23-G25, G36-G37, G45-G98

102 G. Sense organ diseases H00-H61, H69-H93

103 1. Glaucoma H40

104 2. Cataracts H25-H26

105 3. Refractive errors H49-H52

106 4. Macular degeneration H35.3

107 5. Other vision loss H30-H35 (minus H35.3), H53-H54

108 6. Other hearing loss H90-H91

109 7. Other sense organ disorders H00-H21, H27, H43-H47, H55-H61, H69-H83, H92-H93

110 H. Cardiovascular diseases I00-I99

111 1. Rheumatic heart disease I01-I09

112 2. Hypertensive heart disease I10-I15

113 3. Ischaemic heart diseasee I20-I25

114 4. Stroke I60-I69

115 5. Cardiomyopathy, myocarditis, endocarditis I30-I33, I38, I40, I42

116 6. Other cardiovascular diseasese

I00, I26-I28, I34-I37, I44-I51, I70-I99

117 I. Respiratory diseases J30-J98

118 1. Chronic obstructive pulmonary disease J40-J44

119 2. Asthma J45-J46

120 3. Other respiratory diseases J30-J39, J47-J98

121 J. Digestive diseases K20-K92





122 1. Peptic ulcer disease K25-K27

123 2. Cirrhosis of the liver K70, K74

124 3. Appendicitis K35-K37

125 4. Other digestive diseases K20-K22, K28-K31, K38-K66, K71-K73, K75-K92

126 K. Genitourinary diseases N00-N64, N75-N76, N80-N98

127 1. Kidney diseases N00-N19

128 2. Hyperplasia of prostate N40

129 3. Urolithiasis N20-N23

130 4. Other genitourinary disorders N25-N39, N41-N45, N47-N51

131 5. Infertility N46, N97

132 6. Gynecological diseases N60-N64, N75-N76, N80-N96, N98

133 L. Skin diseases L00-L98

134 M. Musculoskeletal diseases M00-M99

135 1. Rheumatoid arthritis M05-M06

136 2. Osteoarthritis M15-M19

137 3. Gout M10

138 4. Back and neck pain M45-M48, M50-M54

139 5. Other musculoskeletal disorders M00, M02, M08, M11-M13, M20-M43, M60-M99

140 N. Congenital anomalies Q00-Q99

141 1. Neural tube defects Q00, Q05

142 2. Cleft lip and cleft palate Q35-Q37

143 3. Down syndrome Q90

144 4. Congenital heart anomalies Q20-Q28

145 5. Other chromosomal anomalies Q91-Q99

146 6. Other congenital anomalies Q01-Q04, Q06-Q18, Q30-Q34, Q38-Q89

147 O. Oral condi tions K00-K14

148 1. Dental caries K00-K04, K06-K14

149 2. Periodontal disease K05

150 3. Edentulism —

151 III. Injuries V01-Y89

152 A. Unintenti onal injur iesf V01-X40, X43-X44, X46-59, Y40-Y86, Y88, Y89

153 1. Road injuryg V01-V04, V06, V09-V80, V87, V89, V99

154 2. Poisonings X40, X43-X44, X46-X49

155 3. Falls W00-W19

156 4. Fire, heat and hot substances X00-X19

157 5. Drownings W65-W74

158 6. Exposure to forces of nature X30-X39159 7. Other unintentional injuries Rest of V, W20-W64, W75-W99, X20-X29, X50-X59, Y40-Y86,

Y88, Y89

160 B. Intentional injur iesf X60-Y09, Y35-Y36, Y870, Y871

161 1. Self-harm X60-X84, Y870

162 2. Interpersonal violence X85-Y09, Y871

163 3. Collective violence and legalintervention

Y35-Y36






Annex Table B First ‐level categories for analysis of child causes of death

GBD cause name ICD-10 code

All causes A00-Y89

I. Communicable, maternal,

perinatal and nutritional

conditionsa

A00-B99, D50-D53, D64.9, E00-E02, E40-E64, G00, G03-G04, H65-H66, J00-J22, J85, N30, N34, N390, N70-N73, O00-P96, U04

HIV/AIDS B20-B24

Diarrhoeal diseases A00-A09

Pertussis A37

Tetanus A33-A35

Measles B05

Meningitis/encephalitis A39, A83-A87, G00, G03, G04

Malaria B50-B54, P37.3, P37.4

Acute respiratory infections H65-H66, J00-J22, J85, P23, U04

Prematurity P01.0, P01.1, P07, P22, P25-P28, P61.2, P77

Birth asphyxia & birth trauma P01.7-P02.1, P02.4-P02.6, P03, P10-P15, P20-P21, P24, P50, P90-P91

Sepsis and other infectiousconditions of the newborn

P35-P39 (exclude P37.3, P37.4)

Other Group I Remainder

II. Noncommunicable diseasesa C00-C97, D00-D48, D55-D64 (exclude D64.9), D65-D89, E03-E34, E65-E88, F01-

F99, G06-G98, H00-H61, H68-H93, I00-I99, J30-J84, J86-J98, K00-K92, L00-L98,M00-M99, N00-N28, N31-N32, N35-N64 (exclude N39.0), N75-N98, Q00-Q99

Congenital anomalies Q00-Q99

Other Group II Remainder

III. Injuries V01-Y89

a Deaths coded to “Symptoms, signs and ill-defined conditions” (780-799 in ICD-9 and R00-R99 in ICD-10) are distributed

proportionately to all causes within Group I and Group II.




Annex Table C Re‐assignment of ICD‐10 codes for certain neonatal deaths.

Cause Recode Cause Recode Cause Recode Cause Recode Cause Recode

A153 P370 D649 P614 I471 P291 J698 P249 K760 P788

A162 P370 D65 P60 I472 P291 J70 P24 K761 P788

A165 P370 D696 D694 I479 P291 J709 P249 K762 P788

A169 P370 D699 P549 I48 P29 J80 P22 K763 P788

A170 P370 E101 P702 I490 P291 J840 P258 K767 P788

A180 P370 E102 P702 I494 P291 J841 P258 K768 P788

A320 P372 E110 P702 I498 P291 J848 P258 K769 P788

A321 P372 E112 P702 I499 P291 J849 P258 K819 P788

A327 P372 E116 P702 I50 P29 J85 P28 K82 P78

A328 P372 E117 P702 I500 P290 J850 P288 K828 P788

A329 P372 E140 P702 I501 P290 J851 P288 K830 P788

A35 A33 E144 P702 I509 P290 J852 P288 K831 P788

A40 P36 E145 P702 I517 Q248 J860 P288 K838 P788

A401 P360 E147 P702 I518 Q248 J869 P288 K839 P788

A402 P361 E149 P702 I519 Q249 J90 P28 K85 P78

A403 P361 E343 P051 I60 P52 J930 P251 K868 P788

A408 P361 E86 P74 I603 P525 J931 P251 K869 P788

A409 P361 E87 P74 I607 P525 J938 P251 K904 P788

A41 P36 E870 P742 I608 P525 J939 P251 K909 P788

A410 P362 E871 P742 I609 P525 J940 P288 K920 P540

A412 P363 E872 P740 I61 P52 J941 P288 K922 P543

A413 P368 E874 P748 I610 P524 J942 P548 K928 P788

A415 P368 E875 P743 I612 P524 J948 P288 K929 P789

A418 P368 E876 P743 I615 P524 J96 P28 N133 Q620

A419 P369 E877 P744 I616 P524 J960 P285 N139 Q623

B00 P35 E878 P744 I618 P524 J961 P285 N17 P96

B000 P352 F322 P914 I619 P524 J969 P285 N170 P960

B004 P352 F328 P914 I620 P528 J980 P288 N171 P960

B007 P352 F329 P914 I629 P529 J981 P281 N172 P960

B008 P352 F439 P209 I632 P529 J982 P250 N179 P960

B009 P352 G91 Q03 I633 P529 J984 P288 N180 P960

B01 P35 G911 Q039 I634 P529 J985 P288 N188 P960

B010 P358 G912 Q039 I635 P529 J986 P288 N189 P960

B011 P358 G913 Q039 I638 P529 J988 P288 N19 P96

B012 P358 G919 Q039 I639 P529 J989 P289 N359 Q643B018 P358 G930 Q046 I64 P52 K220 Q395 N390 P393

B019 P358 G931 P219 I671 I607 K311 Q400 N433 P835

B059 P358 G936 P524 J12 P23 K44 Q79 N883 P010

B060 P350 G952 P025 J120 P230 K440 Q790 R001 P209

B068 P350 I050 Q232 J121 P230 K441 Q790 R011 P298

continued




Annex Table C (continued): Re-assignment of ICD-10 codes for certain neonatal deaths.

Cause Recode Cause Recode Cause Recode Cause Recode Cause Recode

B069 P350 I059 Q238 J128 P230 K449 Q790 R030 P292

B09 P35 I071 Q228 J129 P230 K561 Q438 R040 P548

B25 P35 I080 Q238 J13 P23 K562 Q438 R042 P269B250 P351 I340 Q233 J14 P23 K565 Q433 R048 P548

B251 P351 I348 Q238 J15 P23 K566 P769 R049 P548

B258 P351 I35 Q23 J150 P236 K57 Q43 R05 P28

B259 P351 I350 Q230 J151 P235 K593 Q431 R060 P228

B270 P358 I351 Q231 J152 P232 K625 P542 R064 P228

B370 P375 I352 Q238 J153 P233 K631 P780 R068 P228

B371 P375 I359 Q238 J154 P236 K633 P788 R090 P219

B372 P375 I370 Q221 J155 P234 K65 P78 R092 P285

B373 P375 I379 Q223 J156 P236 K650 P781 R160 Q447

B374 P375 I38 I42 J157 P236 K659 P781 R162 Q447

B375 P375 I42 I42 J158 P236 K660 Q433 R230 Q249

B376 P375 I420 I424 J159 P236 K661 P548 R509 P819

B377 P375 I421 Q248 J16 P23 K720 P788 R568 P90

B378 P375 I422 I424 J18 P23 K729 P788 R571 P741

B379 P375 I429 I424 J180 P239 K732 P788 R58 P54

B509 P373 I442 Q246 J181 P239 K745 P788 R601 P833

B54 P37 I443 Q246 J188 P239 K746 P788 R628 P059

B582 P371 I455 Q246 J189 P239 K750 P788 R629 P059

B589 P371 I458 Q246 J386 Q318 K752 P788 R630 P929

D500 P549 I459 Q246 J439 P250 K758 P788 R638 P929

D609 D610 I460 P291 J69 P24 K759 P788 R75 B24

D62 P61 I469 P291 J690 P249




Annex Table D Country groupings used for regional tabulations

D.1 WHO Regions and Member States

WHO African Region

Algeria, Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad, Comoros, Congo, Côte d'Ivoire, Democratic Republic of the Congo, Equatorial Guinea,

Eritrea, Ethiopia, Gabon, Gambia, Ghana, Guinea, Guinea‐Bissau, Kenya, Lesotho, Liberia, Madagascar,

Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, Sao Tome and

Principe, Senegal, Seychelles, Sierra Leone, South Africa, Swaziland, Togo, Uganda, United Republic of

Tanzania, Zambia, Zimbabwe

WHO Region of the Americas

Antigua and Barbuda, Argentina, Bahamas, Barbados, Belize, Bolivia (Plurinational State of), Brazil,

Canada, Chile, Colombia, Costa Rica, Cuba, Dominica, Dominican Republic, Ecuador, El Salvador,

Grenada, Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru,

Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Suriname, Trinidad and Tobago,

United States of America, Uruguay, Venezuela (Bolivarian Republic of)

WHO South-East Asia Region

Bangladesh, Bhutan, Democratic People's Republic of Korea, India, Indonesia, Maldives, Myanmar,

Nepal, Sri Lanka, Thailand, Timor‐Leste

WHO European Region

Albania, Andorra, Armenia, Austria, Azerbaijan, Belarus, Belgium, Bosnia and Herzegovina, Bulgaria,

Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Georgia, Germany, Greece,

Hungary, Iceland, Ireland, Israel, Italy, Kazakhstan, Kyrgyzstan, Latvia, Lithuania, Luxembourg, Malta,

Monaco, Montenegro, Netherlands, Norway, Poland, Portugal, Republic of Moldova, Romania, Russian

Federation, San Marino, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Tajikistan, The former

Yugoslav Republic of Macedonia, Turkey, Turkmenistan, Ukraine, United Kingdom, Uzbekistan

WHO Eastern Mediterranean Region

Afghanistan, Bahrain, Djibouti, Egypt, Iran (Islamic Republic of), Iraq, Jordan, Kuwait, Lebanon, Libya,

Morocco, Oman, Pakistan, Qatar, Saudi Arabia, Somalia, South Sudan, Sudan, Syrian Arab Republic,

Tunisia, United Arab Emirates, Yemen

WHO Western Pacific Region

Australia, Brunei Darussalam, Cambodia, China, Cook Islands, Fiji, Japan, Kiribati, Lao People's

Democratic Republic, Malaysia, Marshall Islands, Micronesia (Federated States of), Mongolia, Nauru,

New Zealand, Niue, Palau, Papua New Guinea, Philippines, Republic of Korea, Samoa, Singapore, Solomon Islands, Tonga, Tuvalu, Vanuatu, Viet Nam




D.2 Countries grouped by WHO Region and average income per capita*

High income Andorra, Australia, Austria, Bahamas, Bahrain, Barbados, Belgium, Brunei Darussalam Canada, Croatia,

Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany Greece, Hungary, Iceland, Ireland,

Israel, Italy, Japan, Kuwait, Luxembourg, Malta, Monaco, Netherlands, ,New Zealand, Norway, Oman, Poland, Portugal, Qatar, Republic of Korea, Saint Kitts and Nevis, San Marino, Saudi Arabia, Singapore,

Slovakia, Slovenia, Spain, Sweden, Switzerland, Trinidad and Tobago, United Arab Emirates, United

Kingdom, United States of America

Low and middle income WHO African Region

Algeria, Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African

Republic, Chad, Comoros, Congo, Côte d'Ivoire, Democratic Republic of the Congo, Equatorial Guinea**,

Eritrea, Ethiopia, Gabon, Gambia, Ghana, Guinea, Guinea‐Bissau, Kenya, Lesotho, Liberia, Madagascar,

Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, Sao Tome and

Principe, Senegal, Seychelles, Sierra Leone, South Africa, Swaziland, Togo, Uganda, United Republic of Tanzania, Zambia, Zimbabwe

WHO Region of the Americas

Antigua and Barbuda, Argentina, Belize, Bolivia (Plurinational State of), Brazil, Chile, Colombia, Costa

Rica, Cuba, Dominica, Dominican Republic, Ecuador, El Salvador, Grenada Guatemala, Guyana, Haiti,

Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, Saint Lucia, Saint Vincent and the

Grenadines, Suriname, Uruguay, Venezuela (Bolivarian Republic of)

WHO South-East Asia Region

Bangladesh, Bhutan, Democratic People's Republic of Korea, India, Indonesia, Maldives, Myanmar,

Nepal, Sri Lanka, Thailand, Timor‐Leste

WHO European Region

Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Georgia, Kazakhstan,

Kyrgyzstan, Latvia, Lithuania, Montenegro, Republic of Moldova, Romania, Russian Federation, Serbia,

Tajikistan, The former Yugoslav Republic of Macedonia, Turkey, Turkmenistan, Ukraine, Uzbekistan

WHO Eastern Mediterranean Region

Afghanistan, Djibouti, Egypt, Iran (Islamic Republic of), Iraq, Jordan, Lebanon, Libya, Morocco, Pakistan,

Somalia, South Sudan, Sudan, Syrian Arab Republic, Tunisia, Yemen

WHO Western Pacific RegionCambodia, China, Cook Islands, Fiji, Kiribati, Lao People's Democratic Republic, Malaysia, Marshall

Islands, Micronesia (Federated States of), Mongolia, Nauru, Niue, Palau ,Papua New Guinea, Philippines,

Samoa, Solomon Islands, Tonga, Tuvalu, Vanuatu, Viet Nam

* This regional grouping classifies WHO Member States according to the World Bank income categories for the year 2011

(World Bank list of economies, July 2012) and the WHO region.

** Equatorial Guinea is classified by the World Bank as high income, it is kept here with upper middle income to avoid a

regional grouping containing only one country and because its mortality profile is not dissimilar to neighbouring countries.




D.3 World Bank income grouping*

Low income

Afghanistan, Bangladesh, Benin, Burkina Faso, Burundi, Cambodia, Central African Republic, Chad

Comoros, Democratic People's Republic of Korea, Democratic Republic of the Congo, Eritrea, Ethiopia

Gambia, Guinea, Guinea‐Bissau, Haiti, Kenya, Kyrgyzstan, Liberia, Madagascar, Malawi, Mali, Mauritania, Mozambique, Myanmar, Nepal, Niger Rwanda, Sierra Leone, Somalia, Tajikistan, Togo, Uganda, United

Republic of Tanzania, Zimbabwe

Lower middle income

Albania, Armenia, Belize, Bhutan, Bolivia (Plurinational State of), Cameroon, Cape Verde, Congo, Côte

d'Ivoire, Djibouti, Egypt, El Salvador, Fiji, Georgia, Ghana, Guatemala, Guyana, Honduras, India,

Indonesia, Iraq, Kiribati, Lao People's Democratic Republic, Lesotho, Marshall Islands, Micronesia

(Federated States of), Mongolia, Morocco, Nicaragua, Nigeria, Pakistan, Papua New Guinea, Paraguay,

Philippines, Republic of Moldova, Samoa, Sao Tome and Principe, Senegal, Solomon Islands, South

Sudan, Sri Lanka, Sudan, Swaziland, Syrian Arab Republic, Timor‐Leste, Tonga, Ukraine, Uzbekistan,

Vanuatu, Viet Nam, Yemen Zambia

Upper middle income

Algeria, Angola, Antigua and Barbuda, Argentina, Azerbaijan, Belarus, Bosnia and Herzegovina,

Botswana, Brazil, Bulgaria, Chile, China, Colombia, Cook Islands, Costa Rica, Cuba, Dominica, Dominican

Republic, Ecuador, Equatorial Guinea**, Gabon, Grenada, Iran (Islamic Republic of), Jamaica, Jordan,

Kazakhstan, Latvia, Lebanon, Libya, Lithuania, Malaysia, Maldives, Mauritius, Mexico Montenegro,

Namibia, Nauru, Niue, Palau, Panama, Peru, Romania, Russian Federation, Saint Lucia, Saint Vincent and

the Grenadines, Serbia, Seychelles, South Africa, Suriname, Thailand, The former Yugoslav Republic of

Macedonia, Tunisia, Turkey, Turkmenistan, Tuvalu, Uruguay, Venezuela (Bolivarian Republic of)

High income

Andorra, Australia, Austria, Bahamas, Bahrain, Barbados, Belgium, Brunei Darussalam Canada, Croatia,


Israel, Italy, Japan, Kuwait, Luxembourg, Malta, Monaco, Netherlands, ,New Zealand, Norway, Oman,

Poland, Portugal, Qatar, Republic of Korea, Saint Kitts and Nevis, San Marino, Saudi Arabia, Singapore,



* This regional grouping classifies WHO Member States according to the World Bank income categories for the year 2011

(World Bank list of economies, July 2012)






D.4 World Bank Regions

High income

Andorra, Australia, Austria, Bahamas, Bahrain, Barbados, Belgium, Brunei Darussalam Canada, Croatia,


Israel, Italy, Japan, Kuwait, Luxembourg, Malta, Monaco, Netherlands, ,New Zealand, Norway, Oman, Poland, Portugal, Qatar, Republic of Korea, Saint Kitts and Nevis, San Marino, Saudi Arabia, Singapore,



East Asia and Pacific

Cambodia, China, Cook Islands, Democratic People's Republic of Korea, Fiji, Indonesia, Kiribati, Lao

People's Democratic Republic, Malaysia, Marshall Islands, Micronesia (Federated States of), Mongolia,

Myanmar, Nauru, Niue, Palau, Papua New Guinea, Philippines, Samoa, Solomon Islands, Thailand,

Timor‐Leste, Tonga, Tuvalu, Vanuatu, Viet Nam

Europe and Central Asia

Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Georgia, Kazakhstan,

Kyrgyzstan, Latvia, Lithuania, Montenegro Republic of Moldova, Romania, Russian Federation, Serbia,

Tajikistan, The former Yugoslav Republic of Macedonia, Turkey, Turkmenistan, Ukraine, Uzbekistan

Latin America and Caribbean

Antigua and Barbuda, Argentina, Belize, Bolivia (Plurinational State of), Brazil, Chile, Colombia, Costa

Rica, Cuba, Dominica, Dominican Republic, Ecuador, El Salvador, Grenada, Guatemala, Guyana, Haiti,

Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, Saint Lucia, Saint Vincent and the

Grenadines, Suriname, Uruguay, Venezuela (Bolivarian Republic of)

Middle East and North Africa

Algeria, Djibouti, Egypt, Iran (Islamic Republic of), Iraq, Jordan, Lebanon ,Libya, Morocco, Syrian Arab

Republic, Tunisia, Yemen

South Asia

Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, Sri Lanka

Sub-Saharan Africa

Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad,

Comoros, Congo, Côte d'Ivoire, Democratic Republic of the Congo, Equatorial Guinea**, Eritrea,

Ethiopia, Gabon, Gambia, Ghana, Guinea, Guinea‐Bissau, Kenya, Lesotho, Liberia, Madagascar, Malawi,

Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, Sao Tome and Principe,

Senegal, Seychelles, Sierra Leone, Somalia, South Africa, South Sudan, Sudan, Swaziland, Togo, Uganda, United Republic of Tanzania, Zambia, Zimbabwe






D.5 Millennium Development Goal (MDG) Regions

Developed regions Albania, Andorra, Australia, Austria, Belarus, Belgium, Bosnia and Herzegovina, Bulgaria, Canada,

Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland,

Ireland, Israel, Italy, Japan, Latvia, Lithuania, Luxembourg, Malta, Monaco, Montenegro, Netherlands, New Zealand, Norway, Poland, Portugal, Republic of Moldova, Romania, Russian Federation, San

Marino, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, The former Yugoslav Republic of

Macedonia, Ukraine, United Kingdom, United States of America

Developing regions Caucasus and Central Asia

Armenia, Azerbaijan, Georgia, Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan, Uzbekistan

Eastern Asia China, Democratic People's Republic of Korea, Mongolia, Republic of Korea

Latin America and the CaribbeanAntigua and Barbuda, Argentina, Bahamas, Barbados, Belize, Bolivia (Plurinational State of), Brazil, Chile,

Colombia, Costa Rica, Cuba, Dominica, Dominican Republic, Ecuador, El Salvador, Grenada, Guatemala,

Guyana, Haiti, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, Saint Kitts and Nevis,

Saint Lucia, Saint Vincent and the Grenadines, Suriname, Trinidad and Tobago, Uruguay, Venezuela

(Bolivarian Republic of)

Northern Africa Algeria, Egypt, Libya, Morocco, Tunisia

Oceania

Cook Islands, Fiji, Kiribati, Marshall Islands, Micronesia (Federated States of), Nauru, Niue, Palau, Papua

New Guinea, Samoa, Solomon Islands, Tonga, Tuvalu, Vanuatu

South-eastern Asia

Brunei Darussalam, Cambodia, Indonesia, Lao People's Democratic Republic, Malaysia, Myanmar,

Philippines, Singapore, Thailand, Timor‐Leste, Viet Nam

Southern Asia

Afghanistan, Bangladesh, Bhutan, India, Iran (Islamic Republic of), Maldives, Nepal, Pakistan, Sri Lanka

Sub-Saharan Africa

Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad,

Comoros, Congo, Côte d'Ivoire, Democratic Republic of the Congo, Djibouti, Equatorial Guinea, Eritrea,

Ethiopia, Gabon, Gambia, Ghana, Guinea, Guinea‐Bissau, Kenya, Lesotho, Liberia, Madagascar, Malawi,

Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, Sao Tome and Principe,

Senegal, Seychelles, Sierra Leone, Somalia, South Africa, South Sudan, Sudan, Swaziland, Togo, Uganda,

United Republic of Tanzania, Zambia, Zimbabwe

Western Asia Bahrain, Iraq, Jordan, Kuwait, Lebanon, Oman, Qatar Saudi Arabia, Syrian Arab

Republic, Turkey, United Arab Emirates, Yemen




Annex Table E Mapping of India MDS categories to GHE causes

MDS

Cause Million Death Study Cause Category

GHE

causes Comment

Communicable, maternal, perinatal and nutritional conditions

1A01 Tuberculosis 3

1B01 Syphilis 5

1B02 Other sexually transmitted infections

(excl. HIV/AIDS)

9 Other STDs estimated according to GBD 2010 cause fractions

1C01 HIV/AIDS 10

1D01 Diarrhoeal diseases 11

1E01 Tetanus 16

1E02 Measles 15

1E03 Other vaccine preventable diseases 13, 14

1F01 Meningitis/encephalitis 17, 18 Apportioned to according to GBD 2010 cause fractions

1F02 Rabies 32

1G01 Hepatitis 19, 20 Apportioned to according to GBD 2010 cause fractions

1H01 Malaria 22 WHO malaria mortality estimates used

1I01 Protozoal diseases 26

1I02 Leprosy 29

1I03 Arthropod‐borne viral fevers 30

1I04 Trachoma 31

1I05 Helminthiases 34

1J01 Acute respiratory infections 39‐41 Apportioned to according to GBD 2010 cause fractions

1K01 Severe Systemic Infection 37

1K02 Severe Localized Infection 37 Acute bacterial sepsis

1L01 Other infectious diseases 37

1M01 Maternal haemorrhage 43

1M02 Maternal sepsis 44

1M03 Hypertensive disorders of pregnancy 45

1M04 Obstructed labour 46

1M05 Abortion 47

1M06 Other maternal conditions 48

1N01 Low birth weight/preterm 50

1N02 Birth asphyxia and birth trauma 51

1N03 Other perinatal conditions 52, 53 Apportioned using WHO‐CHERG cause fractions

1O01 Protein‐energy malnutrition 55

1O02 Iron, vitamin deficiencies and

nutritional anaemias

56‐59 Apportioned to according to GBD 2010 cause fractions

1P01 Fever of unknown origin Redistributed pro‐rata across infectious disease categories

Noncommunicable diseases

2A Neoplasms 62‐79 Replaced by WHO/IARC cancer estimates

2B01 Diabetes mellitus 80

2C01 Endocrine and immune disorders 81




2D01 Epilepsy 97

2D02 Other neuropsychiatric disorders 83‐93,

95, 96,

98‐101

Apportioned to according to GBD 2010 cause fractions

2F01 Skin diseases 133

2F02 Musculoskeletal disorders 135‐139 Apportioned to according to GBD 2010 cause fractions

2F03 Sense organ disorders 103‐109 Apportioned to according to GBD 2010 cause fractions

2F04 Oral conditions 150

2G01 Rheumatic heart disease 111

2G02 Ischaemic heart diseases 113

2G03 Hypertensive heart diseases 112

2G04 Cerebrovascular disease 114

2G05 Heart failure Redistributed pro‐rata across cardiovascular cause categories excluding

cerebrovascular disease

2G06 Other cardiovascular diseases 115, 116 Apportioned to according to GBD 2010 cause fractions

2H01 Asthma and chronic obstructive

pulmonary disease

118, 119 Apportioned to according to GBD 2010 cause fractions

2H02 Other chronic respiratory diseases 120

2J01 Gastro‐oesophageal diseases 122

2J02 Lliver and alcohol related diseases 86, 123,

125, 154

Apportioned to alcohol use disorders, liver cirrhosis, other

gastrointestinal, and accidental poisoning according to GBD 2010 cause

fractions

2J03 Other digestive diseases 124, 125 Apportioned to according to GBD 2010 cause fractions

2K01 Nephritis and nephrosis 127

2K02 Other genitourinary system diseases 128‐132 Apportioned to according to GBD 2010 cause fractions

2L01 Congenital anomalies 141‐146 Apportioned to according to GBD 2010 cause fractions

Injuries

3A01 Transport accidents 153, 159 Non‐road transport injury estimated using GBD 2010 analysis

3A02 Poisonings 154

3A03 Falls 155

3A04 Fires 156

3A05 Drownings 157

3A06 Venomous snakes, animals and plants 159

3A07 Other unintentional injuries 159

3B01 Self ‐inflicted injuries (suicide) 161

3B02 War, violence and other intentional

injuries 162

3C01 Undetermined Intent Redistributed pro‐rata across intentional & unintentional injury causes.

Symptoms, signs and Ill‐defined conditions

4A01 Senility Redistributed pro‐rata across non‐injury cause categories.

4A02 Other Ill‐defined and abnormal

findings Redistributed pro‐rata across non‐injury cause categories.

4A03 Unspecified deaths Redistributed pro‐rata across all cause categories.




Annex Table F Methods used for estimation of child and adult

mortality levels, and causes of death, by country, 2000‐2011

Mortality method groups:

A: Life tables based on death rates computed from vital registration data.

B: Projection of life table parameters l5 and l60 from adjusted vital registration data, smoothed with

moving average, projected using modified logit system with latest available year's lx as standard;

child mortality from the UN‐IGME.

C: Life tables based on death rates computed from neighbouring regional vital registration data.

D: Life tables based on UNPD’s World Population Prospects – the 2010 revision, and child mortality

estimates from the UN‐IGME.

E: Life tables based on UNPD’s World Population Prospects – the 2010 revision, updated with the

latest HIV/AIDS mortality from UNAIDS and child mortality estimates from the UN‐IGME.

F: Life tables using method E together with unpublished draft updates provided by UN Population

Division (see text).

Abbreviations

VA Verbal autopsy

VR Vital (death) registration

Note (a): WHO and UN Interagency cause‐specific estimates for all Member States as documented in

Section X above.

Country

All -cause

mortalitymethod

Under 5 child cause ofdeath method

Cause of deathmethods fo r ages 5+

Latest

availableVR data

Average

useability2000+

Afghanistan F VA multicause models GBD 2010 plus (a)

Albania A VR multicause models GBD 2010 plus (a) 2004 55%

Algeria D VA multicause models GBD 2010 plus (a)

Andorra C VR multicause models GBD 2010 plus (a)

Angola E VA multicause models GBD 2010 plus (a)

Antigua and Barbuda A VR data GBD 2010 plus (a)

Argentina B VR data GBD 2010 plus (a) 2010 79%

Armenia A VR multicause models GBD 2010 plus (a) 2011 66%

Australia B VR data VR data 2011 95%

Austria B VR data VR data 2011 90%

Azerbaijan A VA multicause models GBD 2010 plus (a) 2007 84%

Bahamas B VR data GBD 2010 plus (a)

Bahrain B VR data GBD 2010 plus (a)

Bangladesh D VA multicause models GBD 2010 plus (a)

Barbados B VR data GBD 2010 plus (a)

Belarus B VR multicause models VR data 2009 88%




Country

All -causemortalitymethod



LatestavailableVR data

Averageuseability

2000+

Belgium B VR data VR data 2009 88%

Belize B VR data GBD 2010 plus (a)

Benin E VA multicause models GBD 2010 plus (a)

Bhutan D VA multicause models GBD 2010 plus (a)

Bolivia D VA multicause models GBD 2010 plus (a)

Bosnia and Herzegovina B VR multicause models GBD 2010 plus (a)

Botswana E VA multicause models GBD 2010 plus (a)

Brazil A VR data VR data 2010 76%

Brunei Darussalam A VR multicause models GBD 2010 plus (a)

Bulgaria B VR data GBD 2010 plus (a) 2011 79%

Burkina Faso E VA multicause models GBD 2010 plus (a)

Burundi E VA multicause models GBD 2010 plus (a)

Cambodia D VA multicause models GBD 2010 plus (a)

Cameroon E VA multicause models GBD 2010 plus (a)

Canada B VR data VR data 2009 94%

Cape Verde A VR multicause models GBD 2010 plus (a)

Central African Republic E VA multicause models GBD 2010 plus (a)

Chad E VA multicause models GBD 2010 plus (a)

Chile B VR data VR data 2009 94%

China F

National VA model basedon subnational Chinesestudies only GBD 2010 plus (a)

Colombia A VR data VR data 2009 89%

Comoros D VA multicause models GBD 2010 plus (a)

Congo E VA multicause models GBD 2010 plus (a)

Cook Islands B VR multicause models GBD 2010 plus (a)

Costa Rica A VR data VR data 2011 87%

Cote d'Ivoire E VA multicause models GBD 2010 plus (a)

Croatia B VR data VR data 2011 87%

Cuba B VR data VR data 2010 90%

Cyprus B VR multicause models VR data 2011 73%

Czech Republic B VR data VR data 2011 88%

Democratic People'sRepublic of Korea D VA multicause models GBD 2010 plus (a)

Democratic Republic of theCongo E VA multicause models GBD 2010 plus (a)

Denmark B VR data VR data 2011 87%

Djibouti E VA multicause models GBD 2010 plus (a)

Dominica B VR data GBD 2010 plus (a)

Dominican Republic A VA multicause models GBD 2010 plus (a)

Ecuador A VR multicause models GBD 2010 plus (a) 2010 59%

Egypt B VR multicause models GBD 2010 plus (a) 2011 61%




Country





Averageuseability

2000+

El Salvador A VR multicause models GBD 2010 plus (a) 2009 58%

Equatorial Guinea E VA multicause models GBD 2010 plus (a)

Eritrea E VA multicause models GBD 2010 plus (a)

Estonia B VR data VR data 2011 94%

Ethiopia E VA multicause models GBD 2010 plus (a)

Fiji D VR multicause models GBD 2010 plus (a)

Finland B VR data VR data 2011 97%

France B VR data VR data 2009 85%

Gabon E VA multicause models GBD 2010 plus (a)

Gambia E VA multicause models GBD 2010 plus (a)

Georgia A VR multicause models GBD 2010 plus (a) 2010 53%

Germany B VR data VR data 2011 87%

Ghana E VA multicause models GBD 2010 plus (a)

Greece B VR data GBD 2010 plus (a) 2010 75%

Grenada B VR data GBD 2010 plus (a)

Guatemala A VA multicause models GBD 2010 plus (a) 2009 73%

Guinea E VA multicause models GBD 2010 plus (a)

Guinea-Bissau E VA multicause models GBD 2010 plus (a)

Guyana A VR data GBD 2010 plus (a)

Haiti E VA multicause models GBD 2010 plus (a)

Honduras D VR multicause models GBD 2010 plus (a)

Hungary B VR data VR data 2011 94%

Iceland B VR data VR data 2009 94%

India AState-level Indian-specificVA model GBD 2010 plus (a)

Indonesia D VA multicause models GBD 2010 plus (a)

Iran (Islamic Republic of) D VA multicause models GBD 2010 plus (a)

Iraq D VA multicause models GBD 2010 plus (a)

Ireland B VR data VR data 2010 94%

Israel B VR data VR data 2010 90%

Italy B VR data VR data 2010 90%

Jamaica A VR multicause models GBD 2010 plus (a)

Japan B VR data VR data 2011 89%

Jordan D VR multicause models GBD 2010 plus (a)

Kazakhstan A VA multicause models VR data 2010 83%

Kenya E VA multicause models GBD 2010 plus (a)

Kiribati A VA multicause models GBD 2010 plus (a)

Kuwait B VR data VR data 2011 87%

Kyrgyzstan A VA multicause models VR data 2010 90%

Lao People's DemocraticRepublic D VA multicause models GBD 2010 plus (a)




Country





Averageuseability

2000+

Latvia B VR data VR data 2010 92%

Lebanon D VR multicause models GBD 2010 plus (a)

Lesotho E VA multicause models GBD 2010 plus (a)

Liberia E VA multicause models GBD 2010 plus (a)

Libyan Arab Jamahiriya D VR multicause models GBD 2010 plus (a)

Lithuania B VR data VR data 2010 94%

Luxembourg B VR data GBD 2010 plus (a)

Madagascar D VA multicause models GBD 2010 plus (a)

Malawi E VA multicause models GBD 2010 plus (a)

Malaysia A VR multicause models GBD 2010 plus (a)

Maldives A VR multicause models GBD 2010 plus (a)

Mali E VA multicause models GBD 2010 plus (a)

Malta B VR data GBD 2010 plus (a)

Marshall Islands A VA multicause models GBD 2010 plus (a)

Mauritania D VA multicause models GBD 2010 plus (a)

Mauritius B VR data VR data 2011 90%

Mexico B VR data VR data 2010 95%

Micronesia (FederatedStates of) D VA multicause models GBD 2010 plus (a)

Monaco C VR multicause models GBD 2010 plus (a)

Mongolia B VA multicause models GBD 2010 plus (a)

Montenegro B VR data GBD 2010 plus (a) 2009 70%

Morocco D VA multicause models GBD 2010 plus (a)

MozambiqueE

VA multicause models GBD 2010 plus (a)Myanmar D VA multicause models GBD 2010 plus (a)

Namibia D VA multicause models GBD 2010 plus (a)

Nauru D VA multicause models GBD 2010 plus (a)

Nepal B VA multicause models GBD 2010 plus (a)

Netherlands B VR data VR data 2011 86%

New Zealand A VR data VR data 2009 97%

Nicaragua D VR multicause models GBD 2010 plus (a)

Niger E VA multicause models GBD 2010 plus (a)

Nigeria A VA multicause models GBD 2010 plus (a)

Niue B VR multicause models GBD 2010 plus (a)

Norway D VR data VR data 2011 89%

Oman D VR multicause models GBD 2010 plus (a)

Pakistan B VA multicause models GBD 2010 plus (a)

Palau A VR multicause models GBD 2010 plus (a)

Panama D VR data VR data 2009 80%

Papua New Guinea A VA multicause models GBD 2010 plus (a)

Paraguay A VR multicause models GBD 2010 plus (a)




Country





Averageuseability

2000+

Peru A VR multicause models GBD 2010 plus (a)

Philippines B VA multicause models VR data 2008 83%

Poland B VR data GBD 2010 plus (a) 2011 74%

Portugal B VR data VR data 2011 82%

Qatar B VR multicause models GBD 2010 plus (a) 2009 74%

Republic of Korea B VR data VR data 2011 85%

Republic of Moldova B VR data VR data 2011 88%

Romania B VR data VR data 2011 92%

Russian Federation E VR multicause models VR data 2010 95%

Rwanda B VA multicause models GBD 2010 plus (a)

Saint Kitts and Nevis B VR data GBD 2010 plus (a)

Saint Lucia B VR data GBD 2010 plus (a)

Saint Vincent and the

Grenadines D VR data GBD 2010 plus (a)

Samoa B VR multicause models GBD 2010 plus (a)

San Marino D VR data GBD 2010 plus (a)

Sao Tome and Principe D VA multicause models GBD 2010 plus (a)

Saudi Arabia D VR multicause models GBD 2010 plus (a)

Senegal B VA multicause models GBD 2010 plus (a)

Serbia B VR data VR data 2011 72%

Seychelles E VR multicause models GBD 2010 plus (a)

Sierra Leone B VA multicause models GBD 2010 plus (a)

Singapore B VR data VR data 2011 74%

SlovakiaB

VR data VR data2010 94%

Slovenia D VR data VR data 2010 89%

Solomon Islands D VA multicause models GBD 2010 plus (a)

Somalia A VA multicause models GBD 2010 plus (a)

South Africa #N/A VA multicause models GBD 2010 plus (a) 2009 68%

Spain B VR data VR data 2011 89%

Sri Lanka A VR multicause models GBD 2010 plus (a) 2006 55%

Sudan D VA multicause models GBD 2010 plus (a)

Suriname B VR data GBD 2010 plus (a)

Swaziland E VA multicause models GBD 2010 plus (a)

Sweden B VR data VR data 2010 89%

Switzerland B VR data VR data 2010 89%

Syrian Arab Republic D VR multicause models GBD 2010 plus (a)

Tajikistan A VA multicause models GBD 2010 plus (a)

Thailand A VR multicause models GBD 2010 plus (a) 2006 48%

The former YugoslavRepublic of Macedonia B VR data VR data 2010 84%

Timor-Leste D VA multicause models GBD 2010 plus (a)




Country





Averageuseability

2000+

Togo E VA multicause models GBD 2010 plus (a)

Tonga A VR multicause models GBD 2010 plus (a)

Trinidad and Tobago B VR data VR data 2008 95%

Tunisia A VR multicause models GBD 2010 plus (a)

Turkey D VR multicause models GBD 2010 plus (a)

Turkmenistan A VA multicause models GBD 2010 plus (a)

Tuvalu A VR multicause models GBD 2010 plus (a)

Uganda E VA multicause models GBD 2010 plus (a)

Ukraine B VR multicause models VR data 2011 96%

United Arab Emirates D VR multicause models GBD 2010 plus (a)

United Kingdom B VR data VR data 2010 93%

United Republic ofTanzania E VA multicause models GBD 2010 plus (a)

United States B VR data VR data 2008 93% Uruguay B VR data VR data 2009 83%

Uzbekistan A VA multicause models VR data 2009 86%

Vanuatu D VR multicause models GBD 2010 plus (a)

Venezuela A VR data VR data 2009 86%

Viet Nam D VR multicause models GBD 2010 plus (a)

Yemen D VA multicause models GBD 2010 plus (a)

Zambia E VA multicause models GBD 2010 plus (a)

Zimbabwe E VA multicause models GBD 2010 plus (a)




Annex Table G Methods used to estimate road traffic deaths for 182

participating countries

Country Group MethodLatest VR

data

Afghanistan 4 Regression estimate

Albania 4 Regression estimate

Andorra 3 Reported deaths (small population)

Angola 4 Regression estimate

Argentina 1 Projected death registration data 2009

Armenia 4 Regression estimate

Australia 1 Projected death registration data 2006

Austria 1 Death registration data 2010

Azerbaijan 1Reported deaths (replacing death registrationestimate)

2007

Bahamas 1 Projected death registration data 2008Bahrain 1 Projected death registration data 2009

Bangladesh 4 Regression estimate

Barbados 1Reported deaths (replacing death registrationestimate)

2008

Belarus 1 Projected death registration data 2009

Belgium 1 Projected death registration data 2006

Belize 1 Projected death registration data 2009

Benin 4 Regression estimate

Bhutan 4 Regression estimate

Bolivia (Plurinational State of) 4 Regression estimate

Bosnia and Herzegovina 4 Regression estimate

Botswana 4 Regression estimate

Brazil 1 Projected death registration data 2009

Brunei Darussalam 1 Projected death registration data 2009

Bulgaria 1Reported deaths (replacing death registrationestimate)

2010

Burkina Faso 4 Regression estimate

Burundi 4 Regression estimate

Cambodia 4 Regression estimate

Cameroon 4 Regression estimate

Canada 1 Death registration data 2010

Cape Verde 4 Regression estimate

Central African Republic 4 Regression estimate

Chad 4 Reported deaths (replacing regression estimate)

Chile 1 Death registration data 2010

China 1 Death registration data (refer to section 3.5) 2010

Colombia 1 Projected death registration data 2008





data

Comoros 4 Regression estimate

Congo 4 Regression estimate

Cook Islands 3 Reported deaths (small population)

Costa Rica 1 Death registration data 2010

Côte d'Ivoire 4 Regression estimate

Croatia 1 Death registration data 2010

Cuba 1 Projected death registration data 2009

Cyprus 1 Death registration data 2010

Czech Republic 1 Death registration data 2010

Democratic People's Republic ofKorea

4 Regression estimate

Democratic Republic of the Congo 4 Regression estimate

Denmark 1 Projected death registration data 2006

Dominica 3 Reported deaths (small population) 2010Dominican Republic 4 Regression estimate

Ecuador 1 Projected death registration data 2009

Egypt 1 Death registration data 2010

El Salvador 1 Projected death registration data 2009

Equatorial Guinea 4 Regression estimate

Estonia 1 Death registration data 2010

Ethiopia 4 Regression estimate

Fiji 1 Death registration data 2010

Finland 1Reported deaths (replacing death registrationestimate)

2010

France 1 Reported deaths (replacing death registrationestimate)

2008

Gabon 4 Reported deaths (replacing regression estimate)

Gambia 4 Regression estimate

Georgia 1 Projected death registration data 2009

Germany 1 Death registration data 2010

Ghana 4 Regression estimate

Greece 1 Projected death registration data 2009

Guatemala 1 Projected death registration data 2009

Guinea 4 Regression estimate

Guinea-Bissau 4 Regression estimate

Guyana 1 Projected death registration data 2008

Honduras 4 Regression estimate

Hungary 1 Death registration data 2010

Iceland 1 Projected death registration data 2009

India 2Regression estimate projected from 2001-2003 data(32, 33)

2010

Indonesia 4 Regression estimate





data

Iran (Islamic Republic of) 2 Projected death registration data 2006

Iraq 4 Regression estimate

Ireland 1 Death registration data 2010

Israel 1 Projected death registration data 2009

Italy 1 Projected death registration data 2009

Jamaica 1 Projected death registration data 2006

Japan 1Projected reported deaths (replacing deathregistration estimate)

2010

Jordan 4 Regression estimate

Kazakhstan 1 Death registration data 2010

Kenya 4 Regression estimate

Kiribati 3 Reported deaths (small population)

Kuwait 1 Projected death registration data 2009

Kyrgyzstan 1 Projected death registration data 2009Lao People's Democratic Republic 4 Regression estimate

Latvia 1 Death registration data 2010

Lebanon 4 Regression estimate

Lesotho 4 Regression estimate

Liberia 4 Regression estimate

Lithuania 1 Death registration data 2010

Luxembourg 1 Projected death registration data 2009

Madagascar 4 Regression estimate

Malawi 4 Regression estimate

Malaysia 4 Reported deaths (replacing regression estimate)

Maldives 1Reported deaths(replacing death registrationestimate) 2008

Mali 4 Regression estimate

Malta 1 Death registration data 2010

Marshall Islands 3 Reported deaths (small population)

Mauritania 4 Regression estimate

Mauritius 1 Death registration data 2010

Mexico 1Projected reported deaths (replacing deathregistration estimate)

2010

Micronesia (Federated States of) 3 Reported deaths (small population)

Mongolia 4 Reported deaths (replacing regression estimate)

Montenegro 1 Projected death registration data 2009

Morocco 4 Regression estimate

Mozambique 4 Regression estimate

Myanmar 4 Regression estimate

Namibia 4 Reported deaths (replacing regression estimate)

Nepal 4 Regression estimate

Netherlands 1 Death registration data 2010





data

New Zealand 1 Projected death registration data 2008

Nicaragua 4 Regression estimate

Niger 4 Regression estimate

Nigeria 4 Regression estimate

Niue 3 Reported deaths (small population)

Norway 1Reported deaths (replacing death registrationestimate)

2010

Oman 1 Death registration data 2010

Pakistan 4 Regression estimate

Palau 3 Reported deaths (small population)

Panama 1 Projected death registration data 2009

Papua New Guinea 4 Regression estimate

Paraguay 1 Projected death registration data 2009

Peru 4 Regression estimatePhilippines 1 Projected death registration data 2008

Poland 1 Death registration data 2010

Portugal 1 Death registration data 2010

Qatar 1 Death registration data 2010

Republic of Korea 1 Projected death registration data 2009

Republic of Moldova 1 Death registration data 2010

Romania 1Reported deaths (replacing death registrationestimate)

2010

Russian Federation 1Reported deaths (replacing death registrationestimate)

2010

Rwanda 4 Regression estimate

Saint Kitts and Nevis 3 Reported deaths (small population) 2008

Saint Lucia 1 Projected death registration data 2006

Saint Vincent and the Grenadines 3 Reported deaths (small population) 2010

Samoa 4 Regression estimate

San Marino 3 Reported deaths (small population)

Sao Tome and Principe 4 Reported deaths (replacing regression estimate)

Saudi Arabia 4 Reported deaths (replacing regression estimate)

Senegal 4 Regression estimate

Serbia 1 Death registration data 2010

Seychelles 3 Reported deaths (small population) 2009

Sierra Leone 4 Regression estimate

Singapore 1 Death registration data 2010

Slovakia 1 Death registration data 2010

Slovenia 1 Death registration data 2010

Solomon Islands 4 Regression estimate

South Africa 1 Projected death registration data 2009

Spain 1 Reported deaths (replacing death registration est.) 2009




data

Sri Lanka 4 Regression estimate

Sudan 4 Regression estimate

Suriname 1 Projected death registration data 2009

Swaziland 4 Regression estimate

Sweden 1 Death registration data 2010

Switzerland 1 Projected death registration data 2007

Syrian Arab Republic 4 Regression estimate

Tajikistan 4 Regression estimate

Thailand 2 Projected death registration data 2008

The Former Yugoslav Republic ofMacedonia

1 Death registration data 2010

Timor-Leste 4 Regression estimate

Togo 4 Regression estimate

Tonga 3 Reported deaths (small population)Trinidad and Tobago 1 Projected death registration data 2007

Tunisia 4 Regression estimate

Turkey 4 Regression estimate

Uganda 4 Regression estimate

Ukraine 1 Death registration data 2010

United Arab Emirates 4 Regression estimate

United Kingdom 1 Death registration data 2010

United Republic of Tanzania 4 Regression estimate

United States of America 1 Projected death registration data 2008

Uruguay 1 Projected death registration data 2009

Uzbekistan 1 Projected death registration data 2005

Vanuatu 4 Regression estimate

Venezuela (Bolivarian Republic of) 1 Projected death registration data 2007

Viet Nam 2 Projected national verbal autopsy survey data 2006

West Bank and Gaza Strip 1 Reported deaths (replacing regression estimate) 2010

Yemen 4 Regression estimate

Zambia 4 Regression estimate

Zimbabwe 1 Reported deaths (replacing regression estimate)

Documents

Global Causes of Death WHO