Child Mortality Estimation Research Articles

Comparison of Lives Saved Tool model child mortality estimates against measured data from vector control studies in sub-Saharan Africa

BackgroundInsecticide-treated mosquito nets (ITNs) and indoor-residual spraying have been scaled-up across sub-Saharan Africa as part of international efforts to control malaria. These interventions have the potential to significantly impact child survival. The Lives Saved Tool (LiST) was developed to provide national and regional estimates of cause-specific mortality based on the extent of intervention coverage scale-up. We compared the percent reduction in all-cause child mortality estimated by LiST against measured reductions in all-cause child mortality from studies assessing the impact of vector control interventions in Africa.MethodsWe performed a literature search for appropriate studies and compared reductions in all-cause child mortality estimated by LiST to 4 studies that estimated changes in all-cause child mortality following the scale-up of vector control interventions. The following key parameters measured by each study were applied to available country projections: baseline all-cause child mortality rate, proportion of mortality due to malaria, and population coverage of vector control interventions at baseline and follow-up years.ResultsThe percent reduction in all-cause child mortality estimated by the LiST model fell within the confidence intervals around the measured mortality reductions for all 4 studies. Two of the LiST estimates overestimated the mortality reductions by 6.1 and 4.2 percentage points (33% and 35% relative to the measured estimates), while two underestimated the mortality reductions by 4.7 and 6.2 percentage points (22% and 25% relative to the measured estimates).ConclusionsThe LiST model did not systematically under- or overestimate the impact of ITNs on all-cause child mortality. These results show the LiST model to perform reasonably well at estimating the effect of vector control scale-up on child mortality when compared against measured data from studies across a range of malaria transmission settings. The LiST model appears to be a useful tool in estimating the potential mortality reduction achieved from scaling-up malaria control interventions.

Identifying Priorities for Emergency Intervention from Child Wasting and Mortality Estimates in Vulnerable Areas of the Horn of Africa

The relation between anthropometric measures and mortality risk in different populations can provide a basis for deciding how malnutrition prevalences should be interpreted. To assess criteria for deciding on needs for emergency interventions in the Horn of Africa based on associations between child wasting and mortality from 2000 to 2005. Data were analyzed on child global acute malnutrition (GAM) prevalences and mortality estimates from about 900 area-level nutrition surveys from Ethiopia, Kenya, Somalia, Sudan, and Uganda; data on drought, floods, and food insecurity were added for Kenya (Rift Valley) and Ethiopia, from Food and Agriculture Organization (FAO) reports at the time. Higher rates of GAM were associated with increased mortality of children under 5 years of age (U5MR), more strongly among populations with pastoral livelihoods than with agricultural livelihoods. In all groups spikes of GAM and U5MR corresponded with drought (and floods). Different GAM cutoff points are needed for different populations. For example, to identify 75% of U5MRs above 2/10,000/day, the GAM cutoff point ranged from 20% GAM in the Rift Valley (Kenya) to 8% in Oromia or SNNPR (Ethiopia). Survey results should be displayed as time series within geographic areas. Variable GAM cutoff points should be used, depending on livelihood or location. For example, a GAM cutoff point of 15% may be appropriate for pastoral groups and 10% for agricultural livelihood groups. This gives a basis for reexamining the guidelines currently used for interpreting wasting (or GAM) prevalences in terms of implications for intervention.

Increased educational attainment and its effect on child mortality in 175 countries between 1970 and 2009: a systematic analysis

In addition to the inherent importance of education and its essential role in economic growth, education and health are strongly related. We updated previous systematic assessments of educational attainment, and estimated the contribution of improvements in women's education to reductions in child mortality in the past 40 years. We compiled 915 censuses and nationally representative surveys, and estimated mean number of years of education by age and sex. By use of a first-differences model, we investigated the association between child mortality and women's educational attainment, controlling for income per person and HIV seroprevalence. We then computed counterfactual estimates of child mortality for every country year between 1970 and 2009. The global mean number of years of education increased from 4·7 years (95% uncertainty interval 4·4-5·1) to 8·3 years (8·0-8·6) for men (aged ≥25 years) and from 3·5 years (3·2-3·9) to 7·1 years (6·7 -7·5) for women (aged ≥25 years). For women of reproductive age (15-44 years) in developing countries, the years of schooling increased from 2·2 years (2·0-2·4) to 7·2 years (6·8-7·6). By 2009, in 87 countries, women (aged 25-34 years) had higher educational attainment than had men (aged 25-34 years). Of 8·2 million fewer deaths in children younger than 5 years between 1970 and 2009, we estimated that 4·2 million (51·2%) could be attributed to increased educational attainment in women of reproductive age. The substantial increase in education, especially of women, and the reversal of the gender gap have important implications not only for health but also for the status and roles of women in society. The continued increase in educational attainment even in some of the poorest countries suggests that rapid progress in terms of Millennium Development Goal 4 might be possible. Bill & Melinda Gates Foundation.

Reducing child mortality in Indonesia

Reducing child mortality in Indonesia

Early Mortality and Loss to Follow-up in HIV-Infected Children Starting Antiretroviral Therapy in Southern Africa

Many HIV-infected children in Southern Africa have been started on antiretroviral therapy (ART), but loss to follow up (LTFU) can be substantial. We analyzed mortality in children retained in care and in all children starting ART, taking LTFU into account. Children who started ART before the age of 16 years in 10 ART programs in South Africa, Malawi, Mozambique, and Zimbabwe were included. Risk factors for death in the first year of ART were identified in Weibull models. A meta-analytic approach was used to estimate cumulative mortality at 1 year. Eight thousand two hundred twenty-five children (median age 49 months, median CD4 cell percent 11.6%) were included; 391 (4.8%) died and 523 (7.0%) were LTFU in the first year. Mortality at 1 year was 4.5% [95% confidence interval (CI): 2.8% to 7.4%] in children remaining in care, but 8.7% (5.4% to 12.1%) at the program level, after taking mortality in children and LTFU into account. Factors associated with mortality in children remaining in care included age [adjusted hazard ratio (HR) 0.37; 95% CI: 0.25 to 0.54 comparing > or =120 months with <18 months], CD4 cell percent (HR: 0.56; 95% CI: 0.39 to 0.78 comparing > or =20% with <10%), and clinical stage (HR: 0.12; 95% CI: 0.03 to 0.45 comparing World Health Organization stage I with III/IV). In children starting ART and remaining in care in Southern Africa mortality at 1 year is <5% but almost twice as high at the program level, when taking LTFU into account. Age, CD4 percentage, and clinical stage are important predictors of mortality at the individual level.

Neonatal, postneonatal, childhood, and under-5 mortality for 187 countries, 1970–2010: a systematic analysis of progress towards Millennium Development Goal 4

Previous assessments have highlighted that less than a quarter of countries are on track to achieve Millennium Development Goal 4 (MDG 4), which calls for a two-thirds reduction in mortality in children younger than 5 years between 1990 and 2015. In view of policy initiatives and investments made since 2000, it is important to see if there is acceleration towards the MDG 4 target. We assessed levels and trends in child mortality for 187 countries from 1970 to 2010. We compiled a database of 16 174 measurements of mortality in children younger than 5 years for 187 countries from 1970 to 2009, by use of data from all available sources, including vital registration systems, summary birth histories in censuses and surveys, and complete birth histories. We used Gaussian process regression to generate estimates of the probability of death between birth and age 5 years. This is the first study that uses Gaussian process regression to estimate child mortality, and this technique has better out-of-sample predictive validity than do previous methods and captures uncertainty caused by sampling and non-sampling error across data types. Neonatal, postneonatal, and childhood mortality was estimated from mortality in children younger than 5 years by use of the 1760 measurements from vital registration systems and complete birth histories that contained specific information about neonatal and postneonatal mortality. Worldwide mortality in children younger than 5 years has dropped from 11.9 million deaths in 1990 to 7.7 million deaths in 2010, consisting of 3.1 million neonatal deaths, 2.3 million postneonatal deaths, and 2.3 million childhood deaths (deaths in children aged 1-4 years). 33.0% of deaths in children younger than 5 years occur in south Asia and 49.6% occur in sub-Saharan Africa, with less than 1% of deaths occurring in high-income countries. Across 21 regions of the world, rates of neonatal, postneonatal, and childhood mortality are declining. The global decline from 1990 to 2010 is 2.1% per year for neonatal mortality, 2.3% for postneonatal mortality, and 2.2% for childhood mortality. In 13 regions of the world, including all regions in sub-Saharan Africa, there is evidence of accelerating declines from 2000 to 2010 compared with 1990 to 2000. Within sub-Saharan Africa, rates of decline have increased by more than 1% in Angola, Botswana, Cameroon, Congo, Democratic Republic of the Congo, Kenya, Lesotho, Liberia, Rwanda, Senegal, Sierra Leone, Swaziland, and The Gambia. Robust measurement of mortality in children younger than 5 years shows that accelerating declines are occurring in several low-income countries. These positive developments deserve attention and might need enhanced policy attention and resources. Bill & Melinda Gates Foundation.

Measuring under-five mortality: validation of new low-cost methods.

There has been increasing interest in measuring under-five mortality as a health indicator and as a critical measure of human development. In countries with complete vital registration systems that capture all births and deaths, under-five mortality can be directly calculated. In the absence of a complete vital registration system, however, child mortality must be estimated using surveys that ask women to report the births and deaths of their children. Two survey methods exist for capturing this information: summary birth histories and complete birth histories. A summary birth history requires a minimum of only two questions: how many live births has each mother had and how many of them have survived. Indirect methods are then applied using the information from these two questions and the age of the mother to estimate under-five mortality going back in time prior to the survey. Estimates generated from complete birth histories are viewed as the most accurate when surveys are required to estimate under-five mortality, especially for the most recent time periods. However, it is much more costly and labor intensive to collect these detailed data, especially for the purpose of generating small area estimates. As a result, there is a demand for improvement of the methods employing summary birth history data to produce more accurate as well as subnational estimates of child mortality. We used data from 166 Demographic and Health Surveys (DHS) to develop new empirically based methods of estimating under-five mortality using children ever born and children dead data. We then validated them using both in- and out-of-sample analyses. We developed a range of methods on the basis of three dimensions of the problem: (1) approximating the average length of exposure to mortality from a mother's set of children using either maternal age or time since first birth; (2) using cohort and period measures of the fraction of children ever born that are dead; and (3) capturing country and regional variation in the age pattern of fertility and mortality. We focused on improving estimates in the most recent time periods prior to a survey where the traditional indirect methods fail. In addition, all of our methods incorporated uncertainty. Validated against under-five estimates generated from complete birth histories, our methods outperformed the standard indirect method by an average of 43.7% (95% confidence interval [CI] 41.2-45.2). In the 5 y prior to the survey, the new methods resulted in a 53.3% (95% CI 51.3-55.2) improvement. To illustrate the value of this method for local area estimation, we applied our new methods to an analysis of summary birth histories in the 1990, 2000, and 2005 Mexican censuses, generating subnational estimates of under-five mortality for each of 233 jurisdictions. The new methods significantly improve the estimation of under-five mortality using summary birth history data. In areas without vital registration data, summary birth histories can provide accurate estimates of child mortality. Because only two questions are required of a female respondent to generate these data, they can easily be included in existing survey programs as well as routine censuses of the population. With the wider application of these methods to census data, countries now have the means to generate estimates for subnational areas and population subgroups, important for measuring and addressing health inequalities and developing local policy to improve child survival. Please see later in the article for the Editors' Summary.

Measuring and correcting biased child mortality statistics in countries with generalized epidemics of HIV infection.

Under Millennium Development Goal4, countries are required to reduce child mortality by two-thirds between 1990 and 2015. In countries with generalized epidemics of human immunodeficiency virus (HIV) infection, standard statistics based on fertility history may misrepresent progress towards this target owing to the correlation between deaths among mothers and early childhood deaths from acquired immunodeficiency syndrome. To empirically estimate this bias, child mortality data and fertility history, including births to deceased women, were collected through prospective household surveys in eastern Zimbabwe during 1998-2005. A mathematical model was then used to investigate the determinants and temporal dynamics of the bias, first in Zimbabwe and then in other countries with different background mortality rates and HIV-related epidemic profiles. According to the empirical data, standard cross-sectional survey statistics underestimated true infant and under-5 mortality by 6.7% and 9.8%, respectively. These estimates were in agreement with the output from the model, in which the bias varied according to the magnitude and stage of the epidemic of HIV infection and background mortality rates. The bias was greater the longer the period elapsed before the survey and in later stages of the epidemic. Bias could substantially distort the measured effect of interventions to reduce non-HIV-related mortality and of programmes to prevent mother-to-child transmission, especially when trends are based on data from a single survey. The correlation between the HIV-related deaths of mothers and their children can bias survey estimates of early child mortality. A mathematical model with a user-friendly interface is available to correct for this bias when measuring progress towards Millennium Development Goal4 in countries with generalized epidemics of HIV infection.

Child Mortality After Hurricane Katrina

Age-specific pediatric health consequences of community disruption after Hurricane Katrina have not been analyzed. Post-Katrina vital statistics are unavailable. The objectives of this study were to validate an alternative method to estimate child mortality rates in the greater New Orleans area and compare pre-Katrina and post-Katrina mortality rates. Pre-Katrina 2004 child mortality was estimated from death reports in the local daily newspaper and validated by comparison with pre-Katrina data from the Louisiana Department of Health. Post-Katrina child mortality rates were analyzed as a measure of health consequences. Newspaper-derived estimates of mortality rates appear to be valid except for possible underreporting of neonatal rates. Pre-Katrina and post-Katrina mortality rates were similar for all age groups except infants. Post-Katrina, a 92% decline in mortality rate occurred for neonates (<28 days), and a 57% decline in mortality rate occurred for postneonatal infants (28 days-1 year). The post-Katrina decline in infant mortality rate exceeds the pre-Katrina discrepancy between newspaper-derived and Department of Health-reported rates. A declining infant mortality rate raises questions about persistent displacement of high-risk infants out of the region. Otherwise, there is no evidence of long-lasting post-Katrina excess child mortality. Further investigation of demographic changes would be of interest to local decision makers and planners for recovery after public health emergencies in other regions.

Levels and trends in under-5 mortality, 1990–2008

As global momentum and investment for accelerating maternal and child survival grows, monitoring of progress at the global and country level has become even more needed. Millennium Development Goal 4 (MDG 4) calls for a two-thirds reduction in the mortality rate among children under the age of 5 years between 1990 and 2015. Generating accurate estimates of under-5 mortality poses a considerable challenge because of the limited data available for many developing countries. In response, experts at UNICEF, WHO, the World Bank, the UN Population Division (UNPD), and members of the academic community formed the Inter-agency Group for Child Mortality Estimation (IGME).1 The IGME aims to source and share data on child mortality, to improve and harmonise estimation methods across partners, and to produce consistent estimates on the levels and trends in child mortality worldwide.

Religious Affiliation, Ethnicity, and Child Mortality in Chiapas, México.

We investigate whether there is a relationship between religious affiliation and child mortality among indigenous and nonindigenous groups in Chiapas, México. Our analysis relies on Brass-type estimates of child mortality by ethnicity and religious affiliation and multivariate analyses that adjust for various socioeconomic and demographic factors. The data are from the 2000 Mexican Census 10 percent sample. Among indigenous people, Presbyterians have lower rates of child mortality than Catholics. However, no significant differentials are found in child mortality by religious affiliation among nonindigenous people. The indigenous health ministry of the Presbyterian Church and the social and cultural transformations that tend to accompany religious conversion may have an impact on child survival among disadvantaged populations such as the indigenous people in Chiapas.

Estimating child mortality due to diarrhoea in developing countries

The major objective of this study is to provide estimates of diarrhoea mortality at country, regional and global level by employing the Child Health Epidemiology Reference Group (CHERG) standard. A systematic and comprehensive literature review was undertaken of all studies published since 1980 reporting under-5 diarrhoea mortality. Information was collected on characteristics of each study and its population. A regression model was used to relate these characteristics to proportional mortality from diarrhoea and to predict its distribution in national populations. Global deaths from diarrhoea of children aged less than 5 years were estimated at 1.87 million (95% confidence interval, CI: 1.56-2.19), approximately 19% of total child deaths. WHO African and South-East Asia Regions combined contain 78% (1.46 million) of all diarrhoea deaths occurring among children in the developing world; 73% of these deaths are concentrated in just 15 developing countries. Planning and evaluation of interventions to control diarrhoea deaths and to reduce under-5 mortality is obstructed by the lack of a system that regularly generates cause-of-death information. The methods used here provide country-level estimates that constitute alternative information for planning in settings without adequate data.

Education is key to reducing child mortality

In 2006, for the first time in recent history, the total number of annual deaths among children under the age of five fell below 10 million, to 9.7 million. This represents a 60-per-cent drop in the rate of child mortality since 1960. Data compiled by the Inter-agency Group for Child Mortality Estimation reveals that progress has been made in every region of the world. Since 1990, China’s under-five mortality rate has declined from 45 deaths for every 1,000 live births to 24 per 1,000, a reduction of 47 per cent; India’s rate declined by 34 per cent. The rates in six countries—Bangladesh, Bhutan, Bolivia, Eritrea, Lao People’s Democratic Republic and Nepal—fell by 50 per cent or more from 1990 to 2006, although under-five mortality rates in these countries remain high. Ethiopia achieved a nearly 40-per-cent reduction during the same period.

Cost-effectiveness of pneumococcal conjugate vaccination in the prevention of child mortality: an international economic analysis

Routine vaccination of infants against Streptococcus pneumoniae (pneumococcus) needs substantial investment by governments and charitable organisations. Policymakers need information about the projected health benefits, costs, and cost-effectiveness of vaccination when considering these investments. Our aim was to incorporate these data into an economic analysis of pneumococcal vaccination of infants in countries eligible for financial support from the Global Alliance for Vaccines & Immunization (GAVI). We constructed a decision analysis model to compare pneumococcal vaccination of infants aged 6, 10, and 14 weeks with no vaccination in the 72 countries that were eligible as of 2005. We used published and unpublished data to estimate child mortality, effectiveness of pneumococcal conjugate vaccine, and immunisation rates. Pneumococcal vaccination at the rate of diptheria-tetanus-pertussis vaccine coverage was projected to prevent 262,000 deaths per year (7%) in children aged 3-29 months in the 72 developing countries studied, thus averting 8.34 million disability-adjusted life years (DALYs) yearly. If every child could be reached, up to 407,000 deaths per year would be prevented. At a vaccine cost of International 5 dollars per dose, vaccination would have a net cost of 838 million dollars, a cost of 100 dollars per DALY averted. Vaccination at this price was projected to be highly cost-effective in 68 of 72 countries when each country's per head gross domestic product per DALY averted was used as a benchmark. At a vaccine cost of between 1 dollar and 5 dollars per dose, purchase and accelerated uptake of pneumococcal vaccine in the world's poorest countries is projected to substantially reduce childhood mortality and to be highly cost-effective.

Understanding the effects of siblings on child mortality: evidence from India

Given the intrinsically sequential nature of childbirth, timing of a child’s birth has consequences not only for itself but also for its older and younger siblings. The paper argues that prior spacing and posterior spacing between consecutive siblings are thus important measures of intensity of sibling competition for limited parental resources. While the available estimates of child mortality tend to ignore the endogeneity of sibling composition, we use a correlated recursive model of prior and posterior spacing and child mortality to correct it. There is evidence that uncorrected estimates under-estimate the effects of prior and posterior spacing on child mortality.

Making deaths count

Death has always been an overriding concern: two of the Millennium Development Goals express targets in terms of mortality and one out of three components of the Human Development Index is a mortality measure. It is therefore surprising that the systematic measurement of mortality is relatively recent: rudimentary life tables to summarize mortality from the 17th century, the use of systematically collected data on vital events to examine mortality differentials and cause of death from the 19th century, and systematic explorations of behavioural risk factors only in the 20th. Today, routinely collected data on vital events provide complete and representative information for only about 40% of the world's countries and a quarter of its population. (1) The planning, implementation and evaluation of health programmes at national and international levels require valid, comparable and timely information on the nature and magnitude of health problems. Information on deaths by cause is the key: the Global Burden of Disease 2000 (2) estimates that 63% of healthy life lost in 2000 resulted from premature deaths. Yet Mathers et al., (3) reviewing cause of death recording among Member States supplying data to WHO for 1990 or later, conclude that quality is high for countries with only 12% of the world's population, medium for 17% and poor for 5%; no data are supplied for the remaining 66%. Only two countries from sub-Saharan Africa, the region with the highest mortality, report causes for 50% or more of their deaths. Global health progress requires action to improve the availability of valid statistics: broader collection and timely dissemination of health statistics, and new methodologies that make better use of existing data or that can collect substitutes for vital statistics records quickly and inexpensively. The articles in this issue address both. Information on child mortality in the 1990s is now available for a large majority of countries. Coordinated programmes such as the Demographic and Health Surveys (DHS) and Multiple Indicator Cluster Surveys have provided information on patterns, levels and trends of child mortality in many developing countries. Population censuses have added indirect observations and small area estimates for local health authorities. The major international agencies have coordinated their efforts, developing a common database and working on methodological improvements, but there is still much we do not know. Existing estimates are not timely: relatively few developing countries provide estimates of child mortality for 2000 or later. Information on cause of death is weak, and existing verbal autopsy approaches identify few causes well. Countries affected by civil strife, with perhaps the greatest problems of elevated child mortality, rarely have recent indicators of child mortality. The situation is worse for adult mortality. …

Understanding the Effects of Siblings on Child Mortality: Evidence from India

Given the intrinsically sequential nature of child birth, timing of a child's birth has consequences not only for itself, but also for its older and younger siblings. The paper argues that prior and posterior spacing between consecutive siblings are thus important measures of intensity of sibling competition for limited parental resources. While the available estimates of child mortality tend to ignore the endogeneity of sibling composition, we use a correlated recursive model of prior and posterior spacing and child mortality to correct it. There is evidence that uncorrected estimates underestimate the effects of prior and posterior spacing on child mortality.

HIV and Mortality of Mothers and Children

The steady decline in child mortality observed in most African countries through the 1960s, 1970s, and 1980s has stalled in many countries in the 1990s because of the AIDS epidemic. However, the census and household survey data that generally are used to produce estimates of child mortality do not permit precise measures of the adverse effect of HIV on child mortality. To calculate excess risks of child mortality as the result of maternal HIV status, we used pooled data from 3 longitudinal community-based studies that classified births by the mother's HIV status. We also estimated excess risks of child death caused by increased mortality among mothers. The joint effects of maternal HIV status and maternal survival were quantified using multivariate techniques in a survival analysis. Our analysis shows that the excess risk of death associated with having an HIV-positive mother is 2.9 (95% confidence interval = 2.3-3.6), and this effect lasts throughout childhood. The excess risk associated with a maternal death is 3.9 (2.8-5.5) in the 2-year period centered on the mother's death, with children of both infected and uninfected mothers experiencing higher mortality risks at this time. HIV impacts on child mortality directly through transmission of the virus to newborns by infected mothers and indirectly through higher child mortality rates associated with a maternal death.

Contemporary patterns of Pacific Island mortality.

The Pacific Island countries are at different stages of the demographic and epidemiological transitions. The availability of accurate and current mortality data is of vital importance for priority setting in health. Available mortality data generally underestimate death rates among both children and adults. In many Pacific Island populations, little is reliably known about levels and causes of death, particularly among adults. The results of two comprehensive approaches to obtaining mortality estimates are reported. First, a systematic review of available life expectancy and infant mortality information reported by countries from 1990 onwards was undertaken and evaluated with respect to quality, and a final "best estimate" was established. Methods were based on registered deaths and indirect demographic methods. The second approach consisted of a demographic evaluation of vital registration data for completeness, with death rates adjusted accordingly, or where vital registration was not available, the application of new model life table methods to generate life tables from estimates of child mortality, as used by the World Health Organisation (WHO). This analysis reveals substantial uncertainty about mortality conditions in Pacific Island populations. In some countries, life expectancy variations of 10 years or more were recorded in the 1990s, depending on the source. Best approaches suggest that life expectancy (at birth) varied considerably, from levels of around 55-60 years in some Melanesian and Micronesian states to levels above 70 years in low-mortality countries. The principal issues with regard to uncertainty around mortality levels include underenumerated vital registration data; annual stochastic fluctuations in mortality in small populations; errors in the imputation of adult mortality from infant and childhood rates; implausible results from indirect demographic methods; use of possibly inappropriate model life tables to adjust death data or for indirect methods; and inadequately described and implausible projections. The WHO model life table method based on adjusted vital registration generally yielded results similar to those suggested by an evaluation of published data, with some exceptions, which are further discussed. This study indicates the urgent need to improve infrastructure, training, and resources for routine mortality estimation in many Pacific Island countries in order to better inform and evaluate health and public policy.

Commentary: On reappraisal of adult mortality.

Due to the development of indirect estimation methods for estimating infant and child mortality from some simple questions asked routinely in censuses and surveys, and collection of detailed birth histories in some sample surveys, there has been marked improvement in the information on levels and trends in child mortality in the developing world. 1 But unfortunately, the same cannot be said about levels and trends in adult mortality. In recent years, the spread of the human immunodeficiency virus (HIV)/AIDS epidemic has brought the paucity of data on adult mortality into sharp focus. Although the United Nations does provide estimates of adult mortality for all the developing countries, in only about onethird of the countries is such information based on vital registration. 2 The estimates derived from indirect information, such as on orphanhood, have been found useable in only 4% of the countries. For about one-third of developing countries the adult mortality levels are arrived at from the estimates of child mortality by assuming age patterns of mortality contained in ‘model’ life tables. The model life tables synthesize the relationship between mortality rates at different ages observed in countries that had reliable data on age- and sex-specific mortality rates. Although they may be useful in approximating the level of adult mortality for the time periods for which estimates of child mortality are available, the assumed high correlation between the levels of adult and child mortality could become too restrictive when ascertaining trends, especially in populations affected by the HIV/AIDS epidemic. It is in this context that Gakidou and colleagues recommend collection and analysis of sibling survival data. 3 It should be noted that there are two alternative methods of estimating adult mortality from this information. The ‘indirect’ method, developed originally by Hill and Trussell 4 and revised subsequently by Timaeus and others, 5 uses data only on the proportion of surviving siblings by respondent’s age. The data collected in the maternal mortality module of the Demographic and Health Surveys (DHS) had additionally included data on current age of surviving siblings, age at death, and time period of those who have died. Data of this type permit the direct estimation of mortality rates by age for any time period in the past, and it is this data and the method that Gakidou and others discuss in their paper. The advantage of the direct method is that it is not affected by assumptions about the age patterns of fertility and mortality made in the indirect method, but could be sensitive to errors in the detailed data used in the estimation. Neither of these methods has been extensively used and hence there is scant information on how well they perform in practice. An issue of particular concern is that as respondents cannot be expected to report on siblings who died before they were born or when they were very young, the method would invariably lead to underestimation of adult mortality. Gakidou and others provide estimates of probability of survival between age 15 and 59 ( 59 q 15 ) derived from the DHS data for 29 populations. Apparently the estimates are for the 5-year period before the survey (the paper is not very clear on this issue). The comparison with the estimates provided by the United Nations for the corresponding period show that sibling survival estimates are lower (the extent of underestimation is difficult to gauge from the data presented by the authors) which appears to confirm the suspicion about the sensitivity of the estimates to selective omission of non-surviving siblings. However, the UN estimates are also not above criticism, as the adult mortality estimates for many of the countries in the data set were probably inferred from the estimates of child mortality using model life tables. Nonetheless, the estimates presented for four countries at two points in time raise doubts about the usefulness of sibling-survival data for monitoring adult mortality trends. Of the four countries, the survey estimates suggest unbelievably large increases in adult mortality in a span of 3‐4 years in the case of two (Indonesia and Malawi), and large declines in one (Peru).