Child Mortality Estimation: A Global Overview of Infant and Child Mortality Age Patterns in Light of New Empirical Data

BACKGROUND: The under-five mortality rate (the probability of dying between birth and age 5 y, also denoted in the literature as U5MR and (5)q(0)) is a key indicator of child health, but it conceals important information about how this mortality is distributed by age. One important distinction is what amount of the under-five mortality occurs below age 1 y ((1)q(0)) versus at age 1 y and above ((4)q(1)). However, in many country settings, this distinction is often difficult to establish because of various types of data errors. As a result, it is common practice to resort to model age patterns to estimate (1)q(0) and (4)q(1) on the basis of an observed value of (5)q(0). The most commonly used model age patterns for this purpose are the Coale and Demeny and the United Nations systems. Since the development of these models, many additional sources of data for under-five mortality have become available, making possible a general evaluation of age patterns of infant and child mortality. In this paper, we do a systematic comparison of empirical values of (1)q(0) and (4)q(1) against model age patterns, and discuss whether observed deviations are due to data errors, or whether they reflect true epidemiological patterns not addressed in existing model life tables. METHODS AND FINDINGS: We used vital registration data from the Human Mortality Database, sample survey data from the World Fertility Survey and Demographic and Health Surveys programs, and data from Demographic Surveillance Systems. For each of these data sources, we compared empirical combinations of (1)q(0) and (4)q(1) against combinations provided by Coale and Demeny and United Nations model age patterns. We found that, on the whole, empirical values fall relatively well within the range provided by these models, but we also found important exceptions. Sub-Saharan African countries have a tendency to exhibit high values of (4)q(1) relative to (1)q(0), a pattern that appears to arise for the most part from true epidemiological causes. While this pattern is well known in the case of western Africa, we observed that it is more widespread than commonly thought. We also found that the emergence of HIV/AIDS, while perhaps contributing to high relative values of (4)q(1), does not appear to have substantially modified preexisting patterns. We also identified a small number of countries scattered in different parts of the world that exhibit unusually low values of (4)q(1) relative to (1)q(0), a pattern that is not likely to arise merely from data errors. Finally, we illustrate that it is relatively common for populations to experience changes in age patterns of infant and child mortality as they experience a decline in mortality. CONCLUSIONS: Existing models do not appear to cover the entire range of epidemiological situations and trajectories. Therefore, model life tables should be used with caution for estimating (1)q(0) and (4)q(1) on the basis of (5)q(0). Moreover, this model-based estimation procedure assumes that the input value of (5)q(0) is correct, which may not always be warranted, especially in the case of survey data. A systematic evaluation of data errors in sample surveys and their impact on age patterns of (1)q(0) and (4)q(1) is urgently needed, along with the development of model age patterns of under-five mortality that would cover a wider range of epidemiological situations and trajectories. Please see later in the article for the Editors' Summary.