Understanding Child Mortality in Africa

This project seeks to quantify and interpret the spatial patterns of child mortality in Sub-Saharan Africa over the last three decades. We apply geo-spatial interpolation methods to data from the Demographic and Health Survey to develop gridded under-5 mortality (5m0) estimates across 28 Sub-Saharan countries. This website provides a platform for interactive data visualization and hosts the data for public download. 

Detailed explanations of methods and our findings are available in the paper:

Marshall Burke, Sam Heft-Neal, and Eran Bendavid.

Understanding variation in child mortality across Sub-Saharan Africa: A spatial analysis.

The Lancet Global Health, 2016, Volume 4, Issue 12, e936-e945

ABSTRACT


Background: Detailed spatial understanding of levels and trends in under-5 mortality is needed to improve the targeting of interventions to the areas of highest need, and to understand the sources of variation in mortality. To improve this understanding, we aimed to map under-5 mortality in sub-Saharan Africa in the 1980s, 1990s, and 2000s and analyze variation.

Methods: We used data from 82 Demographic and Health Surveys in 28 Sub-Saharan African countries, including the location and timing of 3.24 million child births and 393,685 deaths, to develop high-resolution spatial maps of under-5 mortality in the 1980s, 1990s, and 2000s. These estimates were at a resolution of 0.1 by 0.1 degree (roughly 10 km×10 km). We then analyzed this spatial information to distinguish within-country versus between-country sources of variation in mortality, to examine the extent to which declines in mortality have been accompanied by convergence in the distribution of mortality, and to study localiZed drivers of mortality differences, including temperature, malaria burden, and conflict.

Findings: In our sample of sub-Saharan Africa countries from the 1980s to the 2000s, within-country differences in under-5 mortality accounted for 74–78% of overall variation in under-5 mortality across space and over time. Mortality differed significantly across only 8–15% of borders, supporting the role of local, rather than national, factors in driving mortality patterns. We noted that 23% of the eligible children in the study countries continued to live in so-called mortality hotspots; areas in which, if current trends continue, under-5 mortality will fail to reach the target set by the Sustainable Development Goals. In multivariate analysis, within-country mortality levels at each pixel were significantly related to local temperature, malaria burden, and recent history of conflict.

Interpretation: Our findings suggest that subnational determinants explain a greater portion of under-5 mortality than do country-level characteristics. Sub-national measures of child mortality could provide a more accurate, and potentially more actionable, portrayal of where and why children are still dying than can national statistics.