Content area
Abstract
The objective of this research is to develop a methodology for deriving child mortality estimates from the retrospective child survivorship data collected through the EMIS surveys (Infant Mortality Surveys in the Sahel). The EMIS data are of an unusual type for demographic analysis, although such data are prevalent in Africa. Retrospective studies of fertility and mortality are usually based on reports from unconditional random samples of women e.g. a cross-section of the female population at a given time. The EMIS samples are of a different type: all women are interviewed at the time of birth of a child. This type of sample is referred to as conditional (inclusion is conditioned by delivery within the reference period). The research presents a methodology for estimating levels and trends in child mortality applicable to data from conditional samples.
The methodology is tested against U.S. historical data and yields infant mortality trends similar to those recorded by the Civil Registration System. Application to the Bobodioulasso in 1980 is similar to the one experienced by the U.S. White population in the Birth Registration Area by the end of World War I (about 96 per 1000 live births).
Pregnancy histories are routinely collected by some maternity clinics in Sub-Saharan Africa. Bobodioulasso maternity clinic data are shown to be of comparable quality with EMIS data. These data have not been used often to study mortality, partly because the techniques for doing so simply did not exist until very recently. One practical implication of the research is that demographers should increasingly use the facilities of the Health System for the study of mortality in developing countries in general, and in Sub-Saharan Africa in particular, now that the techniques for estimating child mortality from conditional samples exist.
