Direct and indirect estimates of maternal mortality from the sisterhood method.

Abstract

This paper reports the experience of the Demographic and Health Survey Program (DHS) in collecting maternal mortality data with retrospective surveys. Of several potential approaches the DHS questions adult household members about maternal deaths among their sisters. Where past fertility is high this approach yields information on a large number of women from visits to a comparatively small number of houses. The indirect method extends the sibling survivorship method of estimating early adult mortality and poses 4 questions of the respondent: the number of sisters who have reached age 15 or were ever married the number of those sisters who are still alive the number who have died and the number which died of maternal causes. The direct method on the other hand is more detailed and 1st involves listing all brothers and sisters of the respondent. Information is then sought on the survivorship of each the ages of surviving siblings the ages and years ago of deceased siblings and deaths of sisters due to maternal causes. DHS experience with direct and indirect estimates in bolivia Sudan and Egypt is then reported in detail. Reasonable estimates of maternal mortality result from the sisterhood approach. Moreover both indirect and direct methods produced similar results albeit with the direct approach facilitating a much greater depth of data analysis. Greater training interview time and complex data processing and analysis are required in this approach. Nonetheless where detailed estimates are sought regarding the level trends and distribution of maternal deaths according to age and parity the direct approach may be expected to provide a rich source of data.