DHS maternal mortality indicators: an assessment of data quality and implications for data use.

Abstract

This report presents an assessment of the quality of sibling history data used to derive maternal mortality data for internal characteristics mothers parity and time of death. The report includes a compilation of maternal mortality data from 14 Demographic and Health Surveys (DHS) provides a documentation of DHS procedures and discusses the appropriate use and interpretation of DHS maternal mortality estimates. The two estimation methods include the sisterhood method for indirect calculations and the sibling history method for direct calculations. A comparison is made of direct and indirect estimates. The report compares questionnaire modules. Data quality is checked for completeness of reported events omission in reporting of events and plausibility of the patterns of sibling births and deaths. The sibling data are relatively complete for vital status sex age of living siblings and age at death of deceased siblings. Incomplete data appear for the number of years since the occurrence of a siblings death and the timing of death relative to pregnancy childbirth and the postpartum period for adult female deaths. The data are adjusted to account for the incompleteness. It is reckoned that sibling data are probably underestimates due to the large gap between estimates in different reporting periods before the survey. Underestimates are probably greater for females than males. Most countries showed unlikely increases in maternal mortality indicators for the time periods of 7-13 years before the survey and 0-6 years before the survey. These data are not suitable for trend analysis. Maternal mortality ratios provide only approximate levels. Although international conferences call for a 50% reduction in maternal mortality by the year 2000 meaningful monitoring over time is not possible. Three efforts are underway to improve measurement.