Measuring Maternal Mortality: Three Case Studies Using Verbal Autopsy with Different Platforms

BACKGROUND Accurate measurement of maternal mortality is needed to develop a greater understanding of the scale of the problem, to increase effectiveness of program planning and targeting, and to track progress. In the absence of good quality vital statistics, interim methods are used to measure maternal mortality. The purpose of this study is to document experience with three community-based interim methods that measure maternal mortality using verbal autopsy. METHODS This study uses a post-census mortality survey, a sample vital registration with verbal autopsy, and a large-scale household survey to summarize the measures of maternal mortality obtained from these three platforms, compares and contrasts the different methodologies employed, and evaluates strengths and weaknesses of each approach. Included is also a discussion of issues related to death identification and classification, estimating maternal mortality ratios and rates, sample sizes and periodicity of estimates, data quality, and cost. RESULTS The sample sizes vary considerably between the three data sources and the number of maternal deaths identified through each platform was small. The proportion of deaths to women of reproductive age that are maternal deaths ranged from 8.8% to 17.3%. The maternal mortality rate was estimable using two of the platforms while obtaining an estimate of the maternal mortality ratio was only possible using one of the platforms. The percentage of maternal deaths due to direct obstetric causes ranged from 45.2% to 80.4%. CONCLUSIONS This study documents experiences applying standard verbal autopsy methods to estimate maternal mortality and confirms that verbal autopsy is a feasible method for collecting maternal mortality data. None of these interim methods are likely to be suitable for detecting short term changes in mortality due to prohibitive sample size requirements, and thus, comprehensive and continuous civil registration systems to provide high quality vital statistics are essential in the long-term.