Assessment of maternal mortality in Tanzania

OBJECTIVE: To assess maternal mortality. DESIGN: Prospective community-based survey, a sisterhood method survey and hospital data. SETTING: The study was performed in communities in a rural area in Northwestern Tanzania and the local district hospital. RESULTS: Four hundred and forty-seven pregnant women in the community survey were followed up as far as six weeks after delivery; there were 2865 respondents in the sisterhood survey; the hospital study involved 7526 deliveries. Maternal morality ratios derived from the prospective community-based survey, the sisterhood method survey and hospital data analysis were 241, 297, and 845 per 100,000 live births, respectively. CONCLUSIONS: Hospital data tend to overestimate maternal mortality in these communities. The sisterhood method is easier and more convenient than a prospective community-based study.