Objective. To assess maternal mortality.
Design. Sisterhood method survey and hospital data.
Setting. Communities in Kalabo District, a very remote rural area in western Zambia; Kalabo District Hospital.
Results. The number of respondents in the sisterhood method survey was 1,978. The estimated maternal mortality ratio derived from this survey was 1,238 per 100,000 live births. The hospital study involved 2,474 deliveries of 2,374 live babies. The official number of maternal deaths was 13. Further investigation of files revealed an additional 15 maternal deaths, bringing the institutional maternal mortality rate from 548 to 1,179 per 100,000 live births. The major causes of direct maternal deaths were obstructed labor and sepsis. In 71% of all cases substandard care factors contributed.
Conclusions. Maternal mortality in rural Zambia is among the highest as reported in the world. Official hospital data tend to underestimate maternal mortality in the community due to underreporting. The sisterhood method survey is an efficient indirect method to assess maternal mortality in rural areas of developing countries.