Objective: Few prospective studies have been undertaken of maternal mortality in sub-Saharan Africa. National statistics are inadequate, and data from hospitals are often the only source of information available. Reported maternal mortality ratios may therefore show large variations within the same country, as in Mali. This study was designed to produce an estimate of the maternal mortality ratio for the population of Bamako.
Design: Prospective cohort study.
Setting: Bankoni (population 59,000), a district of Bamako (population 700,000).
Population: 5782 pregnant women identified during quarterly household visits.
Methods: After enrolment, two follow up visits, at six weeks and one year after delivery, were performed to collect information on the pregnancy, its outcome, the method of delivery, the puerperium and the first year after birth. Detailed inquiries on deaths were undertaken in the community, the maternity units and the reference hospital.
Main outcome measures: Maternal mortality ratio, late maternal mortality, likely cause of death.
Results: Complete data at follow up were available on 4717 women (82%) (4653 single and 64 twin pregnancies). Most of the women had antenatal care were and delivered in a district maternity hospital. There were 4580 live births (96%). Fifteen maternal deaths were recorded, yielding an overall maternal mortality ratio of 327 per 100,000 live births. Hypertensive disorders and haemorrhage were the main causes of death. Five more deaths occurred within 42 days or one year after delivery.
Conclusions: This study gave an estimate of the maternal mortality ratio for the population of Bamako, and stressed the need of better emergency obstetric care and the importance of late maternal mortality.
PIP: This study assessed the nature and extent of maternal mortality (MM) among a cohort of pregnant women in urban Bamako, Mali. Data were obtained from a sample of 5782 pregnant women identified during March 1989 and September 1992. Interviews were conducted at 6 weeks and 1 year after delivery. Other clinical information was collected from households and medical records in 1993. By 1994, 4717 women had been traced. Over 95% of the sample were married and Muslim. About 16% were primiparous. Over 25% had 4 or more children. Most women had some contact with local health centers during pregnancy. Only 10% delivered at home. 24-55% delivered in maternity units that were different from their source of prenatal care. 4580 had live births; 198 had late abortions or stillbirths. The MM ratio was 327/100,000 (15 deaths). The lifetime risk of maternal death was 2.7%. Inclusion of the 5 late maternal deaths raised the MM ratio to 436/100,000. 13 deaths were due to direct causes. 7 deaths were due to hemorrhage, including 1 abortion and 2 cases of ruptured uterus. 4 were due to hypertensive disorders during pregnancy. 3 died from sepsis after cesarean section, including 2 cases of obstructed labor. 6 women delivered and died at a national teaching hospital. 5 were delivered at a district maternity hospital and died at the referred national teaching hospital. 3 died at home.