Background. To assess causes and characteristics of maternal deaths in a poorly characterized area.
Methods. A health facility- and population-based study, with an audit of case series on maternal deaths, was done in two divisions with 42 villages (population 143 000), in Mbulu and Hanang districts, Arusha region, Tanzania. In 1995, all known deaths of women between 15 and 50 years, who were residents of the study area, were recorded from all the health facilities and villages in the study area. Supplementary data from the health facilities were collected in 1996. Verbal autopsies, health facility and antenatal records, and additional oral information were used to assess each maternal death.
Results. Forty-five maternal deaths were identified, 26 in 1995 and 19 in 1996. Thirteen of the deaths were direct, while 32 were indirect. Hemorrhage was the main cause of direct obstetric deaths and cerebral malaria of indirect death, accounting for 20 cases, with most of them occurring during an epidemic season. Twenty-four of the women, including 10 of those dying of malaria, died postpartum. Most of the deceased women sought help at a health facility as the first place of seeking help, and used the hospital ambulance. Fourteen of the deaths (31%) were considered possibly avoidable under the prevailing circumstances, with delay in seeking help being most common.
Conclusions. Addressing malaria as a major cause of maternal death, both in pregnancy and postpartum, in epidemic-prone areas, and increased community and family awareness of when to seek help could be important factors in reducing maternal deaths.