This study was conducted in the three main hospitals of Dakar, the capital city of Senegal. Each case of 152 maternal deaths identified over a 12-month period was matched with two controls: a safe delivery in the same clinic, and a safe delivery in the same neighbourhood of residence. Controls were matched on age, birth order, place, and time of delivery. The leading causes of death were: puerperal sepsis and other infections (51 cases), haemorrhage (32 cases), eclampsia (29 cases), ruptured uterus (11 cases), and anaemia (7 cases). Results of the case-control study revealed the major risk factors associated with health system failures: medical equipment failure (odds ratio [OR] = 55.0), late referral (OR = 23.2), lack of antenatal visit (OR = 16.9), and lack of available personnel at time of admission (OR = 6.6). Various indicators of maternal status at time of admission (complications, blood pressure, temperature, oedema, haemoglobin level) and of health history prior to admission (previous complications, previous C-section, lack of treatment) were also strong predictors of survival. Lastly, socio-demographic factors also appeared as correlates of maternal mortality, in particular: first pregnancy (OR = 2.3), pregnancy of high birth order (OR = 1.9), rainy season (OR = 2.4), being unmarried (OR = 2.5), and low level of education (OR = 1.6). Implications for policy are discussed.
