Abstract / Summary
Postpartum hemorrhage (PPH) continues to remain the leading cause of maternal mortality globally, with the greatest burden experienced in sub-Saharan Africa. In severe cases where conservative and medical management fails, surgical intervention becomes the definitive life-saving management. This study aimed to determine maternal outcomes and evaluate predictors of maternal mortality among women undergoing laparotomy for severe PPH at a tertiary referral hospital, Dar-es-Salaam, Tanzania. We conducted a retrospective cohort study at Muhimbili National Hospital. Medical records of women referred with PPH between June 2021 and June 2025 were reviewed. Women who required laparotomy for severe or refractory PPH were included. The primary outcome was in-hospital maternal death following surgical management. Descriptive statistics summarized the study population and characteristics. Bivariate and multivariate logistic regression models were used to identify predictors of mortality, reporting adjusted odds ratio (aOR) and 95% confidence interval (CI). A total of 788 women were referred with PPH during the study period, with 217 (27.5%) requiring laparotomy. Among the 217 surgically managed women, 33 maternal deaths were reported, corresponding to an in-hospital mortality rate of 15.2%. The findings showed that unconsciousness at referral (aOR = 5.70, 95% CI:2.24–14.52) and severe anemia (aOR = 4.60, 95% CI:1.00-21.14) were independent predictors of death. On the other hand, referral-to-operation delay ≥4 h (aOR = 1.40, 95% CI:0.61–3.21) was not independently associated with mortality. Among the peri-operative factors, ICU admission was strongly associated with mortality (aOR = 9.31, 95% CI:3.40-28.53), reflecting the severity of illness. Maternal mortality among women undergoing laparotomy for severe or refractory PPH remains relatively high in this tertiary referral setting. Clinical severity at presentation, highlighting unconsciousness and severe anemia particularly in this setting were strong predictor of death. Although referral-to-operation delay ≥4 h was not independently associated with mortality after adjustment for clinical severity, this finding should be interpreted cautiously given that the quality of pre-referral care was not assessed and prolonged delays likely contributed to hemodynamic deterioration before surgical intervention. These findings highlight the need for early recognition, rapid resuscitation, and timely referral from peripheral and lower-level healthcare facilities.