Why mothers continue to die in Pakistan: a nested case-control study of predictors of maternal mortality

BackgroundMaternal mortality ratio (MMR) declined in Pakistan from 276 maternal deaths per 100,000 live births in 2006-07 to 186 (confidence interval: 138-234) in 2019. Despite this decline, reasons why mothers continue to die, the inequity in the burden of maternal mortality and its predictors largely remain unknown. We investigate the levels and predictors of maternal mortality in Pakistan. MethodsThe in-depth analysis of Pakistan Maternal Mortality Survey (PMMS) 2019 was undertaken using the nested case-control design. We matched 147 maternal deaths that occurred during three years prior to PMMS 2019, defined as "cases", with 724 women who delivered during the same period, defined as "controls". We compared socio-demographic background characteristics of cases and controls and performed multivariate regression to investigate the predictors of maternal mortality in Pakistan. ResultsCases and controls were similar on access to antenatal care (ANC) and ANC provider, but differed on age, education, number of pregnancies, type of delivery, tetanus toxoid vaccination during last pregnancy, and ever using a contraceptive method. More of the cases had their last delivery by a skilled birth attendant (83% compared to 63% among controls) and delivered at the government hospital (43% compared to 33% among controls) while home delivery was relatively more common among controls (32% compared to 25% among cases). Odds of maternal death were lowest for women giving birth during age 20-29 (odds ratio - OR: 0.5; 95% CI 0.23-1.07) as compared to women in age 15-19 and those age 30-39 (OR: 1.34; 95% CI 0.61-2.95) and 40-49 (OR: 1.21; 95% CI 0.42-3.45), and among women with secondary or higher education (OR: 0.35; 95% CI 0.17-0.74) compared to women with no education. Surprisingly, women who had their delivery by a skilled birth attendant experienced higher odds of maternal death (OR: 4.07; 95% CI 2.19- 7.57) compared to those who did not. Whereas women having had tetanus injection during last pregnancy (OR: 0.1; 95% CI 0.0-0.96) and ever-used contraceptives (OR: 0.23; 95% CI 0.11-0.39) tend to have reduced odds of maternal death compared to women who did not have the injection and who never used contraceptives, respectively. ConclusionsWomen in certain sub-groups confront higher risks of maternal death. Increasing female education, preventing early and late childbearing through contraceptive use, increasing tetanus vaccination during pregnancy, improving providers skills, and quality of health care are required to eliminate preventable maternal deaths in Pakistan.