Obstetric risk profiles and causes of death: Estimating their association with cesarean sections among maternal deaths in Mexico.

ABSTRACT

BACKGROUND: Maternal mortality is a critical indicator of healthcare quality, and in Mexico, this has become increasingly concerning due to the stagnation in its decline, alongside a concurrent increase in cesarean section (C-section) rates. This study characterizes maternal deaths in Mexico, focusing on estimating the association between obstetric risk profiles, cause of death, and mode of delivery. METHODS: Utilizing a retrospective observational design, 4,561 maternal deaths in Mexico from 2010-2014 were analyzed. Data were sourced from the Deliberate Search and Reclassification of Maternal Deaths database, alongside other national databases. An algorithm was developed to extract the Robson Ten Group Classification System from clinical summaries text, facilitating a nuanced analysis of C-section rates. Information on the reasons for the performance of a C-section was also obtained. Logistic regression and multinomial logistic regression models were used to estimate the relation between obstetric risk factors, mode of delivery and causes of maternal death, adjusting for covariates. RESULTS: Among maternal deaths in Mexico from 2010-2014, 47.1% underwent a C-section, with a significant history of previous C-sections observed in 31.4% of these cases, compared to 17.4% for vaginal deliveries (p<0.001). Early prenatal care in the first trimester was more common in C-section cases (46.8%) than in vaginal deliveries (38.3%, p<0.001). A stark contrast was noted in the place of death, with 82.4% of post-C-section deaths occurring in public institutions versus 69.1% following vaginal births. According to Robson's classification, the highest C-section rates were in Group 4 (67.2%, p<0.001) and Group 8 (66.9%, p<0.001). Logistic regression analysis revealed no significant difference in the odds of receiving a C-section in private versus other settings after adjusting for Robson criteria (OR 1.21; 95% CI 0.92, 1.60). A prior C-section significantly increased the likelihood of another (OR 2.38; CI 95% 2.01, 2.81). The analysis also indicated C-sections were significantly tied to deaths from hypertensive disorders (RRR = 1.25, 95% CI [1.12, 1.40]). In terms of indications, 6.3% of C-sections were performed under inadequate indications, while the indication was not identifiable in 33.1% of all C-sections. CONCLUSIONS: This study highlights a significant overuse of C-sections among maternal deaths in Mexico (2010-2014), revealed through the Robson classification and ana analysis of the reported indications for the procedure. It underscores the need for revising clinical decision-making to promote evidence-based guidelines and favor vaginal deliveries when possible. High C-section rates, especially noted disparities between private and public sectors, suggest economic and non-clinical factors may be at play. The importance of accurate data systems and further research with control groups to understand C-section practices' impact on maternal health is emphasized.