Decomposing the Caesarean section trend in Bangladesh (2012–2025): A multivariate Linear Probability Model analysis

Background While Caesarean section is a life-saving intervention when medically indicated, the World Health Organization recommends an optimal population-level rate of 10% to 15% to prevent iatrogenic morbidities. In Bangladesh, however, rapid clinical integration within a commercialized, dual-practice healthcare architecture has coincided with an unprecedented escalation in surgical deliveries. Because standard cross-sectional models cannot algebraically separate population shifts from changing provider propensity, this study aimed to decompose the compositional and behavioral drivers of the Caesarean section surge in Bangladesh between 2012 and 2025. Methods Nationally representative data were analyzed from the 2012–2013, 2019, and 2024–2025 Bangladesh Multiple Indicator Cluster Surveys (MICS), comprising 26,534 women aged 15–49 years who delivered a child within the preceding two years. Survey skip-logic was harmonized to retain unmedicalized cohorts receiving zero antenatal care (ANC). A survey-weighted multivariate Linear Probability Model (LPM) Oaxaca-Blinder decomposition was applied to a complete-case cohort (N=16,179) to partition the prevalence change into population-level compositional shifts (endowments) and behavioral changes (coefficients). Results National Caesarean section prevalence surged by 32.47 percentage points, rising from 19.48% in 2012 to 36.05% in 2019 and 51.95% in 2025 (p<0.001). Surgical delivery multiplied over sixfold among the poorest wealth quintile (5.53% to 34.37%), rose by 34.83 percentage points among rural mothers (15.66% to 50.49%), and reached 85.12% within private facilities. Multivariate decomposition demonstrated that 87.77% of the overall increase was attributable to compositional shifts rather than behavioral factors (12.23%). The sharp reduction in pregnancies receiving zero ANC visits (35.52% to 8.12%) independently drove 49.04% of the surge, while increased utilization of private clinics and NGOs (18.20% to 54.36%) explained 43.39%. Conversely, biological indications such as fetal macrosomia contributed less than 1%. Conclusions The unprecedented escalation of Caesarean sections in Bangladesh represents a profound pattern of systemic over-medicalization that is independent of baseline clinical risk. The surge is overwhelmingly structural, driven by the disproportionate concentration of deliveries within the commercialized private healthcare sector and supplier-induced demand following integration into formal clinical pathways. Reversing this trajectory necessitates immediate regulatory intervention, specifically the mandate of clinical auditing frameworks such as the Robson classification, to decouple financial incentives from obstetric decision-making in private facilities.