Birth-order based determinants of caesarean sections in Bangladesh: A multilevel mixed effect regression approach

Background The use of Caesarean section (CS) in Bangladesh has increased significantly above the medically justified levels and it varies by socioeconomic background. However, there is little research on the determinants of CS that differ by parity, although there are solid clinical and social arguments that parity-specific effects are likely. This paper used a multilevel-based investigation to explore birth-order-specific determinants of CS in Bangladesh. Methods This study used data from Bangladesh Demographic and Health Survey 2022 to explore the determinants of caesarean section in Bangladesh. The binary outcome variable of the study was whether the respondent had CS in last birth or not. To detect heterogeneity by birth order, the dataset was stratified into two groups; first births and second or higher births. Separate multilevel mixed effect models were then fitted for each group to identify how CS varied by parity. Results The results showed that 45.5% of births were delivered by CS. This rate was higher for first births (52.4%) than for second or higher-order births (41.4%). The community level model performed better for both the stratified models. For first births, age of mothers at birth, higher educational attainment, size of the baby at birth, and antenatal care visits were significant predictors of CS, while rural–urban disparities were reduced after adjustment. For second or higher-order births, higher maternal education, ANC (≥ 4 visits), exposure to the mass media, richer wealth status, and urban residence were important predictors of CS, with notable regional variations (lower odds in Chattogram and Sylhet compared with Dhaka). Conclusion These findings highlight the need for parity-specific, context-specific strategies to reduce unnecessary CS while ensuring access for women in genuine need, particularly by addressing socioeconomic and regional inequities in maternal healthcare.