Institutional delivery in public and private sectors in South Asia: A comparative analysis of prospective data from four demographic surveillance sites

__Background:__ Maternity care in South Asia is available in both public and private sectors. Using data from demographic surveillance sites in Bangladesh, Nepal and rural and urban India, we aimed to compare institutional delivery rates and public-private share. __Methods:__ We used records of maternity care collected in socio-economically disadvantaged communities between 2005 and 2011. Institutional delivery was summarized by four potential determinants: household asset index, maternal schooling, maternal age, and parity. We developed logistic regression models for private sector institutional delivery with these as independent covariates. __Results:__ The data described 52 750 deliveries. Institutional delivery proportion varied and there were differences in public-private split. In Bangladesh and urban India, the proportion of deliveries in the private sector increased with wealth, maternal education, and age. The opposite was observed in rural India and Nepal. __Conclusions:__ The proportion of institutional delivery increased with economic status and education. The choice of sector is more complex and provision and perceived quality of public sector services is likely to play a role. Choices for safe maternity are influenced by accessibility, quantity and perceived quality of care. Along with data linkage between privat