Impact evaluation of contracting primary health care services in urban Bangladesh

BACKGROUND: The Urban Primary Health Care Project (UPHCP) was implemented by the Government of Bangladesh in response to rapid urbanization and growing inequalities in access to and quality of primary health care. The goal of the project was to improve health status of the urban poor living in city corporations and municipalities through the provision of health care services by NGOs that are contracted through public-private partnership. The first phase of the project started in 1998 and the project is currently in its fourth phase covering more urban areas than the first three phases. This study evaluates the impact of the second phase project (UPHCP-II) on health outcomes, mainly child diarrhea, acute respiratory infection, antenatal and postnatal care, skilled birth attendance, breastfeeding prevalence, contraceptive prevalence, sexually transmitted infections, and HIV/AIDS awareness. METHODS: The effect of the project was estimated through propensity score matching between project and non-project areas comparing baseline and endline surveys over a six-year period from 2006 to 2012. An innovation of this study is the recalibration of the sampling weights that allows the use of these two independent surveys in impact evaluation. RESULTS: Over the six-year period, UPHCP-II improved the health status of the population in project areas compared to non-project areas. The study found significant improvement in health outcomes in terms of reduced diarrhea and acute respiratory infection in children, which explains the downward trend in child mortality rate. Moreover, the project also improved antenatal care and skilled birth attendance. Contraceptive prevalence and HIV/AIDS awareness and avoidance increased, and sexually transmitted infections decreased. CONCLUSIONS: UPHCP-II was effective in achieving its health outcome targets, while previous studies show that it was efficient in the delivery of health care and clients were highly satisfied because health facilities were in close proximity, and doctors and staff were perceived as responsive in delivering high quality of care. The results of this study could help inform future design and implementation of urban health interventions that involve contracting primary health care service delivery in Bangladesh and other similar settings. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s12913-019-4406-5) contains supplementary material, which is available to authorized users.