National health insurance and the choice of delivery facility among expectant mothers in Ghana.

The effectiveness of health insurance in removing barriers to the utilisation of maternal healthcare in order to curb maternal mortality especially in developing countries is gaining ground. However, in assessing the effects of health insurance on choice of delivery facilities, previous studies either put all delivery services together and used binary techniques or failed to address endogeneity problem. Moreover, the age of data used for such analysis in Ghana may not tell a convincing story. This study used data from the 2014 to 2008 Ghana Demographic and Health Surveys with a sample of 6319 women and employed multinomial endogenous treatment effects models with Conditional Mixed Process estimator to examine the effects of national health insurance scheme (NHIS) on the choice delivery facility in Ghana. We found that NHIS has varied effects on the use of delivery services across service providers in the health system. Relative to home delivery services, being insured increases the probability of using public hospitals, public clinics and private health facilities for delivery by 20.3 percent, 9.1 percent and 2.3 percent respectively. Moreover, relative to an insured woman who gave birth before 2008, her counterpart who gave birth after 2008 is 6.3 percent, 4.9 percent and 0.77 percent more likely to use public hospitals, public clinics and private health facilities respectively for delivery.