Data Quality of Out-Of-Pocket Payment on Institutional Delivery in India

Estimates of out-of-pocket (OOP) payment on health care are increasingly used in research and policy. In India, estimates of OOP payment are usually derived from health surveys carried out by the National Sample Survey (NSS). The questions on OOP payment on delivery care have recently integrated in the last two rounds of India’s National Family and Health Survey (NFHS-4 & NFHS-5). There are several issues relating to design of questions, reporting and recording of responses that have bearing on reliability of OOP estimates. This paper compares the OOP estimates from latest rounds of two of the large-scale population-based surveys; NFHS-5, 2019-21 and the National Sample Survey (NSS), 2018. We also highlight the type of question canvassed and its implications on OOP estimates of NFHS-5 survey. We used 155,624 births that were reported between in NFHS-5 and a total of 27,664 hospitalised cases for delivery care that were recorded in 75thround of NSS health survey, 2018. We have used descriptive statistics and two-part regression model to examine variations of OOP across surveys. We found large variations in distribution of OOP payment in NFHS-5 and NSS survey. Based on births during the five years preceding the survey, the OOP payment on institutional birth from public health centres in India from NFHS-5 was INR 2,894 (95% CI:2843-2945) compared to INR 2,738 (95% CI: 2644-2832) from NSS. Variations are similar for those availing services from private health centres. Controlling for socio-economic and demographic characteristics, the OOP payment from NFHS was lower among poorest and higher among richest compared to NSS. The variations in OOP across two surveys were larger across states of India. The variations in OOP payment across surveys were possibly due to structure of questions, recall bias, and variations in price level. We suggest to canvass standardised questions across surveys to obtain reliable OOP estimates across surveys.