Measuring respectful maternal and newborn care in Nepal: Comparing linked observation and interview data- prospective...

Respectful maternal and newborn care is the cornerstone of high-quality care, however, measuring experience of respectful care has challenges since it can be subjective, and dependent on expectations. In this study, we assess the concordance between women's reported experiences of respectful maternal and newborn care and independent observation of their care in Nepal.

This is a secondary analysis of a prospective cohort study among 22832 pregnant women conducted in three high volume hospitals in the country: Koshi Provincial Hospital (Hospital A), Bharatpur Hospital (Hospital B), and Lumbini Provincial Hospital (Hospital C) for 18 months between April 2017 and October 2018. The study implemented direct observation during and semi-structured interviews at discharge to evaluate the quality of maternal and newborn care in three large public hospitals. For this analysis, three domains for respectful maternal and newborn care were considered: 1) consent and counselling 2) respect and dignity of care, and 3) care provision. The two data sources (observation checklist and semi-structured interview) were plotted to these three domains to identify common indicators. The level of agreement (LOA) between two measurements was compared using Cohen kappa scores (κ) and Bland Altman plots.

During the study period, 22832 women had both observation and interview completed. For consent and counseling, 78.8% of women reported being informed about routine care while only 47.3% were observed to have been consented and counseled (k, LOA = 59.1%). For respect and dignity of care, 99.0% of women reported being treated with dignity and respect and 96.4% were observed (k, LOA = 95.4%). For care provision, 37.9% reported that the infant was kept in immediate skin-to-skin contact after delivery while only 3.9% were observed (k, LOA = 61.7%).

A significant difference existed between observed and self-reported measures of maternal and newborn care. This study highlights the need for a measurement approach that incorporates independent observations alongside self-reported data. There is also a need to further explore concordance between different sources for progress monitoring.