Abstract
The World Health Organization recognises Routine Health Information System (RHIS) data as integral to data-driven health systems; needed to improve intrapartum outcomes for maternal and newborn health worldwide. However, research in Bangladesh and Tanzania suggests that mode of birth affects register data accuracy, but little is known about why. To address this gap, we undertook qualitative research in these two public-sector health systems. We conducted 44 in-depth interviews in Bangladesh (Sept-Dec 2020) and 35 in Tanzania (Feb-April 2023). Participants included health and data professionals, managers, and leaders from sub-national and national levels. Thematic analysis was undertaken with inductive and deductive coding. Emerging themes were compared/organised using determinants outlined in the Performance of Routine Information System Management (PRISM) framework. Mode of birth affected RHIS data as one part in a multidimensional system; having a caesarean changed the location of birth, availability of health professionals, and the care pathway, impacting data flow and documentation processes at facility-level. Standardised registers were available in the labour wards, but not in all operating theatres. Health professionals in both countries described feeling overwhelmed by duplicative data tasks and competing clinical care responsibilities, especially in labour wards with low staffing ratios. Health professionals perceived electronic data systems to increase duplication (for all modes of birth), along with other organisational factors. In conclusion, mode of birth influenced processes for routine data collection and use because it affected where, what, when, and by whom data were recorded. We found challenges for capturing register data, leading to potential data gaps, especially for caesarean births. Our findings suggest a broader lens is needed to improve the systems, collection, and use of individual-level data for aggregation, not just registers. Co-design of RHIS processes and tools could rationalise the data burden and increase availability and quality of perinatal data for use.