Functionalities of electronic routine health information systems related to newborn data: findings of the IMPULSE study in Uganda, Ethiopia, Tanzania, and the Central African Republic.

Background: Adequate functionality of electronic routine health information systems (eRHISs) is crucial for data use, yet few studies explored it in relation to newborn and stillbirth data in Africa. Methods: We conducted this cross-sectional study between November 2022 and July 2024 in data offices at central and subnational levels in 12 regions and 4 city administrations in the Central African Republic (CAR), Ethiopia, Tanzania, and Uganda. Except for end-user perspectives (collected via interviews), we collected data related to eRHIS functionalities by direct observation following standard operating procedures as for the Every Newborn-Measurement Improvement for Newborn & Stillbirth Indicators (EN-MINI) Tool 3.1, based on the Performance of Routine Information System Management (PRISM) framework. We analysed data according to the PRISM Users' Kit. Results: We assessed 53 data offices in total. All countries used the same software application, the District Health Information Software 2 (DHIS2). Settings were heterogeneous across countries, with a tendency for DHIS2 to offer fewer functionalities to users in the CAR. Overall functionalities for generating facility annual summary reports (100% in all countries) and for calculating percentage of reports received/expected (75.0% in Ethiopia to 88.9% in Tanzania) were widely available. Data integration and data disaggregation, meanwhile, had lower availability. Functionalities for calculating coverage on specific indicators, such as respectful care, were lacking in all countries, those for quality assurance varied across countries, while those related to data visualisation were almost always available in Uganda and Tanzania, but showed specific gaps in Ethiopia (i.e. for early initiation breastfeeding), and most often lacked in the CAR. Most end-users indicated needs for eRHIS improvement (ranging from 37.5% in Ethiopia to 100% in the CAR; P = 0.001), with 17.0% reporting needs for major improvement (from 10.0% in Uganda to 28.6% in the CAR; P = 0.001). Subgroup analyses suggested high within-country heterogeneity and more eRHIS functionalities available at central vs. subnational level. Conclusion: Identified strengths and gaps in existing DHIS2 functionalities can inform the design of context-specific interventions that will enhance data use for reducing neonatal mortality and stillbirth rates.