Maternal health data reliability: Findings from Sidama region, Ethiopia

High-quality routine health information is critical for monitoring maternal health indicators and guiding effective health interventions. However, data quality remains a persistent challenge in low-resource settings like Ethiopia, particularly at sub-national levels. This study assessed the reliability of maternal health service data and its determinants in public health facilities in Sidama Region, Ethiopia.

Institution-based cross-sectional study was conducted from January 5 to 30, 2025 on 517 maternal health service units selected by a multistage sampling method. Data collection tool was adapted from the Performance of Routine Information System Management framework. Data were collected from each unit by using structured interviews with health professionals and reviews of registration books, individual medical records, and monthly reports. Data quality was measured by accuracy, internal consistency, completeness, and timeliness. Multilevel logistic regression was used to identify factors associated with good data quality.

Overall, 63.1% of service delivery units demonstrated good data quality. Register content completeness (66.5%), report timeliness (65.0%), and report accuracy (59.2%) showed relatively high performance, while internal consistency (34.0%) and report content completeness (37.9%) remained low. Factors significantly associated with good data quality included the availability of standardized indicators (AOR = 2.30), user-friendly reporting formats (AOR = 2.94), trained staff (AOR = 2.33), and supervisory feedback (AOR = 2.49). Health posts were significantly less likely to achieve good data quality compared to hospitals (AOR = 0.19).

Maternal health data quality, particularly internal consistency and report completeness, in Sidama Region was below the national standards. Improving data quality should prioritize strengthening technical capacity, enhancing supervision and feedback, promoting staff motivation, and investing in under-resourced health posts.