Assessment of the Quality of post-abortion care records in 20 Public Health Facilities in Uganda: What are the gaps and how can we improve quality?

backgroundEffective strategies for reducing maternal mortality depend on the accuracy of data. In sub-Saharan Africa, however, weak health systems often result in significant gaps in data collection and analysis. These gaps can lead to misreporting, missing data, or double-counting at health facilities, ultimately skewing the aggregated data reported by Ministries of Health. This is a big problem given that the data shapes planning and policy of the country. In this study, we assessed the data quality of Post-abortion Care (PAC) records across 20 health facilities in Uganda.

methodsA two-stage design was employed. In stage 1, DHIS2 data from 80 public health facilities were reviewed for completeness, timeliness, and internal consistency (2018-2021). In stage 2, 20 facilities that met a predefined eligibility criterion (≥ 50% discrepancy or missing data during stage 1) were purposively selected for in-depth validation. Data were collected retrospectively from facility registers for the last 7 months of 2021 and prospectively for the first 6 months of 2022. Data quality was assessed using adapted WHO data quality review (DQR) metrics.

resultsHealth facilities achieved a 100% timeliness rate for report submissions, with all documents provided by the 7th day of each month. Although a statistically significant positive correlation was observed between the number of women who received PAC and those who received post-abortion family planning (PAFP) (r = 0.083, p < 0.001), the association was very weak suggesting limited internal consistency between the two indicators and raising questions about the reliability of data linkages between care and follow-up services in routine reporting systems. Moreover, comparing facility records with DHIS2 data revealed that outpatient records were more accurate than inpatient records, with PAC data being more inaccurate than other maternal health indicators. Further analysis by facility type indicated a higher number of case-load outliers at Health Centre IIIs and IIs (1049) compared to hospitals and Health Centre IVs, highlighting disparities in data quality across different levels of healthcare facilities.

conclusionDespite high submission rates, PAC data accuracy was notably low, highlighting the need for better data management and record-keeping, particularly in lower-level health facilities. Addressing these disparities is crucial for improving maternal health outcomes, emphasizing the necessity for targeted interventions to enhance data accuracy and reliability within Uganda's health system.