TB Data Improvement in Nkembo Health Treatment Center in Libreville, Gabon

Although the estimated tuberculosis (TB) incidence in Gabon is declining, there have been challenges with treatment coverage, HIV status and treatment outcome documentation. Thus, the National TB Program (NTP) conducted an innovative data review at the Nkembo Health Treatment Center in Libreville, which manages more than 70% of Gabonese TB patients. Since our hypothesis was that the Nkembo treatment center was struggling with data mismanagement due to the workload, the objective was to perform a TB data quality review and triangulation exercise at the Nkembo health facility in Libreville, from January to August 2023, and propose recommendations for data improvement. Methods: The study used the data reconciliation method. This is a process that involves comparing and aligning data from multiple sources to ensure consistency, accuracy, and integrity. The primary purpose of data reconciliation is to identify and resolve discrepancies or differences between datasets and make them consistent. Using the “TB onion model”, analysis identified data mismanagement as a key contributor to underreporting. A data review compared TB records to TB registry data and patient folders from January to August 2023 for notification and to the 2022 cohort for treatment results. The study focused on notified TB cases, HIV status and TB treatment outcome documentation. Discrepancies were reconciled, and treatment outcomes re-evaluated. Results: After review, statistically significant increases were observed: +22% for total TB cases (p = 0.0003), +141% for the number of TB cases with known HIV status (p = 0.0017) and +104% for the number of TB cases successfully treated (p = 0.0001), as compared with the previous data. Discussion: This data reconciliation showed the usefulness of triangulation across data sources to improve the completeness of data. Also, current reported data underestimate the number of reported cases, documentation of HIV status, and treatment success. Conclusions: The study shows that data reconciliation can improve TB programmatic data completeness to better reflect program performance.