Data quality assessment of tuberculosis request forms at Upper East Regional Hospital, Ghana, 2025

Introduction: Tuberculosis (TB) diagnosis is a challenge in most resource-limited countries. Some of these countries have adopted the sample referral system to bring TB diagnoses closer to their people and also to improve the turnaround time in TB management. Inaccuracies and incomplete TB request forms can delay diagnosis, misdirect samples and lead to poor patient outcomes. Assessing the completeness of these forms helps identify bottlenecks and enhances data quality in TB surveillance systems. We assessed the completeness of TB request forms to improve data quality and service delivery at the Upper East Regional Hospital. Methods: A Cross-sectional retrospective review of TB request forms at the Upper East Regional Hospital Laboratory was conducted. All TB request forms submitted between January and June 2025 were analyzed. Percentage completeness calculated per form, per field and completeness were analyzed by facilities. Data was analyzed using Microsoft Excel and the results presented in tables and graphs. Results: Of the 685 TB request forms reviewed, 674 (98%) of them had at least one missing parameter. Fields that were commonly left blank were District TB Number, Presume TB Number, Residential Address, Test Requested and Patient Category, representing 88.3%, 46.6%, 23.6%, 19.7% and 9.9% respectively. On facility level completion, 20 out of 24 facilities (83.3%) showed 0% completion. The most common personnel factor affecting request form completion among spokes facilities is adherence to sample referral protocols, which recorded 85% noncompliance, compared to inadequate training and high workload, which recorded 35% and 55%, respectively. Conclusion: Incomplete filling of TB test request forms is a regular occurrence among spokes facilities of the Upper East Regional Hospital Laboratory. Refresher training on the proper completion of the TB request form was organized for facilities based on our recommendation.