Data quality and associations over 4 years of use of WHO’s Clinical Registry - Trauma Module, in Rwanda

Abstract Objective Poor quality of injury care causes avoidable death and disability. Quality data is integral to care-quality improvement. We assessed longitudinal data quality of the Trauma Module (TM) of WHO’s Clinical Registry in Rwanda, towards understanding data collection challenges in low resourced settings. Methods Retrospective analysis of TM data between its implementation in August 2019 and July 2023. We describe trends in overall TM case and variable missingness and their associations with availability of funding and formal training. Variable missingness is also described by each of the 7 domains of the TM. Findings The TM was initiated in 4 hospitals in August 2019 and expanded to 6 in June 2021. Between implementation and July 2023, 13,513 cases were entered. Annual numbers of cases varied per hospital (maximum of 2051, minimum of 259). Case missingness ranged from 60% to 66% across all hospitals. The median annual variable missingness varied between hospitals from 28% to 41%. Over the whole period, across all hospitals, median missingness was 39.63% (IQR 0.65 – 82.46); 12 variables had 100% missingness and 15 variables, 0% missingness. Variable missingness depending on TM domain; highest in “In-ward care & Facility disposition” (77.30%) and “EU disposition” (51.67%). Data collection was dependent on external funding; when funding ceased in 2023, case and variable missingness increased to 100%. Short term increases in data collection were noted following periodic formal trainings. Conclusions Although TM sustainment in Rwanda relied on external funding, even with funding and training, data quality was variable.