Data verification of nationwide clinical quality registries

Background Clinical auditing is an emerging instrument for quality assessment and improvement. Moreover, clinical registries facilitate medical research as they provide 'real world' data. It is important that entered data are robust and reliable. The aim of this study was to describe the evolving procedure and results of data verification within the Dutch Institute for Clinical Auditing (DICA). Methods Data verification performed on several (disease-specific) clinical registries between 2013 and 2015 was evaluated. Sign-up, sample size and process of verification were described. For each procedure, hospitals were visited by external data managers to verify registered data. Outcomes of data verification were completeness and accuracy. An assessment of the quality of data was given per registry, for each participating hospital. Using descriptive statistics, analyses were performed for different sections within the individual registries. Results Seven of the 21 registries were verified, involving 174 visits to hospital departments. A step-by-step description of the data verification process was provided. Completeness of data in the registries varied from 97 center dot 2 to 99 center dot 4 per cent. Accuracy of data ranged from 88 center dot 2 to 100 per cent. Most discrepancies were observed for postoperative complications (0 center dot 7-7 center dot 5 per cent) and ASA classification (8 center dot 5-11 center dot 4 per cent). Data quality was assessed as 'sufficient' for 145 of the 174 hospital departments (83 center dot 3 per cent). Conclusion Data verification revealed that the data entered in the observed DICA registries were complete and accurate.