| Popis: |
Background Congenital heart surgery databases are tools for internal programmatic evaluation, benchmarking institutional results to multi-institutional aggregate data, and research. Therefore, it is essential to ensure the completeness and accuracy of data. This study analyzes the results of ten years of on-site source data verification of the European Congenital Heart Surgeons Association Congenital Heart Surgery Database (ECHSA CHSD). Methods All data forms verified between 2009 and 2018 were analyzed. The data form consists of 12 data elements: dates of birth, admission, surgery, discharge, and death; weight; case category; cardiopulmonary bypass time; aortic cross-clamp time; validation rules; diagnoses; and procedures. Descriptive data calculation and rates of completeness and accuracy were determined. The trend of error rate of seven centers with ≥5 visits was analyzed. Results Sixty-nine on-site verification visits took place at 17 centers. A total of 26,245 cases were verified; 2,841 of these 26,245 cases (10.8%) showed an error. The total mean error rate of centers for all years was 12.3 ± 2.1%. Rates of completeness and accuracy were 99% and 89.2%, respectively. Coded diagnoses and procedure analysis revealed that 716 (2.7%) and 456 (1.7%) datasets were incorrect, respectively. Rates of completeness and accuracy of dates were 100%, and 97.1%, respectively. Validation fields showed no errors. Conclusion Source data verification is an appropriate tool to determine completeness and accuracy of data. The ECHSA CHSD verification analysis of a ten-year period showed a high level of completeness and accuracy. The verified data of the ECHSA CHSD are well-suited for benchmarking and research. |
