BACKGROUND: Since 2008, data concerning pediatric cardiovascular surgeries performed in Japan have been collected in the Japan Congenital Cardiovascular Surgery Database (JCCVSD). We assessed the quality of the JCCVSD data through data verification activities conducted in 2010. METHODS: During 2008 to 2009, 3345 patients with 4327 procedures at 25 hospitals were registered in the JCCVSD. Among them, six sites were selected for data verification. The completeness of case registration was assessed by comparison with original operational logs. Also, data accuracy of patient demographics, surgical outcomes, and processes were assessed with 10% of the registered cases by comparison with medical records. RESULTS: Verification of case registration completeness involved 968 (28.9%) patients and 1279 (29.1%) procedures. As to completeness, we confirmed 1266 (99.0%) of the 1279 procedures. Data accuracy was verified for 129 (3.9%) patients. Accuracy of status of discharge and 30 and 90 days after surgery were very high (99.2%, 100%, and 100%, respectively). Data items with numeric information exhibited lower exact accuracy due to typing error and inconsistent use of rounding; however, the differences between the submitted and the original data were not statistically significant. CONCLUSIONS: High completeness and acceptable range of data accuracy were verified for the data submitted to the JCCVSD in 2008 to 2009. The high accuracy regarding follow-up outcomes was especially noteworthy. The initial success of the JCCVSD should be strengthened through further sophistication of registration protocol, continual training of data managers and auditors, and rigorous expansion of verification activities.
BACKGROUND: Since 2008, data concerning pediatric cardiovascular surgeries performed in Japan have been collected in the Japan Congenital Cardiovascular Surgery Database (JCCVSD). We assessed the quality of the JCCVSD data through data verification activities conducted in 2010. METHODS: During 2008 to 2009, 3345 patients with 4327 procedures at 25 hospitals were registered in the JCCVSD. Among them, six sites were selected for data verification. The completeness of case registration was assessed by comparison with original operational logs. Also, data accuracy of patient demographics, surgical outcomes, and processes were assessed with 10% of the registered cases by comparison with medical records. RESULTS: Verification of case registration completeness involved 968 (28.9%) patients and 1279 (29.1%) procedures. As to completeness, we confirmed 1266 (99.0%) of the 1279 procedures. Data accuracy was verified for 129 (3.9%) patients. Accuracy of status of discharge and 30 and 90 days after surgery were very high (99.2%, 100%, and 100%, respectively). Data items with numeric information exhibited lower exact accuracy due to typing error and inconsistent use of rounding; however, the differences between the submitted and the original data were not statistically significant. CONCLUSIONS: High completeness and acceptable range of data accuracy were verified for the data submitted to the JCCVSD in 2008 to 2009. The high accuracy regarding follow-up outcomes was especially noteworthy. The initial success of the JCCVSD should be strengthened through further sophistication of registration protocol, continual training of data managers and auditors, and rigorous expansion of verification activities.