Validation of the JapanSCORE versus the logistic EuroSCORE for predicting operative mortality of cardiovascular surgery in Yamaguchi University Hospital

Abstract

Recent studies have shown that the European System for Cardiac Operative Risk Evaluation (Euro-SCORE) is a reliable risk model in cardiac surgery. In Japan, the JapanSCORE has been devised as the Japanese original risk model for cardiovascular surgery. We evaluated the validity of the JapanSCORE, especially in comparison with the logistic EuroSCORE METHODS: We calculated the predicted mortality according to two risk models for 523 consecutive patients who underwent cardiovascular surgery within a 6-year period (July 2003 to June 2009) at Yamaguchi University Hospital. We assessed the scores' validity by calculating the area under the receiver operating characteristics curve (C-index) and by the chi-squared test or Fisher's exact test. The C-indexes were 0.688 with the logistic EuroSCORE and 0.770 with the JapanSCORE (P = 0.053). Although the difference was not significant, the JapanSCORE tended to be more accurate. The C-indexes limited to isolated coronary artery bypass grafting (CABG) were 0.564 with the logistic EuroSCORE and 0.790 with the JapanSCORE (P = 0.001). The Japan-SCORE was significantly more valid than the logistic EuroSCORE for isolated CABG. The actual mortality was 4.0%; the mean predictive mortality was 5.6% with the JapanSCORE and 15.1% with the logistic Euro-SCORE. The mortality predicted by the JapanSCORE was closer to the observed mortality. There was a significant difference between the observed mortality and the logistic EuroSCORE (P < 0.0001). The JapanSCORE seems to be a more reliable risk model than the logistic EuroSCORE for patients undergoing cardiac or thoracic aortic surgery at Yamaguchi University Hospital.