Performance of the EuroSCORE II Model in Predicting Short-Term Mortality of General Cardiac Surgery: A Single-Center Study in Taiwan.

Abstract

The latest European System for Cardiac Operative Risk Evaluation (EuroSCORE) II is a well-accepted risk evaluation system for mortality in cardiac surgery in Europe. To determine the performance of this new model in Taiwanese patients. Between January 2012 and December 2014, 657 patients underwent cardiac surgery at our institution. The EuroSCORE II scores of all patients were determined preoperatively. The short-term surgical outcomes of 30-day and in-hospital mortality were evaluated to assess the performance of the EuroSCORE II. Of the 657 patients [192 women (29.22%); age 63.5 ± 12.68 years], the 30-day mortality rate was 5.48%, and the in-hospital mortality rate was 9.28%. The discrimination power of this new model was good in all populations, regardless of 30-day mortality or in-hospital mortality. Good accuracy was also noted in different procedures related to coronary artery bypass grafting, and good calibration was noted for cardiac procedures (p value > 0.05). When predicting surgical death within 30 days, the EuroSCORE II overestimated the risk (observed to expected: 0.79), but in-hospital mortality was underestimated (observed to expected: 1.33). The predictive ability [area under the curve (AUC) of the receiver operating characteristic (ROC) curve] and calibration of the EuroSCORE II for 30-day mortality (0.792) and in-hospital mortality (0.825) suggested that in-hospital mortality is a better endpoint for the EuroSCORE II. The new EuroSCORE II model performed well in predicting short-term outcomes among patients undergoing general cardiac surgeries. For short-term outcomes, in-hospital mortality was better than 30-day mortality as an indicator of surgical results, suggesting that it may be a better endpoint for the EuroSCORE II.