Validation of EuroSCORE II in patients undergoing coronary artery bypass grafting (CABG) surgery at the National Heart Institute, Kuala Lumpur: a retrospective review

Background: The European System for Cardiac Operative Risk (EuroSCORE) II was developed in 2011 to replace the aging EUROScore for predicting in-house mortality after cardiac surgery. Our aim was to validate EuroSCORE II in Malaysian patients undergoing coronary artery bypass graft (CABG) surgery at our Institute. Methods: A retrospective single-center study was performed. A database was created to include EuroSCORE II values and actual mortality of 1718 patients undergoing CABG surgery in Malaysia from 1st January to 31st December 2016. The goodness-of-fit of EuroSCORE II was determined by the Hosmer-Lemeshow goodness-of-fit test and discriminatory power with the areas under the receiver operating characteristics (ROC) curve (AUC). Results: Observed mortality rate was 4.66% (80 out of 1718 patients). The median EuroSCORE II value was 2.06% (Inter Quartile Range: 1.94%) (1st quartile: 1.45%, 3rd quartile: 3.39%). The AUC for EuroSCORE II was 0.7 (95% CI 0.640 - 0.759) indicating good discriminatory power. The Hosmer-Lemeshow goodness-of-fit test did not show significant difference between expected and observed mortality in accordance to the EuroSCORE II model (Chi-square = 13.758, p = 0.089) suggesting good calibration of the model in this population. Cross-tabulation analysis showed that there is slight overestimation of EuroSCORE II in low-risk groups (0-10%) and slight underestimation in high-risk groups (>20%). Multivariate logistic regression analysis showed that gender, age, total hospital stay, serum creatinine and critical pre-operative state are significant predictors of mortality post-CABG surgery. Conclusion: This study indicated that the EuroSCORE II is a good predictor of post-operative mortality in the context of Malaysian patients undergoing CABG surgery. Our study also showed that certain independent variables might possess higher weightage in predicting mortality among this patient group. Therefore, it is suggested that EuroSCORE II can be safely used for risk assessment while ideally, clinical consideration should be applied on an individual basis.