Validation of EuroSCORE II in post-cardiac surgery patients in a tertiary institution in Hong Kong.

Abstract

Introduction: This study aimed to assess the discriminatory ability and calibration performance of the European System for Cardiac Operative Risk Evaluation (EuroSCORE) II, a widely used risk prediction tool, in predicting postoperative mortality among patients undergoing cardiac surgery at Prince of Wales Hospital (PWH) in Hong Kong.

Methods: Complete data from 4180 patients who underwent cardiac surgery at PWH between 2013 and 2023 were available for validation of EuroSCORE II and comparison of its discriminatory ability with the logistic EuroSCORE. Discriminatory performance was primarily assessed using the area under the receiver operating characteristic curve (AUROC). Calibration was evaluated using the Hosmer-Lemeshow test, coefficient of determination (R2), and normalised root mean square error (NRMSE).

Results: EuroSCORE II demonstrated strong discrimination and good calibration for predicting 30-day mortality in the overall cohort (AUROC=0.829; Hosmer-Lemeshow P=0.155) and key subgroups: isolated coronary artery bypass grafting (CABG) [AUROC=0.847; P=0.113], isolated valve surgery (AUROC=0.810; P=0.162), and aortic surgery (AUROC=0.735; P=0.549). More than 85% of the variation in 30-day mortality (R²) was explained across these groups. Compared with the logistic EuroSCORE, EuroSCORE II showed improved discrimination and calibration, with higher AUROC values and lower NRMSE.

Conclusion: EuroSCORE II demonstrates strong discriminatory ability and good calibration for predicting 30-day mortality among patients undergoing cardiac surgery and within key subgroups-isolated CABG, isolated valve surgery, and aortic surgery-in this cohort.