Validation of EuroSCORE II for Predicting In-Hospital Mortality among Pakistani Patients Undergoing Cardiac Surgery

Abstract

Objectives: To evaluate the predictive performance of EuroSCORE II in estimating short-term (in-hospital) mortality among Pakistani patients undergoing cardiac surgery, using observed mortality as the gold standard. Methodology: A total of 294 adult patients (≥18 years) undergoing coronary artery bypass grafting (CABG), valvular surgery, or combined procedures were enrolled through consecutive sampling. Patients undergoing cardiac transplantation or emergency salvage procedures were excluded. Preoperative risk was calculated using the official EuroSCORE II calculator. Predicted mortality was compared with observed in-hospital mortality. Model discrimination was assessed using receiver operating characteristic (ROC) curve analysis and area under the curve (AUC), while calibration was evaluated using the Hosmer–Lemeshow goodness-of-fit test. Diagnostic accuracy measures including sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and overall accuracy were calculated. Results: The cohort was predominantly male (73.5%) with a mean age of 56.36 ± 10.06 years. The median EuroSCORE II predicted mortality was 1.45% (IQR: 1.07–2.09), whereas the observed in-hospital mortality was 6.12%. EuroSCORE II consistently underestimated mortality across all procedure types, particularly in combined surgeries. Using a ≥5% cutoff, sensitivity was 16.7%, specificity 98.6%, PPV 42.9%, NPV 94.8%, and overall accuracy 93.5%. The model demonstrated fair discrimination (AUC = 0.79, 95% CI: 0.70–0.89) and acceptable calibration (Hosmer–Lemeshow p = 0.74), despite underestimation of absolute mortality risk. Conclusion: EuroSCORE II demonstrated acceptable calibration and fair discriminatory ability in predicting short-term mortality among Pakistani cardiac surgery patients. However, it systematically underestimated observed mortality, particularly in medium- and high-risk groups and complex procedures. Caution is advised when applying the model to higher-risk populations in this setting.