Evaluation of accuracy of EuroSCORE II in prediction of in-hospital mortality in patients underwent mitral valve replacement in Egypt

Objective To evaluate the performance of the updated version of European System for Cardiac Operative Risk Evaluation (EuroSCORE II) for prediction of in-hospital mortality after mitral valve replacement. Patients and methods Medical records of 580 Egyptian patients underwent mitral valve replacement in Nasser institute, Suez Canal University, and Minia University hospitals were evaluated in six-year period. Data on demographics, comorbidities, risk factors, operative outcome and postoperative complications were collected with in-hospital mortality as primary end-point. Statistical analysis assessed discriminative power and calibration of EuroSCORE II. Results The study population had younger age, more prevalence of female gender and higher body mass index (BMI) than EuroSCORE II population. The observed in-hospital mortality rate was 2.6% and average of the predicted mortality by EuroSCORE II was 1.60%. EuroSCORE II had low discriminative power (area under the curve 0.52, 95% CI: 0.38–0.66), and poor calibration as determined by Hosmer–Lemeshow (HL) test (P-value = 0.02). On multivariate analysis, other factors not involved in EuroSCORE II were significant predictors for in-hospital mortality including redo-operation, previous valvuloplasty, hypercholesterolemia and chronic liver disease. Conclusion Differences in clinical profile and presence of other local risk factors may explain the reduced discriminative and predictive power of EuroSCORE II in Egyptian patients underwent mitral valve replacement. Other larger prospective studies in this concern are recommended.