Performance of European System for Cardiac Operative Risk Evaluation (EuroSCORE) II in Prediction of Operative Mortality after Treatment of Infective Endocarditis: Should We Use It or Not?

Background: EuroSCORE II is largely used in the European countries to predictive operative mortality after cardiac surgery influencing heart team decisions and patient information. Current literature report satisfying performance results when this score is applied in general cardiac population. However little has known on the prediction ability of postoperative mortality in specific high risk sub-groups. The present study sought to evaluate the calibration and discrimination of EuroSCORE II regarding mortality after surgical treatment of endocarditis.

Methods: Calibration was obtained by regression between observed mortality and predicted probabilities generated by the EuroSCORE II score. Ideal calibration included a slope value of 1 and intercept of 0. For testing the discrimination of the score, a receiver operating characteristic (ROC) curve was plotted. For the generation of ROC curves, observed in-hospital mortality was used as end point for testing the discrimination of the score. Discrimination was considered good if the area under the curve (AUC) was >0.7 and optimal if it was >0.8. Discrimination of EuroSCORE II was than compared with a more specific score through the method of Hanley and Mc Neil.

Results: A total of 260 patients were operated for infective endocarditis between March 2016 and 2020 in a single-center institution. Operative mortality was 20.8% and mean EuroSCORE II was 11.4 ± 5.07 to 9.05 (from 0.7 to 66.2). The AUC of EuroSCORE II was 0.66 (CI 0.60 to 0.72 p = 0.0002). Calibration was showed an intercept of 0,13 (p < 0.0001) and a slope of 0.0063 (p < 0.0001) suggesting significant overestimation. If the discrimination power of the EuroSCORE II was compared with a more specific score for endocarditis such as the infective endocarditis (IE) mortality risk score there was a significant increment of discrimination power in favor the this more specific score (AUC 0,817- p < 0.008).

Conclusion: Caution must be taken when EuroSCORE II is used for mortality prediction after surgery for endocarditis. In this setting a more specific score such as the infective endocarditis (IE) mortality risk score should be preferred.

Publikationsverlauf

Artikel online veröffentlicht:
13. Februar 2024