Comparison between different risk scoring algorithms on isolated conventional or transcatheter aortic valve replacement

Background There are a number of scoring systems for risk evaluation in cardiac surgery, the most important of which are the European System for Cardiac Operative Risk Evaluation (EuroSCORE), The Society of Thoracic Surgeons (STS) score, the ACEF score (acronym for age, preoperative creatinine, and ejection fraction), and more recently, the new EuroSCORE-II. The aim of our study was to analyze and compare the predictive value of these scores in patients undergoing aortic valve replacement (AVR) or transcatheter aortic valve replacement (TAVR). Methods A total of 1,512 consecutive patients undergoing either conventional AVR (n = 1,066) or TAVR (transfemoral, n = 291; transapical, n = 155) were enrolled. Logistic and additive EuroSCORE of all patients were 13.3% ± 13.2% and 7.8% ± 3.3%, on average. The mean STS score, ACEF score, and EuroSCORE-II were 5.7% ± 5.0%, 1.5% ± 0.7%, and 4.2% ± 4.9%, respectively. Results Overall mortality at 30 days was 6.3%. The area under the curve (AUC) was 73.8 for the logistic EuroSCORE and 73.5 for the additive EuroSCORE. The STS score gave an AUC of 70.8. The AUCs for the ACEF and EuroSCORE-II were 63.8 and 71.2, respectively. In the transfemoral TAVR group, AUCs were 59.8 and 59.3 for the logistic and additive EuroSCORE, respectively, 63.2 for the STS score, and 55.9 and 55.4 for the ACEF and EuroSCORE-II, respectively. In the transapical TAVR group, AUCs were 88.0 and 82.8 for the logistic and additive EuroSCORE, respectively, 79.0 for the STS score, and 61.7 and 83.7 for the ACEF and EuroSCORE-II, respectively. Conclusions Overall, 30-day mortality was best predicted by the STS score. Discrimination threshold predicting mortality was equal between all other risk calculators. Surprisingly, the new EuroSCORE-II was not superior to other models in risk prediction for AVR and TAVR patients.