Determination of new risk factors for predicting long-term mortality in patients undergoing TAVI: can the conventional risk scores be used as a long-term mortality predictor?

BACKGROUND: Surgical risk in transcatheter aortic valve implantation (TAVI) patients is determined by conventional surgical risk scoring systems. However, these risk scores have been developed to predict surgical mortality. Because of the insufficient predictive ability of the conventional risk scores, novel risk scores are required for predicting long term mortality in TAVI patients. AIMS: To investigate the predictive value of conventional risk scores for predicting long term mortality. Also, the role of other factors on long term mortality will be evaluated. METHODS: Our study included 121 patients who underwent TAVI. Transfemoral approach was used in all patients. RESULTS: The mean Logistic European System for Cardiac Operative Risk Evaluation (EuroSCORE), EuroSCORE-II and Society of Thoracic Surgeons (STS) were 27.4 (9.7), 7.9 (4.6) and 4.6 (2.4), respectively. In-hospital mortality rate was 1.7%. Any risk scoring system did not predict in-hospital mortality correctly. The STS score corresponded to the mortality rate of approximately the 2nd month, EuroSCORE-II to the mortality rate of the 6th month and Logistic EuroSCORE to the mortality rate of the 30th month. Male gender (OR: 5.668, 95% CI: 1.055-30.446, P = 0.04) and low albumin level before TAVI (OR: 0.109, 95% CI: 0.018-0.654, P = 0.02) were found to be the independent predictors of long term mortality. CONCLUSIONS: Although all conventional risk scores overestimated in-hospital mortality, STS predicted 2-month mortality, EuroSCORE-II 6-month mortality and Logistic EuroSCORE 30-month mortality. The independent predictors of long term mortality were male gender and low blood albumin level before the TAVI procedure.