Frailty as a predictor of delayed discharge and 12-month mortality following transcatheter aortic valve implantation

Transcatheter aortic valve implantation (TAVI) is the principal treatment for aortic stenosis in older and high-risk patients. Whilst important in holistic decision making, frailty indices vary in ease of use, objectivity, and accuracy. We assess prognostic ability of the Clinical Frailty Scale (CFS) and Essential Frailty Toolset (EFT) in predicting delayed discharge and 1-year mortality following TAVI.

Prospective, single centre, observational cohort study of patients undergoing TAVI at a tertiary centre in the UK from 2020-2022. Clinical characteristics, CFS, and EFT were collected pre-procedurally. Primary outcomes were length of hospital stay following TAVI and mortality at 1- year post-procedure. Generalized regression with elastic net is used to assess the additive discriminative ability of frailty scores to predict outcomes. Stepwise regression is used to identify informative outcome predictors.

One hundred and four patients were included with mean age 81.3 ± 5.9 years, 47% female. Median post-procedural stay length was 2 days (IQR 2.75). Use of EFT and CFS resulted in a statistically significant prediction of survival at 1 year post TAVI (EFT: ΔLogL of 2.68, ChiSq P = 0.021; RCFS ΔLogL: 1.42, ChiSq P = 0.092) compared to predictive model using conventional clinical characteristics. EuroSCORE II and TAVI2SCORe were poor predictors of outcomes. CFS and access site (femoral vs. non-femoral) significantly improved prediction of delayed discharge.

EFT informs decision making on survival 1-year post-TAVI, beyond conventional measures. Use of non-femoral access site and high CFS are the best predictors of delayed discharge. Surgical risk stratification scores poorly predict medium-term TAVI survival and delayed discharge.