Background: Heart failure (HF) remains a significant burden following transcatheter aortic valve replacement, adversely impacting survival and quality of life. Identification of patients who may benefit from closer monitoring or adjunctive medical therapy to reduce the risk of HF is an unmet need. The objective of this study was to develop and internally validate a clinical prediction model to determine the 1-year risk of HF hospitalization or death after transcatheter aortic valve replacement.
Methods: Using the Society of Thoracic Surgeons/American College of Cardiology Transcatheter Valve Therapy Registry, we analyzed patients who underwent successful transcatheter aortic valve replacement for aortic stenosis and survived to discharge between 2016 and 2019. Covariates were selected based on expert opinion and prior literature. A hierarchical cumulative odds regression model was used to predict a composite outcome of (1) all-cause death, (2) ≥2 HF readmissions, or (3) 1 HF readmission at 1 year.
Results: Among 78 384 patients (median age, 82 years; 45.6% female), 17.4% experienced the composite outcome, including death (10.9%), ≥2 HF readmissions (1.6%), and 1 HF readmission (4.9%). The model demonstrated good discrimination (C statistic, 0.753 derivation and 0.747 validation) and excellent calibration. Among 1-year survivors, performance in predicting HF readmission as an isolated outcome was similar (C statistic, 0.753). A simplified model, including the top 12 variables from the full model, maintained comparable performance (C statistics, 0.74-0.75).
Conclusions: This prediction model effectively stratifies post-transcatheter aortic valve replacement patients by risk of death or HF readmission, supporting its use to guide clinical surveillance and clinical trial enrollment for adjunctive medical therapies aimed at mitigating this risk.
Keywords: calibration; models, statistical; patient readmission; quality of life; transcatheter aortic valve replacement.