Abstract 13505: Validated Model for Predicting Pacemaker Implantation After Transcatheter Aortic Valve Replacement

Abstract

Introduction: Transcatheter aortic valve replacement (TAVR) has surpassed surgical valve replacement for the management of severe aortic stenosis. High grade AV block is a well-described complication following TAVR. Prior models for predicting risk of pacemaker (PPM) with TAVR lack sensitivity, clinical applicability or cannot be applied pre-procedurally. Hypothesis: To aid planning and patient discussion, we sought to produce a simple tool that can be applied at pre-TAVR clinical visits to stratify risk of PPM associated with TAVR implantation. Methods: We performed an analysis of all patients undergoing TAVR at the University of Colorado. Clinical, valvular, and electrocardiographic data were recorded. Patients were split into a training cohort for development of a predictive model and a testing cohort for model validation. Multivariable stepwise regression was performed on the training cohort to determine patient characteristics most predictive of pacemaker. Results from the initial regression were then applied to the testing cohort for model validation. Results: Of 483 patients analyzed from 2013-2019 in the training cohort, 78 required pacemaker. The need for pacemaker was influenced by 5 pre-procedure variables: RBBB, self-expandable valve, PR interval > 200 ms, ejection fraction > 50%, and albumin > 4.0 g/dL. The regression model produced in our training cohort was highly predictive of need for pacemaker in the validation cohort (Figure). Conclusions: Five pre-procedural characteristics primarily influence risk of PPM after TAVR. These can be used to produce a simple but highly predictive risk model for determining which patients will need pacemaker after TAVR.