Shared Decision-Making and Patient Decision Aids for Percutaneous Left Atrial Appendage Occlusion

Abstract

Importance: Experts recommend a shared decision-making (SDM) process before percutaneous left atrial appendage occlusion (pLAAO) in patients with atrial fibrillation, and the Centers for Medicare & Medicaid Services (CMS) require SDM with the use of a patient decision aid (DA) as a condition for reimbursement. However, little is known about how these guidelines and policies have influenced practice. Objective: To describe overall trends in reported SDM and use of DAs for pLAAO and to identify key patient, operator, and institutional factors associated with their use. Design, setting, and participants: This cohort study analyzed data from October 1, 2022, to June 30, 2024, from the American College of Cardiology's National Cardiovascular Data Registry (NCDR) LAAO Registry. Participants included patients who underwent first-time pLAAO. Exposure: pLAAO implantation. Main outcomes and measures: The proportion of encounters reporting SDM and DA use (SDM plus DA) overall and each month. Hierarchical logistic regression was used to estimate the odds of reported SDM plus DA, the probability of SDM plus DA, and the variance in SDM plus DA associated with operator and institutional levels. Results: A total of 147 296 unique patient encounters (86 593 [58.8%] male; mean [SD] age, 76.6 [7.7] years) were included. Of 830 institutions participating in the NCDR LAAO Registry during the study period, 829 (99.9%) reported on SDM and 817 (98.4%) reported on DA use. In the unadjusted analysis, 95 305 encounters (64.7%) reported SDM plus DA had occurred. Unadjusted rates of SDM plus DA rose steadily during the study period from 62.5% in October 2022 to 75.0% in June 2024. The adjusted analysis suggests that the observed variance in SDM plus DA reporting was large and attributable primarily to the institutional level (median odds ratio, 115.64; 95% CI, 79.71-151.56). The range of estimated probability of SDM plus DA by institution was 0.1% to 76.4%, with a mean (SD) of 52.0% (28.6%). There was no statistically significant difference in odds of SDM plus DA for patients with Medicare vs those without (odds ratio, 1.03; 95% CI, 0.98-1.09). Conclusions and relevance: In this cohort study of patients who underwent pLAAO, SDM plus DA reporting was high, but there was large variation between institutions. Patients with Medicare did not have greater odds of reported SDM plus DA, despite the CMS requirement. These findings exhibit the need for further exploration of institutional barriers and facilitators to SDM and DA use for pLAAO.