A Natural Experiment Of Shared Decision-making For Primary Prevention Implantable Cardioverter-defibrillators In Patients With Heart Failure

Abstract

<h3>Introduction</h3> : In 2018, the Centers for Medicare and Medicaid Services (CMS) mandated shared decision-making (SDM) using a decision aid (DA) prior to implantation of primary prevention implantable cardioverter-defibrillators (ICDs). Little data exists on how this mandate impacts patients' knowledge regarding the ICD or their overall experience of the decision-making process. <h3>Methods</h3> : This was an observational study using a natural experiment. Patients undergoing primary prevention ICD implantation at Emory Healthcare from 2017-2019 (pre and post CMS mandate periods) were surveyed. Survey domains assessed patients' knowledge about ICDs, decisional conflict, values-choice concordance, and engagement in decision-making. Patients exposed to SDM with a DA (post-CMS mandate) were asked a series of questions about the DA. Chi-squared or Fisher exact tests were used to compare survey responses between the pre-mandate and post-mandate groups. <h3>Results</h3> : Of the 101 patients who completed the survey (37% response rate), 45 had an ICD implanted prior to the CMS SDM mandate and 56 after the mandate. No significant differences were noted in composite scores for knowledge about ICDs, decisional conflict, values-choice concordance, or engagement in the decision-making process (Table). Compared to those not exposed to SDM, patients who underwent SDM with a DA reported a better subjective understanding of the benefits associated with ICD implantation (88.2% vs 94.6% p=0.045) but were less able to identify the frequency of minor complications related to an ICD implant (66.7% vs 37.5% p=0.011). <h3>Conclusions</h3> : We conducted an observational study capitalizing on the natural history experiment created by the CMS SDM mandate to explore the impact of SDM using a DA for patients undergoing ICD implantation. Overall, patients in both groups reported high levels of comfort and engagement in the decision-making process, and knowledge about ICDs was low. SDM with a DA did not significantly improve the decision-making process in any of the domains evaluated despite patients' generally positive attitude toward the decision aid. Currently, DA appears to have little impact on patients' experience and their understanding of ICDs. Whether this is evidence that DAs make little difference or that they are not yet implemented effectively is still unclear This raises important questions about whether the goals of the shared decision-making mandate are met and whether there are ways to make it more meaningful.