Abstract 4363436: Changes in decisional conflict among adults with atrial fibrillation viewing a rhythm management decision aid: Results from a single-arm feasibility trial

Introduction: Shared decision-making is recommended for atrial fibrillation (AF) rhythm management. Decision aids improve patient satisfaction, care quality, and health outcomes. There is a lack of decision aids for AF rhythm management, which patients report creates information gaps and lowers decisional quality. Objective: We aimed to evaluate the impact of a novel digital decision aid for AF rhythm management on decision quality (i.e., decisional conflict). Methods: We conducted a single-arm feasibility trial across one urban health system (NCT04993807; R00NR019124, PI: Turchioe). The intervention consists of a web-based decision aid developed by our team describing two common treatment options (e.g., catheter ablation and antiarrhythmic agents) and patient outcomes derived from clinical trials and meta-analyses (af-decisionaid.org). Eligibility criteria included English-speaking adults with scheduled ablation consultations without permanent AF or a prior ablation. Participants were recruited before their consultation via phone or patient portal and sent a link to the decision aid, which they could return to or discuss with their care team if desired. The primary endpoint was the change in Decisional Conflict Scale (DCS) scores from baseline to two weeks post-intervention. Additional survey and qualitative data were collected at two weeks and three months. We analyzed data using descriptive statistics and paired t-tests using R version 4.1.2. Results: A total of 75 adults with AF were recruited with a mean age of 66 (standard deviation [SD] 12) years, 33% were female and 67% were male, 83% were White, 9% were Black or African American, 3% Asian, and 5% preferred not to report their race. Half (47%) had inadequate health literacy. Fifty-seven (76%) participants completed two-week follow-up surveys. Among them, there were no significant differences in overall DCS scores from baseline (mean 24.7 [SD 17.6]) to two weeks (mean 30.2 [SD 25.4]; p=0.155). When stratified by baseline DCS scores, scores significantly decreased for those with high baseline scores, but significantly increased for those with low baseline scores ( Table 1 ). Conclusion: The digital decision aid had differential effects depending on participants’ baseline levels of decisional conflict. Additional trial data may reveal that patients with low initial conflict had limited awareness of treatment options, and that the decision aid prompted deeper reflection, increasing their perceived conflict.