DISCUSSION
This is one of the first studies to report on SDM experiences around AF ablation, perceptions of decision quality, and opportunities to improve quality from the perspective of patients who have undergone ablation and clinicians routinely involved in decision-making and care surrounding AF ablation. Participants reported generally high perceptions of decision quality and low decision regret. However, quantitative and qualitative data also illustrated gaps in information, patient agency, and alignment of goals in the decision-making process. Although clinicians verbally provided information, occasionally complemented by written information and articles, and encouraged patient engagement during consultations, patients and clinicians alike acknowledged the need for information resources and decision aids in more accessible, digestible formats. Moreover, the study highlighted gaps between patient goals and realistic expectations following an ablation; specifically, many patients hoped to be able to discontinue all medications post-ablation, but clinicians supported that in most cases anticoagulants and other medications may need to continue. Together, these findings highlight the need for a decision aid surrounding AF ablation. Towards this end, we also reported on the preferred content, format, and delivery of a decision aid focused on AF ablation suggested by participants.
Decision aids are intended to help translate information about the likelihood of different health outcomes for different patient populations, as well as the goals and subjective experiences of patients, in ways that are accessible in routine care; the goal is to provide informed and patient-centered decisions.23Importantly, decision aids support (but do not replace) the patient-provider relationship. Decision aids objectively improve patient knowledge and agency in decision-making, as well as patient feelings of being informed, particularly when they include detailed information.31IPDAS is a widely recognized set of standards for developing decision aids that emphasize the importance of understanding patient and other stakeholder needs and preferences. However, a recent systematic review found that only half of the studies describing decision aid development included end users in efforts to understand the problem, and only one-third included end users in prototyping and redesign.32
To ensure decision aids are usable and useful for end users, IPDAS recommends defining the scope, purpose, and target audience of the decision aid, establishing a steering group of patients and experts, and conducting several rounds of user-centered design activities including prototyping, testing, and redesign with end users.33An ongoing challenge of user-centered design is balancing end user suggestions with evidence that user-suggested features may actually be counterproductive. For example, participant suggestions to include testimonials in our study run counter to some evidence that testimonials introduce significant bias towards one treatment option over another because decision-making is strongly influenced by self-identification and other emotional experiences.22Typically, decision aids aim to provide information without persuading, and if included, testimonials need to be carefully selected and balanced to avoid portraying one option more or less favorably than another.
In the context of AF, relatively little attention has been paid to supporting shared decision-making around symptom and rhythm management compared to selecting an anticoagulant. Studies of decision aids for anticoagulant selection show they improved patient satisfaction in treatment decisions, clinician communication and satisfaction, and reduced decision conflict.34,35Prior work focused on anticoagulation may provide useful starting points for shared decision-making in other areas of AF care. Common design elements in anticoagulation decision aids include: (1) consideration of the patient’s individual patient risk factors, such as age, sex, and past medical history, (2) predicted risks of events of interest, such as bleeding or stroke, (3) visualizations, especially 100 person icon arrays to portray probabilities of outcomes of interest, (4) and surveys assessing patient values and priorities when choosing a treatment option (i.e. costs, side effects).34–37Most decision aids are web-based, contain interactive visualizations and videos, and include the option to print a decision summary or send it to a clinician or other recipient. Many of these elements were echoed in the suggestions of participants in our study, supporting the inclusion of these elements in future decision aids.
Strengths of this study include alignment with IPDAS recommendations to include end users in understanding the problem space as a first step in decision aid design, as well as the use of standardized scales to objectively measure multiple aspects of decision quality and regret. A limitation of this study is limited generalizability, as this study was conducted with a small convenience sample of patients and clinicians recruited from a single hospital. Of note, the hospital serves a diverse patient population with respect to age, race/ethnicity, and language, which is reflected in our patient sample. Nonetheless, future work should seek to expand the investigation of decision quality surrounding AF ablation in larger, more representative samples of patients, clinicians, and other key stakeholders. These interviews were conducted up to 18 months after the decision, which might have led patients to underestimate the decisional conflict they experienced while making the decision. In addition, we did not interview patients who considered ablation but decided against it. These two limitations together may mean that levels of decisional conflict are substantially higher among patients who are actually encountering this decision.