Implementation of a Patient Decision Aid for Atrial Fibrillation Ablation Improves Patient Procedural Knowledge but Does Not Impact Perceived Involvement With the Shared Decision-Making Process

Abstract

Purpose

Shared decision-making (SDM) is a method for a patient and physician to cooperatively consider a diagnostic or therapeutic option, ultimately empowering the patient to make an informed decision. Atrial fibrillation (AF) ablation is a procedure that would benefit from SDM given the risk of serious adverse events, the high rate of arrhythmia recurrence, and alternative treatment options. Implementing a patient decision aid (PDA) may help facilitate AF ablation SDM by succinctly conveying important information to patients.

Methods

Patients scheduled for initial AF catheter ablation were randomized to a virtual SDM visit utilizing a PDA, which covered procedural risks and benefits, or a virtual control visit with a tool outlining periprocedural processes. Preoperatively, patients completed a questionnaire assessing procedural risk and benefit knowledge, as well as perceived involvement with the decision-making process. Unpaired t-tests were used to compare groups.

Results

The SDM group scored significantly better overall on knowledge-based questions compared to the control group (69% correct [n=34] vs 53% [n=32]; P=0.00013). In particular, the SDM group was significantly more likely to answer questions correctly about stroke risk (P=0.01), anticoagulation (P=0.01), and potential need for additional procedures (P=0.03 and P=0.03). Perceived involvement in the decision-making process was overall not improved with PDA use (4.7 vs 4.6 out of 5; P=0.72).

Conclusions

The addition of a PDA for AF ablation significantly improved procedural knowledge but did not impact patients’ perceived involvement in the decision-making process compared to traditional preprocedural discussion alone.

Shared decision-making (SDM) describes an approach by which patients and providers jointly consider diagnostic or therapeutic options, reviewing evidence, risks, benefits, and patient preferences to ultimately formulate a plan. Recognizing historic shortcomings in provider-patient communications to advance patient autonomy, SDM has been increasingly utilized by cardiologists as a strategy to aid decision-making. In fact, the Centers for Medicare and Medicaid Services (CMS) mandated a formal process of SDM using decision tools prior to insertion of primary prevention implantable cardioverter defibrillator (ICD) and left atrial appendage closure (LAAC) devices as a condition of reimbursement.

Patient decision aids (PDAs) are employed to facilitate SDM. PDAs provide evidence-based educational information relevant to a diagnosis, including treatment options, risks, and benefits. Importantly, PDAs incorporate design elements meant to improve the quality of patient-provider communication while also assisting patients in clarifying their values and goals.¹ Optimally, PDAs should be appropriate for patients of varying ages, cultures, and educational backgrounds. A Cochrane meta-analysis of 105 PDA studies found that their use increased participant knowledge, satisfaction, and patient/provider communication and reduced patient decisional conflict, while requiring minimal additional clinician time.² However, despite the consensus supporting high-quality SDM, there is still concern that a formalized SDM process with PDA deployment may increase provider work without adding meaningfully to clinical outcomes.^1,3

In addition to this growing body of literature, the International Patient Decision Aid Standards (IPDAS) Collaboration has established criteria focusing on the development, content, and implementation of PDAs.⁴ The use of the IPDAS checklists tends to lead to the development of a thorough yet lengthy PDA, which can be effective in a peri-office visit setting but limits their utility during a clinic appointment or tele-visit. Ideally, a decision aid used during an office visit would supplement the traditional pre-procedural counseling, providing only the most important information in a succinct, tangible form.

Atrial fibrillation (AF) ablation is a treatment strategy characterized by multiple alternate therapeutic options, guidelines providing significant latitude, and relatively high arrhythmia recurrence rates, as well as the risk of serious adverse events. Moreover, the Catheter Ablation vs Antiarrhythmic Drug Therapy for AF (CABANA) trial failed to demonstrate an improvement in the primary endpoint (composite of death, disabling stroke, serious bleeding or cardiac arrest) with AF ablation.⁵ The highly complex decision-making with multiple therapeutic choices involved when considering AF ablation requires substantial patient education and focus on individual patient goals of care and values. Therefore, we carried out a prospective, randomized trial testing whether a structured SDM process, which involved an SDM-specific phone conversation with an electrophysiologist utilizing a PDA that reviewed the risks, benefits, and limitations of initial AF catheter ablation, in addition to standard office visit preprocedural counseling, improved patient knowledge and perceived involvement with the SDM process compared to standard pre-procedural counseling alone.

METHODS

Patients over 18 years old with paroxysmal or persistent atrial fibrillation, no prior history of catheter or surgical ablation for AF, and sufficient cognitive capacity to weigh a complex medical procedure’s risks and benefits were considered eligible for this study. Cognitive capacity was considered the ability for a patient to make their own medical decisions and the ability to read. All patients that met these inclusion criteria were identified via the electronic medical record 2–3 weeks prior to their index AF catheter ablation for the electrophysiologists participating in the study. Each patient had previously completed an outpatient visit where AF ablation was recommended by the electrophysiologist; the risks, benefits, and alternatives were explained; any questions or concerns were addressed; and the patient ultimately agreed to proceed with the procedure. Outpatient visits occurred at multiple outpatient clinic sites, and all AF ablations were performed at Froedtert Hospital in Milwaukee, Wisconsin. A random number generator was used to assign patients to the control or experimental group. No attempt was made to determine a patient’s baseline understanding of an AF ablation before randomization. The physician was notified 2–3 weeks prior to each patient’s procedure regarding their candidacy and test group.

The physician initiated a virtual visit during which the experimental group utilized a PDA to complete formal SDM, and the control group used a control document to review periprocedural processes. The patient decision aid tool contained risk and benefit information regarding the AF ablation procedure (Online Appendix A). This tool was designed to emulate existing PDAs for ICD and LAAC implants utilized at our hospital. Compared with traditional PDAs classified by IPDAS, these original tools were designed to complement the classic preprocedural counseling rather than function as an independent resource for patients to use on their own. As such, the physician was able to tailor the discussion surrounding the decision aid, based on heart failure status or anticoagulation use, to more directly apply to each participant. The control tool covered information about the patient experience before and during the ablation, without additional risk and benefit information (Online Appendix B). A control group virtual visit was implemented to avoid the potential confounding impact of an additional encounter, independent of the PDA. Subjects in both groups were encouraged to ask procedural questions. Although subjects had procedural dates, conversations framed the planned ablations as a patient choice. In one case, an experimental patient, after completing the virtual visit, elected to cancel the planned ablation. Upon completion of the virtual visit, the PDA or control document and an informational letter about the research study were mailed to the patient’s address. The decision to randomize patients with scheduled AF ablations was based on institutional practice implementing PDA in ICD and Watchman procedures.

On the day of the procedure, in the preoperative area, patients from both groups completed the same questionnaire with 10 True-False questions about the risks and benefits of AF ablation, as well as 5 Numeric Rating Scale questions assessing their perceived involvement in the decision-making process (Online Appendix C). Basic demographic information was also collected on the questionnaire to further characterize the data. The questionnaire was administered by the preoperative nurse practitioner, electrophysiology fellow, or the electrophysiologist.

Clinicians who participated in the study included 3 established electrophysiologists and 1 electrophysiology fellow. All 4 participated in the design and development of the study, and we thus believed no additional training was necessary to train the providers with the PDA. While the total number of patients enrolled varied between clinicians, two clinicians predominated enrollment.

All study procedures were reviewed and approved by the Medical College of Wisconsin’s institutional research board. Data were collected for 18 months, from August 2020 to February 2022. No compensation was offered to patients for their participation in the study. Protected health information was obtained from the electronic medical record to help track patient progress throughout the project.

Statistical analyses were performed using Excel (Microsoft 365 2020) or SPSS Statistics for Windows (Version 27.0, IBM Corp.). Subject demographic and clinical characteristics between groups were compared using unpaired t-tests. Knowledge and perceived involvement-based questions were analyzed as individual questions and as overall knowledge or perceived involvement scores using unpaired t-tests. Two different linear regression models were used to assess if any demographic or clinical characteristics impacted patient performance. One model assessed control group and phone call length; the other model measured control group and education level above high-school level. P<0.05 was considered significant for all comparisons in the study.

RESULTS

A total of 202 patients were assessed for eligibility (Figure 1). Of the 130 eligible patients, 66 were ultimately enrolled and completed the questionnaire. There were 7 subjects, 3 control and 4 experimental, who were enrolled but did not complete the questionnaire. One patient who enrolled in the experimental group elected to pursue pharmacologic therapy rather than ablation after reviewing the SDM document. All patients were mailed the tool after the virtual visit, except for 1 SDM patient who was provided the tool over email and 2 patients, 1 SDM and 1 control, who were given the tool at a preoperative echocardiogram because of time constraints associated with mailing.

Figure 1.

Participant flow diagram. PDA, patient decision aid; SDM, shared decision-making.

Baseline characteristics between enrolled individuals and those who were randomized but not enrolled showed no difference in age, rates of paroxysmal and persistent AF, patients trialing an antiarrhythmic, or diagnosis to procedure time (data not shown). They did significantly differ in the percentage of women, with a greater proportion of women in the enrollment group (42% [N=64] vs 23% [N=66]; P=0.03).

Table 1 describes the basic demographic and clinical characteristics of the study population. No major differences were observed between the experimental and control groups. Linear regressions demonstrated that the significance of treatment group affecting quiz results was independent of both phone call duration and education level (P=0.29 and P=0.71, respectively).

Table 1.

Demographic Characteristics of Study Population

Characteristic	Control n=32	Experimental n=34
Age, years	66.2 ± 7.5	63.6 ± 8.2
Median	66	64.5
Women	14 (45%)	13 (38%)
Race/Ethnicity
White/non-Hispanic	27 (84%)	32 (94%)
Black/non-Hispanic	2 (6%)	1 (3%)
Hispanic or Latino	0 (0%)	1 (3%)
Asian/Pacific Islander	1 (3%)	1 (3%)
American Indian	0 (0%)	0 (0%)
Other	0 (0%)	0 (0%)
Prefer not to respond	2 (6%)	0 (0%)
Highest level of education
Some high school	1 (3%)	0 (0%)
High school or GED	10 (31%)	11 (32%)
Some college, no degree	6 (19%)	9 (26%)
Associate degree	6 (19%)	2 (6%)
Bachelor’s degree	6 (19%)	7 (21%)
Post-graduate degree	2 (6%)	3 (9%)
Diagnosis to ablation time, years*	3.9 ± 5.1	4.4 ± 6.4
Trialed antiarrhythmic before ablation	18 (56%)	24 (71%)
Paroxysmal AF	23 (72%)	25 (74%)
Persistent AF	9 (28%)	9 (26%)
Phone call length, minutes	9.0 ± 4.0	10.5 ± 4.7

Data are presented as n (%) or mean ± SD. No statistical differences were observed between groups.

^*Time between original diagnosis of atrial fibrillation (AF) and date of AF ablation, in years.

Table 2 depicts the results of the knowledge-based quiz questions. Overall, the SDM group scored significantly better on the knowledge-based questions compared to the control group (69% [n=34] vs 53% [n=32]; P=0.00013). The SDM group scored better on 8 of the 10 questions, with both groups scoring 100% on the other 2 questions. Patients in the SDM group were significantly more likely to correctly answer the questions about the need for continued oral anticoagulation post-ablation driven by underlying thromboembolic risk (P=0.01), AF ablation not decreasing stroke risk compared to medical therapy (P=0.01), the chance for repeat AF ablation being necessary because of failed ablation (P=0.03), and symptoms potentially improving even in patients experiencing recurrent AF after index ablation (P=0.03). Unpaired t-tests comparing control group subjects from the first and second halves of the study showed no significant difference in test scores overall and on individual questions (data not shown). Figure 2 depicts the results of the perceived involvement with SDM quiz questions. Overall, the use of the patient decision aid to facilitate SDM had no significant impact on patient perceived involvement in the SDM process. No question from either group averaged below a 4.5 out of 5. Therefore, perceived involvement with pre-procedural counseling in both groups was high.

Table 2.

Comparison^* of Knowledge-Based Quiz Questions Between Standard of Care (n=32) and SDM Groups (n=34), Organized by Category

Knowledge questions by category	Control %	Experimental %	P
Benefits of the procedure
Improves symptoms	100	100	1.0
Improves quality of life	100	100	1.0
Effect on comorbidity and mortality
Anticoagulation, cannot discontinue	52	81	0.01
Cardiac Arrest, risk not decreased	22	38	0.15
Stroke, risk not decreased	6	29	0.01
Death, risk not decreased	38	53	0.21
Procedural success rate
Short-term success rate	41	62	0.09
Long-term success rate	59	68	0.49
Repeat ablations may be required	66	88	0.03
Symptoms may improve even if failed ablation	47	74	0.03
Overall	53	69	0.00013

^*Unpaired t-tests were used to compare groups, with a P-value of <0.05 being considered significant.

Figure 2.

Comparison of perceived involvement quiz questions between control and SDM groups. To note, x-axis begins at 4, instead of 0, to better visualize differences between groups. Unpaired t-tests were used to compare groups. No significant differences were observed.

Inter-clinician comparisons for knowledge and perceived involvement quiz questions showed no major differences in average scores (data not shown). Notably, however, one clinician had a statistically significant difference in phone call length between groups (Control = 8.5 minutes, N=15, Experimental = 11.1 minutes, N=16; P=0.02).

DISCUSSION

Formal SDM protocols have recently proliferated in clinical medicine, driven by both a desire to improve patient agency relating to medical decision-making and insurer requirements. The literature has supported the efficacy of SDM protocols in improving patient knowledge and satisfaction, while requiring limited additional physician time.²

Despite widespread acceptance of SDM and PDAs among the general medical community, little is known about SDM outcomes in electrophysiology, and to date, guidelines and research have focused on 2 topics. The first topic is stroke prevention in AF. In 2014, ACC/AHA/HRS issued a class I recommendation (level of evidence C) for SDM to initiate long-term oral anticoagulation in patients with AF for thromboembolic prophylaxis.⁶ Since that time, a 2017 systematic review of 7 pharmacological PDAs⁷ and a 2019 meta-analysis of 14 pharmacological and non-pharmacological (LAAC device) PDAs⁸ each showed improved subject knowledge and lower decisional conflict with the use of PDAs. Similarly, a trial of 922 patients with nonvalvular AF found no differences in treatment decisions but greater observed patient involvement in decision-making with the use of a decision aid.³ The second focus of SDM research in electrophysiology is ICD implantation. Many patients with ICDs lack important information about their devices, including the need for eventual battery replacement, activity limitations, potential complications, and end of life issues.⁹ As such, the 2017 ACC/AHA/HRS Ventricular Arrhythmia/SCD guidelines established SDM for ICD implants as a class I recommendation,¹⁰ and CMS now requires SDM in patients undergoing primary prevention ICD implantation. Despite these recommendations, no outcome data have been published validating their use in this population. The DECIDE-ICD study is currently testing the effectiveness and implementation of SDM interventions for ICDs.¹¹

AF catheter ablation is a common electrophysiology procedure, approaching 75,000 cases in the US per year, with significant risks, costs, and association with variable outcomes.¹² The complex decision-making inherent to considering AF ablation suggests that it would benefit from a structured SDM process. However, US guidelines and society statements have not yet endorsed formal SDM use. Our randomized clinical trial found that adding a PDA to standard care during clinical encounters in patients with AF, prior to initial AF catheter ablation, significantly improved patient knowledge, particularly in their understanding of the lack of AF ablation’s impact on stroke risk, the need for ongoing anticoagulation post-ablation based on pre-procedure stroke risk, and the potential need for additional procedures. Importantly, these benefits were observed after expending a modest amount of clinician time during a virtual visit and were independent of patients’ educational levels. The SDM group did not report an increase in patient’s perceived involvement in the decision-making process compared with the control group. One possible explanation for the lack of difference observed between groups may be related to a “trial effect” elevating scores in the control group simply because they participated in the study, as both groups had extra discussion time with a physician prior to their procedure, which may have elevated scores regardless of specific content discussed.

Our study also demonstrates a practical approach to implementing a PDA to guide AF ablation discussions during clinical encounters in a time-constrained practice setting. The strengths of this study include the simplicity and feasibility of PDA implementation. The PDA developed for this study, while not necessarily qualifying as a PDA under the IPDAS recommendations, is easy to use for providers, takes only a few additional minutes of provider time, and quickly provides useful information to patients that may have not been prioritized during traditional pre-procedural discussion. Additionally, if clinicians feel that the burden of a separate virtual visit is too great, then the PDA’s design easily translates to use in the office visit setting as well. Future directions may focus on the implementation of the document in person.

Although the number of patients randomized was relatively small, the significantly better performance of the SDM/PDA group supports the efficacy of the described SDM method to improve patient knowledge. The PDA tool has the advantage of providing patient procedural counseling and education with consistency and clarity. In fact, one patient in the experimental group elected to pursue pharmacologic therapy rather than ablation after reviewing the SDM document. This study also demonstrates a methodology to assess outcomes of various PDAs for electrophysiology procedures—a process that may facilitate the design of PDAs considering individual educational and cultural differences—thus tailoring the presentation of information to optimize patient understanding, engagement, and health outcomes.

Limitations

The study had several limitations. Our study population included only subjects who had already agreed to the AF ablation procedure prior to enrollment, potentially limiting the impact of the PDAs as a tool to facilitate SDM. As a single-center study, enrollment was limited. However, even with a small number of subjects, use of the PDA produced a significant improvement in subject performance on knowledge-based questions. Finally, the outcome surveys were not validated for low levels of health literacy and numeracy.

CONCLUSIONS

In this randomized, controlled study, the use of a guided SDM process that included a patient decision aid tool prior to atrial fibrillation ablation improved patient knowledge related to the procedure, independent of patient educational level. However, it did not directly impact patients’ perceived involvement with the SDM process. Larger studies assessing the benefit of various SDM methods prior to AF ablation should be pursued.

Patient-Friendly Recap.

• Atrial fibrillation (AF) ablation is a treatment strategy with many different therapeutic options, complex guidelines, high arrhythmia recurrence rates, and risk of serious adverse events.

• Patients require much education about these complexities that focuses on individual goals of care and values.

• This study looked at whether a structured shared decision-making process that considered the risks, benefits, and limitations of initial AF catheter ablation improved patient knowledge and perceived involvement with the shared decision-making process, compared to standard counseling alone.

• The intervention significantly improved patients’ procedural knowledge but did not impact patients’ perceived involvement in the decision-making process, compared to standard counseling alone.