Variability in Adoption of Surgical Ablation of Atrial Fibrillation: An International Survey of Cardiac Surgeons

Abstract

Background

Surgical ablation of atrial fibrillation (AF) during cardiac surgery is a guideline-recommended procedure with demonstrated effectiveness in restoring sinus rhythm. Nonetheless, rates of utilization are relatively low, a situation that needs to be better understood.

Methods

This study used an electronically distributed survey to characterize the use, techniques, and clinical indications for concomitant AF ablation among cardiac surgeons.

Results

Of the 362 surveys distributed internationally, 79 responses were received (22%). After excluding 3 incomplete responses, 76 surveys were included in the final analysis. A majority of respondents (67 of 76; 88%) reported performing concomitant surgical ablation of AF. Among surgeons who perform AF ablation, practice variation was observed. Pulmonary vein isolation was the most common lesion set (45%), and over half of surgeons (52%) tailored lesion selection to AF type, favouring more extensive ablation for persistent AF. Regarding patient selection, notable deterrents to concomitant ablation included the following: increased left atrial size, more persistent forms of AF, severe left ventricular dysfunction, reoperation, minimally invasive surgeries, and complex cases with long pump times. Among non-ablating surgeons (9 of 76; 12%), common factors reported to help increase future adoption of AF ablation included stronger clinical evidence and more affordable and accessible equipment.

Conclusions

Our findings reveal that surgical ablation of AF is performed by the majority of research-active cardiac surgeons, yet meaningful variation persists in how and when it is applied. These findings underscore the need for a well designed, definitive trial to guide future care and address the gap between recommendations and practice.

Atrial fibrillation (AF) is the most common cardiac arrhythmia and is associated with morbidity and mortality.¹ Among patients undergoing cardiac surgery, approximately 10%-28% have preexisting AF.^2,3 Cardiac surgery therefore presents a unique opportunity to address AF through concomitant surgical ablation, which aims to maintain sinus rhythm by creating targeted lesions that disrupt disordered electrical conduction.

Concomitant surgical AF ablation has been associated with increased freedom from AF during the first postoperative year, as demonstrated in a meta-analysis of randomized trials.⁴ However, no improvement was observed in patient-important outcomes, such as mortality, thromboembolic events, and neurovascular complications. Due to imprecision and heterogeneity, the overall quality of evidence was rated as moderate to low using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) framework, yielding less confidence in interpreting results.

Despite the limitations of the evidence for patient-important outcomes, the procedure is recommended in contemporary clinical guidelines. The 2023 Society of Thoracic Surgeons (STS) guidelines strongly support concomitant surgical AF ablation, particularly in first-time nonemergent operations, assigning a class I recommendation for both mitral (level of evidence: A) and nonmitral cardiac surgeries (level of evidence: B-nonrandomized), underscoring the procedure’s importance for sinus rhythm restoration and sustained benefit.⁵ Similarly, the 2024 European Society of Cardiology guidelines, in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS), provide a class I recommendation for concomitant surgical AF ablation during mitral valve surgery (level of evidence: A) and recommends consideration during nonmitral procedures (class IIa; level of evidence: B). The 2024 American College of Cardiology (ACC) and American Heart Association (AHA) guidelines also endorse surgical AF ablation, issuing a class IIa recommendation (level of evidence: B-randomized) for its routine use in patients undergoing cardiac surgery to reduce the risk of recurrent AF.⁶

Despite these guideline recommendations across professional societies, the uptake of surgical AF ablation remains limited. Only about 18%-22% of eligible patients in the US undergo concomitant surgical ablation during cardiac surgery.^2,7

This gap between guideline recommendations and clinical implementation raises critical questions regarding the barriers to broader adoption of concomitant AF ablation during cardiac surgery for another indication. In this study, we surveyed cardiac surgeons globally to explore factors influencing the decision to perform surgical ablation of AF in patients undergoing cardiac surgery. We aimed to understand variability in practice patterns so that strategies can be established to bridge the gap between evidence-based guidelines and routine clinical practice in surgical AF ablation.

Methods

The primary objective of this study was to assess the proportion of cardiac surgeons who perform surgical ablation for AF. Among surgeons who perform ablation, we aimed to characterize practice patterns related to ablation techniques. Among those who do not perform ablation, we sought to identify reasons for non-performance and perceived barriers to adoption.

Item generation

Survey items were developed to explore the demographic characteristics of respondents, the factors that influenced the decision to perform AF ablation among those who performed the procedure, and the barriers preventing AF ablation among those who did not. The survey was designed in collaboration with experts in cardiology, cardiac surgery, and survey methods.

Scaling and scoring responses

For surgeons who performed AF ablation, factors influencing their decision were assessed using 5-point scales. A unipolar scale was used to measure the degree of influence each factor had on their decision to perform AF ablation, with response options including no influence, limited influence, moderate influence, important influence, and very important influence. A bipolar Likert scale was used to assess the likelihood of performing AF ablation based on specific factors, with response options including very unlikely to perform, unlikely to perform, neutral, likely to perform, and very likely to perform.

For surgeons who did not perform AF ablation, respondents were asked to select all applicable factors preventing their adoption of the procedure and identify factors that could increase their likelihood of performing AF ablation.

Final instrument

The final survey instrument included one question confirming the respondent’s status as a cardiac surgeon, 9 demographic questions, and additional items related to AF ablation practices. Surgeons who performed AF ablation completed 40 practice-specific questions addressing procedure volumes, ablation methods, lesion-set selection and its dependence on AF type, and factors influencing decision-making. Surgeons who did not perform AF ablation completed 2 questions with multiple-choice response options. The full survey instrument is provided in Supplemental Appendix S1.

Pilot testing

Pilot testing was conducted to evaluate the clarity and feasibility of the survey. The instrument was tested with 1 cardiac surgeon and 2 cardiac surgery residents to identify ambiguities or potential improvements before its wider distribution.

Sampling frame

A convenience sample was drawn from an existing network of cardiac surgeons who had participated in the Left Atrial Appendage Exclusion for Prophylactic Stroke Reduction (LeAAPS) and Left Atrial Appendage Occlusion Study III (LAAOS III) trials as investigators or sub-investigators.^8,9 This representative sample included surgeons from 26 countries where cardiac surgery is performed. These surgeons had previously recruited patients into the trials and provided their e-mail addresses, enabling direct contact. These surgeons were considered likely to participate given their prior engagement in left atrial appendage-related research (but not ablation trials, which would have further biased the sample). In addition, the established communication network facilitated engagement, increasing the likelihood of survey completion.

Survey administration

After research ethics board approval was granted (by the Hamilton Integrated Research Ethics Board, #18258), the survey was administered electronically using REDCap (research electronic data capture).¹⁰ Each potential respondent received an e-mail invitation containing a unique survey link. The survey was available in English, and up to 2 reminder e-mails were sent at 2-week intervals to nonrespondents. Although REDCap required use of e-mail addresses for survey distribution, all responses were de-identified and were not linked to any personal information accessible by the research team.

Statistical analyses

The unit of analysis was the individual respondent. Descriptive statistics were used to summarize the demographic characteristics of respondents and their AF ablation practices. Continuous variables were summarized using means and standard deviations (SDs), and categorical variables were presented as percentages. All statistical analyses were conducted using Python (3.12.8; Beaverton, OR).

Results

From the 362 surveys distributed, 79 responses were received (22% response rate). After excluding 3 incomplete responses, 76 surveys were included in the final analysis.

The geographic distribution of respondents is displayed in Figure 1—41% were located in Canada, 6% in the US, 37% in Western Europe, 8% in Eastern Europe, 5% in Oceania, and 3% in South America. On average, participants reported having completed 22 years (SD = 10) in clinical practice and an annual surgical volume of 190 cases (SD = 73).

Figure 1.

Survey respondents by country. Choropleth map illustrating the number of survey respondents from each country (N = 76). A yellow–green–blue colour scale represents increasing response counts, with blue indicating the highest number of respondents. Countries with no respondents are shown in grey.

Practice focus was diverse, as shown in Table 1. A large proportion of respondents reported a primary focus on coronary artery bypass grafting (CABG; 48of 76; 63%) and aortic valve procedures (47 of 76; 62%). These categories were not mutually exclusive, as some respondents reported multiple areas of focus. The majority of respondents (73 of 76; 96%) reported routinely occluding the left atrial appendage in patients with a history of AF undergoing cardiac surgery for another indication. Among these, the most common techniques were clip occlusion (38 of 73; 52%) and cut-and-sew (27 of 73; 37%).

Table 1.

Characteristics of survey respondents (N = 76)

Characteristics	Value
Years in practice, mean (SD)	22 (10.2)
Surgeries performed annually, mean (SD)	190 (73.4)
Practice focus∗
CABG	48 (63)
Aortic valve	47 (62)
Mitral valve	39 (51)
Aorta	30 (40)
Congenital	8 (11)
Heart failure and/or transplant	17 (22)
Other	14 (18)

Values are n (%), unless otherwise indicated.

CABG, coronary artery bypass grafting; SD, standard deviation.

^∗Categories are not mutually exclusive.

Most respondents (67 of 76; 88%) reported currently performing surgical ablation of AF. Among patients with AF undergoing cardiac surgery for another indication, these surgeons estimated performing surgical ablation in an average of 61% (SD = 30) of cases. The methods of ablation used are shown in Figure 2, with 48% reporting the use of both radiofrequency and cryoablation. Routine lesion sets included pulmonary vein isolation (30 of 67; 45%), posterior box (25 of 67; 37%), and a combination of both (12 of 67; 18%). In addition, 42 of 67 (63%) commonly included a mitral line, and 44 of 67 (66%) added a line between the left atrial appendage and the left superior pulmonary vein. Many respondents (81%) indicated that they would be more likely to perform posterior wall isolation if it could be achieved without opening the heart, such as through the use of novel technologies.

Figure 2.

Methods of surgical ablation of atrial fibrillation. The pie chart shows the percentage of respondents reporting each method as their primary approach to surgical ablation of atrial fibrillation. Percentages are based on the surgeons (n = 67) who reported currently performing the procedure.

More than half of ablation-performing surgeons (35 of 67; 52%) indicated that their choice of lesion set was influenced by the pattern of AF. As shown in Figure 3, respondents reported using a more extensive lesion set—biatrial ablation—more frequently in cases of persistent or longstanding persistent AF. Overall, respondents estimated performing biatrial ablation in 47% (SD = 33) of cases.

Figure 3.

Lesion-set selection by atrial fibrillation (AF) type. Double bar graph illustrating lesion-set selection among 35 surgeons who reported that their chosen approach varies by AF type. Red bars represent biatrial ablation; blue bars represent left atrial ablation.

Figure 4 presents surgeons’ likelihood of performing AF ablation based on various clinical factors. Notable deterrents, as defined by the proportion of surgeons rating the factor as "unlikely" or "very unlikely" to proceed with ablation, included the following: left atrial size > 6.5 cm (40 of 67; 60%), longstanding persistent AF (14 of 67; 21%), severe left ventricular (LV) dysfunction (22 of 67; 33%), reoperation (40 of 67; 60%), minimally invasive surgery (27 of 67; 40%), and complex cases with anticipated long pump times (29 of 67; 43%).

Figure 4.

Clinical factors influencing the likelihood of performing atrial fibrillation (AF) ablation. Horizontal bar chart displaying Likert scale responses from (n = 67) surgeons who currently perform surgical ablation of AF. Responses indicate the perceived likelihood of proceeding with ablation in the presence of various clinical factors, ranging from “very unlikely” to “very likely.” CHF, congestive heart failure; LA, left atrium; LV, left ventricle.

The influence of both clinical and contextual factors on ablation decision-making is shown in Figure 5. Among the strongest motivating factors were adherence to clinical guidelines (limited/moderate: 18 of 67 [27%]; important/very important: 46 of 67 [69%], belief in the procedure’s therapeutic benefit (limited/moderate: 6 of 67 [9%]; important/very important: 60 of 67 [90%]), and patient preference (limited/moderate: 30 of 67 [45%]; important/very important: 26 of 67 [39%]). In contrast, financial incentives were less frequently reported to influence the decision to perform surgical ablation.

Figure 5.

Influence of clinical and contextual factors on the decision to perform atrial fibrillation ablation. Horizontal bar graph displaying Likert scale responses from (n = 67) surgeons who currently perform surgical ablation of atrial fibrillation. Responses reflect the perceived influence of various clinical and contextual factors on their decision-making, ranging from “no influence” to “very important influence.”

A minority of respondents (9 of 76; 12%) indicated that they do not currently perform surgical ablation for AF. Supplemental Table S1 compares the demographic characteristics of these surgeons with those of surgeons who perform ablation. Among non-ablating surgeons, the predominant clinical focus was CABG (8 of 9; 89%). Respondents had a mean of 22 years in practice (SD = 7) and performed an average of 142 procedures annually (SD = 70). Practice locations included Canada (5 of 9; 56%) and Europe (4 of 9; 44%). Among this group, commonly reported factors that would encourage future adoption included stronger clinical evidence (5 of 9; 56%), more affordable or accessible equipment (5 of 9; 56%), additional training (2 of 9; 22%), mentorship (1 of 9; 11%), and an ablation procedure that could be performed faster (1 of 9; 11%; Fig. 6).

Figure 6.

Factors that would increase future adoption of atrial fibrillation ablation. Bubble chart illustrating responses from surgeons not currently performing atrial fibrillation ablation (n = 9). Bubble size reflects the number of respondents selecting each factor.

Discussion

Key findings

This survey was more likely to be completed by surgeons who perform surgical ablation of AF, and notable variation exists in how and when AF ablation is performed. Decisions were strongly influenced by patient factors—including atrial size, AF duration, and left ventricular dysfunction—as well as procedural factors such as reoperation, minimally invasive surgeries, and complex cases with long pump times. Non-ablating surgeons were not well represented. Among those that did respond, common factors reported as helping to increase future adoption of AF ablation included stronger clinical evidence and more affordable and accessible equipment.

This study in context

Our findings demonstrate general alignment between surgeons’ reported practice patterns and current clinical evidence in 2 key areas. First, the vast majority of respondents reported occluding the left atrial appendage in patients with AF undergoing cardiac surgery. This finding is consistent with high-quality evidence from the LAAOS III trial, which demonstrated a 33% reduction in stroke risk and informed a class I guideline recommendation in the 2024 European Society of Cardiology guidelines.^5,9 Second, surgeons were less likely to perform surgical ablation in patients with markedly enlarged left atria or more persistent patterns of AF, both of which are associated with reduced success rates in maintaining sinus rhythm post-ablation.¹¹

Our results are consistent with the multinational 2019 survey by Belley-Côté et al., in which more persistent forms of AF were similarly associated with a lower likelihood of surgeons performing concomitant surgical AF ablation.¹² Procedure type further influenced decision-making, with the likelihood of ablation increasing during valve surgery and decreasing during CABG. Procedural factors also had an influence in our survey, with minimally invasive surgeries and complex surgeries with anticipated long cardiopulmonary bypass times appearing to deter surgeons from performing surgical ablation of AF. Belley-Côté et al. demonstrated that positive surgeon perception correlated with more frequent ablation (Pearson correlation coefficient 0.39), and greater concern about procedural risks correlated with less use (Pearson correlation coefficient –0.21).¹² These findings show parallels to our findings that surgeons were less likely to perform ablation in patients with comorbidities associated with reduced success rates—suggesting how both perceived procedural risks and anticipated patient outcomes shape surgical decision-making.

Singh et al. conducted a population-based cohort study of 22,032 adults presenting to the emergency department with new-onset AF in Ontario between 2010 and 2012. Among the 8161 patients who were assessed by an electrophysiologist, 424 received catheter ablation.¹³ The authors demonstrated that older age (hazard ratio [HR]: 0.80; P < 0.0001) and comorbidities such as heart failure (HR: 0.60; P = 0.001), hypertension (HR: 0.74; P = 0.01), and diabetes (HR: 0.74; P = 0.06) were associated with a reduced likelihood of catheter ablation. In contrast, in our study, a patient’s history of heart failure did not appear to deter surgeons from pursuing ablation. Practice among the electrophysiologists, however, may have changed with the publication of data supporting a benefit of catheter-based ablation in patients with heart failure.¹⁴

Brancato et al. conducted a retrospective analysis of 3124 patients with preoperative AF undergoing cardiac surgery across 21 hospitals in the US between 2014 and 2020.¹⁵ The authors found that prior CABG and valve surgery lowered the odds of receiving concomitant surgical ablation (CABG: OR 0.22, 95% CI 0.08-0.61; valve: OR 0.15, 95% CI 0.08-0.27), which aligns with our findings that surgeons reported being less likely to perform AF ablation during reoperations. Surgeon experience also had an influence: every 10-year increase since graduation reduced the odds by 29% (95% CI: 10%-44%). Our cohort reported an average of 22 years in practice, yet the majority still performed ablation.

Surgeons in our study reported a reduced likelihood of performing ablation in patients with left ventricular dysfunction, despite evidence suggesting that left ventricular dysfunction does not significantly compromise surgical ablation outcomes.¹¹ Stulak et al. conducted a retrospective analysis of 443 patients who underwent the Cox-maze procedure between 1993 and 2002, including 37 with reduced left ventricular function, and reported marked postoperative improvements in systolic function following concomitant surgical ablation.¹⁶ Across their entire cohort, mean ejection fraction improved significantly, from 44% to 54%, early postoperatively (P < 0.05), and this improvement was sustained at late follow-up (54%; P < 0.05). The greatest improvement in systolic function was observed in patients with the most severe preoperative left ventricular impairment (31% to 53%, P < 0.05). Despite these data, hesitation among surgeons to ablate in the context of left ventricular dysfunction may therefore reflect ongoing uncertainty about safety and efficacy in high-risk patients—a concern similarly highlighted by non-ablating surgeons in our study as a barrier to adoption. These findings underscore the need for more robust and targeted clinical evidence to inform surgical decision-making and promote appropriate utilization of ablation in patients who may benefit from it.

No large multicentre trials of surgical AF ablation have been published. Indirect evidence from randomized controlled trials of catheter ablation provides important but incomplete support for the efficacy of surgical AF ablation. The Catheter Ablation vs Antiarrhythmic Drug Therapy for Atrial Fibrillation (CABANA) trial found no reduction in a composite of death, disabling stroke, serious bleeding, or cardiac arrest with ablation vs medical therapy.¹⁷ The Catheter Ablation Versus Standard Conventional Treatment in Patients With Left Ventricular Dysfunction and Atrial Fibrillation (CASTLE-AF) study demonstrated benefit for patients with heart failure and reduced ejection fraction but was limited by small sample size and high loss to follow-up.¹⁸ The more recent Randomized Ablation-Based Rhythm-Control Versus Rate-Control Trial in Patients With Heart Failure and Atrial Fibrillation (RAFT-AF) study showed a nonsignificant trend toward improved outcomes, but recruitment was halted for futility.¹⁹ The Catheter Ablation for Atrial Fibrillation in Patients With End-Stage Heart Failure and Eligibility for Heart Transplantation (CASTLE-HTx) study showed that in end-stage heart failure, catheter ablation reduced the risk of all-cause death, left ventricular assist device implantation, or urgent heart transplantation, compared with medical therapy alone.¹⁴ Collectively, these studies suggest potential benefits of ablation in select populations but remain limited by methodological and generalizability concerns.

Concomitant surgical ablation offers a unique opportunity to address these evidence gaps. Unlike catheter-based trials, in which patients must accept randomization to an invasive procedure vs no procedure, patients undergoing cardiac surgery already face sternotomy and cardiopulmonary bypass, making enrollment more feasible and ethically acceptable. In addition, the surgical setting allows for a more robust control condition than catheter-based trials, which are typically not sham-controlled, as patients and care providers outside the operating room can be blinded to treatment allocation.

A definitive trial of concomitant surgical ablation is therefore well positioned to not only establish the role of ablation in surgical patients but also strengthen the broader evidence base for AF ablation strategies. Such evidence could inform decision-making between catheter-based and surgical strategies in patients eligible for either approach, particularly in settings of clinical equipoise—such as percutaneous coronary intervention vs CABG in multivessel coronary artery disease or transcatheter vs surgical aortic valve replacement in patients with AF—by clarifying the incremental benefit of surgical approaches that permit concomitant AF therapy. The ongoing Surgical Ablation of AF Efficacy (SAFE) Trial is an international, multicentre randomized controlled trial evaluating concomitant surgical AF ablation in patients with paroxysmal or persistent AF undergoing cardiac surgery, and it is designed to address this critical evidence gap (NCT05434819).

Strengths and limitations

This study has several strengths. It included international representation, enhancing the relevance of findings across diverse practice settings. The survey instrument was also developed through a rigorous, iterative process, ensuring clarity and content validity.

Several important limitations warrant careful consideration when interpreting these findings. First, the sampling frame was restricted to surgeons who participated as investigators or sub-investigators in the LeAAPS and LAAOS III trials. Respondents therefore represent a selected group of cardiac surgeons with established interest and experience in AF research and surgical ablation. This group introduces selection bias and limits representativeness relative to the broader cardiac surgery community, particularly surgeons practicing in community or nonacademic settings. As a result, reported rates of AF ablation and familiarity with contemporary techniques likely overestimate real-world adoption and limit insight into barriers among non-ablating surgeons.

Second, the response rate was modest (22%), increasing susceptibility to nonresponse bias, as surgeons with greater interest in or experience with AF ablation may have been more likely to participate. This modest rate reduces the generalizability of the results. In addition, the sample was geographically skewed, with respondents concentrated primarily in Canada and Western Europe and minimal representation from Asia, Africa, and Latin America. Given the small number of participants from several regions, meaningful regional comparisons—particularly regarding access to equipment, institutional resources, and training—are not feasible. English-only survey administration may have further limited participation from non-English-speaking regions, reinforcing these geographic imbalances and constraining generalizability. Finally, the study relied on self-reported practice patterns without external validation.

Conclusion

Together, these findings offer a focused snapshot of contemporary AF ablation practice among academically engaged, research-active cardiac surgeons. Within this group, surgical ablation of AF is commonly performed, yet meaningful variation persists in how and when it is applied. This observed heterogeneity highlights areas of clinical uncertainty and reinforces the need for a well-designed, definitive trial to guide future care and address the gap between recommendations and practice.

Ethics Statement

The Hamilton Integrated Research Ethics Board provided approval (#18258).

Patient Consent

The authors confirm that patient consent is not applicable to this article. This study was a survey of cardiac surgeons and did not involve patients or patient-identifiable data. Participating surgeons provided consent to take part in the survey.

Funding Sources

The authors have no funding sources to declare.

Disclosures

R.P.W. reports receiving grants from AtriCure, Abbott, Cytosorbents, and Boston Scientific; receiving consulting or speaking honoraria from Atricure, Cytosorbents, Johnson & Johnson, and Procyrion; and being supported by a Canada Research Chair. E.P.B.-C.reports receiving grants and consulting fees from Abbott Laboratories; consulting fees from Trimedic Therapeutics; and salary support from the Clive Kearon career award (McMaster University—Department of Medicine) and Population Health Research Institute. W.F.M. reports receiving consulting and speaking honoraria from AtriCure and iRhythm, respectively, and salary support from Heart and Stroke Foundation of Canada and Population Health Research Institute. M.C. reports being a consultant for AtriCure and Medtronic. The other authors have no conflicts of interest to disclose.

Editorial Disclaimer

Given their role as Associate Editor, Saurabh Gupta had no involvement in the peer review of this article and has no access to information regarding its peer review.