Procedural outcomes of single versus double transeptal access for ablation of atrial fibrillation: Data from the NCDR Afib ablation registry

Abstract

Background

The transeptal puncture (TSP) is an obligatory step during catheter ablation (CA) for atrial fibrillation (AF). There is no consensus regarding the utility for a single or double transeptal (TSP) access for CA of AF.

Objective

To evaluate procedural outcomes and in-hospital complications (death, cardiovascular, neurological, peripheral vascular and pulmonary events) associated with CA of AF in patients undergoing single versus double TSP puncture using the National Cardiovascular Data Registry (NCDR).

Methods

Clinical characteristics and procedural outcomes of patients undergoing CA for AF between 2016 and 2024 were evaluated. Inverse probability of Treatment Weighting (IPTW) analysis was developed to adjust for confounding variables between TSP groups.

Results

A total of 212,345 patients were identified. Among patients with paroxysmal AF (PAF) and non-paroxysmal AF (NPAF), single TSP was associated with comparable but lower risk of total procedural complications (PAF: 1.47% vs. 1.73%, OR:0.91, 0.88–0.93), (NPAF: 2.09% vs. 2.58%, OR:0.77 0.72–0.82), hospitalization > 1 day (PAF: 7.15% vs. 7.92%, OR:0.90, 0.87- 0.93), (NPAF: 10.41% vs. 12.13%, OR:0.84, 0.81–0.86) and a similar but higher likelihood of discharge in sinus rhythm (PAF: 95.55% vs. 95.17%, OR:1.09, 1.05–1.14) (NPAF: 93.37% vs. 92.31%, OR:1.17, 1.13–1.22) compared to double TSP technique. The incidence of stroke or TIA was not different between TSP groups.

Conclusion

In a large AF ablation registry, single and double TSP use had similar procedural safety, and comparable likelihood of discharge in sinus rhythm. Single TSP use was not associated with embolism or cerebrovascular events. Operator preference and patient characteristics should be taken into consideration to define optimal TSP strategy.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s10840-025-02196-4.

Introduction

Catheter ablation (CA) for atrial fibrillation (AF) has become a commonly performed procedure with estimates of 35 CA of AF procedures per 100,000 patients/year in the US Medicare population [1]. The transeptal puncture (TSP) can often be a challenging step-in the procedure, requiring intricate anatomic knowledge and advanced technical skills, to avoid potential complications such as cardiac perforation and air embolism. Although recent technological advances such as intracardiac echocardiography and electroanatomic mapping have made the procedure safer and more efficient, there is no consensus regarding whether a single vs double TSP approach will lead to better outcomes.

The double TSP technique requires an additional vascular access and enables the operator to concurrently place one mapping and a separate ablation catheter in the left atrium (LA) allowing immediate visualization of LA intracardiac signals at multiple points, and the ability to perform pacing maneuvers to confirm pulmonary vein isolation (PVI) and assess lines of block [2]. Meanwhile, the single TSP technique simplifies the procedure by eliminating the need for a second vascular access and TSP puncture, and avoids sheath-to-sheath interactions in the LA. However, studies have demonstrated an increased risk of silent cerebral events, possibly caused by the insertion of air in the intravascular system during catheter exchange (3,4). Although there are potential benefits and risks with each technique, its effect on clinical outcomes has not been evaluated in a large US based patient population.

A recent meta-analysis has shown a low overall rate of total procedural complications for CA AF of 4.51%, with most common complications being vascular access 1.31%, pericardial effusion 0.43% and stroke 0.17% [3], events that could be directly affected by TSP technique. But with such a low complication rate, a large dataset of patients is needed to accurately investigate whether there are differences in procedural outcomes between single and double TSP approaches. Therefore, we aimed to use data from a large multicenter registry, namely the National Cardiovascular Data Registry (NCDR) AFib Ablation Registry, to assess for differences among patients stratified into single vs double TSP puncture access in terms of patient characteristics, procedural outcomes, and in-hospital complications.

Methods

Data source

The NCDR AFib ablation registry is a voluntary US nationwide registry that contains information about the index procedure and hospital stay in patients undergoing CA for AF from multiple participating hospitals using a standardized data collection form [4]. Data is collected per admission, allowing for the inclusion of patients multiple times if they undergo multiple CA procedures. The NCDR monitors the data collection process and audits randomly selected participating centers annually to ensure complete, valid, and accurate submissions [5]. For this study, we used version 1.0 of the NCDR AFib Ablation registry [6]. Waiver of written informed consent and authorization for this study was granted by Advarra (former Chesapeake Research Review Incorporated). This study was done according to the Declaration of Helsinki guidelines on human research.

Study population

All patients undergoing CA for AF enrolled in the NCDR registry between January 1, 2016, and August 31, 2024, were included. To adjust for differences within the cohort, a decision was made to divide the analysis into paroxysmal AF (PAF) and non-paroxysmal AF (NPAF) patients. Outcomes were then assessed by comparing patients undergoing a single TSP and double TSP procedure.

Outcomes

Primary in-hospital safety outcome was defined as: rate of total procedural complications (death, procedural complications including cardiovascular events, neurological events, peripheral vascular events, pulmonary events). Secondary efficacy outcomes were defined as 1) percentage of patients discharged in sinus rhythm or atrial paced rhythm, 2) hospital stay > 1 day.

Statistical analysis

Continuous data is presented as mean ± SD or median [Interquartile range] and categorical variables as proportions. Baseline demographic and clinical categorical variables were compared between the two groups using chi-square test, and continuous variables using the Wilcoxon rank sum test or Student t-test as appropriate.

Inversed probability of treatment weighted analysis (IPTW) was performed to control for baseline differences among TSP groups aiming to adjust for confounding variables that could influence outcomes. Therefore, a logistic regression model was performed to estimate the predicted probability of undergoing a single TSP approach within the cohort, by using the following variables: age, gender, race, BMI, insurance type, CHA₂DS₂Vasc score, HAS-BLED score, cryoablation, adjunctive lesions to PVI, use of general anesthesia, and amiodarone use. Additionally, adjustment was made for type of medical institution, (private, university, government), region and number of beds. IPTW was then used to balance baseline characteristics between TSP groups by assigning weights based on the inverse of the estimated propensity score. This approach created a pseudo-population that was adjusted for observed covariates, attempting to simulate a randomized study and estimate an average treatment effect. The baseline characteristics between TSP groups before and after weighting can be found in the Supplemental Tables 1 and 2.

Weighted logistic regressions to obtain odds ratios were used to create 95% confidence intervals (CIs) comparing primary and secondary outcomes between single TSP and double TSP groups. A p-value ≤ 0.05 was considered statistically significant.

A sensitivity analysis was created using propensity score (PS) matching instead of IPTW to adjust for bias inherent to the use of single TSP use that could have affected the results. PS models were built separately for paroxysmal AF and non-paroxysmal AF cohorts, using same covariates as in the IPTW model. We performed 1:1 greedy nearest-neighbor matching without replacement with a 0.25 caliper on the log of PS, to create balanced groups. This method matched each single TSP subject to the closest double TSP subject based on their PS, in order to estimate an average treatment effect on the matched population.

Results

Between January 1, 2016, and August 31,2024, data on 212,345 patients undergoing AF CA were submitted by 100 hospitals. A total of 306 patients were excluded due to missing TSP data and 624 had their CA procedures cancelled due a LA thrombus detected prior to CA. For the final analysis, 5,371 patients were excluded due to missing covariates required for the weighting model. Our final study population consisted of 206,974 patients. PAF was more common in the single TSP group (58% vs. 51.5%, p < 0.001), as well as use of cryoablation (31% vs. 15.5%, p < 0.001) compared to double TSP, respectively. Notably, adjunctive lesions on top of PVI were more common in the double TSP group (62.6% vs. 50.2%, p < 0.001).

Characteristics of paroxysmal AF patients

Amongst the 115,490 patients who underwent CA for PAF, 66% patients underwent a single TSP approach and 33% underwent double TSP access. A higher use of cryoablation was noted in the single TSP group, while the double TSP group had a higher use of adjunctive lesions to PVI and current amiodarone use. The Midwest and Northeast regions had a higher use of single TSP, while hospitals in the South and West regions had a higher use of double TSP. After IPTW analysis, the selected baseline characteristics for single vs. double TSP were balanced, evidenced by smaller standardized mean differences (SMDs). Baseline characteristics before and after balancing confounding variables with IPTW is presented in Supplemental Table 1.

Primary safety outcomes

Single TSP was associated with lower risk of procedural complications compared to double TSP access (1.47% vs. 1.73%, OR:0.91, 95% CI: 0.88–0.93, p < 0.001)), including pericardial effusion requiring intervention (0.33% vs. 0.53%, p < 0.001), pericardial effusion causing tamponade (0.20% vs. 0.29%, p < 0.001), heart failure (0.18% vs. 0.25%, p < 0.001), and bradycardia related adverse events (0.25% vs. 0.32%, p < 0.001). Very small numbers of air embolism were registered, with a higher percentage in the single TSP group (0.03% vs. 0.01%, p = 0.026). No difference in stroke (0.08% vs. 0.09%, p = 0.193), or TIA (0.03% vs. 0.03%, p = 0.404) was noted. Furthermore, reported pulmonary vein damage (0.02% vs. 0.00%, p < 0.001) and phrenic nerve damage (0.14% vs. 0.04%, p < 0.001) were associated with single TSP use (Fig. 1 and Table 1).

Fig. 1.

Procedural complications between single and double transeptal access for catheter ablation of paroxysmal atrial fibrillation. weighted odds ratio of procedural complications in paroxysmal af patients after inverse probability of treatment weighting (IPTW) between single and double transeptal groups adjusting for age, gender, race, BMI, insurance type, CHA₂DS₂Vasc Score, HAS-BLED Score, cryoablation, adjunctive lesions to PVI, use of general anesthesia, amiodarone use, medical center, region, number of beds and profit type

Table 1.

Procedural complications between single and double transeptal access for ablation of paroxysmal atrial fibrillation after inverse probability of treatment weighting analysis

Outcome	Single Transeptal (N = 75,601)	Double Transeptal (N = 37,701)	Weighted Odds Ratio (95% CI)	P value
Total Complications	1.5% (1111/75601)	1.7% (652/37701)	0.776 (0.728—0.828)	<.0001
Death at discharge	0.0% (32/75601)	0.0% (18/37701)	0.831 (0.567–1.219)	0.3448
Cardiac Arrest	0.1% (66/75586)	0.1% (24/37694)	1.153 (0.867–1.534)	0.3280
Myocardial Infarction	0.0% (17/75586)	0.0% (9/37694)	0.909 (0.523—1.580)	0.7355
Air Embolism	0.0% (25/75589)	0.0% (5/37693)	1.962 (1.082—3.556)	0.0264
Bradycardia Adverse Events	0.3% (189/75588)	0.3% (122/37694)	0.652 (0.561—0.758)	<.0001
Cardiac Thromboembolism	0.0% (9/75589)	0.0% (3/37694)	1.068 (0.500—2.282)	0.8645
Heart Failure	0.2% (134/75589)	0.3% (96/37694)	0.628 (0.525—0.751)	<.0001
Pericardial Effusion Resulting in Cardiac Tamponade	0.2% (154/75589)	0.3% (109/37694)	0.658 (0.557–0.777)	<.0001
Pericardial Effusion Req Intervention	0.3% (251/75589)	0.5% (201/37694)	0.600 (0.528–0.681)	<.0001
Cardiac Surgery	0.1% (56/75586)	0.1% (28/37694)	1.015 (0.755—1.366)	0.919
Phrenic Nerve Damage	0.1% (109/75589)	0.0% (16/37693)	1.751 (1.353—2.266)	0.0001
Post Stroke	0.1% (62/75586)	0.1% (37/37693)	0.778 (0.590 −1.027)	0.0765
Post TIA	0.0% (24/75586)	0.0% (11/37693)	0.598 (0.388—0.922)	0.0200
Access Site Bleeding Req Transfusion	0.1% (70/75589)	0.1% (38/37694)	0.816 (0.625—1.064)	0.1329
A-V Fistula Requiring Intervention	0.0% (17/75589)	0.0% (4/37694)	1.754 (0.858—3.589)	0.1237
Deep Vein Thrombosis	0.0% (25/75589)	0.0% (6/37694)	1.791 (1.026—3.127)	0.0405
Hematoma at Access Site	0.2% (151/75589)	0.2% (76/37693)	0.945 (0.786 −1.135)	0.5433
Pseudoaneurysm Req Intervention	0.1% (81/75589)	0.1% (30/37694)	1.284 (0.975—1.692)	0.0754
Vascular Injury Req Surgical Intervention	0.1% (42/76852)	0.0% (16/38647)	1.356 (0.761–2.418)	0.3017
Pneumothorax	0.0% (18/76848)	0.0% (10/38647)	0.883 (0.406–1.921)	0.7543
Hemothorax	0.0% (4/76848)	0.0% (5/38647)	0.426 (0.112–1.617)	0.2099
Respiratory Failure	0.2% (141/76851)	0.2% (96/38647)	0.735 (0.567–0.953)	0.0201
Pulmonary Embolism	0.0% (17/76851)	0.0% (4/38646)	2.418 (0.772–7.571)	0.1294
Pulmonary Vein Damage/Dissection	0.0% (18/76851)	0.0% (1/38647)	11.219 (1.207–104.266)	0.0335

Secondary outcomes

A single TSP was associated with a lower rates of hospitalization > 1 day (7.15% vs. 7.92%, (OR 0.90, 95% CI: 0.87–0.93, p < 0.001)), with a similar likelihood of discharge in sinus rhythm (95.55% vs. 95.17%, (OR 1.09, 95% CI: 1.05–1.14, p < 0.001)) (Table 2).

Table 2.

Secondary outcomes between single and double transeptal access for paroxysmal atrial fibrillation after inverse probability of treatment weighting analysis

Outcome	Single Transeptal (N = 75,601)	Double Transeptal (N = 37,701)	Weighted Odds Ratio (95% CI)	P value
Hospitalization greater than 1 day	7.2% (5408/75601)	7.9% (2988/37701)	0.899 (0.87—0.927)	0.0001
Discharge Atrial Rhythm—Atrial paced	3.3% (2507/75601)	3.2% (1206/37701)	0.985 (0.941—1.032)	0.5288
Discharge Atrial Rhythm—Sinus	95.6% (72,244/75601)	95.2% (35,881/37701)	1.093 (1.050—1.136)	<.0001

Characteristics of Non-Paroxysmal AF patients

Amongst the 91,484 patients who comprised the NPAF cohort. 61% patients underwent a single TSP approach and 39% underwent double TSP access. A higher use of cryoablation was noted again in the single TSP group, while the double TSP group had a higher use of adjunctive lesions to PVI. Again, the Midwest and Northeast region had a higher use of single TSP, while hospitals in the South and West regions had a higher use of double TSP. Baseline characteristics before and after balancing confounding variables with IPTW is presented in Supplemental Table 3.

Primary safety outcomes

Single TSP was also associated with lower risk of procedural complications compared to double TSP (2.09% vs. 2.58%, (OR 0.77, 95% CI 0.72–0.82, p < 0.001)), including pericardial effusion requiring intervention (0.31% vs. 0.47%, p < 0.001), pericardial effusion causing tamponade (0.18% vs. 0.27%, p = 0.001), hematoma at access site (0.23 vs. 0.33, p = 0.021) and heart failure (0.52% vs. 0.64%, p = 0.001). No difference in air embolism was noted between single and double TSP (0.03% vs. 0.03%, p = 0.340), along with TIA (0.03% vs. 0.03%, p = 0.975). Stroke rates were low, with a lower association with single TSP use (0.11% vs. 0.15%, p = 0.004), As noted in PAF patients, phrenic nerve damage (0.15% vs. 0.02%, p < 0.001) was associated with single TSP use (Fig. 2 and Table 3).

Fig. 2.

Procedural Complications between single and double transeptal access for catheter ablation of non-paroxysmal atrial fibrillation. weighted odds ratio of procedural complications after inverse probability of treatment weighting (IPTW) in non-paroxysmal patients between single and double transeptal groups adjusting for age, gender, race, BMI, insurance type, CHA₂DS₂Vasc Score, HAS-BLED Score, cryoablation, adjunctive lesions to PVI, use of general anesthesia, amiodarone use, medical center, region, number of beds and profit type

Table 3.

Procedural complications between single and double transeptal access for ablation of non-paroxysmal atrial fibrillation after inverse probability of treatment weighting analysis

Outcome	Single Transeptal (N = 54,463)	Double Transeptal (N = 36,120)	Weighted Odds Ratio (95% CI)	P value
Total Complications	2.1% (1143/54463)	2.6% (912/35303)	0.769 (0.724—0.818)	<.0001
Death at discharge	0.1% (34/54463)	0.1% (23/35303)	1.178 (0.810—1.713)	0.3917
Cardiac Arrest	0.1% (60/54454)	0.1% (40/35296)	1.091 (0.825—1.441)	0.5411
Myocardial Infarction	0.0% (11/54453)	0.0% (9/35296)	0.874 (0.491—1.555)	0.6469
Air Embolism	0.0% (20/54460)	0.0% (9/35296)	1.319 (0.746—2.332)	0.3409
Bradycardia Adverse Events	0.5% (252/54460)	0.6% (214/35295)	0.703 (0.620—0.798)	<.0001
Cardiac Thromboembolism	0.0% (4/54460)	0.0% (4/35296)	0.598 (0.236—1.514)	0.2780
Heart Failure	0.5% (284/54460)	0.6% (227/35296)	0.741 (0.657—0.836)	<.0001
Pericardial Effusion Resulting in Cardiac Tamponade	0.2% (98/54460)	0.3% (97/35296)	0.625 (0.516—0.757)	<.0001
Pericardial Effusion Req Intervention	0.3% (167/54460)	0.5% (165/35296)	0.664 (0.572—0.771)	0.0001
Cardiac Surgery	0.1% (37/54454)	0.1% (38/35296)	0.733 (0.532—1.010)	0.0573
Phrenic Nerve Damage	0.2% (82/54459)	0.0% (8/35296)	3.268 (2.263—4.719)	<.0001
Post Stroke	0.1% (58/54454)	0.2% (54/35296)	0.682 (0.526—0.885)	0.0040
Post TIA	0.0% (18/54454)	0.0% (13/35296)	0.775 (0.467—1.286)	0.3239
Access Site Bleeding Req Transfusion	0.1% (43/54460)	0.1% (33/35296)	0.814 (0.593—1.117)	0.2031
A-V Fistula Requiring Intervention	0.0% (14/54460)	0.0% (6/35296)	1.096 (0.580—2.069)	0.7783
Deep Vein Thrombosis	0.0% (19/54460)	0.0% (5/35296)	1.714 (0.953—3.084)	0.0720
Hematoma at Access Site	0.2% (124/54460)	0.3% (117/35296)	0.641 (0.538—0.765)	<.0001
Pseudoaneurysm Req Intervention	0.1% (58/54460)	0.1% (32/35296)	1.034 (0.775—1.380)	0.8190
Vascular Injury Req Surgical Intervention	0.1% (32/54460)	0.0% (17/35296)	1.119 (0.752—1.664)	0.5803
Pneumothorax	0.0% (7/54454)	0.0% (10/35296)	0.630 (0.315—1.261)	0.1922
Hemothorax	0.0% (5/54454)	0.0% (4/35296)	0.927 (0.367—2.342)	0.8728
Respiratory Failure	0.4% (195/54459)	0.5% (160/35296)	0.837 (0.724, 0.967)	0.0160
Pulmonary Embolism	0.0% (15/54460)	0.0% (7/35296)	0.975 (0.542, 1.755)	0.9337
Pulmonary Vein Damage/Dissection	0.0% (9/54460)	0.0% (1/35296)	3.045 (0.960, 9.653)	0.0586

Secondary outcomes

A single TSP was associated with a lower rate of hospitalization > 1 day (10.41% vs. 12.13%, (OR: 0.84, 95% CI: 0.81–0.86, p < 0.001)) and with a higher likelihood of discharge in sinus rhythm (93.37% vs. 92.31%, (OR 1.17, 95% CI: 1.13–1.22, p < 0.001)) (Table 4).

Table 4.

Secondary outcomes between single and double transeptal access for non-paroxysmal atrial fibrillation after IPTW

Secondary Outcomes	Single Transeptal (N = 55,364)	Double Transeptal (N = 36,120)	Weighted Odds Ratio (95% CI)	P value
Hospitalization greater than 1 day	10.4% (5671/54463)	12.1% (4285/35303)	0.837 (0.813 −0.862)	<.0001
Discharge Atrial Rhythm—Atrial paced	4.0% (2169/54463)	4.3% (1522/35303)	0.931 (0.888—0.975)	0.0025
Discharge Atrial Rhythm—Sinus	93.4% (50,852/54463)	92.3% (32,587/35303)	1.173 (1.131—1.216)	<.0001

Sensitivity analysis

A PS matching analysis was done in both PAF and NPAF groups. Patient characteristics before and after matching are shown in Supplemental Tables 3 and 4. Procedural outcomes were consistent with our initial analysis in both PAF and NPAF groups, showing a lower association of total procedural complications with single TSP use. Double TSP use was associated with increased length of stay and comparable rates of discharge in sinus rhythm. Supplemental Tables 5 and 6 and Supplemental Figs. 1 and 2.

Discussion

In the largest registry of AF patients undergoing CA to date, we found significant findings regarding the use of one or two TSP access points into the LA. First, a single TSP approach was the most common technique for CA of AF in the United States. Second, the use of single TSP vs double TSP access for CA of AF had comparable procedural outcomes, with single TSP having a minimal lower rate of total procedural complications, including pericardial effusion and access site hematomas. Third, there was no difference in acute procedural success between single and double TSP approaches.

The complication rate found in the NCDR data is in line with a contemporary meta-analysis showing an overall complication rate during CA of 4.51% and a severe complication rate of 2.44% [3]. The large number of patients included in the registry allowed us to detect small differences in complications that prior, smaller studies were not able to identify due to power limitations [7, 8].

Our study is the first to evaluate procedural outcomes of single vs. double TSP use in a large US population undergoing CA for AF. Prior studies comparing both approaches have been smaller, limited in power and likely not generalizable to the US as technical aspects and resources vary by country. A prospective multicenter registry in Germany with 478 patients comparing single and double TSP access, found no difference in procedure duration or complication rate, (Single TSP 10.3% vs. double TSP 10.2%, p = 0.55), with numerically less pericardial effusions (10 vs. 15) and groin complications (n = 9 vs. n = 10) in the single TSP group compared to double TSP respectively [8]. Arrhythmia recurrence after 12 months of follow up was similar in both groups (p = 0.788). The results are in line with our study showing comparable outcomes between techniques, with small differences noted likely due to our large sample size. A randomized study from South Korea with 222 paroxysmal AF patients showed a small increase in minor vascular complications with a double TSP approach (1.4% vs. 6.8%, p = 0.043), while 7% of patients in the single TSP group required a second vascular access and TSP puncture to achieve PVI [9]. A follow-up registry study from the same group in South Korea with 1144 patients found no difference in complication rates among groups, with an increase in total ablation time with the single TSP technique [10]. Our study contributes to the literature by providing an assessment of the performance of single vs double TSP access for CA of AF in the United States population. Overall, our registry was able to show that both TSP techniques are comparable with a difference in procedural complication rates between groups below 1%. Our large sample size was able to detect subtle differences between groups, with double TSP access having an association with higher rate of vascular access site hematomas and pericardial effusion requiring intervention, which could have been related to an extra access site, additional transeptal puncture and more than one indwelling LA catheter during the procedure.

There were no major significant differences between groups regarding cerebrovascular events, suggesting that the risks of catheter exchange during single TSP can be minimized by a wide range of operators with adequate catheter exchange technique. Although there was a small increase in air embolism in the single TSP group in PAF patients, this did not translate in a significant difference in outcomes.

The use of a double TSP approach, with a simultaneous mapping catheter and a second ablation catheter in the LA could guide operators identify areas of reconnection or gaps between ablation lines. However, our data shows a high procedural success rate, with high percentages of confirmed PVI and discharge in sinus rhythm in both groups. The clinical significance of this finding requires further validation with studies that have long term follow-up data.

Notably, the use of a double TSP access for a concomitant LA mapping catheter has recently declined with the advent of pulse field ablation (PFA) delivering rapid and effective atrial lesions. However, in many centers it is common practice an additional TSP access to place an intracardiac echocardiography catheter in LA to assess PFA contact with the atrial wall [11]. The procedural safety and efficacy of this practice needs to be studied further.

Although at a very low occurrence rate, there was an association of single TSP with higher rates of registry reported phrenic nerve injury and pulmonary vein dissection/damage. This finding could be associated with higher use of cryoablation in the single TSP group. A large registry with 2,555 patients showed that phrenic nerve injury with cryoablation is common and can happen in up to 2.5% of cryoablation cases with 0.9% being symptomatic [12]. Moreover, randomized controlled trials have demonstrated higher rates of phrenic nerve injury with cryoablation compared to radiofrequency ablation [13]. In our study, phrenic nerve assessment was more frequently performed in the single TSP (72.2% vs. 51.1%) (Supplemental Table 1) which could underestimate the degree of phrenic nerve injury in the double TSP group. Regarding pulmonary vein damage, intravascular ultrasound studies in pulmonary veins of patients undergoing cryoballoon ablation have shown pulmonary vein dissection in up to 90% of patients [14], which could have explained this association.

Limitations

First, the NCDR AFib Ablation Registry is observational data; therefore, causal inferences cannot be made. Second, the NCDR registry data is limited to the index hospitalization, therefore late complications such as chronic phrenic nerve damage, pulmonary vein stenosis and atrioesophageal fistula were not assessed in this study. Moreover, our study could not determine true clinical benefit and long-term effectiveness between TSP groups, since long term atrial arrhythmia recurrence was not assessed. The use of intracardiac echocardiography or transesophageal echo to guide transeptal access was not captured in the registry and could have affected the results. Due to the observational nature of the study, unmeasured confounder variables such as patient complexity, left atrial substrate, operator’s experience, and center volume could have influenced the decision to perform single versus double TSP access. Although an IPTW analysis was made to adjust for confounding variables including the region and size of the medical institutions, it is plausible that some of the outcomes such as prolonged hospitalization and complication rate were affected by patient characteristics and operator experience rather than the TSP technique itself. Finally, as stated before, our registry results are based on radiofrequency and cryoablation use, further studies are needed to confirm findings with PFA use.

Conclusion

In a large U.S.-based AF ablation registry, single TSP was more commonly used compared to double TSP access. Single and double TSP use had comparable procedural outcomes, with single TSP having a lower rate of acute procedural complications, prolonged hospitalizations and similar likelihood of discharge in sinus rhythm. A single TSP approach was not associated with a significant increased rate of air embolism or cerebrovascular events. Operator preference and patient characteristics should be taken into consideration to define optimal TSP strategy for CA of AF.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AF

Atrial Fibrillation

CA

Catheter ablation

LA

Left Atrium

NCDR

National Cardiovascular Data Registry

NPAF

Non-Paroxysmal Atrial Fibrillation

PAF

Paroxysmal Atrial Fibrillation

PFA

Pulse field ablation

PS

Propensity score

PVI

Pulmonary vein isolation

TSP

Transeptal puncture

Declarations

Conflict of interest

The authors have no conflicts of interest to disclose.