Comparison of procedural outcomes in patients undergoing catheter vs surgical ablation for atrial fibrillation and heart failure with reduced ejection fraction

Abstract Background There is a lack of research comparing procedural outcomes of surgical ablation (SA) and catheter ablation (CA) among patients with heart failure with reduced ejection fraction (HFrEF) and atrial fibrillation (AF). The main objective was to compare the short‐term procedural outcomes of SA and CA in patients with HFrEF. Methods We used the national inpatient sample to identify hospitalizations over 18 years with HFrEF hospitalization and AF, and undergoing SA and CA from 2016 to 2017. Furthermore, the clinical outcomes of SA vs CA in AF stratified as nonparoxysmal and paroxysmal were analyzed. Results A total of 1,770 HFrEF hospitalizations with AF who underwent SA and 1,620 HFrEF hospitalizations with AF who underwent CA were included in the analysis. Hospitalizations with CA had higher baseline comorbidities. The in‐hospital mortality among HFrEF with AF undergoing SA as compared with CA was similar (2.8% vs 1.9%, respectively, adjusted P‐value 0.09). Hospitalizations with SA had a significantly longer length of hospital stay, a higher percentage of postprocedural, and cardiac complications. In HFrEF hospitalizations with nonparoxysmal AF, SA as compared with CA was associated with a higher percentage of in‐hospital mortality (2.4% vs 1%, adjusted P‐value <.05), a longer length of stay, a higher cost of treatment, and a higher percentage of cardiac complications. Conclusion CA is associated with lower in‐hospital adverse procedural outcomes as compared with SA among HFrEF hospitalizations with AF. Further research with freedom from AF as one of the outcome is needed between two groups for HFrEF.


| INTRODUC TI ON
Atrial fibrillation (AF) is the most common cardiac rhythm dysfunction affecting ~5.3 million people across the United States. 1 The varied treatment strategies of AF extend across a wide spectrum, including medical therapy (rate or rhythm control), catheter ablation (CA), and surgical ablation (SA). Traditionally, the standard of care for drug-refractory AF was SA. 2 However, recent advancements with the application of CA for the management of refractory AF have increased significantly owing to its low complexity and adverse events. 3 Hence, CA is approved for drug-refractory AF by European and American guidelines. 4,5 Few randomized clinical trials published comparing SA vs CA have highlighted CA to have less consistent maintenance of sinus rhythm postprocedure. [6][7][8] However, these clinical trials were bound by the limitations in terms of strict patient selection and limited patient population.
Additionally, these trials also excluded patients with an ejection fraction below 45%. [6][7][8] There is a lack of research comparing procedural outcomes of SA and CA among heart failure patients with reduced ejection fraction F I G U R E 1 Flow chart for the selection of the hospitalizations (HFrEF) or systolic heart failure and AF. The evidence is indispensable as CA is performed more frequently among subjects with HFrEF. 9 Therefore, we assessed short-term procedural outcomes comparing SA and CA in patients with HFrEF and AF. We have also expanded our research by dividing our patient population into paroxysmal and nonparoxysmal AF. to December 31st, 2017. We excluded admissions before 2015 since it utilized ICD-9-CM codes. Moreover, ICD-10-CM includes specific codes for the diagnosis and procedures included in the present analysis. The NIS variables provided by the sponsor were used to identify baseline characteristics and demographics. We also included longterm use of aspirin, anticoagulation, and antiplatelet agents using ICD-10-CM codes. The length of prescription and compliance with these medications was unknown, given the nature of the database.
The NIS database has been explained in detail in the past. 11,12 Since the NIS data includes de-identified administrative data with prior ethical committee approval, no additional ethical committee approval was deemed necessary for the present analysis.  The ICD-10-CM codes utilized for these outcomes are detailed in Supplementary Table S1.

| Statistical analysis
All statistical analyses were performed using SAS version 9.4. The statistical analysis was performed strictly in accordance with the Agency for Healthcare Research and Quality and expert consensus recommendations. 15 The analysis accounted for the survey design (SURVEYMEANS and SURVEYFREQ), clustering (HOSP_NIS), and weights (DISCWT).
This is in accordance with the methodological standard required by the sponsor. 15 All tests were two-sided and a level of significance was set to    ( Figure 3).

| D ISCUSS I ON
To the best of our knowledge, this is the first study comparing pro-  Ablation therapy using SA or CA is recommended for patients' refractory to medical therapy. Former randomized controlled trials and meta-analyses demonstrated better symptomatic relief from AF following SA compared with CA. 6,16,17 However, SA is not employed because of its complexity with a steep learning curve, and its association with higher procedure-related adverse outcomes. 16 Alternatively, CA is an alternate option that has slightly lower rates of freedom from AF in contradistinction to SA, however, it is easy to learn and associated with lower procedure-related adverse events. 16  F I G U R E 2 Differences in the clinical outcomes forhospitalizations with atrial fibrillation and heart failure with reduced ejection fraction: Stratified by gender and type of the procedure complications) with SA as compared with CA are likely mechanical and associated with injury to the adjacent structures during the procedure. The increased occurrence of periprocedural complications in SA as compared with CA has been highlighted in former meta-analyses. 16,19 SA requires general anesthesia, which has its adverse effects. These adverse events reflect the longer length of stay and a higher cost of hospitalizations as seen in this study. Besides, these adverse events might lead to worse short-term and long-term quality of life. Former randomized controlled trials and meta-analysis demonstrated SA as the preferred option for previously failed ablation attempts with AF recurrence, as they are associated with better AF-free survival. 16 Nonetheless, CA was associated with less in-hospital outcomes in this study, including both subgroup analysis.
Additionally, prior meta-analyses reported better outcomes with CA as compared with medical therapy. 20 Hence, considering the higher baseline comorbidities among heart failure patients, we suggest that CA should be the preferred first line procedure for the management of treatment-refractory AF, especially in patients with nonparoxysmal AF and HFrEF hospitalization.
Tremendous advancement has been noted in CA techniques.
Cryoballoon pulmonary vein isolation has reported promising results in the treatment of non-paroxysmal AF. 21 The use of radiofrequency balloon catheter and pulsed-field ablation has also shown to improve outcomes post-CA. 22,23 Improved imaging modality such as the use of delayed enhancement cardiac magnetic resonance can detect uncommon anatomy and fibrosis before ablation and hence can help in the better selection of patients, who will possibly benefit from the use of CA. 24 Furthermore, second generation laser balloon can help improve tissue contact and visibility as reported by the MERLIN registry. 25 This study has several limitations that must be taken into con- In conclusion, CA is associated with lower in-hospital adverse procedure-related outcomes compared with SA among HFrEF hospitalizations. The long-term safety, efficacy, and quality of life associated with SA and CA remain to be determined in this population. A future randomized controlled trial comparing clinical outcomes and long-term safety and efficacy in both, paroxysmal and nonparoxysmal, patients are warranted to expand the knowledgebase.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

F I G U R E 3
Differences in the clinical outcomes forhospitalizations with atrial fibrillation and heart failure with reduced ejection fraction: Stratified by race and type of the procedure