Study data were abstracted from the 2016 through 2019 Medicare Fee for Service database (FFS), specifically the Medicare institutional claims files that include inpatient, outpatient, hospice, skilled nursing, and home health agency claims. The Medicare FFS database includes demographic, enrollment and payment information, diagnosis and procedure records, and survival data (12). The Medicare FFS is de-identified and HIPAA-compliant. The Institutional Review Board at Creighton University (Info record number: 2004394) acknowledged this study as Not Human Subjects Research.

The Medicare patient cohort was identified using inpatient and outpatient files from the Medicare FFS institutional claims data. We included patients at least 18 years old carrying an AF diagnosis who underwent a catheter ablation for AF. First, we identified patients carrying a primary or secondary diagnosis of AF using International Classification of Diseases–Tenth Revision–Clinical Modification codes (ICD-10-CM: I48.0, I48.1, I48.2, or I48.91). For the inpatient AF cohort, we identified patients who had a primary discharge diagnosis for a percutaneous intracardiac procedure (MS-DRG: 273, 274) and who underwent an ablation procedure as indicated by a primary or secondary ICD-10 procedure code (ICD-10-PCS: 02583ZZ). For the outpatient AF cohort, we identified patients carrying a Current Procedural Terminology (CPT) code for atrial fibrillation (AF) ablation (CPT: 93656).

To ensure all ablations were specifically for AF, we excluded patients carrying a diagnosis of supraventricular tachycardia, ventricular tachycardia, atrial flutter, other premature beats, Wolf-Parkinson-White, Lown-Ganong-Levine, atrioventricular nodal tachycardia, and a history of pacemaker. Additionally, we excluded patients who underwent pacemaker implantation, implantable cardioverter-defibrillation implantation, and/or open surgical ablation on the same day as their ablation procedure (see Supplementary Table S1 for a complete list of inclusion/exclusion criteria).

For each visit, we extracted patient-specific demographics that included age, biologic sex, and race (white, non-white, unknown). We calculated patient-specific CHA₂DS₂-VASc score that required age, biologic sex, acute myocardial infarction (AMI), congestive heart failure (CHF), hypertension (HTN), stroke, ischemic heart disease, peripheral vascular disease (PVD), and diabetes (13). We also extracted comorbid chronic obstructive pulmonary disease (COPD) and visit-specific characteristics including geographic region (Northeast, Midwest, South, West) (14) geographic division (New England, Middle Atlantic, East North Central, West North Central, South Atlantic, East South Central, West South Central, Mountain, Pacific) (14) and setting (inpatient, outpatient).

The primary aim was to evaluate regional differences in 30-day mortality following AF ablation; 30-day mortality was defined as death from any cause within 30 days after AF ablation. Mortality was confirmed from the Medicare denominator files (Medicare Master Beneficiary Summary File). To allow for consistent, complete 30-day post-ablation follow-up from year to year, the index period was the first 11 months of each calendar year (January 1–November 30). The secondary aim was to evaluate between-race differences in 30-day mortality.

All analyses were assessed separately in the inpatient and outpatient settings as these coded AF ablations represent uniquely different populations.

First, we quantified the frequency of inpatient and outpatient AF ablation procedures among Medicare patients. Next, we stratified demographic and clinical characteristics by region. Categorical variables were presented as frequency and percent, and continuous variables were presented as mean and standard deviation. Between-region differences in categorical variables were compared using the chi-square test or Fisher's exact tests, depending on cell counts. Between-region differences in continuous variables were compared using one-way ANOVA. Unadjusted and adjusted logistic regression models were estimated to evaluate region and race differences in the odds of 30-day mortality. Adjusted logistic regression models controlled for race, comorbid COPD, and CHA₂DS₂-VASc score.

Analyses were conducted using SAS v. 9.4 with two-tailed p < .05 indicating statistical significance.

From 2016 to 2019, 38,477 inpatient and 128,544 outpatient AF ablation procedures met the inclusion criteria. Table 1 presents the frequency of AF ablation by sex, race, geographic region, and geographic division in the inpatient and outpatient setting. The greatest number of AF ablation were performed in the South (inpatient: 17,415, outpatient: 55,932). Within the South, the South Atlantic geographic division performed the most AF ablation at 10,447 inpatients and 30,357 outpatients. AF ablation patients were more frequently male and white (Table 1). Table 2 presents age, sex, race, and comorbidities by geographic region. In the inpatient and outpatient setting, hypertension was the most frequent comorbidity, followed by ischemic heart disease and CHF (Table 2). The average CHA₂DS₂-VASc score was 4.70 (SD: 1.51) and 3.91 (SD: 1.48) in the inpatient and outpatient settings, respectively (Table 2).

Patients receiving inpatient AF ablations in the Northeast and West had lower CHA₂DS₂VASc scores compared to the Midwest and South (p < 0.001; Northeast: 4.62 ± 1.56, West: 4.66 ± 1.55, Midwest: 4.70 ± 1.46, South: 4.76 ± 1.50). Notably, the West had the lowest rate of comorbid HTN, ischemic heart disease, AMI, and diabetes among inpatient AF ablation patients (Table 2). Similarly, in the outpatient setting, patients receiving care in the West had the lowest CHA₂DS₂VASc scores (p < 0.001; West: 3.77 ± 1.51, Northeast: 3.89 ± 1.51, Midwest: 3.83 ± 1.45, South: 4.03 ± 1.46). The rate of COPD, CHF, HTN, ischemic heart disease, AMI, and diabetes were lowest in patients receiving outpatient AF ablations in the West (Table 2).

Table 3 presents the 30-day mortality associated with AF ablation by sex, race, geographic region, and geographic division in the inpatient and outpatient setting. The outcome of biological sex differences was not a focus of our study, however, an increased inpatient 30-day mortality was noted among males compared to females.

Of the inpatient AF ablation patients, 2.28% (N = 879) died within 30 days of their procedure. The Northeast had the lowest mortality rate following inpatient AF ablation at 1.75% (N = 140), followed by the Midwest at 2.23% (N = 175) and the West at 2.30% (N = 120). The South had the highest mortality rate at 2.55% (N = 444). Adjusted for race, comorbid COPD, and CHA₂DS₂-VASc score, patients undergoing AF ablation in the Northeast had 28% lower adjusted odds of 30-day mortality compared to patients treated in the South (NE: 1.75.% vs. S: 2.55%; aOR: 0.72, 95% CI: 0.60–0.88, p = 0.001; Table 4) as well as a 23% lower adjusted odds of 30-day mortality compared to patients undergoing AF ablation in the West (NE: 1.75% vs. W: 2.30%; aOR: 0.81, 95% CI: 0.66–0.99, p = 0.036; Table 4). The adjusted odds of 30-day mortality following inpatient AF ablation were statistically similar between the Northeast and Midwest, Midwest and South, West and South, and Midwest and West (Table 4).

The 30-day mortality following inpatient AF ablation was 2.23% (N = 752) for white patients and 2.77% (N = 110) for non-white patients. After adjusting for region, comorbid COPD, and CHA₂DS₂-VASc score, the adjusted odds of 30-day mortality following inpatient AF ablation were statistically similar between non-white and white patients (Non-white: 2.77% vs. White: 2.23%; aOR: 1.17, 95% CI: 0.96–1.44, p = 0.123; Table 4).

Of outpatient AF ablation patients, 0.19% (N = 243) died within 30 days of their procedure. The Northeast had the lowest mortality rate following outpatient AF ablations at 0.15% (N = 28), followed by the South at 0.16% (N = 92). The West and Midwest had highest 30-day mortality following outpatient AF ablation at 0.25% (N = 71) and 0.20% (N = 52), respectively. After adjusting for race, comorbid COPD, and CHA₂DS₂-VASc score, patients treated in the Northeast were associated with 44% lower adjusted odds of 30-day mortality compared to patients undergoing outpatient AF ablation in the West (aOR 0.56, 95% CI: 0.36–0.87, p = 0.01; Table 4). Patients who received an outpatient AF ablation in the West had 73% greater adjusted odds of 30-day mortality compared to patients treated in the South (aOR: 1.73, 95% CI: 1.27–2.36, p < 0.001; Table 4). The adjusted odds of 30-day mortality following an outpatient AF ablation were statistically similar between the Northeast and South, Midwest and South, Midwest and West, and Northeast and Midwest (Table 4).

The 30-day mortality following outpatient AF ablation was 0.18% (N = 212) for white patients and 0.36% (N = 22) for non-white patients. After adjusting for region, comorbid COPD, and CHA₂DS₂-VASc score, compared to white patients, non-white patients were associated with 73% greater odds of 30-day mortality following outpatient AF ablation (aOR: 1.73, 95% CI: 1.11–2.69, p = 0.016; Table 4).

Despite using inpatient and outpatient data and performing adjustments, the study lacked patient-level data (NYHA class, ejection fraction, medications) and only included older, Medicare FFS patients, which is not fully representative of the AF ablation population. In addition, the Medicare FFS beneficiary files did not include the Medicare Advantage population that has grown significantly in recent years and may limit the generalizability of these findings to the broader Medicare population. The incorporation of Medicare Advantage data in future research would be of value. Regional differences in care could involve other confounding variables and system processes of care that could not be fully adjusted nor measured. Data on non-white patients could not be further sub-categorized due to small numbers. Due to data limitations, race and ethnicity may be conflated. Hispanic patients may be captured in either the white or non-white race categories.