Geographic variation in the use of catheter ablation for atrial fibrillation among Medicare beneficiaries

• Published in: The American heart journal Vol. 169; no. 6; pp. 775 - 782.e2
• Main Authors: Sinner, Moritz F., MD, MPH, Piccini, Jonathan P., MD, MHS, Greiner, Melissa A., MS, Walkey, Allan J., MD, MSc, Wallace, Erin R., MS, Heckbert, Susan R., MD, PhD, Benjamin, Emelia J., MD, ScM, Curtis, Lesley H., PhD
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01-06-2015; Elsevier Limited
• Subjects: Aged; Aged, 80 and over; Atrial Fibrillation - epidemiology; Atrial Fibrillation - surgery; Cardiac arrhythmia; Cardiology; Cardiology - manpower; Cardiovascular; Catheter Ablation - utilization; Catheters; Female; Health Expenditures; Humans; Male; Medicare; Medicare - economics; Older people; Prevalence; Risk Assessment; Terminal Care - economics; United States - epidemiology; United States; United States > US
• ISSN: 0002-8703; 1097-6744
• DOI: 10.1016/j.ahj.2015.03.006

Background Catheter ablation for atrial fibrillation is used increasingly in older patients, yet the risks and benefits are not completely understood. With such uncertainty, local medical opinion may influence catheter ablation use. Methods In a 100% sample of Medicare beneficiaries ≥65 years who underwent catheter ablation for atrial fibrillation between January 1, 2007, and December 31, 2009, we investigated variation in use by hospital referral region (HRR) for 20,176 catheter ablation procedures. Results Across 274 HRRs, median age was 71.2 years (interquartile range 70.5-71.8), a median of 98% of patients were white, and a median of 39% of patients were women. The median age-standardized prevalence of atrial fibrillation was 77.1 (69.4-84.2) per 1,000 beneficiaries; the median rate of catheter ablation was 3.5 (2.4-4.9) per 1,000 beneficiaries. We found no significant associations between the rate of catheter ablation and prevalence of atrial fibrillation ( P = .99), end-of-life Medicare expenditures per capita ( P = .09), or concentration of cardiologists ( P = .45) but a slight association with Medicare expenditures per capita (linear regression estimate 0.016; 95% CI 0.001-0.031; P = .04). Examined HRR characteristics explained only 2% of the variation in HRR-level rates of catheter ablation (model R2 = 0.016). Conclusion The rate of catheter ablation for atrial fibrillation in older patients was low, varied substantially by region, and was not associated with the prevalence of atrial fibrillation, the availability of cardiologists, or end-of-life resource use and was only slightly associated with overall Medicare expenditures per capita.