Primary Prevention of Atrial Fibrillation Using Renin-Angiotensin-Aldosterone System Inhibitors among Medicare Beneficiaries with Hypertension

Abstract

Statement of the Problem: Renin-angiotensin-aldosterone-system inhibitors (RAASIs) have long been associated with reduced risk of new-onset AF in patients with hypertension. However, previous studies have not properly accounted for the presence of competing risks in the usual care setting. Hypothesis: The study was designed to determine the effects of RAASIs on the hazard and cumulative incidence (sub-distribution hazard) of newly documented AF and to test whether the effects were significantly different from those of beta-blockers (BBs) or calcium-channel blockers (CCBs). Methods: A propensity score (PS)-matched retrospective cohort study was conducted in a random 5% sample of the Medicare beneficiary population from 2007 to 2011 with hypertension treated with antihypertensive drug monotherapy, consisting of 50,307 beneficiaries. Beneficiaries on RAASI-based monotherapies were matched 1:1 with beneficiaries on BB-based monotherapies (n=13,242) and CCB-based monotherapies (n=10,843) based on PS. All beneficiaries were free of baseline AF and compelling indications for RAASIs. Competing risk analyses were performed. Cox proportional cause-specific hazard regression was used to estimate the effects of RAASIs on newly documented AF and all-cause mortality without AF (i.e., competing risks). Fine-Gray Models were used to examine whether there was a significant difference in the cumulative incidence of newly documented AF between beneficiaries treated with RAASIs and BBs/CCBs, accounting for all-cause mortality without AF as competing risks. Results: The adjusted cause-specific hazard ratio (95% confidence interval [CI]) in the RAASI vs. BB groups was 0.69 (95% CI: 0.58 to 0.81) for newly documented AF. The adjusted sub-distribution hazard ratio (95% CI) in the RAASI vs. BB groups was 0.69 for newly documented AF (95% CI: 0.59 to 0.81). The adjusted cause-specific hazard ratio (95% CI) in the RAASI vs. CCB groups was 0.55 (95% CI: 0.46 to 0.66) for newly documented AF. The adjusted sub-distribution hazard ratio (95% CI) in the RAASI vs. CCB groups was 0.54 for newly documented AF (95% CI: 0.45 to 0.65). Conclusions: RAASI-based monotherapies were associated with not only a reduced hazard of newly documented AF but also with a reduced sub-distribution hazard of newly documented AF for Medicare beneficiaries with hypertension enrolled in the Part D program.