LAUSR

BACKGROUND The usefulness of β-blockers in heart failure (HF) patients with permanent atrial fibrillation (AF) has been questioned. METHODS AND RESULTS We analyzed data from HF patients (958 patients (801 males, 84%, age 67 ± 11 years)) with AF enrolled in the MECKI score database. We evaluated prognosis (composite of cardiovascular death, urgent heart transplant, or left ventricular assist device) of patients receiving β-blockers (n = 777, 81%) vs. those not treated with β-blockers (n = 181, 19%). We also analyzed the role β1-selectivity and the role of daily β-blocker dose. To account for different HF severity, Kaplan-Meier survival curves were normalized for relevant confounding factors and for treatment strategies. Dose was available in 629 patients. Median follow-up was 1312 (577-2304) days in the entire population, 1203 (614-2420) and 1325 (569-2300) days in patients not receiving and receiving β-blockers. 224 (23%, 54/1000 events/year), 163 (21%, 79/1000 events/year), and 61 (34%, 49/1000 events/year) events were recorded, respectively. At 10-year patients treated with β-blockers had a better outcome (HR 0.447, p < 0.01) with no effects as regards β1selective drugs (53%) vs. β1-β2 blockers (47%). Survival improved in parallel with β-blocker dose increase (HR 0.296, 0.496, 0.490 for the high, medium, and low dose vs. no β-blockers, p < 0.0001). CONCLUSION HF patients with AF taking a β-blocker have a better outcome (with a survival improvement in parallel with daily dose but no differences as regards β1 selectivity) but this does not mean that β-blockers improve outcomes in these patients as we cannot control for all the potential confounders associated with β-blocker use.