Anticoagulation in atrial fibrillation. A large real-world update

INTRODUCTION: In a large nationwide administrative database including ∼35 % of Italian population, we analyzed the impact of oral anticoagulant treatment (OAT) in patients with a hospital diagnosis of non-valvular atrial fibrillation (NVAF). METHODS AND RESULTS: Of 170404 OAT-naïve patients (mean age 78.7 years; 49.4 % women), only 61.1 % were prescribed direct oral anticoagulants, DOACs, or vitamin-K antagonists, VKAs; 14.2 % were given aspirin (ASA), and 24.8 % no anti-thrombotic drugs (No Tx). We compared ischemic stroke (IS), IS and systemic embolism (IS/SE), intracranial hemorrhage (ICH), major bleeding (MB), major gastro-intestinal bleeding, all-cause deaths and the composite outcome, across four propensity-score matched treatment cohorts with >15400 patients each. Over 2.9±1.5 years, the incidence of IS and IS/SE was slightly less with VKAs than with DOACs (1.62 and 1.84 vs 1.81 and 1.99 events.100 person-years; HR=0.85, 95%CI=0.76-0.95 and HR=0.87, 95%CI=0.78-0.97). This difference disappeared in a sensitivity analysis which excluded those patients treated with low-dose of apixaban, edoxaban, or rivaroxaban (41.7% of DOACs cohort). Compared with DOACs, VKAs were associated with greater incidence of ICH (1.09 vs 0.81; HR=1.38, 95%CI=1.17-1.62), MB (3.78 vs 3.31; HR=1.14, 95%CI=1.02-1.28), all-cause mortality (9.66 vs 10.10; HR=1.07, 95%CI=1.02-1.11), and composite outcome (13.72 vs 13.32; HR=1.04, 95%CI=1.01-1.08). IS, IS/SE, and mortality were more frequent with ASA or No Tx than with VKAs or DOACs (p<0.001 for all comparisons). CONCLUSIONS: Beyond confirming the association with a better net clinical benefit of DOACs over VKAs, our findings substantiate the large proportion of NVAF patients still inappropriately anticoagulated, thereby reinforcing the need for educational programs.