AIM:This study aimed to evaluate the risk of major bleeding among two cohorts of nonvalvular atrial fibrillation patients newly initiating a vitamin K antagonist (VKA) or apixaban in a real-world setting in Italy. PATIENTS & METHODS:A retrospective study using a large administrative database of Italian local health units was performed, using data from ten local health units and patients were included from the date of new initiation of apixaban or VKAs from January 2012 to June 2015. RESULTS:Risk of major bleeding was calculated using an adjusted Cox regression model. Compared with VKA, apixaban had a significantly lower risk of major bleeding (hazard ratio = 0.44 [95% CI: 0.12-0.97]). CONCLUSION:In this analysis, apixaban was associated with a lower risk of major bleeding compared with VKA.