Cessation of oral anticoagulation is an important risk factor for stroke and mortality in atrial fibrillation patients.

Oral anticoagulation (OAC) is highly effective preventing stroke and mortality in AF, but withdrawal is common in the elderly, when high bleeding risk and when are difficulties achieving an optimal time in therapeutic range (TTR). We analysed the rate of OAC cessation, predisposing factors to cessation and the relation to clinical outcomes in a large 'real world' cohort of AF patients over a long follow-up period. Consecutive non-valvular AF outpatients clinically stables for six months were recruited. Rates of cardiovascular events, major bleeding and mortality were recorded and related to OAC cessation. We included 1361 patients (48.7 % male; aged 76, IQR 71-81), followed-up for a median of 6.5 years. During follow-up, 244 patients suffered thrombotic events, 250 suffered from major bleeding and 551 patients died. 10 % of patients stopped OAC. After OAC withdrawal, there were 36 thromboembolic events (22 strokes), 10 major bleedings and 75 deaths. OAC cessation was independently associated with adverse cardiovascular events (HR 1.45; 95 % CI 1.01-2.08), stroke/TIA (HR 1.85; 1.17-2.94) and all-cause mortality (HR 1.30; 1.02-1.67). Independent predictors of OAC cessation were age ≥80 (HR 2.29; 1.60-3.29), previous coronary artery disease (HR 0.32; 0.15-0.71), major bleeding (HR 5.00; 3.49-7.15), heart failure (HR 2.38; 1.26-4.47), cancer (HR 5.24; 3.25-8.44) and renal impairment developed during follow-up (HR 2.70; 1.26-5.75). In conclusion, in non-valvular AF patients, cessation of OAC was independently associated with the risk of stroke, adverse cardiovascular events and mortality. Bleeding events and some variables associated with higher bleeding risk are responsible for OAC cessation.