METHOD AND RESULTS:

Thrombo-embolic events were defined as stroke (ischaemic or unknown cause) or systemic embolism (SE). Clinical outcomes were estimated using the Kaplan–Meier method. All-cause mortality and International Society on Thrombosis and Haemostasis (ISTH) major bleeding after events were analysed using a Cox proportional hazards model with time-dependent covariates. Of 18 201 patients in ARISTOTLE, 365 experienced a thrombo-embolic event [337 strokes (ischaemic or unknown cause), 28 SE]; 46 (12.6%) of which were fatal. In the 30 days before and after a thrombo-embolic event, 11% and 37% of patients, respectively, were not taking an oral anticoagulant. During follow-up (median 1.8 years), 22 patients (7.1%/year) had a recurrent stroke, 97 (30.1%/year) died, and 10 (6.7%/year) had major bleeding. Compared with patients without a thrombo-embolic event, the short- and long-term adjusted hazards of death in patients with a thrombo-embolic event were high [30 days: HR 3.5, 95% CI 2.5–4.8; both P< 0.001]. The adjusted hazards of major bleeding were also high short-term (HR 10.37, 95% CI 3.87–27.78; P< 0.001) but not long-term (HR 1.7, 95% CI: 0.77–3.88; P= 0.18).

CONCLUSIONS:

Thrombo-embolic events were rare but associated with high short- and long-term morbidity and mortality. Substantial numbers of patients are not receiving oral anticoagulant therapy before and, despite this risk, after a first thrombo-embolic event.