Background: Stroke prevention with oral anticoagulants (OACs) is the cornerstone for the management of atrial fibrillation (AF). However, data about the use of OACs among patients ≥90 years of age are limited. We aimed to investigate the risk of ischemic stroke and intracranial hemorrhage (ICH) and the net clinical benefit of OAC treatment for very elderly patients with AF (≥90 years of age).

Methods: This study used the National Health Insurance Research Database in Taiwan. Risks of ischemic stroke and ICH were compared between 11 064 and 14 658 patients with and without AF ≥90 years of age without antithrombotic therapy from 1996 to 2011. Patients with AF (n=15 756) were divided into 3 groups (no treatment, antiplatelet agents, and warfarin), and the risks of stroke and ICH were analyzed. The risks of ischemic stroke and ICH were further compared between patients treated with warfarin and nonvitamin K antagonist OACs (NOACs) from 2012 to 2015 when NOACs were available in Taiwan.

Results: Compared with patients without AF, patients with AF had an increased risk of ischemic stroke (event number/patient number, incidence = 742/11 064, 5.75%/y versus 1399/14 658, 3.00%/y; hazard ratio, 1.93; 95% confidence interval, 1.74–2.14) and similar risk of ICH (131/11 064, 0.97%/y versus 206/14 658, 0.54%/y; hazard ratio, 0.85; 95% confidence interval, 0.66–1.09) in competing risk analysis for mortality. Among patients with AF, warfarin use was associated with a lower stroke risk (39/617, 3.83%/y versus 742/11 064, 5.75%/y; hazard ratio, 0.69; 95% confidence interval, 0.49–0.96 in a competing risk model), with no difference in ICH risk compared with nontreatment. When compared with no antithrombotic therapy or antiplatelet drugs, warfarin was associated with a positive net clinical benefit. These findings persisted in propensity-matched analyses. Compared with warfarin, NOACs were associated with a lower risk of ICH (4/978, 0.42%/y versus 19/768, 1.63%/y; hazard ratio, 0.32; 95% confidence interval, 0.10–0.97 in a competing risk model), with no difference in risk of ischemic stroke.

Conclusions: Among patients with AF ≥90 years of age, warfarin was associated with a lower risk of ischemic stroke and positive net clinical benefit. Compared with warfarin, NOACs were associated with a lower risk of ICH. Thus, OACs may still be considered as thromboprophylaxis for elderly patients, with NOACs being the more favorable choice.