BACKGROUND AND AIMS: Key approaches to prevent recurrent stroke in people with Embolic Stroke of Unknown Source (ESUS) include oral anticoagulation for those with atrial fibrillation (AF). Since AF is found in a large proportion, anticoagulation applied to all ESUS patients is being tested. Whilst possibly reducing recurrent stroke costs for those with undiagnosed AF, and reducing costs associated with cardiac rhythm monitoring, bleeding risk could negate these health-economic savings. We assessed the implications of anticoagulating all ESUS patients, including those without AF, in the United States. 
METHOD: Data from Cryptogenic Stroke and Underlying Atrial Fibrillation (“CRYSTAL-AF”) were used to inform the ischemic stroke rate in patients meeting ESUS criteria, but free of AF. Anticipated bleeding rates were sourced from published meta-analyses, whilst, as in the ongoing ESUS studies, we assumed a base-case 30% reduction in ischemic stroke for all NOACs (while testing other rates in sensitivity analysis). Large commercial payor databases and Medicare data were used to understand actual pharmaceutical pricing and private payer costs; and, Medicare costs, respectively. Patient co-pays were also examined.

RESULTS: High-Dose NOACs caused 3.1-28.9 additional bleeds per 100 patients over lifetime. Over projected patient lifetime, NOACs add $5,415.37-21,025.67 of cost while exposing patients to co-payments ranging $1,715.76-12,472.9 depending on agent and patient plan. High-Dose NOACs would remain cost additive even if 100% of strokes were avoided. Low-Dose NOACs would need to reduce embolic stroke risk by 39.55-58.54%.
CONCLUSION: Anticoagulating non-AF ESUS patients may cause excess bleeding and add significant costs for uncertain benefits.

Clinical and Economic Burden of Oral Anticoagulation (OAC) in Patients with Embolic Stroke of Unknown Source (ESUS) Without Atrial Fibrillation (AF) in the United States