RESUMO
BACKGROUND: Ischemic stroke and transient ischemic attack (TIA) standard-of-care etiological investigations include an ECG and prolonged cardiac monitoring (PCM). Atrial fibrillation (AF) detected after stroke has been generally considered a single entity, regardless of how it is diagnosed. We hypothesized that ECG-detected AF is associated with a higher risk of stroke recurrence than AF detected on 14-day Holter (PCM-detected AF). METHODS: We conducted a retrospective, registry-based, cohort study of consecutive patients with ischemic stroke and TIA included in the London Ontario Stroke Registry between 2018 and 2020, with ECG-detected and PCM-detected AF lasting ≥30 seconds. We quantified PCM-detected AF burden. The primary outcome was recurrent ischemic stroke, ascertained by systematically reviewing all medical records until November 2022. We applied marginal cause-specific Cox proportional hazards models adjusted for qualifying event type (ischemic stroke versus TIA), CHA2DS2-VASc score, anticoagulation, left ventricular ejection fraction, left atrial size, and high-sensitivity troponin T to estimate adjusted hazard ratios for recurrent ischemic stroke. RESULTS: We included 366 patients with ischemic stroke and TIA with AF, 218 ECG-detected, and 148 PCM-detected. Median PCM duration was 12 (interquartile range, 8.8-14.0) days. Median PCM-detected AF duration was 5.2 (interquartile range, 0.3-33.0) hours, with a burden (total AF duration/total net monitoring duration) of 2.23% (interquartile range, 0.13%-12.25%). Anticoagulation rate at the end of follow-up or at the first event was 83.1%. After a median follow-up of 17 (interquartile range, 5-34) months, recurrent ischemic strokes occurred in 16 patients with ECG-detected AF (13 on anticoagulants) and 2 with PCM-detected AF (both on anticoagulants). Recurrent ischemic stroke rates for ECG-detected and PCM-detected AF groups were 4.05 and 0.72 per 100 patient-years (adjusted hazard ratio, 5.06 [95% CI, 1.13-22.7]; P=0.034). CONCLUSIONS: ECG-detected AF was associated with 5-fold higher adjusted recurrent ischemic stroke risk than PCM-detected AF in a cohort of ischemic stroke and TIA with >80% anticoagulation rate.