Detecting Atrial Fibrillation in Patients With an Embolic Stroke of Undetermined Source (from the DAF-ESUS registry)

Atrial fibrillation (AF) causes a substantial proportion of embolic strokes of undeterminded source (ESUS). Effective detection of subclinical AF (SCAF) has important therapeutic implications. We conducted a prospective study to determine the prevalence of SCAF in patients with ESUS through of a 21-day Holter monitoring. In an early-monitoring group, Holter was initiated immediately after hospital discharge. The results were compared with a previous cohort of patients in whom the Holter was initiated at least 1 week after hospital discharge (late-monitoring group). We included 100 patients (50 each group; 69 ± 13 years, 56% male). Mean time from ESUS to Holter was 1.2 ± 1 day in the early-monitoring group and 30 ± 15 days in the late-monitoring group. SCAF was detected in 22% of patients in the early-monitoring and 6% in the late-monitoring group (p <0.05). Patients with SCAF were older (77 ± 9 vs 67 ± 11 years, p <0.05), with a higher rate of left atrial enlargement (50% vs 20%, p<0.05), renal impairment (28% vs 5%; p<0.01), and a slower mean heart rate (55 ± 6 vs 70 ± 6 beats/min; p<0.001). On multivariate analysis, the presence of persistent bradycardia (≤60 beats/min) in the 21-day Holter was a powerful and significant risk factor for SCAF. In conclusion, the sooner 21-day Holter electrocardiogram monitoring is initiated after ESUS, the more likely SCAF can be detected. Sinus bradycardia is a powerful predictor of SCAF in patients with ESUS.