Atrial Fibrillation in Patients With Cryptogenic Stroke

Study Questions:

How often is clinically unapparent paroxysmal atrial fibrillation (AF) detected in patients with cryptogenic stroke?


The subjects of this study were 572 patients (mean age 72 years) with an ischemic stroke or transient ischemic attack (TIA) of unknown cause within the prior 6 months. A standard neurological evaluation, electrocardiogram, echocardiogram, and 24-hour Holter monitor were performed at the time of the stroke/TIA. Patients known to have AF were excluded. The patients randomly were assigned to undergo monitoring with a 30-day event-triggered loop recorder (n = 286) or with an additional 24-hour Holter monitor (n = 285). The primary outcome was detection of AF ≥30 seconds in duration within 90 days of randomization.


AF was detected in significantly more patients in the loop recorder group (16.1%) than in the Holter monitor group (3.2%). The AF lasted >2.5 minutes in 9.9% of loop recorder patients compared to 2.5% of the Holter monitor patients.


The authors concluded that clinically silent AF is detected in patients with cryptogenic stroke/TIA much more often by 30 days of ambulatory monitoring than by a 24-hour Holter monitor.


In patients without an apparent cause of an ischemic stroke/TIA, it is reasonable to assume (but not proven) that clinically silent AF detected by monitoring is a causative factor. Current guidelines recommend a 24-hour Holter monitor in patients with cryptogenic stroke. This important study makes it clear that AF is much more likely to be detected with a 30-day loop recorder than a 24-hour Holter monitor, allowing a greater proportion of patients to potentially benefit from anticoagulation for secondary prevention.

< Back to Listings