Two recently published trials have provided evidence in favor of longer ECG monitoring among patients with cryptogenic stroke (CS). In the CRYSTAL-AF trial, new atrial fibrillation (AF) was detected in 12.4% of patients with implantable monitor when compared with 2% among those receiving standard follow-up. A similar result was observed in the EMBRACE trial in which AF was detected in 16.1% of patients who received 30-day event recorder (3.2% in controls). These data are compelling in convincing us that long-term ECG technologies have superior sensitivity for the detection of AF in CS; however, clinical specificity for the definition of CS etiology of such findings cannot be established and can be lower than expected, leaving open questions about the etiologic weight of AF in CS. The causative role of AF in this subpopulation remains to be proven, and diagnostic routes cannot be solely unbalanced toward the research of AF.
Keywords: atrial fibrillation; cryptogenic stroke; holter ECG; loop recorder; thromboembolism.