Continuous Stroke Unit Electrocardiographic Monitoring Versus 24-Hour Holter Electrocardiography for Detection of Paroxysmal Atrial Fibrillation After Stroke

• Published in: Stroke (1970) 2012-10, Vol.43 (10), p.2689-2694
• Main Authors: Rizos, Timolaos, Güntner, Janina, Jenetzky, Ekkehart, Marquardt, Lars, Reichardt, Christine, Becker, Rüdiger, Reinhardt, Roland, Hepp, Thomas, Kirchhof, Paulus, Aleynichenko, Elena, Ringleb, Peter, Hacke, Werner, Veltkamp, Roland
• Format: Article
• Language: English
• Subjects: Aged; Atrial Fibrillation - complications; Atrial Fibrillation - diagnosis; Atrial Fibrillation - physiopathology; Biological and medical sciences; Electrocardiography, Ambulatory - instrumentation; Electrocardiography, Ambulatory - methods; Headache. Facial pains. Syncopes. Epilepsia. Intracranial hypertension. Brain oedema. Cerebral palsy; Hospital Units; Humans; Inpatients; Ischemic Attack, Transient - etiology; Ischemic Attack, Transient - prevention & control; Male; Medical sciences; Middle Aged; Monitoring, Physiologic - instrumentation; Monitoring, Physiologic - methods; Nervous system (semeiology, syndromes); Neurology; Outpatients; Prospective Studies; Risk Factors; Secondary Prevention; Stroke - etiology; Stroke - prevention & control; Time Factors; Vascular diseases and vascular malformations of the nervous system
• ISSN: 0039-2499; 1524-4628
• DOI: 10.1161/STROKEAHA.112.654954

Cardioembolism in paroxysmal atrial fibrillation (pxAF) is a frequent cause of ischemic stroke. Sensitive detection of pxAF after stroke is crucial for adequate secondary stroke prevention; the optimal diagnostic modality to detect pxAF on stroke units is unknown. We compared 24-hour Holter electrocardiography (ECG) with continuous stroke unit ECG monitoring (CEM) for pxAF detection. Patients with acute ischemic stroke or transient ischemic attack were prospectively enrolled. After a 12-channel ECG on admission, all patients received 24-hour Holter ECG and CEM. Additionally, ECG monitoring data underwent automated analysis using dedicated software to identify pxAF. Patients with a history of atrial fibrillation or with atrial fibrillation on the admission ECG were excluded. Four hundred ninety-six patients (median age, 69 years; 61.5% male) fulfilled all inclusion criteria (ischemic stroke: 80.4%; transient ischemic attack: 19.6%). Median stroke unit stay lasted 88.8 hours (interquartile range, 65.0-122.0). ECG data for automated CEM analysis were available for a median time of 64.0 hours (43.0-89.8). Paroxysmal AF was documented in 41 of 496 patients (8.3%). Of these, Holter detected pxAF in 34.1%; CEM in 65.9%; and automated CEM in 92.7%. CEM and automated CEM detected significantly more patients with pxAF than Holter (P