Incidence of atrial fibrillation in Acute Ischemic Stroke detected by long duration holter monitoring.

Bertrand C, Pico Fernando, Alamowitch S, Faisandier Y, Leclercq JF.


BACKGROUND AND PURPOSE: our aims were to quantify the yield of long duration Holter monitoring for detection of atrial fibrillation in asymptomatic patients with Stroke and Transient Ischemic Attack (TIA).

METHODS : 121 patients with cryptogenic stroke having a first negative 24H holter monitoring, have been followed by a long duration Holter monitoring (3 weeks of continuous recording) to detect paroxysmal atrial fibrillation (PAF). PAF is defined by a continuous Atrial Fibrillation (AF) lasting at least 30 sec. 9% of patients had TIA and 91% of patients had stroke.
RESULTS : 132 patients (60+ 17 years, 38% of women) were included and 132 Holter were recorded. The mean number of Holter recorded was 1,09 by patient and the mean duration of recording was 20,5 + 4,5 days. Atrial vulnerability was positive in 6,8% with atrial fibrillation and atrial flutter. The duration of PAF and atrial flutter was from 4 minutes up to 9 days in continuous. The mean duration of atrial vulnerability was 1,72days. Episode of arrhythmia were present in 72,7% for the 1st week, , 27,2% for the 2nd week, and 0,9% for the third week. Intermediate variability was present in 22 ,7% of the patients and they have been considered as positive in 43,3% if we considered the duration of atrial extra beat and the mean rate of the runs and the aspect of multiple localizations of ischemia in the brain RMI. Total positive cases was 15,1%. If we had recorded continuously only 1 week, 3 significant AF and flutter would have been not detected because they occurred between 8 and 21 days.

PAF is present in 6,8% of stroke and TIA patients, with a mean duration of 1,1days. Atrial vulnerability was considered as positive in 15,1%. A 3 weeks continuous Holter monitoring is more efficient to detect PAF than a 1 or 2 weeks recording.