Effectiveness of cardiac Holter monitoring for more than four weeks for the detection of occult atrial fibrillation in embolic stroke of undetermined source

Abstract

Background

Hidden atrial fibrillation (AF) is considered a major contributor to embolic strokes of undetermined source (ESUS). Prolonged cardiac monitoring is effective for detecting AF, especially when started soon after stroke onset.

Methods

This prospective multicentre study evaluated the rate of AF detection in ESUS patients under real-world clinical conditions using wearable Holter cardiac monitoring for 30 days.

Results

A total of 352 patients were included. The mean age was 67.8 years, and 46% were women. The AF detection rate was 17.1%, with a mean episode duration of 401.7 min. Compared with patients without AF, those with AF were older (72.2 vs 67.0 years; mean difference, 5.16 years; 95% CI, 2.24–8.09; p = 0.0012), had a higher frequency of large vessel occlusion (35.9% vs 17.0%; OR, 2.66; 95% CI, 1.40–5.05; p = 0.002), showed larger left atrial diameter (40.4 vs 38.1 mm; mean difference, 2.25 mm; 95% CI, 0.56–3.95; p = 0.013), and more often had excessive supraventricular ectopic activity on 24-hour Holter monitoring (37.5% vs 20.5%; OR, 2.33; 95% CI, 1.26–4.32; p = 0.006). Notably, 73.3% of AF cases were detected during monitoring initiated within the first 6 months after stroke onset.

Conclusions

These findings support the potential clinical utility of 30-day prolonged cardiac monitoring to detect AF in ESUS patients, highlighting its value in etiological stroke assessment in real-world clinical practice, with greater efficacy during the first 30 days, but also in cases where monitoring is initiated months after the event.