Twenty-eight day Holter monitoring is poorly tolerated and insensitive for paroxysmal atrial fibrillation detection in cryptogenic stroke.

This pilot study in a prospective cohort of 20 cryptogenic stroke patients showed that a significant proportion has paroxysmal atrial fibrillation undetected by 24-h Holter monitoring. However, longer monitoring with 28-day Holter was poorly tolerated and still insufficiently sensitive for paroxysmal atrial fibrillation detection. Further studies are urgently needed to elucidate the optimal timing, method and duration of cardiac rhythm monitoring following ischaemic stroke.