The aim of the study is to assess the incidence of patients suffering from ischemic stroke or transient ischemic attacks who have underlying asymptomatic paroxysmal atrial fibrillation.

Patients who have suffered an ischemic stroke or transient ischemic attack, without a history of atrial fibrillation, are planned to be included. Starting within 14 days of the ischemic stroke, participating patients are asked to perform 10 second ECG recordings using a handheld ECG device twice daily (mornings and evenings) during 30 days. These recordings are transmitted via telephone to a secure encrypted Internet site. Within these 30 days the participants also perform an ambulatory 24 hour Holter recording. Handheld ECG recordings are evaluated continuously. In case of atrial fibrillation the patient is informed and offered treatment with anti coagulant medication (Warfarin). The investigation is a comparison between 24 hour continuous ECG recordings and short intermittent ECG recordings twice daily over a longer time period to determine which method is the best to detect atrial fibrillation in this patient group.

Hypothesis: Short Intermittent ECG recordings over a longer time period is more efficient, compared with continuous 24 hour ECG recordings, in detecting silent paroxysmal AF in patients with an ischemic stroke and without a history of atrial fibrillation.

Inclusion Criteria:

• No previously diagnosed atrial fibrillation.
• Acute ischemic stroke or transient ischemic attack verified by computed tomography or clinical diagnosis using the National Institute of Health Stroke Score within 14 days of enrollment.
• Ability to understand study instructions both verbal and written.
• Ability to perform and transmit ECG recordings according to study design.

Exclusion Criteria:

• Previously diagnosed atrial fibrillation.
• Hemorrhagic stroke
• Low compliance
• Pacemaker carrier

Patients from hospital strokewards.