To determine the diagnostic yield of a 30-day smartphone electrocardiogram recording compared to 24-hour Holter monitor for detecting occult paroxysmal atrial fibrillation in patients with a recent ischemic stroke or transient ischemic attack of undetermined etiology after completion of a standard clinical stroke work-up.

Study outcomes

Primary:

Detection of one or more episodes of atrial fibrillation or atrial flutter ≥30 seconds as assessed at the 30 day follow-up.

Secondary:

1. Proportion of patients prescribed oral anticoagulation, as assess at the 30-day follow up.
2. Patient adherence with 30-day smartphone ECG monitoring, and percentage of patients using the monitor >75% of the target in 30-day period
3. Time to first detection of AF within 30 days (days)
4. One year rate of recurrent ischemic stroke or transient ischemic attack, death, hemorrhagic stroke, major adverse bleeding events, detection of AF outside the study protocol.

Exploratory:

To explore the feasibility and cost effectiveness of 30-day smartphone ECG recording for detecting occult paroxysmal atrial fibrillation

Inclusion Criteria:

• i. Age 55 years or older. ii. Diagnosis of the index event made by neurologist or general physician of an acute ischemic stroke or TIA (WHO definition) of undetermined etiology occurring within the previous 6 months (180 days). The event must be either:

  1. An ischemic stroke confirmed by neuroimaging; or
  2. A transient ischemic attack, defined as involving a focal unilateral motor deficit, speech/language deficit or hemianopia, with symptom duration <24 hours (note: amaurosis fugax/ transient monocular blindness, pure sensory spells, isolated vertigo spells, etc. do not qualify for enrolment given the potential for misdiagnosis of such events).

iii. Patient meets the following:

1. At least one 12-lead ECG has already been obtained as part of the routine clinical post-stroke/TIA work up, and not ECGs have shown any episodes of AF or atrial flutter.
2. A Holter monitor has already been obtained as part of the routine clinical post-stroke/TIA work-up, and does not show any episodes of AF or atrial flutter ≥30seconds.

iv. The patient is being actively investigated for the etiology of the stroke/TIA event and additional cardiac monitor is desired to screen further for the possibility of AF or atrial flutter.

v. The following diagnostic test have already been completed as part of clinical routine post-stroke/TIA:

1. Brain imaging with CT or MRI
2. Vascular imaging of the extracranial and intracranial circulation with either CT angiography or MRI angiography to exclude significant large vessel occlusion disease as the most likely mechanism for index ischemic event (carotid Doppler ultrasound is acceptable for those presenting with anterior circulation ischemic events).*
3. Transthoracic (or transesophageal) echocardiography to exclude thrombus or other structural heart disease that in the opinion of the investigator is the most likely cause for the stroke/TIA events.* *(if a baseline investigation cannot be obtained clinically after the index event and prior to study enrolment, then it is acceptable for study purposes for investigations to be obtained after patient enrolment into the study but prior to the 90-day follow-up visit.) vi. Informed consent from the patient (or from a legally authorised representative if the patient is not competent, due to stroke-related cognitive impairment, aphasia, or anosognosia).

vii. The patient is expected to survive at least 12 months.

Exclusion Criteria:

• i. Any previously documented atrial fibrillation or atrial flutter (a remote history of transient AF during perioperative period is not exclusionary).

ii. Exclusively retinal stroke or retinal TIA event. iii. A most responsible etiological diagnosis for the qualifying stroke/TIA event has already determined i.e. cervicocephalic artery dissection, venous sinus thrombosis, hypercoagulable states, or other known cause.

iv. Planned carotid endarterectomy within 90 days. v. Any finding on echocardiography for which there is already an evidence-based indication for long-term anticoagulation (e.g. mechanical heart valve, thrombus, etc) vi. Inability to use the AliveCor smartphone ECG monitor upon enrolment into the study (if patient is randomised into interventional group).* vii. Participating in a clinical trial involving investigational medication. viii. Endocarditis. ix. Pregnancy.

The final decision is depending on the agreement among the co-investigators.