Stroke remains the leading cause of mortality and morbidity worldwide. Patients surviving the first ever stroke remain at high risk of stroke recurrence. While the cause of stroke recurrence is multifactional, atrial fibrillation (AF) has been recognized as one of the most important factors for stroke recurrence. Despite the fact that AF related stroke is highly preventable with long-term oral anticoagulation therapy particularly the non-vitamin K oral anticoagulants (NOAC), the arrhythmia is often not diagnosed until stroke recurrence due to its paroxysmal and asymptomatic nature.

Diagnosing AF before stroke recurrence has been recognized as one of the most important objectives for stroke management. Strategies to detection AF in stroke survivors have been recommended including 7-day or 14-day Holter monitoring at the early post-stroke period. There's also study trial exploring the clinical application of insertable cardiac monitor to detect AF in patients with recent cryptogenic stroke.

In the past decade, advance in ECG technology has made possible to record ECG using handheld smartphone accessory devices in household setting. Together with the rapid developing artificial intelligence-based ECG diagnosis and mobile communication, it is possible to remotely monitor hundreds of thousand ECGs from patients at risk of AF. Few randomized trials have assessed the effectiveness of handheld ECG recording device for AF detection in patients with history of stroke. Here we test the hypothesis that long-term home-based ECG monitoring will be more sensitive than standard care in detecting AF in patients with history of stroke but no documented AF for 24 months. Secondarily, we will investigate whether early detection of AF might confer a benefit on longer-term clinical outcomes.

The trial will primarily be conducted in Hong Kong SAR, Macau SAR, and mainland China. It is a prospective, multi-centered, randomized controlled, parallel-group study to compare the time to detect AF with portable handheld single lead ECG recorder versus conventional care in patients with history of ischemic stroke without documented AF. Patients will be randomly assigned in a 1:1 ratio to Home-based AF screening group and Control group.

The home-based AF screening system comprises (1) a handheld single lead electrocardiogram (ECG) recorder (Comfit Healthcare Devices Limited, Hong Kong SAR, China), and (2) a patient-facing smartphone application specially designed for the study. Patients randomized to the home-based AF group will be instructed to record a 30-second single lead ECG using the handheld ECG device every morning or when symptomatic. In addition, all study participants will be instructed to measure and report their blood pressures in the morning and evening through the patient-facing smartphone application. In addition, all study participants will be encouraged to input their routine blood tests from the public hospitals through the smartphone application.

All remotely obtained data will be automatically transmitted in real-time through the study smartphone application to a secured cloud hosting and displayed to display on a web-based dashboard at the clinicians' offices for review. All ECG recordings will be analyzed using an artificial intelligence based diagnosed system. When a diagnosis of AF is made with the system, the patients will be called back for a formal cardiology consultation and standard 12-lead ECG to confirm the diagnosis within 1 week. This will be followed by echocardiogram and blood tests to determine the possible options of long-term anticoagulation therapy.

Inclusion Criteria:

• age >= 18 years
• history of ischemic stroke within 5 years
• voluntarily agrees to participate by providing written informed consent

Exclusion Criteria:

• previously documented AF and/or atrial flutter
• long-term anticoagulation therapy
• short life expectance (< 1year) due to concomitant medical condition(s)
• cardiac implantable electronic device
• inability or refusal to provide inform consent
• lack of skills in operating simple electronic devices
• unavailability of a mobile network service in the place of residence

Dataset used during the study will be available from the corresponding author on reasonable request from date of publication for 2 years.

Dataset used during the study will be available from the corresponding author on reasonable request from date of publication for 2 years.

Dataset used during the study will be available from the corresponding author on reasonable request from date of publication for 2 years.

Independent research officers and data staff responsible for data entry will be blinded from randomization assignment.