Screening for High Stroke Risk Atrial Fibrillation: Assessing a New Paradigm

Abstract

Background: The Loop trial suggested that short episodes of atrial fibrillation(AF) do not increase the risk of stroke and do not require long-term anticoagulation. In patients with CHADS-Vasc score 2, AF episodes of 24 hours or longer should be considered high risk AF(HR-AF) as opposed to shorter episodes that are low risk AF(LR-AF). With this new paradigm, the AF duration needs to be determined when screening for AF. Objectives: Determine accuracy of AF screening modalities for detecting HR-AF while minimizing detection of LR-AF. Methods: ASSERT II trial data and published LOOP trial data were used to calculate the rate of HR-AF and LR-AF detection when screening a hypothetical population of 1000 elderly patients, 40 with HR-AF, using different modalities over one year. Results: The AF burden due to LR-AF is twice the burden of HR-AF. A first annual 12-lead ECG detects mostly HR-AF but the second annual ECG detects mostly LR-AF. Twice daily 30 second ECGs for 14 days detects 14 HR-AF, 33 LR-AF. One-day and 30-day Holters after an ECG showed low HR-AF detection rates. Twice a week 30 second ECGs found 37 HR-AF and 20 LR-AF patients. A smartwatch used hourly for a year detects 39 HR-AF and 1 LR-AF patients. Conclusions: After the first ECG, annual ECGs would detect mostly LR-AF. Twice weekly ECGs for a year detects most HR-AF patients but also detected a high number of LR-AF patients. A Smartwatch used every hour for a year detects almost all HR-AF and very few LR-AF patients..

Keywords: Arrhythmia detection, smartwatch, rhythm monitoring, ECG monitoring