CE-452772-3 SELF-PULSE EXAM FOR ATRIAL FIBRILLATION SCREENING: A RANDOMIZED CONTROLLED TRIAL

Abstract

Atrial fibrillation (AF), the most commonly treated cardiac arrhythmia, is associated with significant morbidity including the potential for stroke. Because AF is often asymptomatic, there is growing interest in AF screening. Trials of AF screening using external monitors, smartwatches, or other personal electronic devices have shown that screening increases detection of AF. Due to the cost and low yield of screening with insufficient evidence of clinical benefit, the US Preventative Services Task Force has not recommended comprehensive screening for AF at the current time. However, no study has evaluated whether teaching patients to detect irregularity on self-pulse exam is a feasible and effective screening tool for AF. To conduct a randomized controlled trial of self-pulse exam for detection of AF in patients who would be at high risk for stroke. We recruited 524 patients with no history of AF or atrial flutter (AFL) who would be at high risk for stroke (CHADS2VASC score ≥ 2 in men or ≥3 in women) for participation in a study of self-pulse exam for detection of AF. Participants were recruited from local primary care and cardiology clinics and randomized to either usual care or to viewing a 5-minute educational video designed to teach patients to self-detect pulse irregularity. Participants were asked to self-check twice daily for 14 days. Those detecting an abnormal pulse were then offered a confirmatory 14-day external cardiac monitor. The primary endpoint was time to detection of AF or AFL and patients were followed for one year. The one-year rate of AF or AFL detection was increased (5.5% vs. 2.4%, p=0.07), while time to AF/AFL diagnosis (121.8 days vs. 140.6 days, p=0.07, Figure) and time to anticoagulation initiation (127.8 days vs. 239.0 p=0.3) were decreased in the intervention vs. control groups, respectively. Multivariate analysis demonstrated a significant independent association of the intervention with rate of detection of AF/AFL (OR 3.04, p=0.04) and decreased time to detection of AF/AFL (HR 9.75, p=0.01), after adjusting for potential confounders. Patients can be taught a self-pulse exam via a simple educational video. This easily implemented screening tool is independently associated with a greater than three-fold increased rate of detection of AF/AFL, as well as significantly decreased time to AF/AFL diagnosis.