Quarterly versus annual ECG screening for atrial fibrillation in older Chinese individuals (AF-CATCH): a prospective, randomised controlled trial

BACKGROUND: Screening for atrial fibrillation before onset of symptoms and the subsequent initiation of oral anticoagulants could prevent stroke and death. The most cost-effective strategy to screen for atrial fibrillation in a population at high risk aged 65 years and older is unknown. Therefore, we aimed to investigate whether more frequent electrocardiography (ECG) recordings would significantly improve the detection of atrial fibrillation compared with annual ECG screenings. METHODS: We did a randomised controlled trial that compared different screening frequencies of 30 s single-lead ECG (AliveCor Heart Monitor) in the detection of atrial fibrillation in Chinese residents (≥65 years) in five community health centres in Shanghai, China. Only participants without history of atrial fibrillation and without atrial fibrillation rhythm at baseline were eligible for inclusion in the trial. Random assignment was done with the use of a random number table and stratified for study site. Participants were randomly assigned in a 1:1 ratio to annual or quarterly screening groups. The quarterly screening group was further randomly assigned in a 3:1 ratio to subgroups of quarterly screening and quarterly screening plus (which involved ECG screening once per week for the first month of follow-up, then quarterly for the remainder of follow-up). The primary outcome was the detection rate of atrial fibrillation. The intention-to-treat analysis was done for all randomly assigned patients who had at least one ECG recording during follow-up. This trial was registered at ClinicalTrials.gov, NCT02990741, and terminated on Oct 31, 2020. FINDINGS: Between April 17, 2017, and June 26, 2018, 8240 participants were randomly assigned to annual screening (n=4120), quarterly screening (n=3090), and quarterly screening plus (n=1030), with a mean number of ECG recordings of 1·6 (SD 0·5) for annual screening, 3·5 (1·5) for quarterly screening, and 5·2 (2·9) for quarterly screening plus during a median of 2·1 years follow-up (13 284 person-years). 73 incident cases of atrial fibrillation occurred: 26 in the annual screening group (4·1 per 1000 person-years) and 47 in the quarterly screening group (6·7 per 1000 person-years. Quarterly screening was associated with a significant increase in the detection rate of atrial fibrillation, compared with annual screening (hazard ratio [HR] 1·71; 95% CI 1·06-2·76; p=0·029). 40 incident cases were detected in quarterly screening (7·2 per 1000 person-years; HR compared to annual screening, 1·83; 95% CI 1·12-3·00; p=0·017) and seven in the quarterly screening plus group (4·8 per 1000 person-years; HR compared with annual screening, 1·24; 0·54-2·86; p=0·61). No significant difference was noted between quarterly screening and the quarterly screening plus group (HR of quarterly screening plus compared with quarterly screening, 0·68; 0·30-1·52; p=0·35). INTERPRETATION: Quarterly 30 s single-lead ECG screening was associated with a significantly higher detection rate of incident atrial fibrillation compared with annual screening, but additional once per week screenings in the first month did not yield an added predictive value. Quarterly screening might be considered in a general population at a high risk of atrial fibrillation, such as those aged 65 years and older. FUNDING: Bayer Healthcare Company.