MADRID — A simple home-based screening strategy that involved wearing an ECG patch for 14 days increased the diagnosis of atrial fibrillation (AF) and associated use of anticoagulants in patients at increased risk for the arrhythmia, according to the results of the AMALFI trial. 

“This remote approach was clearly very acceptable” and could be “highly scalable,” said Louise Bowman, MD, a professor of medicine at University of Oxford, United Kingdom, and senior author of the study. 

However, the researchers found a higher-than-expected rate of AF unrelated to the use of the patch, so the difference between the intervention and control groups was smaller than had been anticipated, Bowman said. 

In addition, of the patients with arrhythmias detected by the patch, over half had an AF burden of less than 10%, with just one third showing persistent AF.

“These findings make us wonder how much impact atrial fibrillation screening might have on reducing stroke risk,” Bowman said, adding that she and her colleagues are conducting longer-term follow-up and economic analyses to gain more insight into this question. 

Bowman presented the AMALFI trial today at the 2025 annual congress of the European Society of Cardiology. The trial was simultaneously published online in JAMA.

Isabelle van Gelder, MD, of the University Medical Centre in Groningen, the Netherlands, called AMALFI “a very relevant, timely, and innovative trial targeting important gaps of evidence for screening.”

The “simple design of this trial is essential in the elderly higher-risk population and shows that screening in this way is feasible, with 84% of participants having worn and returned the patch,” Gelder added. “Results suggest that short-term screening is sufficient, but whether screening reduces stroke and is cost effective is still a gap of evidence.” 

In her presentation, Bowman explained that detecting AF is important because starting oral anticoagulation early can reduce stroke rate quite substantially. 

“But the problem is we often don’t know that people have got atrial fibrillation, so these preventive treatments just don’t get started,” she said. 

Historically, the means of detecting AF were cumbersome and not comfortable for patients to wear for prolonged periods. But digital technologies have advanced the field in recent years, Bowman noted, and there are now discreet options available that people can wear for many days at a time to look for arrhythmias over a long period. As a result, the question of whether screening for AF is an important way of preventing stroke is more pressing than ever, Bowman noted.

However, trials to definitively answer the question would need to enroll more than 100,000 people, Bowman estimated. 

In the meantime, the AMALFI trial set out to evaluate whether screening a high-risk population could enhance the detection of AF. 

The trial randomized 5000 individuals age 65 years and older with risk factors for AF — CHA₂DS₂-VASc score of 3 or higher (men) or 4 or higher (women) — with no previous AF or atrial flutter. Patients were drawn from 27 UK primary care practices.

“One of the key aspects of this trial was making it exceptionally straightforward for the participants to take part. And we did that by using electronic healthcare data. We scoured UK practice information, doing electronic search for eligible patients, and mailed out invitations to 22,000 people with an excellent response rate of about 25%, the majority of whom were eligible,” Bowman said. 

Participants were randomly assigned to receive an ECG patch monitor (Zio XT, iRhythm) in the mail, or to receive usual care. Patients in the patch group were asked to mail the device back to the researchers after wearing it for 14 days. The data were forwarded to primary care doctors who made their own decisions about appropriate preventative treatments.

Follow-up data were collected electronically from general practices and supplemented by central NHS data. Participants were asked to complete quality-of-life questionnaires to contribute to the health economic analyses. 

“Throughout the trial and this approach to screening, there was no need for any contact with a healthcare professional whatsoever. And I think that’s a real asset to this study and potential for a future screening program,” Bowman said. 

The participants had a mean age of 78 years, 47% were female, and the median CHA₂DS₂-VASc score was 4. A total of 2126 participants (84.4%) wore and returned the patch. 

Bowman highlighted the “really very impressive” adherence rate in the study, with 84% of individuals randomly assigned to the patch arm having worn it for almost all the designated time and returned it as requested. 

Results showed that AF was detected by the patch in 89 participants (4.2%). Of these, 55% had an AF burden less than 10%, and just over half had the arrhythmia detected on the first day of wearing the device.

The patch also detected other conditions including atrioventricular block and sustained ventricular tachycardia, and these were reported back to the primary care physician urgently so that they could act appropriately. 

After 2.5 years of follow up, a record of AF was present in 172 individuals (6.8%) in the intervention group vs 136 (5.4%) in the control group (ratio of proportions, 1.26; P = .03), with consistent results in prespecified subgroups. 

Mean exposure to oral anticoagulation by 2.5 years was 1.63 months in the intervention group and 1.14 months in the control group (difference, 0.50 months; P < .001). 

The study was not powered for stroke outcomes, but stroke occurred in 69 participants (2.7%) in the intervention group and 64 (2.5%) in the control group, a nonsignificant difference. 

Bowman said a higher-than-expected level of AF was detected through routine care, which meant the difference between the groups was smaller than the investigators were expecting. 

At an ESC press conference, Rohan Wijesurendra, DPhil, of University of Oxford, and the lead investigator on the trial, called the increase in detection with the patch “modest,” as a result of the higher-than-expected detection rate in routine care.

But earlier detection in the patch group could still be important, he added. “If we want to reduce stroke risk, this is dependent on having an actionable time window, a greater time period during which preventive treatments can be given,” Wijesurendra told Medscape Medical News. 

Wijesurendra said the key question of whether this approach is worth rolling out on a larger scale is not clear. 

“I think we need more data. We’re currently working on health economic analysis, which could be key, and we are going to follow patients up to 5 years to obtain longer-term data,” he said. 

Wijesurendra said the approach used in this study differs from that using smart watches to try to detect AF. 

“This patch is a validated device for the clinical diagnosis of atrial fibrillation. It actually analyses the ECG, so further testing is not needed,” he said. “In contrast, the watch-type devices are used by the general population and detect an irregular pulse which may flag individuals who need to go on for further investigations, which could involve ECG analysis with devices such as this patch.” 

The AMALFI study was funded by the NIHR Oxford Biomedical Research Centre and the British Heart Foundation. iRhythm provided Zio XT ambulatory monitoring patches, electrocardiogram analysis, and cardiac technician data review. Bowman reports receiving grants from Novartis, Novo Nordisk, and the British Heart Foundation outside the submitted work. Wijesurendra has disclosed no relevant financial relationships.