AMALFI: Modest AF Detection With ECG Patch Screening

MADRID, Spain—The use of a noninvasive ECG patch modestly increases the detection of atrial fibrillation (AF) and treatment with anticoagulant therapy among patients at moderate to high risk for stroke, according to the results of the AMALFI trial.

Investigators, who presented the data last week during a Hot Line session at the European Society of Cardiology Congress 2025, said the relatively small difference in detection rates with the ECG patch versus usual care, coupled with the low burden of AF, make it challenging to see a difference in hard clinical outcomes.

“Over half of the AF that we detected with the patch had a burden below 10% altogether,” lead investigator Rohan Wijesurendra, MB BChir, DPhil (University of Oxford, England), said at a press conference. “This means that AF screening in this setting may have limited impact on stroke events at 2-and-a-half years.”

AMALFI, which was published simultaneously in JAMA, was an investigator-initiated, parallel-group, unblinded study that was conducted without physical study sites or in-person visits. The remote trial was designed to be a simple, low-cost intervention that could possibly be expanded to a national screening program.

“The concept would be to take a population that’s at risk, screen them, find atrial fibrillation that was previously undiagnosed, [and] institute treatment with anticoagulation and perhaps other drugs to reduce clinical risk and prevent stroke and other cardiac events,” said Wijesurendra. With the introduction of new digital technologies, there are more discrete and less cumbersome ways of screening for AF than a traditional Holter monitor, he added.

Higher-Risk Population

In the US, population-wide screening for AF is not recommended. The United States Preventive Services Task Force has concluded that the current evidence is not strong enough to assess the balance and risks of screening for AF in those without signs and symptoms of the arrhythmia. In Europe, population-based screening with a prolonged, noninvasive ECG has a class IIa recommendation in selected patients, such as those 75 years and older (and those 65 years and older with risk factors).

There have been several trials investigating screening, including STROKESTOP, LOOP, and STROKESTOP II. In the STROKESTOP studies, screening for AF in older patients had a small net clinical benefit compared with usual care, while there was no reduction in stroke risk seen in the LOOP trial.

AMALFI differed in that investigators focused on a higher-risk population than previous studies. More than 22,000 invitations were sent out to patients 65 years and older with a CHA₂DS₂-VASc score or 3 or higher (men) or 4 or higher (women). In the end, 5,116 patients (mean age 77.7 years; 47% female) were randomized to usual care or screening with a noninvasive, single-lead ECG patch (Zio XT; iRhythm) for 14-day continuous ambulatory cardiac monitoring. The patch was sent in the mail and the results were sent to family physicians, with doctors then allowed to act on the results if they wished. In total, 84% of participants wore and returned the patch, with participants wearing the patch for a median of 14 days.

The rate of AF detection at 2.5 years was 6.8% in the ECG patch arm and 5.4% in the usual-care arm (P = 0.03). These results were consistent in subgroups prespecified by age and sex. More than half (57%) of AF was picked up on the first day the patients wore the patch. Oral anticoagulation was prescribed to 14.4% of patients who received the ECG patch and 12.8% of those treated with usual care (P = 0.08). The mean exposure to oral anticoagulation at 2.5 years was 1.6 months in those randomized to the patch and 1.1 months in controls (P < 0.001).

“This suggests that GPs were clearly responding appropriately to the patch findings and starting the preventative treatments that were suitable,” said senior investigator Louise Bowman, MBBS (University of Oxford), during the Hot Line presentation.

The investigators estimated that usual care would catch roughly 1.7% of patients with AF, but the observed rate was 5.6%. Similarly, they estimated use of the ECG patch would identify 4.4% of patients with AF, but the rate was higher at 7.0%.

“There was a lot of AF detected through routine care,” said Bowman.

The study, she added, wasn’t powered to detect differences in clinical outcomes. A trial designed to assess whether AF screening could prevent stroke would require 100,000 to 150,000 patients, said Bowman. The researchers plan to follow patients for several more years and perform quality-of-life and economic analyses.

Isabelle Van Gelder, MD (University Medical Center Groningen, the Netherlands), the discussant for AMALFI, praised the innovative trial, saying that it shows the remote screening strategy is feasible. “Whether screening reduces stroke and is cost-effective is still a gap in the evidence,” she said.

Thomas Pilgrim (Bern University Hospital, Switzerland), one the session’s panelists, questioned whether the data were strong enough to justify routine AF screening, even in this higher-risk population. “I am a little concerned about bleeding as a downside of the detection of atrial fibrillation,” he said. Bowman said their group plans to track long-term outcomes, including bleeding, to determine the risk-benefit profile of this strategy.