Smartwatch for AF Detection: Has the Time Come?

Damian McNamara

November 15, 2019

The time for routine use of smartwatches and other wearable technologies to identify atrial fibrillation (AF) may be drawing near.

In a large-scale "virtual study" that had more than 400,000 participants, investigators at Stanford University in California found that the Apple Watch flagged irregular heart rhythms in 0.5% of the study population.

In participants who were subsequently monitored with an electrocardiogram (ECG) patch, there was a positive predictive value of 0.84.

"The study results demonstrate that this algorithm on Apple Watch can identify periods of irregular pulses and prompt subsequent clinical evaluation that confirms the irregularities to be atrial fibrillation," lead study author Marco V. Perez, MD, told Medscape Medical News.

The study was published online November 14 in the New England Journal of Medicine.

"Unprecedented Study"

In previous research, AF has been associated with a fivefold increased risk for stroke. The condition affects about 6 million people in the United States, including an estimated 700,000 people with undiagnosed AF, the investigators note.

"As we know in the medical community, atrial fibrillation is a leading cause of stroke and hospitalization in the US that often remains hidden because many people don't experience symptoms," Perez said.

Traditional continuous heart monitors and implantable devices detected AF in studies of people at high risk, the researchers note, "but have limited monitoring periods and require either invasive procedures or activation by the user."

In search of a more pragmatic approach, Perez and colleagues studied the potential of the Apple Watch to alert people through a linked iPhone app.

"The advent of wearable technology has prompted important questions about its potential use in improving healthcare," Perez said. "In order to help answer some of these questions, the Apple Heart Study, an unprecedented virtual study with over 400,000 enrolled participants, was designed to evaluate the ability of wearable technology to identify atrial fibrillation."

The researchers enrolled 419,297 adults between November 29, 2017, and August 1, 2018. Participants had no history of AF or current anticoagulant use at baseline.

The Apple Watch features a photoplethysmography sensor that can detect changes in blood flow using light-emitting and light-sensitive diodes. When the watch detected pulses that varied over 1 minute, the app prompted wearers to consult a telemedicine expert.

A total of 20 participants with urgent symptoms were directed to go to urgent care or the emergency department. This group included 18 patients with AF and ventricular rates greater than 200 beats per minute lasting more than 30 seconds. In one participant, a pause lasted more than 6 seconds, and another experienced nonsustained ventricular tachycardia for more than 6 seconds.

Those with nonurgent symptoms received an ECG patch in the mail and were instructed to wear it for up to 7 days. Trained technicians evaluated the data upon return of the patches. Those with serious arrythmias were instructed to seek immediate medical care.

Only 0.52%, or 2161, of the 491,297 participants received notification of an irregular pulse during the study.

Among the 450 participants who wore and returned an ECG patch, 84% (95% confidence interval, 76 – 92) of subsequent smartwatch notifications were confirmed as representing AF. Patches were worn an average of 6.3 days.

High Accuracy

"We were very encouraged by the outcomes more than anything. Comparison between irregular pulse detection on Apple Watch and simultaneous ECG patch recordings showed that the pulse detection algorithm has an 84% positive predictive value," said Perez.

"This means that if a participant had an irregular pulse detected on the watch, 84% of the time, this was confirmed to be atrial fibrillation on a simultaneous ECG patch," he added.

Participants whose watch signaled an irregular pulse tended to be older and were more likely to be male and white. In addition, this group was more likely to be at higher risk for a stroke, with a CHA2DS2-VASc score of 2 or higher when compared to the overall study population.

"The study also showed how digital health alerts can enhance engagement with the healthcare system overall," Perez said.

The investigators sent out follow-up surveys at 90 days to the 2161 participants who were notified about an irregular pulse. They received 1376 responses (64%).

Of this group, 57% said they had contacted a healthcare provider outside the study, 36% received a recommendation for further testing, 33% were referred to a specialist, and 28% were prescribed a new medication.

Overall, 76% of this group, or 1041 participants, followed up with a visit to a study physician, a healthcare provider not affiliated with the study, or both.

"This suggests that many actively sought medical attention as a result of an irregularity identified by their Apple Watch," Perez added.

Of the individuals notified by the app, 44% said they received a new AF diagnosis. In contrast, only 1% of those who were not notified reported receiving a new AF diagnosis.

The research provides insights for future research, Perez said.

"The study's completely virtual nature eliminated the need for patients to be physically present and allowed for the implementation of a massive recruitment strategy in a relatively short period of time," he said.

The study was not a randomized clinical trial, which would provide greater diagnostic information. It was also not designed to test the utility of this strategy for screening entire populations for AF, said Perez.

"It is our hope that the study provides a foundation for further research on the health implications of wearable technology," he added.

Pros, Cons of Digital Medicine

"Both the power and the limitations of digital innovation in medicine are evident in a report by Perez and colleagues," Edward W. Campion, MD, and John A. Jarcho, MD, note in an accompanying editorial.

"The study was sponsored by Apple, and in 8 months it managed to enroll some 419,000 participants through the use of a smartphone application (app)," note Campion, who is executive editor and online editor of the New England Journal of Medicine, and Jarcho, who is affiliated with the Division of Cardiovascular Medicine at Brigham and Women's Hospital in Boston, Massachusetts.

Having an iPhone and an Apple Watch were entry requirements, so the study participants were customers of the sponsor, they note.

The study population was skewed toward younger participants, they add; 52% were younger than 40 years, and only 6% were aged 65 years or older, "which may be the opposite of a desirable age profile for a study of atrial fibrillation," the editorialists note.

"It's difficult to draw any conclusions about the true frequency of atrial fibrillation, since only 21% of those with irregular pulse notifications based on monitoring by the smartwatch subsequently returned the ECG patch for analysis," they point out.

Nevertheless, they write, "the results of the Apple Heart Study could be very valuable. The study challenges us to reassess the relation of atrial fibrillation to stroke. The data on that relation have been based on traditional, less-sensitive approaches, such as ECG and shorter-term monitoring of patients with symptoms."

Campion and Jarcho add that "easy-to-use, wearable devices will facilitate research and allow more immediate, reliable patient reports than are available with traditional interviews."

Apple Inc sponsored the study and owns the data. The analyses were performed by Stanford quantitative scientists independent of the sponsor. Perez received grant funding for the study and consulting fees from Apple Inc. Campion and Jarcho have disclosed no relevant financial relationships.

N Engl J Med. Published online November 14, 2019. Abstract, Editorial

Follow Damian McNamara on Twitter: @MedReporter. For more Medscape Neurology news, join us on Facebook and Twitter.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: