Hey there! I’m excited to be back sharing original Research Corner columns after a productive August away doing some reading, reporting and continuing my internal medicine residency. I’m currently on a rotation with our cardiology service, so it feels like the perfect time to dive into a question I’ve been wondering about since I bought my Apple Watch earlier this summer. Are the heart health monitoring features of this watch helpful or harmful?

PS. In case you missed it, the Research Corner newsletter made a crossover appearance on the Tradeoffs podcast last week. I had a great time talking with Dan about how writing this column is shaping my thinking as a new doctor — and vice versa. I hope you’ll give it a listen.

How Smartwatches Could Lead Doctors to Overtreat Patients

I love the idea of a smartwatch warning me if my sleep cycles are off or if my heart rhythm is going out of whack in a dangerous way. As a doctor, I can see how that information — if accurate — could help catch problems sooner. 

At least that was the rosy view of these gadgets I had in mind when I bought my watch a month ago — joining the 1 in 5 Americans who now own a smartwatch or fitness tracker. But a recent study has the doctor in me seeing these devices in a new and worrisome light.

Doctors are quick to react to patients’ Apple Watch notifications

The study Patrick Demkowicz, Rohan Khera and their colleagues published in last month’s American Heart Journal asked 95 doctors how they would respond to different kinds of patients sharing that their Apple Watch had caught some irregular heart activity (but no other symptoms). This irregular heart activity can often be a sign of a common medical condition known as atrial fibrillation or AFib. Though an AFib episode isn’t usually harmful by itself, without proper management ongoing episodes can quintuple a patient’s risk of stroke or heart failure.

Over the past several years, evidence has emerged that the Apple Watch can detect a high percentage of AFib cases correctly. In light of this data, the U.S. Food and Drug Administration cleared the use of the Apple Watch to help diagnose and monitor the condition.

What’s less clear, however, is whether using smartwatches to screen for irregular rhythms is good for people’s health overall. There’s no conclusive evidence (at least not yet) that this type of screening decreases rates of heart disease or death. Consumer devices of this sort also catch lots of false positives, which can lead to unnecessary and potentially harmful treatment.

Still, despite the lack of firm evidence that the benefits of using these watches to screen for AFib outweighs the risks, Demkowicz and colleagues found in their study that:

• In 99.5% of cases, physicians said they’d order at least one follow-up diagnostic test in response to the smartwatch’s notification, ranging from a low-cost electrocardiogram to a more intense procedure, like a stress test.
• Only 1 out of the 95 respondents explicitly said they’d need to first review the evidence for Apple Watch-detected AFib before ordering a test or drug treatment.
• 1 in 4 said they’d consider prescribing medication for patients who receive these alerts.

The study was limited, in that it was based on clinical vignettes instead of real-world cases and clinicians could select multiple answers at the same time (for example, choosing to order tests and start medications without specifying whether they’d do one first or both together).

Relying on smartwatches to catch AFib could waste time, or even harm patients and worsen health disparities

The growing prevalence of these devices has opened a Pandora’s box — they’re now on millions of patients’ wrists, creating a seemingly infinite number of data points. And doctors can’t stop that.

Instead, some scholars I’ve spoken with urge more research and more clinical guidelines to help physicians know what to do (or not do) with this data.

Ishani Ganguli, an assistant professor of medicine at Harvard and Tradeoffs Advisory Board member, has noted that without clearer guidelines, wearables like the Apple Watch can waste clinicians’ and patients’ time and money by prompting a cascade of inappropriate tests and treatments that have questionable benefit. Ganguli pointed to evidence of similar consequences resulting from other diagnostic tools like MRIs and echocardiograms.

All those false alarms could also raise anxiety levels for patients. An upcoming clinical trial aims to investigate the size of this problem.

Finally, there’s the risk that these technologies will worsen health inequities among patients.

“The Apple Watch uses light that passes through the wrist’s skin to detect AFib. But what if your skin is darker?” warned Rohan Khera, a cardiologist and assistant professor of medicine at Yale, when I asked him about the possible usefulness of smartwatch data.  Khera sees parallels to problems that surfaced early in the pandemic around the use of pulse oximetry in Black people.

None of the Apple Watch studies so far, Khera said, enrolled enough patients with darker skin to determine whether the device’s AFib detection works equally well in all groups.

While I’m personally still excited by the potential of these devices to help patients understand and improve their health, this study reinforced for me just how early we are in this era of patient-generated health data. To fully realize its potential — in a safe, equitable and cost-effective way — we’re going to need a lot more evidence.. Until we have that, it seems best to proceed with a healthy dose of caution.

Tradeoffs’ coverage of diagnostic excellence is funded in part by the Gordon and Betty Moore Foundation.