Skip to Main Content

NEW ORLEANS — New results of a gigantic study testing how well the Apple Watch detects important heart-rhythm changes offer a tantalizing glimpse of how consumer devices might change medicine and lead to new ways to conduct science — but little more.

Unfortunately, the data don’t tell us much about whether anyone should buy an Apple Watch, or whether the device is good for the population at large.


“At least it is not an epidemic of false alarms,” said Dr. Eric Topol, director and founder of the Scripps Research Translational Institute.

The Apple Heart Study is being released at the annual meeting of the American College of Cardiology here. Full results will be presented before thousands of doctors at the opening of the convention Saturday morning. STAT was given a press release in advance containing some data from the study, which was shared with the researchers who commented for this story.

The study recruited 419,297 people who had an Apple Watch. It tested technology in the watch that uses an optical sensor to detect the wearer’s pulse. Of the cohort, 0.5 percent — or 2,161 people — were notified that they might have atrial fibrillation, a relatively common arrhythmia that can lead to feelings of weakness and increases the risk of stroke and other conditions.


“The important thing here is that only 0.5 percent of people were getting notified,” said Dr. Marco Perez, an associate professor at Stanford University and one of the lead investigators of the study.  “If you look at the young people, it was 0.16 percent. The notification rates are really low. These are needles in a haystack.”

What happened next: 945 people reported the notification to the researchers and had a visit with a telemedicine company, American Well. Some were told to go to see a doctor immediately, or excluded because they already had been diagnosed with atrial fibrillation. Then 658 had a patch shipped to them that could perform an electrocardiogram, or EKG, and 450 patients actually used the EKG patch. When the watch again signaled an atrial fibrillation rhythm, it agreed with the EKG 71 percent of the time. But the watch would not have notified the patient until it received five such rhythms; such notifications matched up with the EKG patch 84 percent of the time. (Apple announced EKG capabilities for the watch last year, but these were not part of the study.)

“The technology has arrived. I am not one of those cardiologists that are blaming Apple for giving us too much information,” said Dr. Rajat Deo, an associate professor of medicine at the University of Pennsylvania. It’s doctors’ job, he said, to figure out how to use this information. He said that 84 percent rate was low enough to be “concerning,” but he expects it to improve over time. The study, therefore, represents a promise of what might be — but it is not yet a device doctors should trust on its own.

“We are not talking about a medical-grade technology, we’re talking about a fun technology we go running with to check our heart rate,” Deo said. “I would caution clinicians not to act based on Apple Watch data alone.”

That mix of hope and skepticism was generally the reaction of doctors given a preview of the data. Some were harsher: Dr. Steven Nissen, the chairman of cardiology at the Cleveland Clinic, worried that the data are still “unpublished, unverified, unreviewed, and not tested by peers.”

And the unknowns could be expensive.

“How much money is it going to cost the health care system to follow up on thousands of patients who have an incidental atrial fibrillation?” Nissen asked. This study doesn’t provide an answer, but it is not an idle question: Millions of watches have been sold. And treatments given to patients with atrial fibrillation are not without risk.

If the 84 percent concurrence rate holds and the population in the study is similar to that of Apple Watch buyers (that is a very big “if”), one would expect that for every 400,000 watch buyers, there would be roughly 2,000 people notified they have atrial fibrillation. Of those, 1,600 would have atrial fibrillation, and 400 would not.

“If we had everybody go to CVS or Walgreens and get a blood pressure check, we’d probably save a lot more lives and it would cost a lot less money,” said Dr. Wilber Su, chief of cardiac electrophysiology at Banner Health in Phoenix.

Su said he’s seen about a hundred patients who have come to him after getting a scary result from the Apple Watch, but he doesn’t remember any of them needing a major intervention. He could foresee using the watch to track some patients who have received treatment for atrial fibrillation.

“Most of us have mixed feelings about it,” Su said. “It’s not uncommon for patients to go into complete panic thinking they’re going to have a stroke the next moment because this is positive.”

But for many experts, the study isn’t really about whether the Apple Watch can have an impact on public health now. It’s about embarking on an entirely new way of doing clinical research and practicing medicine someday.

Dr. Jordan Safirstein, an interventional cardiologist at Morristown Medical Center, said the result was “amazing,” not because of the results, but because researchers were able to recruit a staggering 419,000 patients in a very short period of time, and follow up on them only with telemedicine.

“It’s a lovely demonstration of how we can enroll patients completely virtually, which I think is a huge extension for us in terms of patient engagement and enrollment,” Safirstein said.

Dr. Mintu Turakhia, one of the Stanford University researchers who designed and conducted the study with Apple, said this was one of the major points of the study: not to reach an immediate conclusion on the Apple Watch — even there, follow-up data on exactly how patients were treated will be needed — but to learn how to use wearable devices in medical research. “We’ve enrolled 400,000 people in just over eight months,” Turakhia said. He hopes that what they learned will allow them to do more ambitious studies in the future.

“I don’t think any of us view this study as the end in terms of using apps and devices to monitor cardiac performance or any number of other things,” said Dr. Lloyd Minor, dean of the Stanford University School of Medicine. (He owns a small amount of Apple stock.) “But we think it’s an exciting beginning.”

Consumer-facing technologies, Minor argued, could change medicine just as they have changed every other aspect of our lives. But that still doesn’t tell you whether it’s a good idea to buy an Apple Watch.

  • Apple should designed an alarm in the watch that can alert the patient by sound and also direct the abnormality to the patient’s iPhone and then directly to the patient’s cardiologist who can then read the EKG and notify the patient promptly to seek emergency help.

    This would help the patient, the cardiologist and emergency room, as well as be beneficial to the investment and finances of Apple.

    I see this watch as beneficial to any stress environment in normal healthy individuals such as athletes as well as people who work in stressful situations such as police officers.

    As a physician, I certainly would enjoy being involved in the further development of such a sophisticated and helpful contribution to health care.

    I also think a research grant from a federal agency would stimulate further research and development and probably even lower healthcare and insurance costs.

    Respectively submitted,
    Alan Knopf MD
    [email protected]

    • I am 67 and have been diagnosed with atrial fibrillation and take daily medication for this. My doctor recommended I purchase a kardia, which I did to verify if I am having afib or not, for me I know because my heart rate runs 130 and higher. It would be great however to have a watch that keeps up with and lets you know when it appears symptoms are beginning to exhibit themselves. My kardia is always spot on every time and I can take my reading (strips) in for my physician to see. Please Apple Watch get this technology down. 💕

  • Thank you, Mr. Herper. Solid reporting. As we embark, inevitably, on the journey to more “wearable devices” their implications for clinical decision-making will evolve and mature. Time for Cleveland Clinic to be less of a stick in the mud and more of a leader on this journey.

  • As I’ve said before, the Apple Watch 1, 2, and 3 used the optical sensor for pulse. Pulse is not EKG. You can have congestive heart failure and have a normal EKG. The Apple Watch 4 has an actual EKG, which should provide much more reliable information about the electrical activity of the heart, which is what you want for detecting atrial fibrillation.

    Another important difference is the optical sensor used in the older Apple Watch versions was operating all of the time. The EKG in the Apple Watch 4 only operates when the user initiates a reading, as would occur when you feel something in your chest that seems concerning. That’s good for confirming atrial fibrillation in people who can feel these events, but it doesn’t do anything for those cases where the feeling is below perception.

Comments are closed.