WHAT THIS POST IS ABOUT. I comment on the results of Stanford’s Apple Heart Study. What’s going on is much more than the accuracy of a watch.


To borrow the vernacular of popular Fitbits and other health-focused “wearables”, the recently-released results from the Stanford Apple Heart Study are just the first “10,000 steps” toward truly understanding the promising world of consumer medical and healthcare devices.

In this particular case, the Stanford Apple Heart Study, funded by Apple, was chartered to determine the accuracy of the Apple Watch in detecting irregular heart rhythms (FAQ). Simple enough.

Depending on the media account, the Stanford study revealed a “mixed bag” (Wall Street Journal) and the Apple Watch “might not live up to the hype” (Business Insider). Other reports were just as tepid.

While it is important that the Apple Watch is accurate, especially when considering what’s a stake, that is literally the first “10,000 steps” toward a much bigger picture.

My “Participation”

I have to be careful here because I might be full of shit.

Credit: Stanford Medicine

I didn’t participate in the Stanford study, which attracted nearly 420,000 American adults who owned an Apple Watch and an iPhone. The study began in 2017 and concluded in August 2018.

At the time, I was ineligible to participate because I owned a Series 0 (early original) Apple Watch, not a Fall 2016-and-beyond watch with the more advanced optical technology and newer WatchOS. I chose not to upgrade because, by the start of the study, I was waiting for the rumored Series 4 (with ECG capability) to be released.

My participation didn’t really matter to me since I’ve been using the available technology in an Apple Watch from the beginning. I’ve checked heart rates. I’ve recorded numbers, steps, calories, whatever data I can find (and, yes, not all of this would be germane to the Stanford study).

Out of curiosity, at every single opportunity, I tested in real time my watch’s results against the equipment in doctor’s offices and hospitals. How did my recordings from the watch stack up against their measurements? I won’t get into it now, but suffice to say “surprisingly well.”

Enough that the information generated from the Apple Watch, how I collected it, interpreted it, and shared and discussed it with cardiologists and electrophysiologists figured significantly into three potentially life-saving procedures (which I’ll discuss later).

For some quick background, I was diagnosed in my 30s with an irregular heart rate, atrial fibrillation, and have been learning to live with it (and the just-as-flummoxing atrial flutter) ever since.

I feel I’ve done more behind-the-scenes work than most people with Afib or Aflutter and have pushed the consumer medical/healthcare technology further than most doctors and professionals.

The Stanford Study

Back to the Stanford study.

Business Insider contends that the study’s findings “call into question the Apple Watch’s accuracy and effectiveness as a medical tool.” It also noted that results revealed the watch did detect Afib “but the majority of warning notifications ended up being false alarms” and “about 2,000 participants were notified about irregular heart rhythms.”

Of the 2,000, 450 participants received and returned “EKG patches” (ePatch) to confirm their Afib. (Again, the FAQ.) I’m not sure what these patches were, exactly, or how they worked. I assume maybe a 12-or-24-hour wearable heart monitor?

In any event, Business Insider stated that “only 34 percent of the people who returned the patches were confirmed to have Afib — a far cry from the accuracy needed to deem the Watch a worthy substitute for existing medical devices.” (My bold italics.)

This screenshot is from the Series 4. The Stanford study used early watches, but the result is similar: Participants got notifications of Afib.

The Wall Street Journal cited the same statistics and concluded: “The remaining two-thirds had no confirmed atrial fibrillation during the time they wore the patches — raising questions about the watch’s accuracy.” (Again, my bold italics.)

Doesn’t look promising, does it?

An Important Question

How was the Apple Watch’s accuracy tested?

It’s more than reading the FAQ. Why? Afib occurs when the upper chambers, the atria, beat irregularly. If the atria beat regularly but super fast, like a hummingbird’s wings, that’s Aflutter — not the focus of the Stanford study but another irregular heart rate that can be unofficially “detected” by the Apple Watch (read my account here).

Afib and Aflutter are immensely frustrating and mysterious because they come and go. Afib may be detected but not show up again for years. Or it shows up on a Sunday afternoon but not again for another month. The Apple Watch may sense atrial fibrillation, send a notification, but it may not occur again for another hour, day, week, month, year, or 10 minutes.

I assume Stanford took this erratic behavior into account when developing the study. Once a participant received a notification of atrial fibrillation, he or she would then contact whomever to set in motion the next phase — using the ePatch to record more heart rhythms and, ultimately, determine whether the individual actually has Afib.

I also assume the participant, upon receiving an atrial fibrillation notification, could bypass the study and go directly to his or her doctor to be examined, which probably impacted the final results of the study. Who knows.

All of this is to say: The Apple Watch is just a tool. The device itself and the technology is still in its infancy and — like all consumer medical/healthcare devices — is expected to improve as people use them and technology matures.

So it’s not surprising at all that the Stanford study returned less-than-stellar numbers and publications warn that “people should be cautious about relying on the technology as a diagnostic tool,” so says the WSJ.

My Disappointment

What I find disappointing is that Stanford (apparently), maybe Apple, and those interpreting the results are not seeing the bigger picture: that the Apple Watch and its ability to detect irregular heart rhythms is actually a process that needs to be explained and improved. This can actually be said about any consumer medical or healthcare device like Fitbits and products from Withings.

What I mean by process I’ll explain soon. So much to write about.

For now, I take some comfort in the words of Sumbul Desai, a Stanford professor and Vice President of Health at Apple, who was quoted recently by the WSJ saying: “I view the Apple Heart Study as just a first step.”

The first 10,000 of many, many more to come.


If you made it this far, thank you.

Apple Watch & Afib Timeline

February 28, 2019

This connects where I started, I (Heart) Notion, with Your (Apple) Watch is Correct as together I was given a surprising diagnosis. I had Aflutter, not Afib, and suddenly it all made sense.

February 21, 2019

This focuses on a recent opportunity to test the accuracy of my Series 4 Apple Watch with professional equipment used in a doctor’s office and hospital.

February 19, 2019

How I started using the app Notion to collect, organize, and present the huge amount of data produced by the Apple Watch and supporting apps.

Leave a Comment

Your email address will not be published. Required fields are marked *