The Exchange

Commentary and Observations from
the Medical Front Lines

Was the Apple Heart Study Virtually Useless?

Was the Apple Heart Study Virtually Useless?

My wife absolutely loves her Apple Watch and would admit that it is like her crack cocaine: she feels compelled to close all of her rings daily, even when we are on vacation. Clearly, it can be a great motivator for many. But can it also be a tool for preventive medicine?

When Apple included in the watch the capability of providing public health apps in detecting things such as falls and irregular pulses, many began to ponder its actual utility. Stanford University researchers asked if the watch could detect and notify the user of atrial fibrillation (AFib) based upon an algorithm, and the Apple Heart Study was implemented and enrolled nearly 420,000 people in less than one year, a remarkable accomplishment. In short, the answer is, yes, the algorithm and watch can detect irregular pulses. But is this just a gadget or is this a real medical tool with significant utility? Like it or not, consumer-driven health “gadgets” are not only here, but are likely here to stay.

Many are arguing that this trial will provide only a glimpse into what the most recent wearable devices can do. The trial had numerous flaws that would be unacceptable in traditional rigorous clinical trials. The selection bias was huge: who was enrolled? This might be better answered by asking who wants—and can afford—an Apple Watch? Apple is, by some accounts, a plaything for the young and affluent. Participants tended to be young (more than 80% under 55 years) and white (more than 70%), and this likely implies healthy people with regular access to healthcare. As expected, these individuals had low CHADS scores (87%), hence a low risk for AFib. Also, Apple Watches are typically recharged overnight, meaning that they are worn only about two-thirds of the time, a huge gap in monitoring. Additionally, 15% who signed up for the study had one or more pre-existing episodes of AFib, enriching the likelihood of picking up AFib.

Even with those biases in favor of finding AFib, the results were fairly dismal. There were 2161 individuals (0.5%) who had an irregular pulse detected by the watch, yet only 945 (44%) actually engaged with a provider to follow-up. Even in those who did, and had a wearable ECG patch sent to them, only 450 wore and returned the patch for further analysis. Nearly 80% of those who were notified of an irregular pulse were lost to follow-up, unacceptable in a traditional research study. In the end, 72 individuals (0.84%) had documented AFib, dramatically lower than what is generally accepted as a baseline rate of AFib. One important finding of the Apple Heart Study was that 20% of the AFib discovered was greater than 24 hours in duration.

So while many may claim this to be a new age for medicine, and while there are lessons to be learned for future applications, there is little ground-breaking new information here.

Reference
  • ACC News Story. Apple Heart Study identifies AFib in small group of Apple Watch wearers. www.acc.org/latest-in-cardiology/articles/2019/03/08/15/32/sat-9am-apple-heart-study-acc-2019. Accessed April 5, 2019.

Share this page:

Filed under: Cardiometabolic, Preventive Medicine

Development Widget