The Apple Watch has an FDA approved accessory that can diagnose you with an arrhythmia. But how accurate is it? The evidence says not very. While it’s pretty good at telling you if you’re normal, a diagnosis of atrial fibrillation isn’t accurate at all. Read the full story here. Â https://montrealgazette.com/opinion/opinion-why-the-apple-watch-might-be…
*************************************************
Given that I’m not someone who is overly fond of new gadgets, the arrival of the new Apple Watch 3 didn’t really make that much of an impression on me. But a new feature of this latest Apple Watch is interesting because, unlike other devices, it can measure the electrical activity of your heart.
Heart rate monitors are not new, of course. A study in JAMA Cardiology found that four of the commercially available wrist-worn heart rate monitors (Apple Watch, Fitbit Charge HR, Mio Fuse and Basis Peak) were reasonably accurate at measuring heart rate, although not quite as good as a device that included a chest strap heart rate monitor. The novelty of the Apple Watch, though, is that it doesn’t just measure your heart rate; it also measures the electrical waveform of your heart rhythm and can detect if you are having an arrhythmia.
At least that’s the idea. Researchers have compared the Apple Watch to a standard ECG, which despite being a 100-year old piece of technology defiantly remains the standard of care. In a group of 51 patients who were scheduled for a cardioversion (a procedure to reset the electrical rhythm of the heart with a small electrical shock), the Apple Watch did quite well. Its positive predictive value (you actually having atrial fibrillation when the Apple Watch said you did) was 91 per cent, while its negative predictive value (you having a normal heart rhythm when the Apple Watch interpreted the rhythm as normal) was 98 per cent.
Unfortunately, the Apple Watch performed less well in a larger general cohort of patients. In this group of 1,617 patients, of whom only about four per cent had atrial fibrillation, its results were frankly disappointing. When the Apple watch identified your heart rhythm as normal, it was correct 98 per cent of the time. However, when it identified rhythms as atrial fibrillation, it was correct eight per cent of the time. Therefore, it was wrong 92 per cent of the time. Flipping a coin would have been better.
Tests tend to perform better in sicker patients and less well in healthier ones. The reasons for this are mathematically fascinating, but I will spare you the proof. Suffice it to say, that any time you apply a test to a large number of otherwise healthy people, you have to be ready for some false positives.
Atrial fibrillation is an important condition to diagnose, as giving high-risk patients anti-coagulants can lessen the risk of stroke. But in otherwise young health patients, it remains unclear how much benefit there is in going out to look for it. In a 2007 BMJ study, researchers found that even in a group of just under 15,000 patients, aged 65 and older, systematically screening people for atrial fibrillation was not terribly productive. They found it was just as useful to check someone’s pulse during their doctor’s visit and then send them for an ECG if it was irregular.
The danger with any new medical technology is that in the enthusiasm to be an early adopter, for what is an admittedly cool gadget, we risk engaging in a population-wide screening strategy that is decidedly flawed. In fact, the U.S. Preventative Services Task Force has cautioned against ordering ECGs in low-risk patients because the benefits don’t seem to outweigh the risks, which usually involve ordering further tests for minor abnormalities that are not clinically significant. The problem with the Apple Watch is that with a 92 per cent false positive rate in the general population, it will be sending lots of healthy people running to the doctor thinking they have an arrhythmia, when in fact they do…
