Why Doctors Shouldn’t Dismiss the Apple Watch’s New ECG App

Apple recently announced a new feature to the Apple Watch: the latest version will be able to measure heart rhythms and notify patients about abnormal, and potentially harmful, patterns. Doctors, however, are skeptical. Their biggest concern is that the feature hasn’t been rigorously tested and could provide unreliable data, creating a false sense of risk among users and leading patients to ask for unnecessary tests.

While these concerns are valid, doctors shouldn’t be too quick to dismiss the new feature, particularly as it appears amidst growing consumer enthusiasm for wearable devices that measure health behaviors. The Apple Watch has the potential to provide valuable data that benefits the entire health care community.

Monitoring the heart

The ECG app, debuting later this year in the new Apple Watch Series 4, allows users to touch the digital crown to generate an ECG in 30 seconds. ECG stands for electrocardiogram, a recording of the electrical activity of the heart. While the ECG in the Apple Watch is much simpler than one performed in a doctor’s office, it can be used to detect certain abnormal heart rhythms. It’s currently being billed as primarily detecting atrial fibrillation—the most common arrhythmia that affects about 1% of Americans, mostly older adults, putting them at risk for complications such as stroke.

When initially developed, ECGs were exclusively interpreted by physicians, but advances in computing technology have made it possible for computers to interpret ECG data. Although computerized interpretations of ECG data aren’t perfect, they have now been a part of routine health care for decades. A 1991 study showed that computerized algorithms were almost as accurate as cardiologists (and in some cases, better) in detecting cardiac rhythm anomalies on ECGs, and ECG technology has only been improving over the last 30 years.

But atrial fibrillation has been historically difficult to accurately detect with computer algorithms, and the ECG app will be no exception. One study of the Apple Watch ECG showed that about one third of the ECG recordings it took were “noninterpretable” by the algorithm (i.e., the app couldn’t determine what type of heart rhythm was present), and when it was interpretable, the algorithm still made errors.

The FDA has “cleared” the app for commercial use (which is not the same as formally approving it for medical use), but says that it is not intended to provide a diagnosis. It has stated that patients will still need to see their doctor to determine whether they truly have atrial fibrillation, and if so, if they need treatment—a potentially difficult and costly task. The Apple Watch ECG will be wrong in many instances and incorrectly inform patients that they may have atrial fibrillation, which is why doctors are concerned about the stress that the ECG app’s false-positive tests could place on their patients, their practice, and the health care system.

The benefits of more data

While doctors are right to be concerned about a rise of false positive tests, the ECG app could still be hugely beneficial to the health care community. The sheer number of individuals who purchase Apple Watches – an estimated 18 million in 2017 – could generate unprecedented data.

The most obvious benefit is that the watch may detect atrial fibrillation in patients not known to have the disease, which could lead to earlier treatment and reduction of stroke risk. But there are other ways in which the ECG could ultimately be used. In theory, for patients with chest pain, emergency physicians could pull up an ECG taken by a patient before that patient arrived to the hospital, helping the physician make a better diagnosis and expediting treatment, perhaps by identifying specific heart rhythms that are seen in conditions like heart attacks. For patients with chest pain, shortness of breath, dizziness, and related conditions, such information could be valuable to physicians.

The data could also be useful in the case of sudden cardiac death – an important, often unheralded, and sometimes addressable cause of death among young adults, precisely the population likely to use the watch. If patients had recorded an ECG before they died, medical examiners could also review the data and gain more insight into the cause of death, including, for example, whether it was a heritable condition.

Aside from utility in single patient encounters, data could be aggregated across watch users, linked to clinical information, and analyzed to better understand how often abnormalities of the heart are the root cause of patient morbidity and mortality. For example, if ECG data and particular ECG abnormalities can be linked to adverse health outcomes, it’s possible that algorithms to detect these abnormal rhythms may be developed. We already know that underlying heart rhythm abnormalities can signal downstream adverse health outcomes – like heart attacks, sudden cardiac death, mortality from lung clots, etc. – but if data on these outcomes can be linked to ECG data from the Apple Watch, it is possible that these adverse health outcomes could be predicted.

Collecting and sharing large amounts of health data can benefit everyone. We’ve seen this in other areas of health. In 2012, for example, The Global Lung Function Initiative, released results from a database of almost 100,000 normal breathing tests from patients of different age, sex, height, and ethnic background, allowing doctors to make better, more personalized diagnoses for patients with shortness of breath. The Apple Watch could record millions of ECGs to build a massive database spanning different populations, allowing for a future with highly personalized ECG interpretation.

Researchers could also use ECGs collected from Apple Watches to study changes in the heart’s electrical activity in large numbers of patients who have other non-cardiac diseases, who are taking specific medications, who live in certain areas, who eat certain diets, and so on—the possibilities are difficult to predict. New algorithms could detect new patterns or abnormalities that are too small or too subtle for human interpreters to see, but that can be picked up by a computer, allowing physicians to find early signs of disease they currently cannot detect.

The Apple Heart Study, which is ongoing at Stanford University and will end in 2019, will look at data collected from older models of the Apple Watch that detect heart rate but do not collect ECGs. Future studies that include ECG data from the newest Apple Watch could take years to perform but have the potential to provide useful information and ultimately improve patient care.

Mobile devices, apps, and other ways to measure health behaviors are growing rapidly. Doctors are going to have grapple not only with the Apple Watch’s ECG app, but with various other technologies that patients are using to monitor their health. As doctors reasonably question whether these technologies will lead to unnecessary doctor’s visits, testing, and patient anxiety, it’s important to recognize the opportunities and to push for the creation and sharing of the unique data that they generate.