Fitbit Study: Devices Can Identify COVID-19 Early

An August 19 blog post on Fitbit’s website discussed preliminary results of the company’s COVID-19 study suggesting that Fitbit devices can detect signs of the disease before symptoms are noticed. The 100,000-person study found more than 1,100 positive cases of COVID-19, of which nearly 50% were detected at least one day before participants reported symptoms.

The preliminary data suggest that breathing rate, resting heart rate, and heart rate variability are useful metrics for indicating the onset of COVID-19. Fitbit trackers showed that heart rate variability often decreased in people with symptoms of illness, while resting heart rate and breathing rate were often elevated. In some cases, early warning signs were detected nearly a week before symptoms were reported. Although Fitbit’s algorithm produced a 30% “false positive” rate, alerting users to the need to get tested sooner would allow those infected to isolate and potentially reduce the spread of the disease.

The study also offered insight into common symptoms, severity, duration of illness, and the symptoms most likely linked to hospitalization. The most commonly reported symptom was fatigue, followed by headache, body aches, decrease in taste and smell, and cough. Mild cases had a median duration of 8 days; moderate cases had a median duration of 15 days; and in severe cases requiring hospitalization, the median duration of illness was about 24 days, with a wide variance.

Fitbit is continuing to work with its research partners to further validate the technology and hopes to make the COVID-19 algorithm available to consumers as soon as possible.

This post was written by Joseph P. DiBella, a law school graduate who is not yet admitted to the New York State Bar.