Fitbit Claims COVID Detection Success

Posted on August 20, 2020 by Paul Thurrott in Fitbit with 13 Comments

Fitbit said this week that its fitness wearables could detect COVID-19 in their users before they show any sign of symptoms.

“In May, we announced the launch of the Fitbit COVID-19 study aimed at building an algorithm that detects COVID-19 before symptoms start,” the firm explains. “In just over two months, more than 100,000 Fitbit users across the US and Canada have enrolled, with more than 1,000 positive cases of the virus reported. This study presents an exciting opportunity to see how the power of the Fitbit community will help us better understand this new and complex disease.”

Fitbit has submitted its early research for publication in a peer-reviewed journal, but it is sharing some of the preliminary findings ahead of that publication. Among them is the claim that Fitbit’s devices can detect almost “50 percent of COVID-19 cases one day before participants reported the onset of symptoms with 70 percent specificity.” The firm notes the importance of getting this data so quickly because people can transmit the virus before they have symptoms.

“Our study also reinforces that breathing rate, resting heart rate and heart rate variability (HRV) are all useful metrics for indicating onset of illness and are best tracked at night, when the body is at rest,” Fitbit continues. “Our research shows that HRV, which is the beat-to-beat variation of the heart, often decreases in people who are exhibiting symptoms of illness, while resting heart rate and breathing rate are often elevated. In some cases, those metrics begin to signal changes nearly a week before participants reported symptoms.”

Fitbit also says that it findings confirm some other theories about COVID-19, including that being older, male, or having a high body-mass index (BMI) increases the likelihood of severe outcomes. And that shortness of breath and vomiting are the symptoms most likely to predict that someone with COVID-19 will need to be hospitalized. Sore throat and stomachache were among the symptoms least likely to predict a need for hospitalization, Fitbit says.

Interesting stuff.

Join the discussion!


Don't have a login but want to join the conversation? Become a Thurrott Premium or Basic User to participate

Comments (13)

13 responses to “Fitbit Claims COVID Detection Success”

  1. docpaul

    That's amazing!

  2. TFinch

    Rolled my eyes at first -- Fitbit touts a lot of science but regularly comes up short in controlled research -- but not counting it out until I see the research in print. Let's see what journal accepts the work ("peer reviewed" as a descriptor is increasingly meaningless as industries create their own journals to confuse a public rightfully too busy to track reputable vs shyster research outlets) and how they arrived at their results.

  3. Pbike908

    Wow a 50% success rate. Great PR, however, of little practical value. Unless of course Fitbit was perhaps willing to give FREE devices away....Or perhaps FREE devices to say LOW INCOME folks...Yeah, right...

    • bluvg

      In reply to Pbike908:

      "of little practical value."

      I totally disagree. This is not a COVID test and its accuracy shouldn't be gauged against one (they're nowhere near 100% accurate either). This is a data point, a warning to the wearer, and in aggregate, very useful data that is otherwise difficult or infeasible to obtain.

    • abillimore

      In reply to Pbike908:

      It's not a 50% success rate. They are reporting a 50% detection/prediction rate. So 500 people in this study would potentially get tested and isolate earlier. That has practical value, even if the other 500 slip through.

    • ontariopundit

      In reply to Pbike908:

      You scoff but 50% early detection before even the onset of symptoms would be HUGE. People will not necessarily go to get tested the day symptoms begin. If they have their Fitbit notifying them that they should consider a test that would help curb transmission, and if necessary, allow for much earlier medical intervention.

      That would reduce the likelihood of transmission by a number of days for up to half of all diagnosed cases.

      What I'd like to know is what percentage of false alarms (false positives) does the Fitbit generate? And, on what day of the disease does the Fitbit algorithm actually detect the disease? Is it really one day before onset of symptoms or is it retrospective based on the data that's been collected?

      As for the complaint about low income access to the technology: simply because a diagnostic tool is unavailable to everyone means we shouldn't use it? It's not Fitbit's responsibility to ensure everyone has a Fitbit if it is really that useful, that's society's responsibility!

  4. brandonmills

    This has been true with every fitness tracker ever, for anyone who knows their base heart rate. I'm a regular exercise guy, so I've noticed this for years. You go to the gym, you check your base heart rate, and its abnormally high. Sick every single time. Your body is fighting something, whether you have noticed it yet or not. Noticed the same thing with my Microsoft Bands and Garmin watches. Shocked it is taking people this long to catch on. ( also most people don't know their base heart rates. )

  5. Kevin

    I just want my fitbit to push up the message, "consider time off, you might need it soon".

  6. rfeeley

    Hoping Fitbit's claims have at least some legitimacy. Every little help brings us closer to getting back some normalcy, control.

  7. randallcorn

    So it looks at symptoms and suggests you may have Covid? Then go get a test done to confirm?

    The symptoms are so similar to other things. But at least it will mean go get yourself tested.

    • lvthunder

      In reply to randallcorn:

      You need to read the next to last paragraph again. They are trying to detect it before you get symptoms. The last paragraph said if you have these symptoms you have a higher chance of needing to go to the hospital.

  8. BrianEricFord

    I’m super skeptical about these claims but assuming they’re true one hopes Fitbit treats this as an open source obligation rather than a selling point for their devices.

    Edit to add that my suspicion is that this is going to turn out to be a “correlation is not causation” situation.

  9. tobyburnett41

    I've had a Fitbit for almost 2 years, and can attest that my resting heart rate, which it measures daily, tracks my health and well-being very well. I signed up for this study, and I'm now hoping that they will make some of these other metrics available to users as well.

    I've asked several medical professionals if they were aware of this indicator, which is much more sensitive than body temperature, and gotten blank stares--it they can't measure it, it doesn't exist.