Can Your FitBit Detect Covid?

This does sound interesting.

A few atypical complications come to my atypical mind. Will it detect early infections your body can just naturally fend off? Can it be made to ignore such benign infections? Can it differentiate those infections you can naturally fight from those which will just make you miserable or those that might maim or kill you? Will hypochondria run rampant? Will hypochondriacs lose the remainder of their immune system?

It struck me because I had not heard of anyone else going this way. Not being a Dr. Fauci I couldn't tell you that much about the bodies specific immune response to a given infection. Maybe they are all the same. Maybe the body can differentiate. Is our sensor tech up to the task? Can machine learning AI learn to tell things apart?

That was what I thought was cool is this is looking at working With our immune system, trying to read what its telling us.

At least someone is looking at this angle and put the initial money up for the research and published it for review so others could start looking down this path as well. Exudes confidence and initiative. Does not scream I'm doing this for the money.

I learned something new!
The fact that the body is infected and developing an immune response up to a week before we "feel" it was a new thing to me. I just figured it was much faster.

I have always been a proponent of working with the bodies already amazing abilities.

Battles are waged, won and lost under our skin that we oftentimes have no idea they even happened. Factories churning out war machines to go do battle, invaders raping and pillaging the landscapes, building their forces . . . .

O.K., maybe I'm being a little melodramatic.

Yes, but I like it! Great visual.

my forces are led by general De Gaul Bladder.

The most useful differentiation would certainly be telling if the infection was contagious. I doubt the bio-signs alone can do that differentiation but combining Bluetooth proximity contact data to a medical database might make this possible. A highly contagious disease, measles, where the pathogen can viably linger hours after the infected spreader leaves a location could not be identified by these devices.

Getting everyone to wear a Bluetooth tracking device does have some Orwellian overtones. Then again if you include Candy Crush or Angry Birds for free...

I don't need some expensive watch to tell me if I'm sick or not....and even the laboratory tests are not conclusive....useless

Competing Interest Statement

All authors are employed at Fitbit Inc.

Funding Statement

All authors were funded by Fitbit Inc.

There's one plus side to the USA having so many infections, there is heaps of data to go over.

For Covid where testing can be (was/is) in short supply and sometimes takes days to get results something like this could have been useful in helping to determine who gets tested and whether their sample moves up in testing order.

Remember early on when only symptomatic patients could get tested? Meanwhile asymptomatic people were passing the virus on unawares. Someone health conscious enough to wear a fitbit could be made aware they are potentially infected (with something) and take steps to protect themselves and others.

A lot more data needs to be studied to refine the diagnostic tool but it seems worthwhile to me.

The way we tell here if you're sick, is you don't want to hit the surf....you must be sick...

Beach Report: No change....No rising Sea Level to report...70 years, no change...

I have several worries with this theme. The first is that the studies have been conducted on Covid patients without controls, so there is no way of knowing whether other fevers and infections also reduce heart rate variability (false positive Covid diagnosis). The second is that pre-existing diseases, such as diabetes and heart failure, may of themselves reduce HRV, and the Covid effect would not be noticed (false negative). Given that there were only just over 1000 patients in the study, reporting statistics to 3 decimal places seems an unjustified precision.

I get the feeling that they went as far as they could and are now putting it out for those with more expertise to possibly take it to the next step. I found it interesting since it was using a very different approach than the silver bullet/vaccine folks. It acknowledges that we may not ever have a vaccine and will need to learn to manage it in the population.

There's a bit of going off at tangents here. The reported study had nothing to do with vaccines, nor indeed with any aspect of treatment. All it did was present a possible way of making a diagnosis. The evidence that it is an effective and reliable method of making a diagnosis is along the lines of "We've got a gadget that measures this. I wonder if it can measure that?" As a scientific study it's rubbish.

Indeed the vaccine folk are using a different approach. It's based on established scientific method. They are testing a variety of vaccines in 3 phases on thousands of experimental subjects. One or two of that variety of vaccines will be good.

I think you may have missed the point. No one is presenting this as anything more than an interesting new possibility that has never been fully explored. I find it encouraging because it takes on the unpleasant task of looking for answers to the questions that will come when we can't eradicate or vaccinate against this and have to learn how to control and manage it within our society. That possibility is statistically more likely than a vaccine btw.

I also am quite impressed that they took it this far with their (fitbits) limited abilities and then turned it public so others better equipped and suited to the task can take it up instead of trying to monetize it like the big pharma folks are doing.

That whole Midwestern ethos about helping your neighbor just because its the right thing to do and not putting all your eggs in one (vaccine) basket. Granted it doesnt seem like many of us are actually left in the midwest. Sigh.

Oh dear. Let us take this step by step. The claim is that the Fitbit can possibly detect Covid-19 before symptoms appear, by noting changes in heart rate variability. This is a very bold claim, given that the Fitbit is one of the less accurate wearables when it comes to measuring the basic heart rate, see here .

Now if it's bad at overall heart rate what chance does it have of providing an accurate heart rate variability?

When you have sorted that out, then you have to ask whether a reduced HRV is diagnostic of Covid-19 infection. Well, it isn't, it may accompany any number of other infections, so as a measure of the likelihood of Covid infection it's useless.

Now I must question your assertion that the possibility that you cannot eradicate or vaccinate against Covid is "statistically" more likely than a successful vaccine. Where on earth did you get that from? Would you like to give me the actual statistical value and its source?

And do tell me what the whole Midwestern ethos has for alternative egg baskets. Are we going to smoke weed as an alternative to vaccination or is it back to Donald Trump's hydroxychloroquine?