Dr. Snyder: Wearables are an area that's starting to have power, but even the medical establishment is still fairly resistant to wearables, I would argue, but they're gaining traction.

They started out as fitness trackers and the way we got involved was when they started out as fitness trackers, Apple Watch didn't exist. We said well, gosh, these are pretty good you know, physiological monitors they're measuring, as you say, heart rate, skin temperature, more limited number of things at that time. Now they measure all kinds of things heart rate, heart rate variability, respiration, even your blood oxygen, even though it's not accurate, it measures your changes pretty good, and blood pressure. Yeah, all kinds of different things you can get from a smartwatch and many of those same things from a ring, the Oura ring, for example.

So they're very, very powerful. And if you think about it, they're measuring you 24/7 and they're taking hundreds of thousands of measurements every day. Some of them will make millions of measurements every day. And so they're really getting a detailed view of your physiology and we think that's super powerful. And so what we discovered early on, we put these on folks, and once again, that's probably the second most important thing I learned from monitoring myself was when I got Lyme disease. I figured out from a smartwatch and something called a pulse ox, it measures your blood oxygen. And the story there, if you'd like to hear it, was I was...you can tell I measure everything. I mean, I'm wearing four smartwatches right now. I normally have a ring, but I lost it. And glucose monitors, the whole shebang.

So anyway, with these smartwatches...or sorry, with the Lyme disease, I was helping my brother put up fences in rural Massachusetts. And then two weeks later, I was flying to Norway through Frankfurt, actually, and on this last flight, I measured myself as you know. I always wear pulse ox because it turns out your blood oxygen drops on airline flights. Most people don't know that, most pilots do, most flight attendants don't. But anyway, your blood oxygen does drop. It's not well documented how much it drops we've now documented all that. But it was pretty clear to me when I flew from Frankfurt to Norway, my blood oxygen dropped abnormally low, it dropped at 90 when it would normally drop on that kind of a plane to 96. And it never came back to normal and I saw my heart rate was running high. And then when...yeah, and same thing, it stayed high. And then I later learned my skin temperature was up too, all measured from my smartwatch.

And then I had no symptoms at that point but a day later, I started getting some mild symptoms and they didn't go away. So I went to a doctor in Norway, I warned him it might be Lyme because of the timing was two weeks. And he drew blood. So my immune cells, called monocytes, were up. He said, "Yep, we've got a bacterial infection." He recommended I take penicillin. I said no, I need doxycycline, which is what you use for Lyme. Was a little tense moment there because, you know, doctors don't like their patients telling them what to do. And he was no exception, that's for sure. He did give in because I was about to go above the Arctic Circle, and I did not want to be sick.

And he gave it, and it cleared up right away, you do take it for two weeks. And when I got back, I got tested. And sure enough, I was Lyme-positive, my antibodies even had some antigen, some proteins from the Lyme that were still there. And it's well-controlled experiment because I had given blood three days before I left. And sure enough, I was negative so I seroconverted during that time. So the key part of that story is actually figured out when I was first getting sick from my smartwatch and a pulse ox and it was how I detected my Lyme before symptoms.

So with that, we realized that these are really powerful devices for measuring when you get ill. And so I had two years of data at that point on my smartwatch. And so we went and looked over all the data. And it turns out I was ill four times during that period. One was the Lyme case, two times were viral infection, and the fourth time I was asymptomatic, but I know I was sick because there's a protein called C-reactive protein that was elevated, and it was just as high as my viral infections and Lyme. So there were four times I was ill. We looked at the data and every single time, my resting heart rate and my skin temperature was up and was up before the symptoms appeared.

So we wrote an algorithm...it works for heart rate, doesn't work for skin temperature. We wrote an algorithm that follows your baseline and looks for a jump in your heart rate, resting heart rate I should emphasize, and it worked. So retrospectively, we could show every single time I got ill my heart rate jumped up early and it was advanced symptoms. It worked on me and it worked on three other people who also were ill and were wearing the same smartwatch. One of them got sick twice. Every single time, we could see the jump up in heart rate before symptoms. So that really showed that these are incredible health monitors. And these weren't expensive devices, these were like...especially at that time, I think it was something like a $150 watch that was doing all this. And I know you can do it for a $100 watch.

So as you might imagine that when COVID came along, we had been building out improving our algorithms, building an infrastructure to do this at scale. I'll come back to what that means in a minute. But what we can do now is...and when COVID came along, we quickly enrolled, opened up the study partner with Fitbit, Garmin, and launched a study to try and first show if we could detect COVID with a smartwatch. And then the second part, which we're in now, is alerting people if their heart rate goes up.

And so what we're doing for the first part, we showed that with a smartwatch, we can see people's resting heart rate jump up. We had 32 people wearing a Fitbit. At the same time, they had COVID and they had a diagnosis date and a symptom date. And we showed that for 26 of the 32, we could detect the jump up in resting heart rate, and in basically in nearly all cases it was at or before symptoms. So 81% of the time, we can see people's resting heart rate jump up with COVID. And then we have now...and it turns out, by the way, the very first case we had, it was 10 days early. Somebody...is a very clear signal, you can see this person's heart rate jumped up 10 days early. On average, it's about four days. So people's heart rate will jump up four days before their symptoms if they have COVID, it turns out, and that we can pick up with a smartwatch.

So we're now...we've written some algorithms to do this in real-time, comes back to what I was telling you before, so you follow your baseline and you look for this elevation in resting heart rate. We have three different algorithms to do this. And so we'll profile you...I don't know if you've signed up for this study yet, Rhonda, I hope you have. So anyway, it builds an hour-by-hour measurement of you. And when you jump up, you know, pretty high for extended periods, say 6...these days to keep the false positive, it's more like 12 or 24 hours. If you're up for a while and it's statistically, you know, unlikely, it's not a random fluctuation, it's up there, we send an alert. And it turns out it works pretty well. So 70% of the time, we can detect illness.

We had 63 people as of the end of January who had COVID, and 44 of them, we could detect in real-time and alert them before or at the time of symptoms. So it's not perfect, it needs a lot of tuning still, that's why we want people to enroll in our study, we want to basically improve the algorithms. And we don't...it's not just specific for COVID -- other illnesses will trigger it. And also other things will trigger it, like too much alcohol, not a drink or two, but if you really tie one on, you know, it will send your heart rate up. Hiking in the mountains will do it too. So you have to contextualize it. But it does work in general for detecting respiratory illnesses. And so with improvement in algorithms, I want to be able to tell the difference between, you know, drinking too much versus the respiratory viral infection. I don't yet know if we'll tell the difference between flu and COVID, we'll have to see.

Dr. Patrick: Where can people go to sign up for this COVID tracking study with their wearables?

Dr. Snyder: Yeah, go to innovations.stanford.edu/wearables. So innovations.stanford.edu/wearables. And we'd love to have you enroll, and it does work.