If you read this post from Wednesday about using smart thermometers to insta-detect local outbreaks of COVID-19, you already know exactly what this is. If you didn’t, stop here and read it. It’s worth it.
Here’s their new website, as promised: healthweather.us. Bookmark it. All of the data on the map comes from Kinsa smart thermometers, which log temperatures of users and then relay that data to Kinsa HQ to be aggregated and logged. The company can “see” clusters of fever among its customers instantly, in other words. And now we can too.
You can use the map to view “observed illness” or “atypical illness,” which compares the current rate of fever to what the rate is in normal times based on historical data. (Scroll down beneath the map and you’ll see a graph showing just how far the current rate deviates from the normal rate.) You can also search by zip code or county to see what your own backyard looks like, or view the whole United States if you like.
New York is the focus of most of the country’s attention because of the soaring number of positive tests there, but if Kinsa is right then we have a big — big — problem brewing in Florida. Which is the last place anyone wants a big coronavirus problem given the disproportionate number of senior citizens there.
One problem with the map is that there’s insufficient data for many counties. Not everyone has a Kinsa at home, of course, and poorer places may be less likely to have them since they run a little more expensive than the cheapest options on the market. I assume the company’s customers also skew younger since older people who are less tech-savvy might shy away from smart thermometers. Nor does the map (yet) have a feature that lets you see growth in atypical illness in real time. A week from now, it’d be nice to be able to view at a glance which counties have experienced an increase in atypical illness over the past seven days.
The little graph beneath the map does show change over time, though. Check this out:
King County was the site of one of the biggest outbreaks in America earlier this month. You can see the rate of illness rise out of the “typical” range … and then begin to fall, to the point where it’s now tracking back in range. I repeat my question from yesterday: Has Seattle flattened the curve?
Here’s the data for Manhattan. As you can see, the rate of illness is very much atypical — but:
That rate is coming down. I checked the other four boroughs of NYC and they’re all following the same trend, atypical but declining back towards typical rates in the past few days. Social distancing seems to be working. For now.
The Kinsa site could be a valuable resource, and the more people who use their product, the greater it’ll be. Good luck finding thermometers for sale anywhere right now, though. I’ll leave you with this thought: According to the Johns Hopkins COVID-19 tracking website, there are a little over 14,000 confirmed cases in the United States. According to the best estimates of modelers, though, there were something like 80,000 actual cases as of three days ago, in line with the official numbers from China.