Trends in the Spread of COVID-19

Let’s face it. Public Health centered testing for the coronavirus has been a rolling disaster that is slow to improve in the U.S. Testing may improve with the passage of time although at this rate the value of testing will have disappeared entirely. So, for now we’re flying blind. 

No one really knows when the virus landed in the U.S. or how widespread the virus was and is now today. Some of those infected are spreading the coronavirus without themselves showing any symptoms. Truth-be-told, if I’m right about my own health, I may have unintentionally spread the coronavirus myself (here).

Trending Warmer, Hotter

HOWEVER, Kinsa Health, a private sector digital thermometer company (here), is collecting and reporting body temperature across the country (here). I won’t bore you with the details of their data collection and algorithms (here).  But the map below reports on counties and regions across the country where body temperature is higher than expected toward the tail-end of a normal flu season.

Rising Body Temperature Heat Map
Kinsa Health

The colors are self-explanatory. The bright spots on the east coast align with media reports about the greater NYC metro area and points northeast. Florida looks like it could burst into flame at any time. And the trend toward the “cooler” yellows on the west coast affirm the effect social distancing has had in northern California, Washington, and Oregon.

Remember, not everyone with coronavirus presents with a fever. Those who are ill do not present a fever throughout their infection; and some never do. So, the map above is an underestimate of the spread of the virus. In fact, it’s believed a lot of those with the virus never show any symptoms. Consequently, they’ll never appear on these heat maps. Worse yet, we don’t know where they are today and where they may have been—the testing failure, again. It’s believed by public health officials—again, in the absence of any credible testing—that what we can’t see is five-to-ten times worse than what we can. Let’s hope that’s incorrect.

A bit surprising is the southeast, Michigan, and Indiana. The trend toward warmer colors suggests a NYC response, unless social distancing proves effective. That’s a bit more doubtful in the southeast, since state governors in the region have been slow to issue stay-at-home orders.

There’s concerning viral activity in parts of Colorado, Oklahoma, and Kansas—Texas, too.

Trending Cooler

The Kinsa Health map below reports on the most recent seven-day trend. The deeper blue the county or region, a decreasing rate of infection—at least where reported body temperatures are concerned.

Body Temperature Cooling Map
Kinda Health

It helps to live in remote areas where low population density brings a natural version of social distancing. Yet, there are many higher density population areas where actively practiced social distancing must be having a positive effect.

Unfortunately, some of the more remote areas in the northeast along the Canadian border reflect an increasing rate of infection. Indeed, the Atlantic coast from northeastern Massachusetts through Maine reflect a slightly less impressive rate of infection but are still areas of concern.

The region of the country just east of the confluence of the Mississippi and Missouri Rivers is troubling. While many of the parts of the country are trending toward cooler body temperatures, this area is doing the exact opposite. That’s not good for southeastern Illinois, southern Indiana, and the western parts of Kentucky and Tennessee.

Surprisingly, the greater New York City metropolitan area is trending toward a decrease in the spread of the virus—again, based solely on body temperature alone. These maps exclude any infected persons whose symptoms do NOT include an elevated body temperature. 

That Testing Failure, Again

The untracked and unmapped include people who cough, experience difficulty in breathing, for example, but not a fever. And, the map doesn’t reflect the so-called silent spreaders who are asymptomatic—people who look healthy and safe but aren’t.

What we don’t know about this virus and where it’s been, where it is, can still do us serious harm.

Final Comments

Kinsa Health claims to have approximately two million of their digital thermometers in the U.S. For a population of 320 million, that’s a per capita rate of 6.25 thermometers per thousand of population. So, the obvious question is this: Do the Kinsa heat maps reflect a representative sample of U.S. citizens?

First, per capita is per person—not per family. Thermometers are household purchases more than they are individual purchases. Depending on the source of information, the number of households is estimated somewhere in the range of 120 million, give or take a million here, a million there. Assuming no household has more than one of Kinsa’s thermometers, then roughly 1.7 percent of households is using and reporting body temperature data. Are those households representative across the country?

Second, a digital thermometer is not as high a high-tech purchase today as it might have been, say, five years ago. However, to report the data from a household to Kinsa Health requires downloading and using an app on a digital device. That makes a Kinsa digital thermometer a high-tech purchase. And, high-tech purchases skew toward wealthier—relatively speaking—and better educated heads of households. Some might say that favors urban, higher density population centers. “Some” might be wrong, however. For example, farming can be very high-tech involving GPS, crop management systems, and sophisticated software right down to steering a farm tractor as it seeds a crop field. Same for water management for a crop in the field. So, while it is generally true the digital divide separates rural from urban areas, it isn’t a truism that high-tech is hard to find on the ground in, say, Iowa, Nebraska, Kansas, or other rural states. 

At the end of the day, Kinsa Health is engaged in an observational study. It would be helpful if the people who study these things, at the CDC, research organizations, and institutions of higher education, shared their collective opinion on the representative nature (or not) of Kinsa Health’s data and algorithms, and by extension, their heat maps. Until they do, please note my inferences and statements must be taken with at least a dose of skepticism.

I know that I apply the same standard to the writings of others. You live by the sword, you die by that same sword.

Copyright 2020, Howard D. Weiner