USA Fever Heatmap: Isolation working since Mar 20

In addition to the nation-wide fever numbers, Kinsa tracks numbers by state and county.

Take for example, Forsyth County, GA, due north of Atlanta. The Heat Map shows Forsyth has as many current Atypical Fevers (per capita) as every other county adjacent to Atlanta.

Georgia is not a state in lockdown, and is below normal in testing. Individual counties in GA are in lockdown. This includes South Fulton county which includes Atlanta. It does NOT include the cities of Sandy Springs, Roswell, Alpharetta, Johns Creek, and Milton which comprise North Fulton county, and Forsyth county north of them.

The Fever Heatmap shows the folly of this situation. Politicians are driving thru the rear-view mirror. Atlanta has the most cases Now. But the heatmap suggests all those location mentioned above have as many people (per capita) actively spreading the disease.

The case could be made that lower population density reduces the need for lockdown. But such an argument needs to be made and supported, not assumed. Currently it is not even addressed.

This company Kinsa sells internet connected thermometers. Their "US Health Weather Map" (that I called a Fever Heatmap) shows areas with more fevers than normal. Under the map note the chart of fevers vs Date.

Normally (sans epidemic) these fevers are from flue and colds. On the chart, the blue line extending thru May is the number of fevers expected. The Orange line is the Observed number, and Red shows fevers exceeding the expected range.

This data is recorded by Kinsa's servers as their customers measure their temperature. As such it is 5-10 days (my estimate) ahead of the case numbers reported by doctors and hospitals. This is because people wait until they feel really sick before calling a doctor.

Therefore, these numbers are timely, but harder to interpret. Remember these are all fevers, not just SARS-COV2. Also this is most certainly NOT a random sampling of the population. The sample must be small, and skewed by income and age.

That said, what the chart shows is that in the USA, fevers exceeded the expected range about Mar 9, then fell back into the expected range about Mar 20, and are now below expected.

The conclusion drawn is that these Atypical Fevers represent people sick with Covid-19. And that Social distancing quickly reduced the spread of ALL viruses. Fevers dropped below expectation because isolation reduces the spread of influenza, rhinovirus, and coronavirus.

I have been plotting current cases and deaths for this epidemic all month, hoping to see a break in the exponential growth. That has not yet happened. But this cart is the first good news I've seen.

(Note: You must track Current Cases, not Cumulative Cases to see the real exponential growth. Recovered & dead patients do not spread the disease. So the graph of Cumulative Cases will fall below exponential growth early due to attrition, not effective isolation.)

This has been some of the most interesting data to look at in this whole season. I'm not sure whether I'm right in feeling bolstered by what it's saying but it seems that in most areas, the "measured fever" population is on par or lower than expected -for totally normal situations?- since "social distancing" has been encouraged.

Even if I'm way off in interpreting that trend line, it sure takes the edge off to see someone point the trend line down-and-to-the-right for once