Trying to Make Sense of Coronavirus Numbers

By March 30, 2020 Commentary

According to the Covid Tracking Project site, link in other posts, in the United States, out of 857,000 coronavirus tests, 142,000 had positive results.  That is a 16.5% positive rate.  New York is very much an outlier in positive rate and there is large variation across the country.  Minnesota’s positive test results have hovered around 3%.  There is likely very wide variation in who is getting tested in various states and some just seem bizarre, like California has a very low number of reported test results.  In any event, two things to keep in mind.  Lots of infected people are not getting tested, probably multiples of the number that are.  Some are asymptomatic, some may have difficulty accessing a test and some just don’t feel sick enough to go get a test.  If more infected people in these groups get tested, the positive percent of results will obviously rise, but the percent of seriously ill will decline.  On the other hand, if we did some population testing, and we really should, we would get a clearer picture of what percent of the population is actually infected, and I suspect it is going to be well under the 16.5% positive rate for current tests.

According to Worldometers, link also in another post, there have been a total of 770,000 cases, with 37,000 deaths.  That looks like a scary 4.8% of cases.  But on a population basis and on a likely infected persons basis it is much less scary.  By “case” they appear to mean a positive test result.  There are 573,000 active cases, of which 29,400, or 5%, are critical or serious, which I believe means hospitalized.  According to this site, the United States has had a total of 155,000 total cases, 148,000 are active and 3,400 are critical.  Our critical rate is only 2.3%.  If my math is accurate, always questionable, if the infection rate were 15% of the population (I don’t think it probably is), we will eventually have 49,500,000 cases.  If the critical illness rate is 2.3% of cases, we would have over a million seriously ill people.  A lot of sick people if it happened all at once, which it obviously won’t.  If you drop the infection rate to a more likely 10%, the number of critically ill people drops by a third.  And if the number of positive test results climbed as a result of widespread testing, we will have more “active” cases, but the percent of serious/critically ill in those total cases will also drop dramatically, I suspect to 1% or less.  So now instead of a million seriously ill people you have 330,000; which is roughly equivalent to the annual number of hospitalizations for flu.  So you can see why more population testing to establish true rates and numbers of infected persons (“cases” at Worldometers) becomes so important to understand what the burden on medical resources will be.

Leave a Reply