Finding True COVID-19 Infection and Fatality Rates, Mid-April

VirtualWright Covid-19

What are the true fatality and infection rates of the Wuhan virus? I think we can come to some pretty fair conclusions based on the data we have in the U.S. and it’s right in line with what the worldwide data have been showing for some time.

Up to now, the only number we could have some confidence in was the number of deaths, although even that is changing. As of today, the ration of deaths to cases is 4.24%. For the past couple of weeks, the total case and the fatality rate have been rising in lock step linear fashion.

I can only conclude that we are way behind in testing; in fact, I think it’s pretty clear that we are, with few exceptions, only testing people sick enough to be admitted to the hospital, especially in the big cities where the great majority of cases have been (some exceptions being doctors, nurses, police, firefighters, politicians, celebrities). I have a relative in the Chicago area who had all the symptoms (cough, high fever, difficulty breathing) but the doctors didn’t test him and he recovered at home. They said they knew it was COVID-19.

Here’s the number of cases plotted logarithmically:

It sure looks like the total number of cases is going to top out at around 1,000,000. I also expect around 33,000 total deaths as described previously.
If these numbers are accurate it would mean a fatality rate of 3.3% of those hospitalized. We also have a pretty good idea from worldwide data and doctors treating people that probably only 20% of those infected get symptoms serious enough to be hospitalized. If true, it would mean that the fatality rate based on those getting the disease is only 0.7% (0.2 x 3.3). That would be about the same as a bad outbreak of the flu, something Dr. Fauci himself speculated March 16 in the New England Journal of Medicine.

What about the infection rate? This is clearly much worse than the flu, otherwise we wouldn’t have seen this explosion of hospitalizations.

The maximum death rate was about 1,800 per day about 4 weeks after the president and states started shutting everything down. The average incubation period is around 5 days and the typical time to death is about 3 weeks after onset of symptoms which shows that the shutdowns worked and that the 1,800 deaths per day rate was the rate of infections without being influenced by the shutdown. We hit it with a sledge hammer. Or did we? Wouldn’t we expect to see a sharp deviation in the curve at this time (day 41 below)? Is it possible it’s all a coincidence and that the shutdown had no effect at all? No sharp deviations or obvious inflection anywhere. Hmmm….

Here’s the lasted CDC report of flu and pneumonia showing the big spike due to COVID-19. And note how similar it looks to the 2017-2018 flu season. They report flu and pneumonia together because secondary pneumonia is the big killer if you get the flu (and COVID-19):

How does it compare to the flu? The ramp up in the flu this season prior to the COVID-19 showed a maximum death rate due to flu and pneumonia of about 556 deaths/day:

So that means it looks like COVID-19 is 1,800 / 556, or ~3.2 times as infectious as the typical seasonal flu with a fatality rate as high as 0.7%. The flu is estimated to have an infection rate of about 1, which means that if you have it, on average you infect 1 other person. So with COVID-19 it looks like you on average might infect 3.2 people.

But the only way to really know the infection and fatality rates is to do widespread antibody testing which would show how many people actually had the disease. This is different than the current testing which only shows who currently has the disease.


Richard J. Wright, a chemist and information technology specialist, has been tracking the official Covid-19 (Corona-SARS) data and assessing our government’s disease progression modeling, plus responses, at virtualwright.com.

Speak Your Mind