Covid-19 Mortality Rate Data
31 March 2020 10:40am - initial revision
The mortality rate, or case fatality rate (CFR), of Covid-19 may be vastly different from what is being reported in the media. I am not a doctor, but I am a data science professional and I understand basic math. Anyone can see this.
The formula the WHO and CDC are using, and what is being reported in the media is cumulative current total deaths / current confirmed cases.(source: https://www.worldometers.info/coronavirus/coronavirus-death-rate/) So, as of 3/31/202 the worldwide number would be 39,032/803126 or about 4.8 percent. They assume some percentage of the population have no symptoms an guess at a mortality rate of roughly 3 or 4 percent.
There is a huge logical flaw in this formula. Deaths/current cases does not account for the fact that for most current cases, the disease has not yet fully run its course. A closer estimate of the true fatality rate would be too look only at cases which have had an outcome- either death or recovery. CFR=deaths/(deaths+recoveries).
Imagine you have bet on the outcome of 100 football games, but only 10 of the games have finished. In five games you won, and you lost 5. The WHO is saying your loss rate is 5 out of 100 games, 5 percent. I hope this illustrates the absurdity of the WHO calculation. Since 90 out of 100 games haven't finished yet, it would be better to say your loss rate so far is 5 out of 10... 50 percent.
So what do the fatality rates look like when we look at only the cases with outcomes, excluding those cases which have not yet reached an outcome?
Worldwide data:
Coronavirus Cases
803,126
view by country
Deaths:
39,032
Recovered:
172,396
World CFR=39,032/(39032+172396)=18.6%
US data:
Coronavirus Cases:
164,435
Deaths:
3,175
Recovered:
5,507
US CFR=3175/(3175+5507)=36.6%
Italy Data
Coronavirus Cases:
101,739
Deaths:
11,591
Recovered:
14,620
Italy CFR=11591/(11591+14620)=44.2%
This tells us that of known cases with an outcome, worldwide 18.6 percent of people with Covid19 have died. In the US, of known cases with an outcome, 36.6% of people have died. In Italy, of known cases with an outcome, 44.2% of people have died.
This is why we have shut down entire nations.
Yes, there is some number of people who have contracted Covid-19 and recovered without ever being tested. We have ZERO data about that and it is irresponsible to assume this is some large number just to produce a fatality rate that lets people sleep at night. We should be reporting these numbers. People SHOULD BE SCARED. Maybe then they will stay home.
Another problem here is that there are so many unknown data points- how many cases are not recorded? How many deaths attributed to other causes are actually Covid-19? (Germany is not recording Covid-19 as teh cause of death when patience have any other conditions). Many suggest the data from China is totally fabricated. I've heard the same thing about data from New York. By adjusting these variables one could produce any fatality rate from 0.5% to 50% or more. The only way to really understand what is happening is to do surveilancee testing- testing the general public for Covid-19- not just people who report symptoms. That means lots and lots of tests.... which we don't have.
