This Post was last updated on March 15, 2020.

An important question regarding the etiology of the COVID-19 virus is outstanding… what’s the average case fatality rate?  

The case fatality rate is the number of folks that die from an illness divided by the number infected. The WHO has released a statement suggesting that the case fatality rate is 3.4%. This is inaccurate. Other sources have suggested that it’s 2%. This also is almost certainly a gross over-estimate of the average case fatality rate.

So are there any good data out there to get a better estimate?

Perhaps the best source is the S. Korean CDC, which has done robust testing. S. Korea has done far more testing than we have here in the US and have a much more broad testing protocol than we do.  Up until this week in the US, only severely ill patients with no alternative diagnosis, symptomatic folks with known contact with a case, or symptomatic people with travel history to an epidemic region are in the testing protocol.

As of this morning, S. Korea has 8,162 confirmed cases and 75 deaths (case fatality rate of 0.9%). Bottom line- the best data I can find is from S Korea and it suggests the average case fatality rate is 0.9%.

But the case fatality rate varies a lot with age. Of the deaths in S. Korea, 25 have been over 80 years old, and 28 between 70-79 y/o. Fourteen were between 60-69 and 6 deaths have been among people 50-59 y/o.

Seventy-one percent (71%) of the deaths in S Korea are among people 70 years old and above and 89% are among those over 60. The case fatality rate for those over 80 is about 9%, and for all those over 70 it’s about 6.7%.

However, for people under 40 years it’s just 0.04% or 4 in 10,000. You can see from those data that COVID-19 is far more serious among those over 79 years old, and that for those under 40 it’s really not all that lethal.

______

To get a better picture of the percentage of confirmed cases that are asymptomatic let’s take a look at what happened on the Princess Cruise Ship – perhaps the best place to look for which there’s a good denominator (as they were a captive population).

A brand new study from that incident estimated that the delay-adjusted asymptomatic ratio of the positive COVID-19 infections on board the ship was 34.6%. In other words, about 35% of the folks on the Princess ship that were infected with the virus were asymptomatic. That’s an important factor, because it suggests that the S Korean CDC data (perhaps the best that’s out there right now) is an overestimate of the the case fatality rate.

We’ll get a better picture of how lethal the COVID-19 virus is in the coming weeks- but I felt compelled to include this data from S Korea and the cruise ship this week because there are a lot of people out there that are hanging their hat on the 3.4% and 2% case fatality rate numbers- and those are overestimates.

Also, remember that this discussion is about the average case fatality rate.  The fatality rate for certain sub-populations like the elderly and people with pre-existing conditions that put them at more risk will have a much higher case fatality rate than the average case fatality rate.