I recently was playing with the data for the state of Washington to try and see the impact of age on cases and deaths. I was able to interpolate for the full data as of May 26.
Given the percentages reported by demographics, plus the total number of cases and deaths, we can get a look at the actual numbers of cases and deaths by age, as well as the chances of dying given that one tests positive for the Coronavirus.
What this shows is that if you’re under 59, your chances of dying of coronavirus if you are sick enough to be tested is below 2%. This is an important footnote; I’d bet that the actual probability of dying if you get the virus is considerably lower, especially if you’re younger. Why? There’s lots of evidence that the younger you are, the less likely you are to get really sick, and the less sick you get, the less likely you’ll bother to get tested. This effect thus isn’t a simple multiple; I’d bet if you’re in the 80+ category you’re very likely to get sick and tested. If you’re 17, you’re probably unlikely to do so. Thus the denominator problem persists – how many people in the 20-39 group get tested? Maybe half? Then halve the p(death|positive test) to get p(death|infected), the number we’d all really like to know.
On the other hand, this shows how really deadly this virus is to 80+ year olds – over 1/3 chance of death if infected.
UPDATE: I’ve added similar numbers for hospitalization. These are more daunting. 6% of young adults and 13% of 40-60 year olds had a really bad case – bad enough to need hospitalization, oxygen, and/or intubation. Worth avoiding, especially because having a serious respiratory infection can lead to lifelong problems.