Commenter Tom Davies pointed to a paper that in turn referred to a paper by Verity and others. Instead of looking at data for the number hospitalized relative to the number of tested cases, which is skewed by the fact that people without severe symptoms are less likely to be tested, this paper seems to assume that hospitalization rates are proportional to death rates. Accordingly, they estimate a much lower rate of hospitalization for those in the lower age brackets. 1.1% for age 20-29, 3.4% for 30-39, and 4.3% for 40-49.
I prefer their method to the one that seemed to show a 12 percent hospitalization rate for the 20-44 age group.
Any number derived from number of tested cases is suspect in my eyes because of the severe constraints on testing. The only numbers I’m paying attention to are deaths and (where available) number of hospitalizations. Why aren’t the public health folks collecting and reporting hospitalization stats including demographic data? Or if they are, please tell me where to find it.
surely south korea has posted data on this — given they don’t have the same testing constraints the RoW is struggling with, it should be our best estimate