This so totally confirms my priors that of course I believe it. :-) But it's also almost by definition a more meaningful result than the ones that have come out of California and other low-infection places
This implies that actual infections are about 13X confirmed cases, and that the infection fatality rate so far (using the NYC estimate that includes probably coronavirus deaths) is 0.84%
Which is in the same general ballpark as recent serology results from Switzerland, the Netherlands and Germany, and in a completely different one from the conclusions (if not the actual survey findings) of the Santa Clara County study that has sparked so much criticism
The low-end estimate from the Los Angeles County study implies a 0.36% infection fatality rate, so also in the same general ballpark.
A lot of people (including me) have been comparing these IFRs with a 0.1% fatality rate for influenza, but that's actually the fatality rate for *symptomatic* flu cases, and influenza can be asymptomatic as well. 0.04% may be better estimate
So the emerging evidence from these serological surveys seems to be that the coronavirus is 10X+ deadlier than seasonal influenza, although it can of course vary with demographics, hospital capacity, etc.
Also, this @richardneher chart, which simply estimates prevalence based on an IFR of 0.5%, seems like it might be a fair representation of the spread in other hard-hit regions
One big caveat: at this point I think all we know of the study is what Cuomo just said about it on TV. It's based on a small, imperfect sample of New Yorkers, and actual prevalence of the disease in NYC almost certainly isn't exactly 21.2%.
But just to reiterate something from yesterday: Such tests are generally going to deliver much more reliable results in places with high percentages of infected people (NYC) than in places with low percentages (California)
Finally, after reading the critiques of the Santa Clara County study, which took a 1.5% positive rate in the testing and somehow concluded that between 2.49% and 4.16% of the county was infected, my guess is that 1% of SC County is infected. Which gives an IFR of … 0.5%
