Interesting. Some issues, for example the Icelandic continuous testing pretty explicitly pointed out that their estimate was not an IFR (PCR testing obvs only catches current infections). Also would be useful to put into a MA model rather than just a median
-
-
This Tweet is unavailable.
-
Actually, that's not true. Many of those studies are published in news reports in foreign languages. Some of them have no information on the denominator, making it hard to meta-analyze, and some of them have only been published since the beginning of May
1 reply 0 retweets 0 likes -
The benefit of a preprint however - we can update and see if there are any more studies that met our criteria prior to submission
1 reply 0 retweets 0 likes -
This Tweet is unavailable.
-
I'm pretty sure some of the numbers are also just wrong? Like the IFR of pregnant women from NYC is 0.53% - that study didn't include anyone who died, and was not a population estimate anyway. Can't really include it in an estimate of IFR. At least 2 more similar issues
1 reply 0 retweets 0 likes -
And including a sample of just high school students? Several of just blood donors? Makes no sense for an overall IFR estimate
1 reply 0 retweets 0 likes -
This Tweet is unavailable.
-
Nope, it's because the fatality rate will be much lower. We only included population estimates for this reason - lumping them all in together will lower your IFR significantly
1 reply 0 retweets 0 likes -
This Tweet is unavailable.
Nope they really aren't. Can't just extrapolate like that, and the adjustment in the table is pretty opaque - I have a hard time understanding how any of this would make sense together
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.