Ok, another COVID-19 IFR estimate. This time from a Danish seroprevalence with random sampling, indicating 0.5-1.8% infected in early May This puts the IFR in Denmark somewhere between 0.48-1.7(!)%
-
-
As a very crude estimate, I usually take the deaths from a week or so after the tests were done, but I think it'd be really interesting to develop a more robust model to correct for right-censoring from seroprevalence studies
-
I am not sure that right sensoring is uniformly an issue that underestimates deaths. Here's a sensitivity analysis I did on Geneva moving the numerator (deaths) back and forth. Intuitively should estimated IFR drop over the course of three weeks? Last estimate is pretty robust.pic.twitter.com/j3RSsLyfNr
- Show replies
New conversation -
-
-
Median might do a lot of obfuscating if there’s a fat tail of lag and the outbreak is quickly contained. Both of which I think are true. But it’s surprisingly hard to find a full cohort analysis of the distribution of days to death. Would love to see it if anyone has it.
-
Best I’ve seen was from a few hospitals in NYC which looked at everyone admitted in March and reported their status on April 28. ~40% were still hospitalized, ~40% dead and 20% released.
- Show replies
New conversation -
-
-
This Tweet is unavailable.
-
This Tweet is unavailable.
- Show replies
-
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.