I've tried to point out that the study design is biased and could explain this observation. Emerging serology from elsewhere suggests the true infection fatality rate is somewhere between 1/200 and 1/100 1/n
Again and with tribute to @jbakcoleman who first pointed this out to me...
If the mortality rate were as low as @sudha_lakshmi suggests - if every single man woman and child in NYC had been infected it would still not explain the death toll there.
Look at the numbers 2/2
If the mortality rate were as low as @sudha_lakshmi suggests - if every single man woman and child in NYC had been infected it would still not explain the death toll there.
Look at the numbers 2/2
Following up on the conversation overnight. We shouldn't be distracted by the (IMO) stretched estimate of the IFR in the paper, but focus on the serology. For the reasons laid out in this tweet from @viciykevin
