Those CDC numbers are really weird btw. 0.4% of symptomatic cases is like a 0.2-0.3% IFR.
0.3% of the New York City population has already died of corona-virus. So assuming everyone in New York was infected that gives us an IFR of .3%. But if a more realistic 20% of the population was infected then the IFR could be as high as 1.5%.
I tried going through all the most recent research and almost all of it came up with an IFR in the 0.5-1.2% range. And the few that had a lower rate were looking at anti-body tests in places with very low base rates.
Maybe the 20% of the population was more susceptible? It seems that the outbreak was concentrated in long-term care facilities, healthcare workers (who received a larger viral load than average), and the poor (who tend to have more pre-existing conditions). If the 20% is not randomly distributed the math can still work out.
The math could work out, but if that was true then that means if NYC got 5 times as many infections the number of fatalities would barely change.
I would find that hard to believe and I also doubt they would have enough data to validate that type of hypothesis. We're still guessing at the number of people infected, not to mention how old are they, what pre-existing conditions did they have, how do those pre-existing condition affect Covid-19 mortality in different age groups, what viral load did they receive, etc..
Fair points! You could probably make an estimate with a stratified approach but if 0.3% have already died there’s no wiggle room for additional deaths.
An interesting question is this: which 20% got it? A representative sample, or a localized group with unusual risk characteristics, like retirement homes or prisons?
0.3% of the New York City population has already died of corona-virus. So assuming everyone in New York was infected that gives us an IFR of .3%. But if a more realistic 20% of the population was infected then the IFR could be as high as 1.5%.
I tried going through all the most recent research and almost all of it came up with an IFR in the 0.5-1.2% range. And the few that had a lower rate were looking at anti-body tests in places with very low base rates.