[briffle]

When the virus hit China, it had around a 4% death rate. Now, in the US, we have 5.6M confirmed cases, and 175k deaths. That is 3% death rate. (although, deaths lag behind diagnosis) You're right, its less than they first feared, but not much. And still 30 times the death rate of the flu. There have been some groups that keep looking at the number of deaths, and compare it to the population as a whole, to make the numbers look much smaller. But those are purposefully deceptive.

[ghufran_syed]

and how many unconfirmed cases? I personally know 10 people who have had symptoms that are likely covid, but not serious enough to go get tested. The fact that the least symptomatic patients are also least likely to get tested means that the calculation you presented gives us, at best, an upper bound. A better way to get the true measure would be based on population-based antibody testing.

More info available at the WHO website: https://www.who.int/news-room/commentaries/detail/estimating...

[nradov]

The actual best estimate of death rate in the US is 0.65%, not 3%.

https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scena...

[jashmatthews]

Sadly 3% is the right number to use for that comparison between observed fatality in the US and earlier estimates https://coronavirus.jhu.edu/data/mortality

The estimated IFR is quite different and can’t be just swapped.

[nradov]

You're comparing one meaningless number to another meaningless number. What's the point?

[jashmatthews]

CFR isn't a meaningless number and the estimated IFR is useful for making predictions.

[nradov]

In what sense is CFR meaningful? What can you do with that number? There is little consistency between countries in testing rates or the criteria of what counts as a "case".