Comment

Comment by mabbo

Let's presume this death rate is accurate to within an order of 2- 1.33% to 0.33%. We can use that to do some interesting estimates on the true infection rate.

https://www.worldometers.info/coronavirus/ (trust it if you like) says that the USA has (as of 2020-04-05T15:13:00Z) had 9,119 deaths- and I was going to say 8,486 but I refreshed the page and it jumped to 9,119 because of course it did.

9,119 / 0.33% = 2,763,333 infections.

9,119 / 1.33% = 685,639 infections.

But, we also have to consider that death isn't instant- so add in everyone who became infected in the last 5-10 days. That'll be a bigger number. The reported number of infections because of limited testing is 320,828.

I think it's not unreasonable to say that probably 1m Americans have now been infected by this virus. And the longer ago this comment was made, the more likely that estimate is correct.

Edit to add: I trust the number of deaths more than the number of infections because it's easy to say "don't come to the hospital to get tested if you don't need to" but it's hard to say "let's pretend that dead body died of something else".

parent mabbo 6 years ago

Discussion

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton