Thread by @Noahpinion

1/Some quick back-of-the-envelope coronavirus math.

Currently, confirmed cases are around 4200, probably 4500+ by end of day.

worldometers.info/coronavirus/co…

2/Recent models suggest the true number of cases is 10x-20x higher than the confirmed cases, so let's say we have 50k-100k actual cases in the country. 100k to be on the safe side.

https://twitter.com/Noahpinion/status/1239580446918234112

3/Now, the confirmed case number has been doubling every 3 days. But studies suggest the true doubling rate is more like 6 days. So if these trends continue, the ratio of confirmed cases to true cases is cut in half every 6 days or so.

4/That makes sense given that we're ramping up testing fairly quickly, with tests highly variable but generally doubling every 2 days.

docs.google.com/spreadsheets/d…

5/So in about 3.5 weeks, if we continue to push hard for more testing, the confirmed case number will mostly catch up with the true case number, and we will have a good idea of how many people actually have coronavirus in the U.S.

6/Now, let's think about ventilators.

The U.S. has about 100,000 ventilators total, and about 1/7 coronavirus patients need ventilators. So if and when we get to 700,000 true (rather than confirmed) cases of coronavirus, our system will be overwhelmed.

fivethirtyeight.com/features/coron…

7/Of course, ventilators aren't all in the places we'll need them to be, and take time to ship around, so some areas will be overwhelmed before the total national true case # gets near 700,000, while some will never be overwhelmed.

But let's just say 700,000...

8/If we have 100k cases today and the true case doubling time is 6 days, then in a little less than 3 weeks we'll start saturating the hospital system.

9/Thus, we have 3 weeks in which to use shutdowns and social distancing to dramatically reduce the rate of infection. And during this time, we still won't have a great idea of how many true cases we have.

How long will shutdowns take to bend the curve?

10/Italy has been in lockdown for a week now, and their curve is maybe slightly starting to bend right now. (Of course, this doesn't take into account the possibility that *Italian* testing has been improving during this time.)

11/So the upshot of this is: If we #ShutItDown this week, we might have a chance of saving our hospital system from being overwhelmed (except in a few hotspots).

But if we don't #ShutItDown this week, we're probably headed for a lot of deaths.

(end)

Some quick addendums to this thread:
1. Some ventilators need to be used for non-coronavirus patients.
2. We have some extra ventilators in storage.
3. It generally takes some time (2 weeks?) from when people get infected to when they need a ventilator.