Thread by @jeremyfaust

Q: Why have so many SARS-CoV-2 models failed?
A: Failure to account for asymptomatic spread.

My collaborators @biomathematicus+@JacobBAguilar are among the world's great minds on this.

medrxiv.org/content/10.110…

That's why they found an *Ro of 12* in San Antonio.

Let's discuss🧵

The main finding in our new paper is that in order to get the R down below 1, a place like San Antonio would have to do a good shutdown for up to 83 days.

Which might ring a bell!

That's about how long the places that beat this thing stayed closed.

More on that later.

First, the Ro (R-naught) or basic reproduction number must be understood.

It's a combination of biology (things about the virus) and behavior (how we live).

The earlier estimates missed asymptomatic spread.

So they said Ro=2.68.

That was very, very wrong.

As @biomathematicus has said, a pathogen can not spread throughout the world in a few months with such an Ro, especially given the contagion windows we know about.

The CDC later upped the Ro to 5.7

Still not high enough to explain these rapid outbreaks

AND add the fact that we are missing tons of cases (lack of testing!)

For those of you who say "No way! Ro can't be 12."

We say, really?

Why not?

Airborne viruses have Ro of 15 or higher.

SARS-CoV-2 is at least SOMEWHAT in the air, have we not come to that agreement?

Ro of 12 seems down right reasonable to us.

Now we come to the real news:

The EFFECTIVE reproductive number is the number of days it takes (once mitigation strategies like shelter-in-place and masks) to get the R under 1.

R < 1 means that each infected person infects, on average, FEWER than 1 person.

It's how pandemics end.

Fewer new cases.

What this model does is to take these strategies (and a realistic Ro like 12) into account.

It can then tell you how long your area must close in order to get the spread to a safer number (R<1 being safest).

We are even making a calculator that spits out various options.

So you could say "Ok, we are willing to do a complete shutdown, but for only 14 days."

Fine.

Do that. But then you must do a modified shutdown for X number of days.

We can tailor this to the needs and, frankly, the appetite of your jurisdiction.

Not one-size-fits-all.

Some places might want a longer HARDCORE shutdown if that means they can get the R to close to 1 sooner.

Others will say, No we can't do that but we'll do modified shutdowns for longer.

It can be tweaked.

Then you monitor in real time and adjust as needed.

Data driven.

We USE MATH.

But factor in human nature and political will.

*That* is how we can do better this Fall and Winter.

Please reach out to us.

We are here to help save the country and the planet.

To buy time until the vaccine comes!

@JacobBAguilar @biomathematicus @LWestafer