Dr. Birx on CV-19 R-Naught values

[Shermann]

Dr. Birx discussed the virus models and R-Naught (R0) values at the 59:00 mark in Thursday's presser.

R-Naught without mitigation : 6 (up significantly from earlier estimates of 3)
R-Naught with social distancing mitigation : 1.3 to 1.5

This is being presented as a major victory for the American people. To a degree that is true, but this is also a bit of a bombshell to me. It suggests that the incidences of new cases are not going to go down with social distancing, but rather plateau or rise slightly for an extended time. As soon as the lock-down is removed, it will take off again. It does not appear you can bring the R-Naught value below the magic value of 1 without draconian measures similar to what China did. Any experts here? Am I missing something?

[Bernardo de La Paz]

1. If you bring it down to 1.3 to 1.5, then aggressive contact tracing will bring it down even more

At that level, you are not overwhelmed trying to contact trace.

[Shermann]

4. I don't believe those are related in the way you describe

Contact tracing can only be done when the total number of cases is below a certain number.

Once the total number of cases is too high (like is currently the case), it will not go down with a R0 above 1.

So we're stuck.

[Bernardo de La Paz]

5. Question of resources. You're right that at some point the case load exceeds tracing capabilities.nt

[Shermann]

9. Also the virus seemed to go right around contact tracing in Washington State

They were detecting community spread almost from the beginning. That was when there was presumably a finite number of Patient Zeros.

Now you've got infected people coming and going from neighboring counties and states in every direction.

[BGBD]

17. It seems to me

that the high number of contagious asymptomatic cases is going to make contact tracing impossible without some kind of significant and random testing regime.

Also, it's only reduced to 1.3 or so with a LOT of social distancing going on. As soon as that lets up it's going back up to 6ish.

Until there is a vaccine, we aren't going to be able to stop SD without returning to exponential growth.

[Shermann]

18. You mean FASTER exponential growth. 1.3 still grows exponentially! nt

[BGBD]

21. ok.

fair enough, let me rephrase that.

we would return to FASTER exponential growth.

That's a pretty big difference in the number of people who will die though, 1.3 to 6

Standing by my original statement. Social distancing will have to remain until a vaccine is available or enough people have longterm immunity to reduce R0 to >1 without SD. Since we don't know how long immunity will last after an infection I'm not sure how much we can rely on it.

[uponit7771]

36. +1, ... or until we have a proper amount of sample testing / isolation and tracing done like

... in S Korea et al.

The testing has be a weekly rolling test too

[Celerity]

33. Contact tracing with airborne respiratory viral diseases is also extremely hard, it is not at all

like tracing the contacts of an AIDS patient, for instance.

[Azathoth]

2. 1.5 is horrendous with this level of lockdown

People just don't understand what happens with exponential growth.

1.5 will only stretch the curve out a few more infection cycles before it spirals out of control.

If we cant get this thing under 1.0, we're in trouble.

[Shermann]

3. Seems that way

So they touched on this briefly, and moved on.

The press is fixated on the death counts. The fact that those estimates are going down seems to be masking this darker aspect of where we are. Where our economy is.

We're just stuck.

[Pobeka]

6. This is not true. R0 is not the only factor.

R0 is initially the controlling factor, true. But with a low enough R0, there is time for herd immunity to build, and the growth is contained and does not spiral out of control. Once the R0 * (herd immunity) < 1, the epidemic fades away. Herd immunity factor is basically (susceptible population) / (recovered population).

[Azathoth]

7. It's very true and you're proving my point

Herd immunity is meaningless till you're over half the population, or in our case, about 160,000,000 people.

If you hit 160 million cases at 1.5 exponential growth, you will have *millions* getting sick each day at the height. Orders and orders of magnitude beyond our healthcare capacity.

In order to have a peak in the curve, we have to get to 1.0 or below NOW. That's the entire point of taking these extreme measures.

[Pobeka]

11. We'll just have to disagree.

But I recommend you look at the wikipedia site for SIR models to see the basics of how epidemics work. Then you'll see herd immunity is not meaningless at levels below 50%, it also controls the rate of rise of the curve.

[Azathoth]

14. OK, since we're all stuck at home, here is what it looks like

https://imgur.com/hafZmLL

This is a rough SIR model for COVID19 with R{0} = 1.5 and a 14-day infectious period.

Pay attention to the blue curve. That's the infected population at time t. At its peak, the US is looking at almost 25 million infected people, increasing by somewhere between 500,000 to almost a million a day. The country has less than 100,000 hospital beds in total.

Herd immunity kicks in meaningfully where the other two lines cross.

R{0} of 1.5 *while under lockdown* is very, very not good.

[Shermann]

15. R{0} of 1.5 is probably optimistic

The conservatives are getting ansy about the lockdown. FXN isn't helping matters.

This Apple/Google tracking app will never take off here.

1.5 is probably the best the US can do and will probably start to slide through that 200 days.

I do have to wonder, how much would an ample supply of N95 masks for the general population affect the R0? China has had those from the very beginning. You have to wonder.

[Shermann]

24. How did you calculate the Beta and Gamma values?

Last edited Sat Apr 11, 2020, 10:11 AM - Edit history (2)

I want to run a simulation with R0 = 1.3 but I'm unclear how you got Beta = .12 and Gamma = .08 from R0 = 1.5

14 day infectious period divided by 200 total days = .07 which is close

[Azathoth]

28. R naught is beta over gamma

Gamma is the recovery rate per unit of time. 13-day infectious period means 1/13 = 0.077 recovery per day. Beta then would be R naught times 0.077.

I'm not an epidemiologist and I don't pretend to be an expert with these models, but any significantly positive R naught during lockdown is really really bad.

The idea at the this point is to reduce R0 below 1 so that infection rate plummets and the curve dips back to the ground. Then you gradually reduce lockdown but maintain minimum precautions to keep the R0 at a manageable level, occasionally reinstituting lockdown as needed.

A positive R0 during lockdown means the strain on our health system is going to get progressively worse from *where we are now*, and there is nothing we can do because we're already using the nuclear option.

[uponit7771]

32. THIS !!!! This should be its own OP, its great to have smart people. IF we're R0 1.2 during lockdown

... then even half back to normality gets us screwed.

I was looking at IHME and it sounds like they assume NY/NJ level lock down

[Shermann]

25. The peak of the SIR model seems to be 4 months out

Last edited Sat Apr 11, 2020, 10:15 AM - Edit history (2)

I wonder why there's such a disparity with the current view that it is in the upcoming weeks?

Perhaps "peak" is being defined differently? Is it peak of rate of acceleration increase? As opposed to peak of active cases?

I would think peak of active cases is the most significant metric.

[uponit7771]

34. +1, I was looking at IHME and it looks at peak deaths not rate of increase just deaths

[Azathoth]

37. I haven't looked at the details of the current models, but I'm guessng one of three things

Either those models are assuming

1) The R{0} under lockdown is actually <= 1

or

2) They're assuming the decrease in R{0} from the sudden lockdown will lead to a localized (short-term) decline in cases at the start of an overall steady climb

or

3) They're not considering R{0} and instead are fitting curves based on the data from China (where the R{0} was quickly brought *way* below 1). I believe IHME is doing something like this.

[Shermann]

38. Here is a simulation with R{6}

With no mitigation we would have close to 100% recovered after only three months. That doesn't seem to line up with reality.



R{3} is similar. I'm not sure what to think of this.

[uponit7771]

31. I think the poster has ***FACTORED IN*** 160 million people, that will take 1 yr is to get to

... with our current SIP measures IINM

[Shermann]

8. Herd immunity is not the way...ask the UK

So if nobody is immune, the herd immunity factor should be 1. If everybody is immune, it is 0.

So to get 1.3 down to 1.0, you need an immunity factor of .76.

I don't believe any country is anywhere near that. I concede that we don't have antibody test data, however you'd have to be wildly optimistic to believe that's in the cards this year.

Worse, 1.0 doesn't beat the virus. It just plateaus forever. It really needs to be lower than that.

There are even worse scenarios if the acquired immunity isn't for life.

[Pobeka]

12. No one knows about herd immunity levels, until there is adequate testing.

I'm not proposing it either.

For COVID19 the number of actual cases per tested positive is now estimated at 2:1, it may go higher yet.

In the abscence of a vacine, there are only two viable options to control the disease, social distancting and herd immunity. The no social distancing for for fast herd immunity approach isn't working in sweden.

[Shermann]

13. Trump won't wait for herd immunity. Trump won't wait for a vaccine.

Trump won't wait for the green light from Fauci.

[BGBD]

19. The good news is

Trump isn't the one who's going to get to make the decision. He conceded that power when he let the governors order the lockdowns instead of doing a national one.

[Shermann]

20. Yeah the press asked for clarification on this "decision" that is supposedly his

His wasn't able to articulate an answer very well.

[BGBD]

22. Maybe he thinks

he can bully governors into ending them?

He's used to getting sycophants to do whatever he wants but I'm not so sure he can on this. They aren't likely to go so far as to be willing to sacrifice a couple hundred thousand of their citizens for him....outside of maybe DeSantis and Kemp, they probably would be.

[DeminPennswoods]

10. Just posted this in another thread, but well worth the read

link: https://www.newyorker.com/magazine/2020/04/06/how-does-the-coronavirus-behave-inside-a-patient

[JCMach1]

16. Or it will gut punch us once Fall hits

[HarlanPepper]

23. Love the pop up expert motif

Not a fan of the scarves either.

[Squinch]

26. Let's toss this little wrinkle into the conversation:

www.democraticunderground.com/100213266741

South Korea is obviously no slouch on understanding the science of this. The explanation for these re-infections has up to now been faulty testing and human error. But I imagine that in 91 cases, they double checked the results pretty thoroughly.

[samnsara]

27. sounds like shes deliberately talking over trump*s head...

[uponit7771]

30. She is also trying to talk over the reporters heads and filibustering when it comes to testing

[uponit7771]

35. K&R, SIP with sick people only gets us more sick people howbeit slower. There were no isolation

... measures before SIP orders went into place.