Returning to the dance studio safely (covid-19)

[lucia_l]

On the topic of modelling, have you seen that the Imperial College model is up on GitHub?

GitHub - ImperialCollegeLondon/covid19model: Code for modelling estimated deaths and cases for COVID19.

Code for modelling estimated deaths and cases for COVID19. - ImperialCollegeLondon/covid19model

github.com

Yep. Saw it. But... welll... James Annan and Julie Hargreave's model here is better:

BlueSkiesResearch.org.uk: Model calibration, nowcasting, and operational prediction of the COVID-19 pandemic

New manuscript! Now up on github too. Title as post. Yes, this is us dipping our toes into epidemiology. Turns out that calibrating a ...

julesandjames.blogspot.com

They've just submitted. It might be rejected. It's still better.

 

[Warren J. Dew]

(after the data come in and do what I predict! )

I just want to take this opportunity to point out here, that the data are already doing what I predicted back in March, when everybody was upset at me for saying that school closures and masks would have a sufficient effect! I'm not sure of all the places the moderators banished my posts to, but you can also see the predictions in my Quora post back on March 23 when everything was still growing exponentially, with my "I told you so" updates in comments:

What did South Korea do to stop the Covid-19 spread that was so successful that the US did not do?

Warren Dew's answer: First, it should be mentioned what South Korea did not do that the US did do: stop travel from China. The US stopped travel from China early, while South Korea delayed. As a result, exponential growth of cases in South Korea started earlier than in the US - three weeks earlie...

www.quora.com

So, Dance Forums, I told you so!

My prediction now is that we'll still need masks, spring/summer temperatures, or some other additional mitigation to reopen ballroom studios and competitions. And yes, I was still right that if we'd gone with face masks and school closures in the first place like South Korea did, we would never have needed dance studio or ballroom competition closures in the first place.

I don't look at case counts. I think that data is to strongly affected by changes in the number of test kits, limitations of of access. In contrast, dead is dead.

I agree there is some sensitivity of case counts to testing availability and strategy. However, I think this is a big effect only in the early stages when testing is ramping up and also when the medical system gets overwhelmed and deprioritizes testing.

I also think there are some difficulties with death counts, though. In particular, I think the delay to death is bimodal: there are some early deaths of people who don't get to treatment in time; ventilators then keep people who do get to treatment in time alive for a couple extra weeks until the half of people that ventilators can't save die. As a result, I no longer think death counts are more reliable once mitigations start having an effect since that effect has different delays on the two types of death. If there were numbers that differentiated between those types of deaths, that would solve that problem, but I haven't found them.

Without a vaccine, we won't be able to go back to status quo ante. But we may not need to have the level of stay-at-home we currently have. We still need the vaccine.

Agreed. But to me, switching schools - and businesses where possible - to remote access, plus universal use of masks outside the home, could get us pretty close, with everything else reopening.

Regarding the need for additional mitigations, like masks, to support business reopenings:

For full reopening. Partial... maybe not so much. Shut down doesn't have to be all of nothing. Yep. We would want to keep "R0<1".

I think our difference of opinion here is mostly because we differ on where R is for Illinois right now. If R=0.5 as you think, yes, partial reopening could be supported without additional mitigation. If R=1.0 as I think, then you need mitigations to compensate for any reopenings.

But even if R=0.5, do you think ballroom studios would be among the partial reopenings? What about ballroom competitions? Those are the things I actually care about here!

 

[David_In_CA]

"The basic concept is to use the model to invert the time series of reported deaths back through the time series of underlying infections in order to discover the model parameters such as the famous reproductive rate R. It’s actually rather simple and I am still bemused by the fact that none of the experts (in the UK at least) are doing this."

Whaaaaaaaat? Do they have another blog that describes how the others *are* calculating R?

 

[Warren J. Dew]

I would not place any faith in hand sanitizer and masks making social dancing safe. Those measures can be effective when combined with "social distancing", but "social distancing" and social dancing are incompatible concepts. The cloth masks that people use will be pretty worthless in dance frame. A more effective mask will restrict airflow and make the elevated breathing required to dance difficult.

I disagree. The evidence from comparison of East Asian countries with Western countries is that masks are more effective than maintenance of a 6 foot distance.

Even in a closed ballroom hold, a mask stops a sneeze before it hits one's partner.

 

[Warren J. Dew]

"The basic concept is to use the model to invert the time series of reported deaths back through the time series of underlying infections in order to discover the model parameters such as the famous reproductive rate R. It’s actually rather simple and I am still bemused by the fact that none of the experts (in the UK at least) are doing this."

Whaaaaaaaat? Do they have another blog that describes how the others *are* calculating R?

There's a fundamental difference in the types of models here, which the blogger is ignoring or doesn't understand.

The Imperial College model is a simulation model. "Inverting the time series" can give you a parameter for the total effectiveness, but it can't apportion that effectiveness between the various types of mitigations. The whole point of the Imperial College model was to guide policy decisions on which mitigations to use, so the single parameter you get from inverting the time series is pretty much useless.

The Washington State model is a curve fitting model. Using the single parameter from "inverting the time series", it can give progressively more accurate estimates of future behavior, assuming a continuation of current policy. However, it can't be used for the "what if" analyses required to judge between, say, shutting down large events like ballroom compettions, or shutting down small continuing businesses like ballroom studios.

The Imperial College estimates for R are based on previous research, both from Covid-19 and from earlier research back to the previous century. To refine the parameters in that model, though, you need to have comparative data from places that have different conditions and took different policy decisions - and you can only get limited data from those comparisons, since governments don't set policy by varying only one factor at a time for maximum scientific knowledge.

 

[Warren J. Dew]

I don’t follow this. Who is/would be expecting everyone to lose a year of their lives, and do you mean “lose“ as in “put their lives on hold until a vaccine is developed” or something else?

Under stay at home orders, I would say we are merely surviving, rather than actually living.

Thus, if we have to stay at home for a year until a vaccine is available, we're losing a year of life.

 

[lucia_l]

Warren,

So, Dance Forums, I told you so!

I still don't think masks only would have been a sufficient response in March. I still think we needed to slow the exponential growth fast. Then once the number of infected are down, then we can relax to a lesser response. If we'd only done masks, the death behavior would be more like my brown curve-- which would still be increasing and higher than we have now. And we would have had continued high deaths. That's a lot of death. How much economic and social cost we get for each death can be debated. But it's a lot of death.

My prediction now is that we'll still need masks, spring/summer temperatures, or some other additional mitigation to reopen ballroom studios and competitions. And yes, I was still right that if we'd gone with face masks and school closures in the first place like South Korea did, we would never have needed dance studio or ballroom competition closures in the first place.

Korea did test/track/trace, not just masks and school closures. We couldn't do what Korea did because the CDC screwed up the test kits. By early March we knew we couldn't do it because we knew the test kits were screwed up. I think you need to acknowledge that context of conversation-- that is: that choice was precluced.

I agree there is some sensitivity of case counts to testing availability and strategy. However, I think this is a big effect only in the early stages when testing is ramping up and also when the medical system gets overwhelmed and deprioritizes testing.

March was early stages. Testing is still ramping up. We had cases rising before we had the ability to test.

I also think there are some difficulties with death counts, though.

Sure. Deaths is imperfect.

I think it's still better than case counts when case counts are dramatically influenced by extreme shortages of test kits, or when the shortage is changing. Case counts will be much worse than death counts until we have a surplus of test kits.

I think our difference of opinion here is mostly because we differ on where R is for Illinois right now. If R=0.5 as you think, yes, partial reopening could be supported without additional mitigation. If R=1.0 as I think, then you need mitigations to compensate for any reopenings.

I'm not sure what you mean by mitigation.
I think we don't want R0 to go above 0.5. It will go up if we change lighten the lock downs. We don't want it to go above 1. If we fully reopen, without vacinnes, Ro will go back to 3, deaths will start rising exponentially again, which we don't want.

I think all of the following three need to be kept true:

• Behavior needs to correspond to a reproduction rate of Ro < 1 . (Note, I am keeping the subscript.)
• And the daily death rate needs to already have dropped to 50% of its peak value.
• It needs to be observed to then either be constant or dropping for at least a week straight on the week before emerging.

If all three are kept true, I think we can start reopening things, track and trace, and watch to see if the first and third remain true. Some people will still be dying. But I think if all those three are true, then the general public threat is contained. People can make individual choices. But if monitoring shows deaths increasing, then .... bad. We need to go back to stricter controls.

I'm not Illinois Governor though. But that's what I think.

 

[lucia_l]

There's a fundamental difference in the types of models here, which the blogger is ignoring or doesn't understand.

The blogger understands that model. He's criticizing them for not comparing their model to

The Imperial College estimates for R are based on previous research, both from Covid-19 and from earlier research back to the previous century. To refine the parameters in that model, though, you need to have comparative data from places that have different conditions and took different policy decisions - and you can only get limited data from those comparisons, since governments don't set policy by varying only one factor at a time for maximum scientific knowledge.

IOTW: To refine the parameters you need to do what Annan and Hargreaves propose doing which is to:

• Get data from different areas.
• Do a Bayesian analysis to refine the estimate of the parameters based on data from those area.
• See what they are
• Project based on what the refined parameters show.
• Compare future data to projections. (Which they did. Obviosly not for too long-- but I saw projection and follow on data at their blog.)

No: you don't need to know what the policy is to estimate the values of the parameters actually are on the ground during a period with a more or less constant policy (e.g. "pre-lockdown"). You just find it. It's going to be different in different countries. That doesn't prevent you from finding the value in "pre-lockdown UK" or "pre-lockdown France" or Wuhan or wherever.

Also: Government's policies only dictate the size of one of the parameters. Government policy doesn't affect the time constant it takes for a person to go from being exposed from infected. They also don't dictate how long it takes an infected person to become non-infectious. Both are human biological response to the virus. The only thing lockdowns, masks or government policy affect is the rate at which infectious people infect susceptible people. That affects Ro, but not the other two.

 

[Warren J. Dew]

Lucia, while I think discussion of models and how they impact the potential for ballroom studio and competition reopenings is definitely on topic here, I think the modeling details you are discussing in the last two posts drift away from the issue of when we can expect ballroom activity to resume. I'm therefore answering them in the DA forum here:

Covid-19 NON DANCE debate

It's almost like a PR campaign for the government's mishandling of this. yes - exactly - ridiculous

www.dance-forums.com

 

[MerchantMe333]

You've got to be kidding. How are people going to get tested when even people who are exhibiting symptoms can't get testing?

And even if someone manages to wrangle a test, it will be worthless since the turnaround time is so long that someone could have been infected while waiting for the test results.

Let me reiterate: If a studio is worried about the coronavirus, they should require everyone to get tested. If a studio is not worried, then they should not. Wearing masks and gloves and whatever is a good preventative measure, but when you are literally hip to hip with someone, there is only so much it can do. Not to mention studios often have food for people to take, people will be sweating and touching tables, chairs, etc. and for rhythm and latin, sometimes people touch each others arms and stuff. Unless you take extremely preventative measures, there is no way to stop it from infecting everyone.

 

[raindance]

Let me reiterate: If a studio is worried about the coronavirus, they should require everyone to get tested. If a studio is not worried, then they should not. Wearing masks and gloves and whatever is a good preventative measure, but when you are literally hip to hip with someone, there is only so much it can do. Not to mention studios often have food for people to take, people will be sweating and touching tables, chairs, etc. and for rhythm and latin, sometimes people touch each others arms and stuff. Unless you take extremely preventative measures, there is no way to stop it from infecting everyone.

How worried a studio might be has little effect on availability of testing. Maybe you're from a country where "anyone that wants a test, gets a test," but that certainly is far from the case in the USA, which was snapdancer's point.

And if everyone is tested the first week, what about the next week, and the one after that? How frequently are you imagining they would "require everyone to get tested"?

Lastly, this virus is not is not spread by sweat. I agree though, that with a lot of people in close quarters, one infected person among the group is potentially bad news for the rest.

 

[snapdancer]

A dance studio could demand that everyone jump over the moon before they could come in. A dance studio could demand everyone be tested for COVID-19 that day before they could come in. Either way, they would have no customers who died of COVID-19 and for the same exact reason.

 

[lucia_l]

It might not hurt to take everyone's temperature. That won't stop asymptomatic spreaders, but at least it will catch a few infectious people.

 

[Cal]

i don’t know enough about it, but wouldn’t you need to know each person’s ”base” temperature before you could determine whether they had a fever or not? My “base” temperature runs between 97.6 and 97.7. So, my fever point might be a lot lower than that of someone whose base temperature is the standard 98.6. Yes? or maybe not - I don’t know.

 

[lucia_l]

Cal,
Maybe. I have no idea what my base temperature is.

But taking temperature is at least quick and easy. My understanding is the standard turn around for current antibody tests is about 28 days, so that would be rather useless as a test. Virus tests are potentially fast, but they aren't readily available right now. Even doctors don't have access to enough. A studio won't be able to do that until fairly inexpensive quick tests exist and they don't right now. (Or they don't unless things have changed. Things are changing quickly>)

Temperature might be coarse, but at least you can get a result quickly.

I'm all for having widespread antibody test eventually. But I don't think any business will be able to screen people at the door.