[Muscle.Coach]

Get yer muzzle back on before you contract the dreaded ‘long-covid’.

[Oldmanmatt]

Get yer muzzle back on before you contract the dreaded ‘long-covid’.

No, you!
😝

[Fultonius]

Finally starting to feel a little better, some more focus. I think I must have had a mild re-infection but will never know!

Very interesting article in Scientific American:

https://www.scientificamerican.com/article/what-we-know-so-far-about-how-covid-affects-the-nervous-system/

[lagerstarfish]

The impact on health and welfare outweighs the benefits.

Over what time period are you thinking?
In terms of deaths and lost QALYs, by how much do you estimate this?
(I would appreciate a brief explanation of the method used to arrive at the estimate)

[slab_happy]

Thought I'd dump these links here, as the government rules become increasingly arcane (and detached from the actual science) -- I've found all these very helpful for understanding the current scientific consensus re: the relative risks of different situations and activities.

https://www.bmj.com/content/370/bmj.m3223

https://www.erinbromage.com/post/the-risks-know-them-avoid-them

http://www.ezekielemanuel.com/writing/all-articles/2020/06/30/covid-19-activity-risk-levels

https://elemental.medium.com/the-most-likely-way-youll-get-infected-with-covid-19-30430384e5a5

[tomtom]

Nice guide to aerosol transmission here from El Pais (in English).

https://english.elpais.com/society/2020-10-28/a-room-a-bar-and-a-class-how-the-coronavirus-is-spread-through-the-air.html

[Stu Littlefair]

Worth pointing out that article should come with a huge flashing caveat. The model is calibrated based upon super spreading events... This model shows you what would happen if you are unlucky enough to be in a super-spreading event in various places.

Most people (perhaps 8/10) do not cause super spreading events, so we might roughly say that 8 times out of ten putting yourself in the scenarios shown will NOT lead to the outcome suggested...

[Johnny Brown]

Sure, but in these situations we don't know who is infected, and we don't know who is super-spreader. So it's best to proceed on the basis that everyone is a potential super spreader. It has given me food for thought on how we run our classes.

[slab_happy]

This is a pre-print, but models transmission to suggest that there's no need to postulate particular individuals who are "super-spreaders" in order to have super-spreading events:

https://www.medrxiv.org/content/10.1101/2020.10.21.20216895v1.full.pdf

[Oldmanmatt]

Worth pointing out that article should come with a huge flashing caveat. The model is calibrated based upon super spreading events... This model shows you what would happen if you are unlucky enough to be in a super-spreading event in various places.

Most people (perhaps 8/10) do not cause super spreading events, so we might roughly say that 8 times out of ten putting yourself in the scenarios shown will NOT lead to the outcome suggested...

True, but...

I installed a massive extraction system at the wall. Far bigger than was recommended or than my own calculations called for (I have a fair amount of experience designing HVAC systems for ships), but, looking at those numbers, I want to run their calculator ( it’s very glitchy on a pad, I will try on a Chrome book later). Nevertheless, I think I might buy another fan even so, because I can bring the place to a virtually “outdoor” standard.
I went in last night, and though we don’t require people to wear masks whilst climbing/exercising (all other times except sat in cafe eating/drinking), many were actually wearing masks to climb.
Almost everybody in last night came in on an NHS entry or membership, which might be telling...

[Stu Littlefair]

There's no "but" here - the article is great for informing best practice etc, but it gives a totally misleading idea of the risks of being in a room with 6 other people...

[Stu Littlefair]

This is a pre-print, but models transmission to suggest that there's no need to postulate particular individuals who are "super-spreaders" in order to have super-spreading events:

https://www.medrxiv.org/content/10.1101/2020.10.21.20216895v1.full.pdf

That's sort of missing the point. The paper shows that you don't need to have an index case with viral loads at the upper end of the distribution to have a super-spreading event. Fine.

But it's perfectly obvious that the majority of scenarios like the ones in the el-pais article don't lead to the outcomes shown. In every one of those scenarios the index case causes 5+ secondary cases, even given moderate controls (e.g reduced bar capacity).

How then was R0 only ~3 before any controls were introduced? It's simply not possible if the modelled scenarios are typical.

[RobK]

Nice guide to aerosol transmission here from El Pais (in English).

https://english.elpais.com/society/2020-10-28/a-room-a-bar-and-a-class-how-the-coronavirus-is-spread-through-the-air.html

"a conclusion backed by the European Center for Disease Control and Prevention’s (ECDC) observation that not a single case of fomite-caused Covid-19 has been observed"

Hang on, I feel like this is fairly major or I am missing something blindingly obvious (the latter is more likely). There are no known cases of surface Covid tranmission...at all?!

[danm]

This is a pre-print, but models transmission to suggest that there's no need to postulate particular individuals who are "super-spreaders" in order to have super-spreading events:

https://www.medrxiv.org/content/10.1101/2020.10.21.20216895v1.full.pdf

That's sort of missing the point. The paper shows that you don't need to have an index case with viral loads at the upper end of the distribution to have a super-spreading event. Fine.

But it's perfectly obvious that the majority of scenarios like the ones in the el-pais article don't lead to the outcomes shown. In every one of those scenarios the index case causes 5+ secondary cases, even given moderate controls (e.g reduced bar capacity).

How then was R0 only ~3 before any controls were introduced? It's simply not possible if the modelled scenarios are typical.

I'm probably being incredibly dumb here: how can you relate the number of people infected in these scenarios to a R number in the population? I don't spend all my time in the pub, in an office with other people etc, so for most of the time these models wouldn't apply to me?

[tomtom]

Here’s an example of (I think - not checked it out fully) a classic CV19 FAIL by peer review! SOTEN (the journal it’s in) is a really well renowned publication - and not especially easy to get a paper in!

https://retractionwatch.com/2020/10/29/amulets-may-prevent-covid-19-says-a-paper-in-elsevier-journal-they-dont/

[Stu Littlefair]

This is a pre-print, but models transmission to suggest that there's no need to postulate particular individuals who are "super-spreaders" in order to have super-spreading events:

https://www.medrxiv.org/content/10.1101/2020.10.21.20216895v1.full.pdf

That's sort of missing the point. The paper shows that you don't need to have an index case with viral loads at the upper end of the distribution to have a super-spreading event. Fine.

But it's perfectly obvious that the majority of scenarios like the ones in the el-pais article don't lead to the outcomes shown. In every one of those scenarios the index case causes 5+ secondary cases, even given moderate controls (e.g reduced bar capacity).

How then was R0 only ~3 before any controls were introduced? It's simply not possible if the modelled scenarios are typical.

I'm probably being incredibly dumb here: how can you relate the number of people infected in these scenarios to a R number in the population? I don't spend all my time in the pub, in an office with other people etc, so for most of the time these models wouldn't apply to me?

Because we do have a fairly decent idea of how long people are infectious for, and what they were doing with their time before the pandemic. Suppose you are infectious for a day or two. The average infected person before controls were in place would find themselves in one of these environments - enclosed transport, offices, pub, climbing wall.

The only way you could have the scenarios shown be typical and still have R0~3 is if people were only infectious for a few hours, so there’s a high chance they weren’t in such situations when they were infectious.

[Oldmanmatt]

This is a pre-print, but models transmission to suggest that there's no need to postulate particular individuals who are "super-spreaders" in order to have super-spreading events:

https://www.medrxiv.org/content/10.1101/2020.10.21.20216895v1.full.pdf

That's sort of missing the point. The paper shows that you don't need to have an index case with viral loads at the upper end of the distribution to have a super-spreading event. Fine.

But it's perfectly obvious that the majority of scenarios like the ones in the el-pais article don't lead to the outcomes shown. In every one of those scenarios the index case causes 5+ secondary cases, even given moderate controls (e.g reduced bar capacity).

How then was R0 only ~3 before any controls were introduced? It's simply not possible if the modelled scenarios are typical.

I'm probably being incredibly dumb here: how can you relate the number of people infected in these scenarios to a R number in the population? I don't spend all my time in the pub, in an office with other people etc, so for most of the time these models wouldn't apply to me?

Because we do have a fairly decent idea of how long people are infectious for, and what they were doing with their time before the pandemic. Suppose you are infectious for a day or two. The average infected person before controls were in place would find themselves in one of these environments - enclosed transport, offices, pub, climbing wall.

The only way you could have the scenarios shown be typical and still have R0~3 is if people were only infectious for a few hours, so there’s a high chance they weren’t in such situations when they were infectious.

Ok...

Isn’t that backwards? I mean, not that (in that scenario) were only infectious for a few hours, but that they were only in those situations for a few hours?
Edit: or at least, in those situations long enough, ie a lesson usually being an hour or less.
The rest of the time their virus went to “waste”?

[Stu Littlefair]

It’s not backwards, it’s just playing with the only variable not fixed in these simulations.

Whoever ran the simulations took a good online tool for simulating aerosol infections.

https://cires.colorado.edu/news/covid-19-airborne-transmission-tool-available

There are many inputs to this model, but the El Pais looks at fixed scenarios - eg the class has a fixed room size and no ventilation etc. The main unknown variable is the amount of virus expelled by the infected person. You can download the tool and play with it - to get these levels of infection the modellers have tweaked this number to it’s upper limits.

For more reasonable values and a class of 50mins you’d get a 0.5% chance of infection per class member, which is much more realistic.

But suppose you take the El Pais numbers as representative. And suppose someone is infectious for several days. They would find themselves in at least one of these scenarios in that time and infect around 15 people, which is miles away from the 2-3 we actually saw.

So there are only two possibilities

1) people aren’t infectious for 2-3 days, but only a short period and so probably didnt attend class, or an office meeting whilst infectious.

2) the el Pais simulations use settings applicable to super spreading events where the index case expels a lot of virus.

And, if you go and play with the tool you see the second one is exactly what has happened.

So the article is great in that it shows you what you need to do to make classes/restaurants/family christmas as safe as possible. It’s terrible for giving you an indication of how risky that event actually is.

[tomtom]

I’d agree but modify your last sentence Stu - The article is not good for telling you how risky an event is likely to be. Its better for informing you about how risky an event could be.. (in a bad case)

[Stu Littlefair]

Yes. Well put

[tomtom]

Rishi loses some of his gloss: Eat out to Help out led to fresh clusters of cases

https://news.sky.com/story/coronavirus-eat-out-to-help-out-accelerated-second-wave-of-covid-19-study-says-12118285

[seankenny]

More details on that report here:

https://twitter.com/fetzert/status/1322078576133525504?s=21

Wasn’t this a rather obvious consequence of the policy?

[Sidehaas]

Rishi loses some of his gloss: Eat out to Help out led to fresh clusters of cases

https://news.sky.com/story/coronavirus-eat-out-to-help-out-accelerated-second-wave-of-covid-19-study-says-12118285

This has been blatantly obvious to me for almost exactly 2 months (when all the cases in our area started rising exponentially, after seeing the state of the local restaurants for the previous 4 weeks).

[Will Hunt]

I don't think anybody was under any illusions that the scheme wouldn't lead to more cases, but literally any policy less severe than chaining everybody to their radiators for the whole summer would have had to same effect to a greater or lesser degree. The question is, was it worth it?
Given the better climatic conditions over summer I think it was right to let people live a little because, EOTHO or no, a tightening of restrictions over winter was inevitable. Having said that, the government should have been quicker with restrictions this time round and should also have been stricter about behaviour in restaurants/pubs over the summer.

We took advantage of it by going out for lunch to places where we could sit outside, and we bought more coffees/ice creams-to-go than we otherwise would have done, but seeing photos of multiple generations crowded round a table in a curry house seemed daft.

[dunnyg]

Having the extra ice cream is very good  of you x