Are The Lockdowns Working/How Effective Are They?

[jim360]

As is usually the case, the answers are "possibly" and "fairly, maybe?". But this is also an interesting video to see how statistical analysis can be used to even try to answer the question.

[Peter Pedant]

zm stop carping !
why bother - because it is interesting to do
because it is there
because it gives you an idea of the future....

zacks - - no one is making you read this
if it offends you - read something else

[fiction-factory]

I will watch again tomorrow but it just occurred to me that the green line might have started falling because of measures that were being ramped up (where have I heard that before?) in the run up to the lockdown. If other countries acted like we did in UK, there was a gradual tightening in the run up to lock down (events cancelled, pubs closed, schools closed, discouraging family visits -eg for mothers day here) and maybe these were working but the time lag in the infection and testing, admission, deaths, meant it wasn't apparent the measures were having some effect already.

And yes, I also liked the way normal distributions were done from each day's data and then aggregated.

[jim360]

I'd imagine Cary knows about the SIR model but is viewing this as a stats exercise -- ie, given recorded cases can we infer actual date of infection -- which in my understanding would sort of bypass SIR. Same with my approach, where I'm currently just fitting a log-normal distribution to the data. There's not enough info about the tail yet to see if my fit is any good, but I'm getting reasonable results up to around the peak. Taking it literally, we'd probably be looking at 26,000+/- 5,000 deaths in English hospitals on the current trend by the "end" of the outbreak, but there's still a sizeable uncertainty here so I wouldn't take the central figure too seriously at the moment. Also, just to be clear, this isn't taking care home figures into account, and is subject to NHS England's data being accurate (and they freely acknowledge that it is continually updated and based on positive tests only).

[jim360]

It's not impossible that the Government panicked when they didn't need to, to be sure. By definition we can never know -- Sweden's model at least partly benefits from lower population, lower population density, and so on, and as the number of deaths per million there is still growing it's not clear that they have really done better anyway.

Compare instead the "crushing" of the curve in New Zealand, Australia, and South Korea, and it's clear, and indeed common sense, that strict lockdown measures absolutely slow the spread to a trickle.

[BilSuth]

Keep up the good work, Jim. You will see that I believe that we need both elements in the equation to work. The R figure apparently in now 0.62 for transmissions but I don't think we have the discipline needed to keep human interaction at a low level. There's too much backsliding already and the awaited phased easing of restrictions will put paid to hopes of a spectacular drop in infections.

[jim360]

Thanks, although I'm not sure that "keep up the Good work" is entirely fair. When it comes to Covid I'm just an interested observer.

[New Judge]

//…but I've noticed things slipping in the last week - e.g. I've seen a lot more family visits/overnight stays at neighbours , teenagers gathering in groups again and people going shopping twice a day for something to do,//

Indeed. Take a look at this thread:

https://www.theanswerbank.co.uk/ChatterBank/Question1704416-4.html#answer-12400388

The usefulness of the lockdown will reach a natural conclusion by about the end of May. The “breakouts” that you describe will become more widespread and those who choose to will simply ignore it. No amount of mathematical modelling will alter that.

//…there's a risk things may not fall as much as hoped for unless we have a short strong tightening to shift things downwards more quickly.//

And how, pray, will that be achieved? It is not possible to enforce the current regulations, let alone anything more severe. Where observance is taking place it is happening by consent not because of enforcement. It is simply not possible to enforce any sort of lockdown without the public’s consent and that is very likely to be withdrawn in large quantities soon. Yes, there are people who support the lockdown and they may continue to do so. But there are lots who do not.

//Sweden's model at least partly benefits from lower population, lower population density, and so on, and as the number of deaths per million there is still growing it's not clear that they have really done better anyway.//

Sweden may not have done any better (though that’s yet to be determined) but they certainly haven’t done any worse. The UK currently has a little over eight times the number of hospitalised infections and more than nine times the number of deaths though it only has around six times the population. Of course, as you mention, population density must play a big part but one thing Sweden has not done is to virtually paralyse its economy to the point of trashing it and has taken comparatively very few lockdown measures.

[fiction-factory]

>And how, pray, will that (a short tightening) be achieved?

I'm sure I covered that. I mentioned close all shops except maybe pharmacies, ban all travel without a permit, for example.

But it's just an idea to be shot down. No more outlandish than those who just want to go the other way and return to 'normal' even though NHS couldn't cope

[Peter Pedant]

latest from St Thomasina's and the London hospitals
on modelling and not really on 'and I have noticed how the tarts are out on the streets again: that is a good sign' side of things

video consultations dont really work if the patient is ill

admissions are fed into the real time GOMBOOLIAST database[great observer medical on line optimising ..... something else] to find out 1)who will need admission rather than tmt at home, 2)who will go to ITU and 3)of those who will survive.

kinda useful to know the answers to these qq.....
and it is finding nothing - there are no identified predictors of 1, 2 or 3.

they cant predict who will get to hospital, who will need ITU, who will need ventilation and who survives other than around 70% die on a ventilator. I am not sure they can show that stopping at CPAP is better.

so they are stuck at: people who do badly are male, fat and over 60 with heart disease. which we have known for a few months

[jno]

children may have their own problems, PP

https://www.theguardian.com/society/2020/apr/27/what-is-the-new-illness-affecting-children-and-is-it-linked-to-coronavirus

[fiction-factory]

That's concerning doctors, I know, jno, because there is no obvious explanation, but thankfully the number of cases needing treatment in ICU is very small - a dozen cases over the last 3 weeks- and as far as I know none have been fatal.

[Peter Pedant]

Is this the thread where someone suggested that if the population was spaced out or isolated then the epidemic would peter out ( pun intended)#
and I was er rather rude about it?

Times article today on this very point - counties with a lodda sheep and not many people were spared - wiltshire was an example
so I was going to post ( here) and say sorry - it is a valid point ....

but then the Skye outbreak - yeah really same day - became news - everyone in a care home in ultima Thule is now testing positive. Hacks writing the article were NOT aware of Panums results on the measles outbreak in Faroe Islands in er 1860
(they all died)

[fiction-factory]

Yes, that was one of my musings. The Skye one may be different though- it only needed one visitor/carer with the infection to come into the care home and it could spread like wildfire

[fiction-factory]

Thanks for the "What a load of cobblers!" warning before you posted that