Coronavirus Briefing with Whitty and Vallance (21/09)

[Freightmaster]

Unfortunately I can't post the graphs, but it seems that this analysis of cases by date of sample don't seem to be supporting Wittys exponential growth.

Here is a graph I saw posted on another forum:

Cases are clearly doubling every seven days...





MARK

 

[trebor79]

TBF I believe they were referring to the PHE data when they said cases were doubling every 7 days. But it's still a very large leap of faith to state that's the trajectory we are on - after a one week increase after weeks and weeks of stability.

 

[Domh245]

The issue with that analysis however is that it seems to ignores the fact that test capacity has begun to max out, and the number of tests taken was artificially limited as a result. The utilisation of pillar 2 (general public) tests was over 100% between the 11th and 14th (based on publish date), and the average across both pillar 1 and 2 has been 93% since the 11th - it'd be naïve to think that isn't having some sort of distortive effect on the numbers of cases detected, and when.

 

[trebor79]

The issue with that analysis however is that it seems to ignores the fact that test capacity has begun to max out, and the number of tests taken was artificially limited as a result. The utilisation of pillar 2 (general public) tests was over 100% between the 11th and 14th (based on publish date), and the average across both pillar 1 and 2 has been 93% since the 11th - it'd be naïve to think that isn't having some sort of distortive effect on the numbers of cases detected, and when.

PHE testing is pillar 4.

 

[Domh245]

PHE testing is pillar 4.

Correct - pillar 4 is the PHE/ONS surveillance testing which has shown a rise for the last few weeks it has been published (notably, they're published a week behind the data they've used), my post was in relation to the analysis linked in posts 119 & 121

 

[seaviewer]

I have just been looking at the data – from the European Centre for Disease Prevention and Control. I have no idea where the "doubling every seven days" comes from. It certainly does not apply to the UK as a whole, where a better fit to the data is obtained with a linear relationship. To be fair, they did say it was a projection (I forget the exact wording), but it seems far more pessimistic than is justified by the current trend.

 

[Yew]

Looking at the growth rate on worldometers, it seems like since the 15th, the increase in cases has reduced in speed, rather than exhibited exponential growth. I'd imagine this would be awkward if you'd predicted 50000 cases in a few weeks.

 

[Bletchleyite]

Looking at the growth rate on worldometers, it seems like since the 15th, the increase in cases has reduced in speed, rather than exhibited exponential growth. I'd imagine this would be awkward if you'd predicted 50000 cases in a few weeks.

This is because of limited test availability. The figures predicting this apparently came from a prevalence study.

 

[DB]

This is because of limited test availability. The figures predicting this apparently came from a prevalence study.

But if it was as high as some are suggested, we would be expecting to see a significant increase in hospital admissions by now - that doesn't seem to be happening.

 

[trebor79]

But if it was as high as some are suggested, we would be expecting to see a significant increase in hospital admissions by now - that doesn't seem to be happening.

If you're looking at the "patients admitted" number on the government coronavirus website here:

UKHSA data dashboard

Overall summary of the respiratory viruses in circulation within the UK

coronavirus.data.gov.uk

that is a very out of date number. If you click on "more about healthcare" you will see that admissions in England have more than doubled over the past 10 days or so.
It's not panic stations armageddon yet, and prerumably some of those will have some other condition that the PCR test has picked up as COVID, but it's undeniable there is an increase taking place.

The government website is very good, except for the summary page being so completely misleading in a number of ways.

 

[duncanp]

I'm not sure what value the 10clock kerfew will be - but it is at least relatively limited in scope (hopefully that's last orders, rather than chucking out ?).

The 10pm curfew IS when people have to leave the premises unfortunately, so last orders will have to be about 9:30pm.

 

[jtuk]

If you're looking at the "patients admitted" number on the government coronavirus website here:

UKHSA data dashboard

Overall summary of the respiratory viruses in circulation within the UK

coronavirus.data.gov.uk

that is a very out of date number. If you click on "more about healthcare" you will see that admissions in England have more than doubled over the past 10 days or so.

It would not surprise me if a decent percentage of these are people who have gone in for something completely unrelated to Covid and then tested false positive

 

[david1212]

@david1212 I strongly disagree; what's your proposal? Make everyone redundant and stay at home until a vaccine is rolled out?

That is not was I intended to imply at all as I did not mention any change to businesses allowed to trade.

Rather simply that the guidelines to keep 2m distance from everyone except your household / bubble as much as possible and where not possible the absolute minimum to be 1m with extra precaution e.g. a face covering were widely ignored by the end of July.

My view is that if these had been followed the level of transmission i.e. ' R ' would now be significant lower, perhaps for England overall 0.8 - 1.1 rather than the latest range of 1.1 - 1.4.

 

[Islineclear3_1]

I don't know what value 10 o'clock closing time would offer - is the virus gonna lie in wait and ambush at chucking out time? I did not know the virus could tell the time

 

[IanXC]

For all those pondering the 50,000 number, take a look at the first graph in this article...

Coronavirus: Last throw of dice before the really tough call

New restrictions are being introduced - but there is fear this will not be enough to curb the spread.

www.bbc.co.uk

BBC News article by Nick Triggle which shows the 50,000 figure claimed by Whitty and Vallance overlaid with the actual experience in France and Spain, and showing that if our infection rate proceeded as per theirs, we'd be at 10,000 by mid October, and not the 50,000 figure.

 

[initiation]

we'd be at 10,000 by mid October, and not the 50,000 figure.

Of course the Gov can then claim the measures have been sucessful. If they don't work, it's because of the public.

 

[big_rig]

For all those pondering the 50,000 number, take a look at the first graph in this article...

Coronavirus: Last throw of dice before the really tough call

New restrictions are being introduced - but there is fear this will not be enough to curb the spread.

www.bbc.co.uk

BBC News article by Nick Triggle which shows the 50,000 figure claimed by Whitty and Vallance overlaid with the actual experience in France and Spain, and showing that if our infection rate proceeded as per theirs, we'd be at 10,000 by mid October, and not the 50,000 figure.

This is a really shocking article, isn’t it. The message I took from it is essentially that Whitty and Vallance decided to deceive/lie to the public to try scare people into changing behaviours to meet what they see as a terrifying threat. The ends justify the means for them.

They gave themselves plausible deniability by couching it as just a ‘potentials scenario and not a prediction’ but it isnt like they’re going to be crying into their Weetabix this morning because the PM misinterpreted their potential ‘what if’ and introduced knee jerk restrictions when he shouldn’t have. They knew full well what they were doing.

 

[DelayRepay]

They gave themselves plausible deniability by couching it as just a ‘potentials scenario and not a prediction’ but it isnt like they’re going to be crying into their Weetabix this morning because the PM misinterpreted their potential ‘what if’ and introduced knee jerk restrictions when he shouldn’t have. They knew full well what they were doing.

I expect any crying will be because they feel they weren't listened to, and the Prime Minister's announcement yesterday didn't go far enough. I think whatever measures they recommended were probably much more restrictive than those announced, e.g. ban on having visitors at home. Perhaps they are wishing they used the 'quadrupling every two days' scenario instead...

 

[Bletchleyite]

For all those pondering the 50,000 number, take a look at the first graph in this article...

Coronavirus: Last throw of dice before the really tough call

New restrictions are being introduced - but there is fear this will not be enough to curb the spread.

www.bbc.co.uk

BBC News article by Nick Triggle which shows the 50,000 figure claimed by Whitty and Vallance overlaid with the actual experience in France and Spain, and showing that if our infection rate proceeded as per theirs, we'd be at 10,000 by mid October, and not the 50,000 figure.

The graph there is the national figures which are incorrectly "capped out" by the lack of tests. The prevalence tests (random testing of particular numbers of people nationally and extrapolating) are where the figure they used came from, so it isn't flawed in the way people are saying.

Like most of this, before rubbishing something you need to spend the time to understand it.

 

[hwl]

The graph there is the national figures which are incorrectly "capped out" by the lack of tests. The prevalence tests (random testing of particular numbers of people nationally and extrapolating) are where the figure they used came from, so it isn't flawed in the way people are saying.

Like most of this, before rubbishing something you need to spend the time to understand it.

Exactly - public request based testing in July / August was estimated to be finding ~50% of community cases based on the parallel prevalence testing stream. The problem currently is that request based testing can only do a third (Harding last week in select committee) of what is being requested is September, hence the ~50% is likely to have fallen so the real case number multiplier to get from +ve test numbers to an estimate of total cases will almost certainly have increased and that is the interesting number! The problem is that number is also a very good metric to judge the performance (or not) of government testing

 

[island]

I don't know what value 10 o'clock closing time would offer - is the virus gonna lie in wait and ambush at chucking out time? I did not know the virus could tell the time

Of course not. But closing earlier means people get less drunk, and it has been observed that people observe physical distancing and other precautions much less when they get drunk.

 

[Bletchleyite]

Of course not. But closing earlier means people get less drunk, and it has been observed that people observe physical distancing and other precautions much less when they get drunk.

And 10pm is late enough that many will just go home after, on last buses etc which are often between 10 and 11pm if there is an evening service. Had it been 9 or even 9:30 I reckon it'd have prompted more afterparties at home.

I wonder will lock-ins return?

 

[takno]

And 10pm is late enough that many will just go home after, on last buses etc which are often between 10 and 11pm if there is an evening service. Had it been 9 or even 9:30 I reckon it'd have prompted more afterparties at home.

I wonder will lock-ins return?

I think this is pretty fanciful, but even if it does work out then we can presumably expect some exciting super-spreading events on these packed last buses

 

[jtuk]

Of course not. But closing earlier means people get less drunk

If someone is going to get pissed, they will get pissed, be it by drinking earlier, faster or harder

 

[Bletchleyite]

If someone is going to get pissed, they will get pissed, be it by drinking earlier, faster or harder

That assumes they are setting out to do that. Many people who roll out of pubs drunk at midnight do so because of "oh, just another one?" Shorten the session by 2 hours and that's 2-3 fewer pints.

 

[Islineclear3_1]

Of course not. But closing earlier means people get less drunk, and it has been observed that people observe physical distancing and other precautions much less when they get drunk.

I wouldn't be so sure. Either people will venture into pubs earlier or just buy more drinks when they're there and then drink faster before they're chucked out

That assumes they are setting out to do that. Many people who roll out of pubs drunk at midnight do so because of "oh, just another one?" Shorten the session by 2 hours and that's 2-3 fewer pints.

Or, these people will just go into pubs 1-2 hours earlier if they can...

 

[Ianno87]

That assumes they are setting out to do that. Many people who roll out of pubs drunk at midnight do so because of "oh, just another one?" Shorten the session by 2 hours and that's 2-3 fewer pints.

Ahhh, that takes me back to the pre-Kids days...

 

[IanXC]

The graph there is the national figures which are incorrectly "capped out" by the lack of tests. The prevalence tests (random testing of particular numbers of people nationally and extrapolating) are where the figure they used came from, so it isn't flawed in the way people are saying.

Like most of this, before rubbishing something you need to spend the time to understand it.

That doesn't change the fact that other than their 'worst case line' all the other data is comparable. Their argument went, this is what's happened in the UK, we know what the risk is because we have seen what has happened in France and Spain, here's a graph with a line that gets to 50,000, and used that as their argument. We do indeed know what happened in France and Spain and it was not the line they presented in that graph.