Coronavirus

[johnb]

I see that The Guardian is again referencing the IHME report (that of the 2,932 peak and 66,314 deaths) in today's paper. To me this seems both irresponsible and lazy journalism.

IHME appears to be using a general model (i.e. not designed specifically for the UK) into which, at some point in the past, it fed the parameters for 140 countries, regions within countries and all US states. As far as the UK is concerned it *appears* that the model has only been updated by replacing forecast deaths with the actual numbers (to 5th April) but the the starting parameters don't appear to have been amended to reflect the current situation. There is also the issue of how accurate those initial parameters were, especially when the UK was just one of the 140 regions being modelled.

Something of the accuracy of IHME model and the figures quoted by The Guardian can be judged by comparing the model's projections for the 7th April with the actual data:

(The IHME gives mean, lower and higher projections. The Guardian has only quoted the mean projection.)

7th April: IHME vs Actual

Hospital beds for Covid-19 patients
IHME ("covidbeds needed by day"): mean 56,379, lower 40,911, higher 75,770
Actual: 19,438 excluding NI

Intensive Care Beds
IHME ("ICU covid beds needed by day"): mean 12,468, lower 9,164, higher 16,821
Actual: 3,067 excluding NI

Deaths
IHME: mean 1,233, lower 480, higher 2,614
Actual: 938, but the average of the past 3 days might be a better figure: 721

Rate of increase of the deaths as at 7th April
IHME: mean 20%, lower 8%, higher 35%
Actual: 13.3% (average of the last 3 days)

So as far as deaths are concerned the UK might be somewhat higher than the IHME lower projections but markedly lower than the mean projections quoted by The Guardian. Time will tell.

[LHC]

I see that The Guardian is again referencing the IHME report (that of the 2,932 peak and 66,314 deaths) in today's paper. To me this seems both irresponsible and lazy journalism.

IHME appears to be using a general model (i.e. not designed specifically for the UK) into which, at some point in the past, it fed the parameters for 140 countries, regions within countries and all US states. As far as the UK is concerned it *appears* that the model has only been updated to include the actual deaths but the the starting parameters don't appear to have been amended to reflect the current situation. There is also the issue of how accurate those initial parameters were, especially when the UK was just one of the 140 regions being modelled.

Something of the accuracy of IHME model and the figures quoted by The Guardian can be judged by comparing the model's projections for the 7th April with the actual data:

(The IHME gives mean, lower and higher projections. The Guardian has only quoted the mean projection.)

Hospital beds for Covid-19 patients
IHME ("covidbeds needed by day"): mean 56,379, lower 40,911, higher 75,770
Actual: 19,438 excluding NI

Intensive Care Beds
IHME ("ICU covid beds needed by day"): mean 12,468, lower 9,164, higher 16,821
Actual: 3,067 excluding NI

Deaths
IHME: mean 1,233, lower 480, higher 2,614
Actual: 938, but the average of the past 3 days might be a better figure: 721

Rate of increase of the deaths as at 7th April
IHME: mean 20%, lower 8%, higher 35%
Actual: 13.3% (average of the last 3 days)

So as far as deaths are concerned the UK might be somewhat higher than the IHME lower projections but markedly lower than the mean projections quoted by The Guardian. Time will tell.

The IHME Report (at least with respect to countries other than the US) appears largely to be garbage, and I am surprised that the Graun is still giving it credence.

[Dave2002]

The talk of plateaus and peaks is just bonkers. The curves that are shown in press conferences are rate curves. Now we're at 1000 deaths/day.

To get the total number of deaths integration is needed, which is the area under the curve. For a given shape of curve, a doubling of the height of the peak represents a quadrupling (x4) of the number of deaths. It's not quite elementary mathematics, but many people in the UK ought to understand this. Flattening out the peak does protect the situation and capacity in the health service and its ability to cope, but overall it's the area under the curve which represents the number of people who die. That's why I say this has to be stamped on vigorously.

It's no good saying "we're over the hump" - this is not a cycle race. It needs to be dealt with very hard indeed, to avoid a worse outcome.

[Dave2002]

Deaths
IHME: mean 1,233, lower 480, higher 2,614
Actual: 938, but the average of the past 3 days might be a better figure: 721

Rate of increase of the deaths as at 7th April
IHME: mean 20%, lower 8%, higher 35%
Actual: 13.3% (average of the last 3 days)

So as far as deaths are concerned the UK might be somewhat higher than the IHME lower projections but markedly lower than the mean projections quoted by The Guardian. Time will tell.

Sadly the comment about deaths now seems incorrect, as also in my earlier post #1437. It now appears to be over 1000 per day. Of course this is a lagged figure - a couple of weeks behind the infection onset data, but nevertheless a considerable concern.

I find comments about scientific use of data very suspect. For one thing, there is no control sample - and we don't want one either. That would imply a country with similar characteristics to the UK just letting things go rampant. Secondly, waiting for good "baseline" data to get more accurate results about whether interventions are effective! Absolutely not the right approach. I'll leave it there for the moment.

Regarding a control sample, the best we can do, or seem to have done, which was not unreasonable, was to use a computer model. While at one point the revised model seemed scary, now it seems to be very modest in its predictions - sadly. Thank goodness it was revised, as otherwise the situation would be a lot worse than it is now.

[LHC]

The talk of plateaus and peaks is just bonkers. The curves that are shown in press conferences are rate curves. Now we're at 1000 deaths/day.

To get the total number of deaths integration is needed, which is the area under the curve. For a given shape of curve, a doubling of the height of the peak represents a quadrupling (x4) of the number of deaths. It's not quite elementary mathematics, but many people in the UK ought to understand this. Flattening out the peak does protect the situation and capacity in the health service and its ability to cope, but overall it's the area under the curve which represents the number of people who die. That's why I say this has to be stamped on vigorously.

It's no good saying "we're over the hump" - this is not a cycle race. It needs to be dealt with very hard indeed, to avoid a worse outcome.

Based on the experience of Italy and Spain, who are a few weeks ahead of us, I think this weekend, and perhaps the first half of next week, are expected to be the peak period, after which we will hopefully start to see the daily death rate start to fall. However, having around 1,000 deaths a day for the next few days will be devastating for all those families, and for the people in the NHS treating them. And as you say, there must be no relaxation of the lockdown until we see a significant and consistent fall in the numbers dying each day.

Seeing the report on the BBC that Manchester police had to deal with 494 house parties - some with DJs, fireworks and bouncy castles - and 166 street parties last weekend makes you wonder at the selfishness and stupidity of some of our fellow citizens.

[Joseph K]

Seeing the report on the BBC that Manchester police had to deal with 494 house parties - some with DJs, fireworks and bouncy castles - and 166 street parties last weekend makes you wonder at the selfishness and stupidity of some of our fellow citizens.

What happened to herd immunity?

[Dave2002]

What happened to herd immunity?

Nobody knows right now if it exists in any way, but the idea that the UK could have just let the virus run rip through the country and herd immunity would somehow save the day was booted out weeks ago.

Re my post #1454 above, I may have overestimated the daily number of deaths, whereas previously it was perhaps underestimated. The data I see comes in from various sources, and it is somewhat erratic. Today the official NHS data was below 900, though as explained at the press conference, the NHS data is quite robust and based on people who actually have had the virus. It thus represents a lower bound for the data each day.

There is an interesting report in the New York Times that the outbreak there was spread by people coming from Europe, not from Asia. That is interesting for several reasons, one of which seems to be that the scientists and doctors in the USA are now able to differentiate between different strains of CV-19 with some certaintly. I don't know if this claim is actually true, but it is an interesting thing to note.

https://www.nytimes.com/2020/04/08/science/new-york-coronavirus-cases-europe-genomes.html?action=click&module=Spotlight&pgtype=Homepage

[oddoneout]

What happened to herd immunity?

The Netherlands still favours that approach apparently. https://www.bbc.co.uk/news/world-europe-52135814

[Dave2002]

The Netherlands still favours that approach apparently. https://www.bbc.co.uk/news/world-europe-52135814

Sweden also appears to be somewhat "relaxed", though I have contacts there, and could try to find out more. It's possible that individual people will do their own social distancing, even though the government there hasn't AFAIK imposed severe lockdowns yet.

[muzzer]

The response of the US has been even worse than here in the U.K. The scariest aspect of the whole thing is not being able to tell true from false. I realise this sounds alarmist and simplistic but that’s how I feel. So I’m steeling myself for another 3 weeks of this minimum. But even then of course the things that keep me sane that aren’t books - concerts, galleries, restaurants and pubs - won’t be open anytime soon. At least it isn’t winter!

[ardcarp]

Hope you don't remain stolen too long!

I'm a bit cheesed off with TV News. There must be other things...serious things...going on in the world other than Covid and which are just squeezed out of the news agenda. Today Channel 4 News (which I usually respect) made great play of a sole person on a vast empty beach being told by the fuzz to move on. What's happening in Syria, for instance, and what's Putin up to? The RN had to be alerted the other day because Russian warships were manoeuvring close to our shores in the N. Sea.

[DracoM]

Hope you don't remain stolen too long!

I'm a bit cheesed off with TV News. There must be other things...serious things...going on in the world other than Covid and which are just squeezed out of the news agenda. Today Channel 4 News (which I usually respect) made great play of a sole person on a vast empty beach being told by the fuzz to move on. What's happening in Syria, for instance, and what's Putin up to? The RN had to be alerted the other day because Russian warships were manoeuvring close to our shores in the N. Sea.

Could not agree MORE!!!

I'm more or less trying to avoid ANY news at the mo, and most of the BBC R4 progs if it comes to that who carry at least one if not several items whatever the umbrella subject matter for that prog usually is. However, one thing I have also stopped listening to is the wretched / one bl.... track-obsession - The Archers. Blimey!

I think we got it that this is a crisis, but to go on blackening the picture with ever-deepening misery stories is just NOT helping the national mood - or it isn't in my neck of t'woods.

[LMcD]

Hope you don't remain stolen too long!

I'm a bit cheesed off with TV News. There must be other things...serious things...going on in the world other than Covid and which are just squeezed out of the news agenda. Today Channel 4 News (which I usually respect) made great play of a sole person on a vast empty beach being told by the fuzz to move on. What's happening in Syria, for instance, and what's Putin up to? The RN had to be alerted the other day because Russian warships were manoeuvring close to our shores in the N. Sea.

The 6.00 p.m. news on Radio 4 (Monday to Friday) and BBC World News bulletins are somewhat less parochial.

[Joseph K]

I think we got it that this is a crisis, but to go on blackening the pitcure with ever-deepening misery stories is just NOT helping the naitonal mood - or it isn;t in my neck of t'woods.

Well, here's a feel-good story... from my neck of the woods.

https://redditchstandard.co.uk/news/101-year-old-keith-beats-covid-19-after-two-weeks-in-the-alex/

[MrGongGong]

Could not agree MORE!!!

I'm more or less trying to avoid ANY news at the mo, and most of the BBC R4 progs if it comes to that who carry at least one if not several items whatever the umbrella subject matter for that prog usually is. However, one thing I have also stopped listening to is the wretched / one bl.... track-obsession - The Archers. Blimey!

I think we got it that this is a crisis, but to go on blackening the picture with ever-deepening misery stories is just NOT helping the national mood - or it isn't in my neck of t'woods.

Linda will rise on Sunday
(Boris seems to have done it three days early )