Coronavirus

[Count Boso]

Getting back to recovery rates, it was announced on 19 May (by The Sun, an article which I found yesterday but now can't) that the daily recovery figures would be published daily from (a date in) June.

The story was picked up (quoting The Sun) by The Express:

https://www.express.co.uk/news/uk/1284030/coronavirus-recovery-rate-uk-covid-19-figures-death-toll-return-to-work-lockdown-johnson

and France News

https://www.fr24news.com/a/2020/05/number-of-britons-who-beat-the-coronavirus-to-publish-daily-to-get-people-back-to-work-the-sun.html

'A Downing Street source added, “We have been asking for recovery rates for weeks, but the Department of Health just hasn’t kept the numbers.

“It is important for us to provide the public with a wide range of statistics so that they get the full picture, not just the most striking ones. “

This decision is possible because scientists are now receiving much more data on the epidemic, thanks in large part to a large infection survey conducted by the Office of National Statistics.

Until recently, NHS chiefs were only able to give recovery rates for patients with coronavirus admitted to hospital.

The vast majority of other less serious cases have not been documented because the victims were unable to get tested.'

[Cockney Sparrow]

..................
'A Downing Street source added, “We have been asking for recovery rates for weeks, but the Department of Health just hasn’t kept the numbers.

“It is important for us to provide the public with a wide range of statistics so that they get the full picture, not just the most striking ones. “

This decision is possible because scientists are now receiving much more data on the epidemic, thanks in large part to a large infection survey conducted by the Office of National Statistics..............'

Oh yeah..... of course. Bear in mind the Downing St "source" will have been a Cummings appointee, if not the man himself. Seems finding the correct perspective on available data is absorbing a lot of time for Johnb. For myself, with no refined statistical skills, I read Johnb's posts and a couple of papers & Private Eye (demurring at the cost of the FT online) watch Newsnight and form the best impression I can.

As far as I'm concerned anything coming from no 10 or any departmental source is just "presentation" so pretty suspect. All communications emanate via Cummings' SPADs so straightforward honesty is in the past - its all part of the drive to distract us from the awful truth we have handled the pandemic almost as poorly as any other country (it doesn't feel good at all to think we might end up ahead of Trump's USA or Putin's Russia).

[johnb]

'A Downing Street source added, “We have been asking for recovery rates for weeks, but the Department of Health just hasn’t kept the numbers.

“It is important for us to provide the public with a wide range of statistics so that they get the full picture, not just the most striking ones. “

This decision is possible because scientists are now receiving much more data on the epidemic, thanks in large part to a large infection survey conducted by the Office of National Statistics.

Until recently, NHS chiefs were only able to give recovery rates for patients with coronavirus admitted to hospital.

The vast majority of other less serious cases have not been documented because the victims were unable to get tested.'

This type of briefing is infuriating.

If hospital recovery rates were available why on earth haven't they been published?

Of course the "vast majority of other less serious cases have not been documented" - that's bleeding obvious. It is also totally irrelevant.

There are recovery rates for different groups of infected patients of course:

Patients in hospitals. If this is available it should have been published.

Those in care homes who contracted the virus. The recovery rates in care homes would probably make very depressing reading indeed, but it should have been collated.

Those in the general population. These *might* be gauged by the ONS surveys otherwise there is obviously no information.

Incidentally, the other week I attempted to get a rough idea of the recovery rate of hospital patients in England - using:

daily admissions data for England (from Press Briefing datasets)
daily number of beds occupied for England (from Press Briefing datasets)
- to calculate the discharges
daily hospital deaths by date of death (from NHS England)
- to calculate the percentage of patients who survived using a 7 day rolling average to ameliorate any timing mismatches

This is only a rough and ready guestimate but I came up with:

Around 50% hospital survival at the peak of the pandemic in early April

steadily increasing to:

Around 80% hospital survival a week ago.

Please take these guestimates with a whole bucket full of salt. They might be way out. I might have got the whole thing hopelessly wrong.

Later edit.
One *possible* reason for the increase in the percentage of discharged hospital patients who have recovered might be that recovery can be slow and for some patients it can take many weeks. This would mean that, at the peak of the pandemic, patients were dying much sooner than those who eventually recovered - hence the lower recovery rate at the peak.

[Count Boso]

This type of briefing is infuriating.
<snip>
Please take these guestimates with a whole bucket full of salt. They might be way out. I might have got the whole thing hopelessly wrong.

At least they're a bit more reassuring than was reported a month ago!

https://www.citymetric.com/fabric/epidemiologist-update-uk-reports-notably-low-recovery-rate-5077

"While the daily confirmed cases and deaths are decreasing in most European countries, the UK experienced a noticible(!) increase in daily new cases on 6 May. While the number of daily confirmed death is still decreasing in the UK, the total recovered cases in the UK is very low at about 960 cases. Compared to the global recovery rate of 32%, UK's rate of 0.46% is concerning and shows that UK still has far to go until full recovery."

[DracoM]

Enjoy this - with a wince:

https://www.facebook.com/185180654855189/videos/541864546502790

[MrGongGong]

https://www.theneweuropean.co.uk/top-stories/uk-population-being-fooled-by-government-1-6684939?utm_source=Facebook&utm_medium=Social_Icon&utm_campaign=in_article_social_icons&fbclid=IwAR0X0MVfVMGKS-VsO4BVa7J-ZDvH3YVueC6Fa-WT8xJyg5oQ738b9d64-qI

[johnb]

At least they're a bit more reassuring than was reported a month ago!

https://www.citymetric.com/fabric/epidemiologist-update-uk-reports-notably-low-recovery-rate-5077

"While the daily confirmed cases and deaths are decreasing in most European countries, the UK experienced a noticible(!) increase in daily new cases on 6 May. While the number of daily confirmed death is still decreasing in the UK, the total recovered cases in the UK is very low at about 960 cases. Compared to the global recovery rate of 32%, UK's rate of 0.46% is concerning and shows that UK still has far to go until full recovery."

That is an interesting article but I am a bit puzzled about the 0.46% for the UK recovery rate. That is roughly 1 in 200 and just cannot be correct. They might have got the decimal point in the wrong place ... yet ....

The article was published on the 8th May.

On the day before that, the 7 day rolling average for the hospital cases recovered was 967 according to my possibly flawed calculations.

This roughly ties in with the article saying "the total recovered cases in the UK is very low at about 960 cases".

But, over the same days, the 7 day rolling average for the total hospital discharges was 1,268 (once again, according to my possible flawed calculations).

This would give a hospital recovery rate of 76% at that date. Not 0.46%.

Admittedly what I am calculating is the percentage of the people who are discharged from hospital that have recovered, i.e. who have been discharge alive. That is not the same as the percentage of the people who are admitted to hospital who eventually recover - though the two figures should oven out over time.

[LeMartinPecheur]

...the two figures should oven out over time.

Is this some new half-baked statistical concept johnb?

[Flay]

Listen to R4 More or Less and weep

[oddoneout]

Is this some new half-baked statistical concept johnb?

Oven-ready statistics to go with the oven-ready deals?

[DracoM]

Listen to R4 More or Less and weep

A programme that has become even more of a must-hear - a light in darkness.

[Count Boso]

Your calculation is consistent with that. 0.46% is utterly absurd!!

There were a number of references to that figure but, worldlywebwise, they all seem to stem from CityMetric, whatever that is. However, one thing puzzles me - thinking of the high recovery numbers reported by other countries - the vast majority of 'reported cases' never go to hospital at all, do they? So how do they compute the number of those who recover at home? (Some of the published figures are from samples, so are calculations rather than measurements.)

For some countries, if you add up the number of recoveries and the number of deaths the sum comes to something not far off from the total number of cases.

[Dave2002]

Silent spreaders - an interesting article here - https://www.bbc.co.uk/news/uk-528407...5D-%5Bisapi%5D

The detective work in this article is interesting, as it shows several things, which seem generally now more widely known.

These are:
* presymptomatic transmission
* asymptomatic transmission
* transmission based on location - surfaces or air in a space previously occupied by someone with the virus

Also the discovery that some health workers may be completely asymptomatic, so it is important to test workers in health care settings. In the longer term it may be important to test as large proportion of the population as possible. There may be plans for city wide testing in some UK cities, such as Norwich.

[Frances_iom]

Silent spreaders - an interesting article here - https://www.bbc.co.uk/news/uk-528407...5D-%5Bisapi%5D

Interesting that for a poster who has queried privacy is happy to post a link with an apple tracker embedded - for BBC links just remove all from the ? onwards

[Count Boso]

Interesting that for a poster who has queried privacy is happy to post a link with an apple tracker embedded - for BBC links just remove all from the ? onwards

Also, I posted this last week - without the apple tracker !

http://www.for3.org/forums/showthread.php?21399-Coronavirus&p=795043#post795043