Coronavirus

[teamsaint]

Something curious, which has been alluded to before by others.

We are told that the number of cases in the UK is rising - and that is true.

I might well be wrong but when I calculate the number of positive cases per 100,000 processed tests, the 7 day average has been more or less flat from the 28th July, i.e. the increase in the number of cases appears to be due to the rise in the number of tests rather than an increase in overall prevalence even though there are a number of spikes across the country.

That was professor Heneghan’s point. Hard to see past it, right now.

John, could you have a stab at where all the current( stubbornly high) non- hospital deaths are coming from ?
Has it got anything to do with the increased testing over the last couple of months, and deaths of people who tested positive, but later died of non covid causes?

In other news, we really need to pay a LOT more attention to the number of excess deaths since lockdown that were non Covid. I think I saw a figure of 20k somewhere. It is certainly very high, and certain to continue rising.

[cloughie]

That was professor Heneghan’s point. Hard to see past it, right now.

John, could you have a stab at where all the current( stubbornly high) non- hospital deaths are coming from ?
Has it got anything to do with the increased testing over the last couple of months, and deaths of people who tested positive, but later died of non covid causes?

In other news, we really need to pay a LOT more attention to the number of excess deaths since lockdown that were non Covid. I think I saw a figure of 20k somewhere. It is certainly very high, and certain to continue rising.

...and were the deaths undiagnosed Covid or other causes that have gone neglected through people thinking they should bother their GP because Covid was a priority?

[johnb]

John, could you have a stab at where all the current( stubbornly high) non- hospital deaths are coming from ?
Has it got anything to do with the increased testing over the last couple of months, and deaths of people who tested positive, but later died of non covid causes?

The trite answer to the question is that I don't believe the very large number of apparent non-hospital deaths exist, at least not on the scale indicated by the difference between the hospital deaths and figures collated by PHE and published by the government. But then, there is the question "What is a death due to Covid-19?"

Looking at England, there are three sources for Covid-19 deaths:

- NHS England hospital deaths (where there has been a positive test and also where there has not been a test but CV19 was recorded on the death certificate)

- ONS (based on whether CV19 is mentioned on the death certificate)

- The figures compiled by PHE/DHSC

The NHS England and ONS figures roughly tally, with the ONS deaths being somewhat higher as one would expect as not everyone dies in hospital.

Both NHS and ONS figures seem to be fairly reliable but the PHE/DHSC deaths are very significantly higher, suspiciously so, especially as they do not include deaths recorded on the death certificate as CV19 but where there hasn't been a positive test.

See the chart in post #3383: http://www.for3.org/forums/showthrea...904#post803904

- As you surmised, part of the difference is probably deaths due to other causes, of people who have previously tested positive, e.g. the example of falling under a bus after recovering from CV19. But the person who fell under the bus might have done so because of post-viral fatigue and people dying from other illnesses might have been made more susceptible to those illnesses because of having had CV19. The definition of a CV19 death can be fuzzy but even so I would guess that the PHE figures include deaths that most people wouldn't have classed as due to the virus.

- The other aspect is how PHE arrives at the deaths. It collects data on deaths from NHS hospitals, from the ONS and from numerous other sources. It then melds the whole lot together using the NHS number or the patient's name, date of birth and post code when the data doesn't have the NHS number. At some point in the process it runs the deaths against the database of people who have tested positive at some time in the past (also by NHS number or name/DoB/Post Code/etc), with no elapsed time cut off - so, unlike in Wales, Scotland, etc, anyone who falls under the proverbial bus after recovering is counted as dying from CV19. It also runs a process to de-duplicate the deaths. What could possibly go wrong?

On the "Coronavirus Dashboard" it states "The definitions used for deaths are currently under review. It is likely that revised data, which more accurately reflect the current and overall burden of COVID-19 will be published shortly."

By the way, I do very much agree about the excess deaths.

[johnb]

It is:
https://www.bbc.co.uk/news/uk-51768274

(One I prepared earlier. )

PS The chart should say "Processed Tests", not "Tests". (Over the same period number of mailed kits that were processed was only 15% those recorded as sent out.)

[teamsaint]

I won't quote the post, but thanks for your reply #3416 Johnb.

Re the question of non Hospital covid deaths currently, the government really does seem to be making a rod for its own back. Unless of course they are " happy" with a figure of 50-100 deaths per day to help keep the right conditions for whatever it is they want to implement.

Something that really concerns me is the very low level of media reporting around issues that have the potential to affect us long after covid becomes ( hopefully) just a bad memory. Loss of democratic accountability, changes in access to healthcare, concentration of business in ever fewer hands, the new building laws , etc etc.

The liberal media need to look at these mainstream issues much more comprehensively,but I fear that other areas will continue to dominate their agendas.

[teamsaint]

(One I prepared earlier. )

PS The chart should say "Processed Tests", not "Tests". (Over the same period number of mailed kits that were processed was only 15% those recorded as sent out.)

So I'm guessing that today's increased number of confirmed infections, over 1000, reflects the testing on August 07 and 08. when over 180k tests were processed, as presumably results come in over the next 24/48 hours ?
Still, not much room for complacency.

[Frances_iom]

The liberal media need to look at these mainstream issues much more comprehensively,but I fear that other areas will continue to dominate their agendas.

This is a classical dictatorship in the making - assume (or be given) power for some necessary but short term measure but hang on to said powers when the need has gone away (the WW1 DORA is a good example) but DC has an agenda whereas BJ only wants the trappings of power but has no real intellectual depth nor ability to handle details and is thus easily led

[johnb]

So I'm guessing that today's increased number of confirmed infections, over 1000, reflects the testing on August 07 and 08. when over 180k tests were processed, as presumably results come in over the next 24/48 hours ?
Still, not much room for complacency.

That chart includes the tests reported today. On the data file it gives the date of "Activity" as the 8th. The dates are for when the tests were processed rather than when the specimens were taken. (There is another data file that gives that data but there is a lag and it hasn't been updated for a few days.)

[teamsaint]

That chart includes the tests reported today. On the data file it gives the date of "Activity" as the 8th. The dates are for when the tests were processed rather than when the specimens were taken. (There is another data file that gives that data but there is a lag and it hasn't been updated for a few days.)

Ok, thanks.
Interesting numbers of patients in hospital in Scotland, around one third of the England total, and remarkably stable around the 260 mark. They must be keeping in some recovering patients an awfully long time. Or something .........

[johnb]

This is a classical dictatorship in the making - assume (or be given) power for some necessary but short term measure but hang on to said powers when the need has gone away (the WW1 DORA is a good example) but DC has an agenda whereas BJ only wants the trappings of power but has no real intellectual depth nor ability to handle details and is thus easily led

When does crony-ism and pursuing free market liberalism at a cost to the nation shade into corruption and incompetence?

This week we have had the truly bizarre spectacle of Robert Jenrick announcing a policy to please the construction industry (who just happen to be Tory party donors) while only weeks ago he tried to grease the wheels for Richard Desmond's construction project - and guess what, he is also a Tory party donor.

Then we have all the contracts that have been awarded to commercial companies in relation to Covid-19, regardless of whether they are capable of fulfilling them as required. A level of private involvement in the health service that might well become entrenched.

[johnb]

I've updated the Excel file I posted a link to in #3410: http://www.for3.org/forums/showthrea...166#post804166, so I thought I might as well upload it (fingers crossed).

The new file can be downloaded from: https://app.box.com/s/7ugl5c8iecft9o1cy0uf2x7nnyuqq222

(It needs macros to be allowed to run otherwise the labelling, etc won't work.)

Screen grab:


Most of the "useful" functionality will be lost in OpenOffice or Apple Number, i.e. the labelling of the data points with the area name, without which it isn't much use.

The file presents a scatter chart of the 170 local authorities with the highest populations, a similar chart, though animated, is on post #3409 http://www.for3.org/forums/showthrea...164#post804164.

The previous version calculated the number of increased cases by comparing the week with the average of the two previous week. It now compares it with just the previous week.

There is a slider bar which enables one to scroll through from w/e 31/05 to w/e 03/08, day by day.

There is a check box to enable dynamic labelling of the data points based on your criteria. If all data points are labelled the thing becomes a mess - so the dynamic labelling will automatically adjust the labelling as one scrolls from one day to the next.

There is also button to re-do the labelling at any stage.

There is a second check box which, when ticked, expands the X & Y axes ranges to encompass Leicester, which otherwise is way off the scale on certain dates. (I hadn't realised how seriously affected Leicester was until I did this.)

The file size is almost 1/10th of the size of the previous version (I've deleted the redundant raw data).

I'm not sure how much use it is to anyone. It keeps me off the streets if nothing else.

PS The cases/100,000 I've calculated uses the number of cases over 7 days. When cases/100,000 are quoted in the media or by government it is often 14 days of cases that are used so the figures in on the chart would need to be doubled to be comparable.

(Good heavens, I do ramble on .... )

[Globaltruth]

I've updated the Excel file I posted a link to in #3410: http://www.for3.org/forums/showthrea...166#post804166, so I thought I might as well upload it (fingers crossed).

The new file can be downloaded from: https://app.box.com/s/7ugl5c8iecft9o1cy0uf2x7nnyuqq222

(It needs macros to be allowed to run otherwise the labelling, etc won't work.)

Screen grab:


Most of the "useful" functionality will be lost in OpenOffice or Apple Number, i.e. the labelling of the data points with the area name, without which it isn't much use.

The file presents a scatter chart of the 170 local authorities with the highest populations, a similar chart, though animated, is on post #3409 http://www.for3.org/forums/showthrea...164#post804164.

The previous version calculated the number of increased cases by comparing the week with the average of the two previous week. It now compares it with just the previous week.

There is a slider bar which enables one to scroll through from w/e 31/05 to w/e 03/08, day by day.

There is a check box to enable dynamic labelling of the data points based on your criteria. If all data points are labelled the thing becomes a mess - so the dynamic labelling will automatically adjust the labelling as one scrolls from one day to the next.

There is also button to re-do the labelling at any stage.

There is a second check box which, when ticked, expands the X & Y axes ranges to encompass Leicester, which otherwise is way off the scale on certain dates. (I hadn't realised how seriously affected Leicester was until I did this.)

The file size is almost 1/10th of the size of the previous version (I've deleted the redundant raw data).

I'm not sure how much use it is to anyone. It keeps me off the streets if nothing else.

PS The cases/100,000 I've calculated uses the number of cases over 7 days. When cases/100,000 are quoted in the media or by government it is often 14 days of cases that are used so the figures in on the chart would need to be doubled to be comparable.

I retract my confident statement re Apple Numbers - there is no equivalent to macros in Numbers, instead you have to use AppleScript. This means Numbers doesn't automatically show the labels, which makes it less than useful.
..I will be more than happy to be corrected on this by a Mac superuser.
If I get round to doing the conversion later then I will post a link, but for my purposes this current Excel version is fine to look at in Box and to send to various people all of whom, it turns out, use Excel. (a polite way of saying I hope you carry on maintaining for a while longer John)

[oddoneout]

When does crony-ism and pursuing free market liberalism at a cost to the nation shade into corruption and incompetence?

This week we have had the truly bizarre spectacle of Robert Jenrick announcing a policy to please the construction industry (who just happen to be Tory party donors) while only weeks ago he tried to grease the wheels for Richard Desmond's construction project - and guess what, he is also a Tory party donor.

Then we have all the contracts that have been awarded to commercial companies in relation to Covid-19, regardless of whether they are capable of fulfilling them as required. A level of private involvement in the health service that might well become entrenched.

Kakistocracy in action... I find myself working on the basis of 'who would be the most unsuitable(whatever the criteria) candidate' for any given post/contract. That way any surprise tends to be 'well it could have been worse'.
It sums things up rather when PM Johnson uses the term 'moral duty' (in connection with schools going back in September) without a shred of self awareness or irony.

[Dave2002]

This TED talk is quite interesting - particularly the first 20 minutes or so - then it tails off a bit - https://www.ted.com/talks/bill_gates...he_near_future

[Simon B]

References are being made in various places to the observation that while confirmed cases are going up, associated hospitalisations are going down. This appears to have been going on for long enough to not be explainable away by the lag between these previously well-correlated variables.

However, I've not seen any proper data on the age distribution of recent confirmed cases. Rhetorical allusions yes, data with verifiable sources, no. Does anyone know of a source of this information in digestible format?

If, as seems plausible, current infections are heavily biased towards younger cohorts, this is exactly what would be expected. A toll of severe illness and worse could then come weeks or even months later when an outbreak inevitably propagates into higher risk groups...