Dilemma

Statistics or convenient story

Didn't find your answer?

Sitting at my desk waiting for information a thought struck me that I felt I needed to get rid of by sharing it with AWEB

In the beginning when the virus had just struck we had a low % of positive cases

NOW

We have a high %

REASON

We're getting better at and doing more tests

QUERY 

What % of the population tested was positive at the beginning compared to now and is it better or worse ??

 

Ho Hum

 

 

 

Replies (33)

Please login or register to join the discussion.

By ireallyshouldknowthisbut
06th Jan 2021 09:08

The ONS has been considering this point at length in their (numerous) announcements. Hake a poke around on their site if you are looking for genuine efforts to work it out.

Otherwise the best proxy is really hospital admissions, but they are flattened "at the top" by virtue of capacity. That is to say, people who would be admitted when they are quiet are not when at peak. The other issue with hospital numbers is the mix of cases. Early doors, there was lots in the older population due to the "seeding" of care homes by sending covid patients into care homes who are more likely to end up in care. it seems to be much wider amongst the younger persons now.

In short no-one knows, but lots of decent guess work has occurred.

What I find odd is, despite all the testing, it seems to make little difference in that people who have it are clearly still spreading it. That is to say either its not quick enough, or people are not isolating properly when found positive. That after all is the whole point of testing, to then isolate, plus all your contacts to isolate. Testing is not an end to itself.

Thanks (0)
Replying to ireallyshouldknowthisbut:
Psycho
By Wilson Philips
06th Jan 2021 09:26

I think the issue is that people are not isolating properly - period. By that I refer to those that, for whatever reason, believe it is OK to ignore the rules and hold large house parties, pub lock-ins etc etc. It's no use for the politicians to beg and say "you mustn't do this". The fact is that that people will. And in fact, IMO, the tighter the restrictions, the more determined will certain factions be to resist. As long as such behaviour persists - and it will - no amount of shutting down premises for no good reason is going to stop the spread. At least when pubs and restaurants are open there is a semblance of a track and trace system - there is no such system to monitor household gatherings. They have effectively shifted part of the problem from a controlled environment to a completely uncontrolled (except for the new "dob your neighboour" request) environment.

From the beginning it has been clear that London and the SE is the issue - nothing wrong with the people that live there, there are just too many of them. Even at the height of lockdown, the Tube was still rammed (according to a relative living there). It is inevitable that once infection rates reach a certain level there it is going to spread very quickly to the rest of the country. I would hope that in the future as soon as a flare-up is detected (or predicted) firm and immediate measures are put in place, rather then the dithering and close-the-stable-door (in)action that we have seen to date. If those measures were to include prioritisation of vaccination for densely-populated areas I'd be fine with that.

Thanks (0)
Replying to ireallyshouldknowthisbut:
paddle steamer
By DJKL
06th Jan 2021 09:36

Re your last point, spreading, this is possibly occurring because some individuals display no symptoms when they have it.

Regarding your earlier comment re positive rates, whilst I have not really looked into any data my suspicion would be that maybe the sampled populations are not comparable, for instance maybe right now not as many people are going abroad so are not seeking tests when they have no symptoms, last year both my kids were tested before coming to stay at our house, my son was tested before flying to the US, so maybe earlier more people with no symptoms were being tested whereas now it is more those with symptoms.

One of the most valuable lessons I learned about statistics was from the late Professor David Kerridge at Aberdeen, whilst we learned all the calculations, the significance of various distributions etc (none of which I now really remember) he also in tutorials taught us to question samples, populations, bias in selection, in fact a few of his published lectures in later life bewailed how statistics tends to get taught at universities with its emphasis on the maths and its brushing over how the data sampled was collected.

https://www.qualitydigest.com/dec07/departments/spc_guide.shtml
https://www.qualitydigest.com/jan08/departments/spc_guide.shtml

Thanks (0)
Replying to DJKL:
avatar
By gphemy
06th Jan 2021 11:57

Thanks for that diversion, DJKL, like you I can remember virtually nothing of the statistics I learned save that there is something fishy about distributions which aren't normal! (Sorry)

I tracked down the original Kerridge paper referred to in the first link, and for others of like mind, it can be downloaded from:

Thanks (0)
avatar
By bernard michael
06th Jan 2021 09:17

Oh dear - still waiting and another statistical aberration pops up - the daily death statistic
This includes anyone who has died less than 28 days after testing positive for the virus. That includes deaths by traffic accidents, flu,falling down stairs, heart attacks etc
How many have actually died because of the virus and only the virus ??

Thanks (0)
Replying to bernard michael:
avatar
By Tax Dragon
06th Jan 2021 09:35

"Excess deaths" probably provides a more realistic estimate... and it runs slightly ahead of the reported 28-day deaths, which suggests if anything the official number is too low.

That's partly, presumably, because not everyone who dies of COVID does so within the 28-day window. (But, if that window was bigger, your being-hit-by-a-bus point would carry more weight. For the most part, you'd hope that didn't happen to people whilst they were self-isolating, for example.)

Thanks (0)
Replying to bernard michael:
RLI
By lionofludesch
06th Jan 2021 12:32

bernard michael wrote:

This includes anyone who has died less than 28 days after testing positive for the virus.

This was only changed when some professor pointed out that you could never, ever recover from Covid.

The original mark was if you died after having Covid, no matter how long ago, you counted.

Quality of data - crucial.

There's no doubt that the number of cases has increased by increased testing. I wouldn't have thought I had Covid without a test. I didn't have enough symptoms. For me, the most telling statistic is how many folk are in hospital. And even then, that's risen because hospitals have got better at dealing with Covid so folk don't die as quickly, so they clog up beds.

You can conclude what you like from the data with a bit of spin.

Thanks (0)
Replying to lionofludesch:
avatar
By Tax Dragon
06th Jan 2021 13:09

Folk not dying as quickly... see my comment above on the 28-day measure. (Damn the NHS for doing such a good job, eh?)

A bit of spin... there's a lovely table of which countries are 'best' at sequencing the RNA of the virus. The UK is (or was) 8th, behind Vietnam (7th).

Last time I looked (I took a snapshot, saddo that I am), Vietnam had sequenced 8% of its reported cases. The UK had sequenced only 7.4%. However, the UK had sequenced vastly more samples than any other country. These numbers will be well out of date now, but my shot shows UK 157,439 samples sequenced to Vietnam's 113.

Thanks (0)
Psycho
By Wilson Philips
06th Jan 2021 09:35

And another thing. Whilst I do not wish ill on any elderly person (I have elderly and vulnerable relatives myself) I do wonder about the wisdom of prioritising the oldest. It will sound heartless, and perhaps it is, but I wonder how many of those being vaccinated at present will be likely to see this year out in any event. It does sound like a bit of a waste of vaccine, while supplies are limited, to me - in my view (and I know that this is the approach in at least one country) it would be far better to focus on the the front-line workers and then those most likely to spread it. Then get teachers and schoolchildren to the front of the line - that immediately gets the schools open again. Next in line would be students - I remember the first thing that happened in Freshers' Week was a BCG test and vaccination for anyone that failed. Stop the spreaders and you stop the spread.

Thanks (0)
Replying to Wilson Philips:
avatar
By Tax Dragon
06th Jan 2021 09:37

Presumably this is why we've switched to a single-dose approach, initially. Get the result you suggest without sounding heartless whilst doing it.

Thanks (0)
Replying to Wilson Philips:
avatar
By Ginger Tom
06th Jan 2021 10:19

Why vaccinate those who will not die if they get it? Surely the absolute priority must be given to saving lives by vaccinating those over 60 and those with underlying health conditions. Why waste the vaccine on the under 60's? The vaccine does not stop you from spreading the virus.

One thing this pandemic has highlighted is just how self centred and selfish the younger generations are. If this virus killed young people and not older people there would be a very different attitude by younger people.

Thanks (3)
Replying to Ginger Tom:
Psycho
By Wilson Philips
06th Jan 2021 10:54

How do you know that someone will not die if they contract it? Younger people may be less likely to do so, but as we've seen some of them (without underlying conditions) do.

Vaccination may not completely stop an individual from spreading it. But to spread it you first need to pick it up from a surface or be infected with it. Vaccination may have no direct effect on the first, but it has been shown that even though vaccination may not stop you being infected it dramatically reduces the chances of you spreading it - either to others directly or onto surfaces.

Younger people, most of whom have many years in front of them and who are the ones that are going to have to deal for years with the economic and other impacts, are perfectly entitled to wonder why the vaccine is being 'wasted' for the sake of a few more months of life of someone in a care home. And which few months' existence is unlikely to be significantly affected by an unecessarily prolonged shutdown of the economy. I generalise, of course.

Thanks (0)
Replying to Wilson Philips:
Psycho
By Wilson Philips
06th Jan 2021 15:30

Woo hoo - my first pink post :¬)

Thanks (0)
Replying to Wilson Philips:
avatar
By Tax Dragon
06th Jan 2021 17:53

I was going to say that you must have passed your probation, welcome to the pink club (we get special privileges).

But it seems you shouldn't have bragged... you have been "de-pinked" and are out of the club again!

Thanks (0)
avatar
By Tax Dragon
06th Jan 2021 10:11

Can I ask... you say "convenient story".

For whom is it convenient? Sure, Jeff Bezos has done alright out of it, but I find it hard to believe he's in cahoots with the British government - or indeed governments throughout the world that are struggling to manage the situation.

Thanks (0)
Replying to Tax Dragon:
paddle steamer
By DJKL
06th Jan 2021 10:20

Its all that lot in the human suits with the zips up the back, the Slitheen I believe, if not them then has to be the Illuminati or some other cabal working by dark forces to enslave the whole of mankind.

Thanks (0)
Replying to DJKL:
avatar
By bernard michael
06th Jan 2021 10:27

DJKL wrote:

Its all that lot in the human suits with the zips up the back, the Slitheen I believe, if not them then has to be the Illuminati or some other cabal working by dark forces to enslave the whole of mankind.


Close to what I was thinking but of a different nationality ( can I say that??)
Thanks (0)
Replying to bernard michael:
avatar
By Tax Dragon
06th Jan 2021 10:44

You're suggesting (I think) that our press and media are controlled by a foreign power.

Donald Trump really has got inside your head.

Thanks (0)
Replying to Tax Dragon:
paddle steamer
By DJKL
06th Jan 2021 10:49

Surely forgetting foreign and thinking alien is the way to go. Did V not also have zip back aliens. (I only watched a couple of episodes)

Thanks (0)
avatar
By Ginger Tom
06th Jan 2021 10:26

People are not able to isolate. They need to buy food and have zero chance of getting a delivery slot. This also applies to those advised to shelter. A sheltering letter, or a positive Covid test, should automatically entitle that person to a guaranteed delivery slot, in the case of a positive test for the next 3 weeks, in the case of a sheltering letter until the end of sheltering advice.

Not everyone has family or neighbours to shop for them.

Thanks (0)
Replying to Ginger Tom:
paddle steamer
By DJKL
06th Jan 2021 10:36

We have had no issues getting a delivery slot once a week since last April. (well we did miss the week between Christmas and New Year, my daughter collected milk and bread and we did not need anything else)

Thanks (0)
Replying to Ginger Tom:
avatar
By Tax Dragon
06th Jan 2021 10:38

I didn't expect to find myself agreeing with you on much (I don't normally), but I can't deny that I have long thought something like:

Ginger Tom wrote:

A sheltering letter, or a positive Covid test, should automatically entitle that person to a guaranteed delivery slot, in the case of a positive test for the next 3 weeks, in the case of a sheltering letter until the end of sheltering advice.

It's probably not really workable. And the lack of this provision has - in some places - been made up for by a communal approach (Big Society?), which, where it exists, probably* gives a better outcome.

*Unsubstantiated guess, not measured observation.

Thanks (0)
avatar
By whitevanman
06th Jan 2021 10:36

I'm not a great believer in statistics (though I have quoted them on a few occasions) but;
Some weeks ago, mass testing commenced in Liverpool. In about 1 week, 200,000 people were tested (and we must assume it was a fairly "random" sample). Result? They found 800 new cases. That's about 0.4%.
We now have government figures suggesting infection rates of around 2%. There are said to be 1.3m infected out of (say) 65m people.
So, it would appear infection rates have gone up.
The real question is "does that mean anything" ?
Answers on the back of a fag packet please (that seems to be Boris's approach).

Thanks (0)
Replying to whitevanman:
avatar
By Tax Dragon
06th Jan 2021 10:40

whitevanman wrote:

The real question is "does that mean anything" ?

Ask the NHS.

Thanks (0)
Replying to Tax Dragon:
avatar
By whitevanman
06th Jan 2021 11:03

Whilst I agree with what I assume to be, the sentiment behind that, I would say that there is no direct evidence of anything having changed recently. Yes I have seen the numbers from the latest slides. The problem is that we do not have directly comparable figures for last year. The outbreak really became apparent in (say) March when the winter was almost over.
Winter is a time when greater pressure is always on the NHS.
I am of course only addressing the statistics point raised by the OP and not expressing any opinion on the virus or how best to deal with it. I will say however that one will always get different opinions, especially when faced with one's own mortality or finances.
The old want to live, the young want the freedom to "live", those in the middle have got mortgages to pay and kids to teach.
The only certainty is that few, if any, will feel the government are doing the right things or at the right time!

Thanks (0)
Replying to whitevanman:
avatar
By Tax Dragon
06th Jan 2021 11:10

whitevanman wrote:

Whilst I agree with what I assume to be, the sentiment behind that, I would say that there is no direct evidence of anything having changed recently.

Not really sure what you mean by changed (or what the significance of change might be). The virus has obviously changed. It is now even more spreadable. It's Pandemic 2, but starting from a higher base. Our response has to be the same as it was to Pandemic 1, except this time we have a vaccine to add to the mix (something else that's changed?)

Thanks (0)
Replying to Tax Dragon:
avatar
By whitevanman
06th Jan 2021 14:17

As usual, I have only addressed issues directly relevant to the OP.
The statistics revealed on Monday show that there is an increase in the number of people, apparently, infected with covid. I say apparently because there is poor evidence available and I myself gave an example. The rate in Liverpool of 0.4% is not really useful because, when you look behind it, this was mass testing of people who had no symptoms (and from what I see locally, a lot of such testing is ongoing). So it ignores those who do have symptoms.
What is the basis of the governments latest 2%? We don't really know. If it is based on actual numbers from tests on people without symptoms, one can conclude that the rate of infection has, apparently, risen. But is that true of the Liverpool city region? Some stats suggest there has been a rise there but nothing like the overall "rise" (to 2%).
The rate of infection in London and SE has increased massively (or so it seems) and that may well be distorting the results for elsewhere.
We did not have wide spread testing of people without symptoms in London until very recently and no such testing anywhere before (say) November. So again, comparisons are not available.

If we want to know what is happening, we need to be able to compare like with like and as we don't have such info, the statistics are in some ways, meaningless.
The same is true of the NHS numbers. There are many factors that mean the numbers in hospital now are higher than (say) May last. My point again is that we are not comparing like with like so the stats don't really help.

Ignoring the OP however, what we can say for certain is that the biggest issue is whether the NHS can cope. The current number of people in hospital is such that the answer to that question is rapidly moving towards "No" so inevitably something had to be done. The issue of vaccines only complicates matters because no-one knows how effective a single dose will be and in order to vaccinate the first 15m people by mid-Feb, a lot of NHS resource will have to be diverted, at a time of year when things are traditionally difficult.
It is probably better to focus on such practical issues than to keep putting out graphs and tables of meaningless stats. But that probably wouldn't play so well for TV and media and politicians might have to face up to the uncomfortable truths (statistically first!).

Thanks (0)
Replying to whitevanman:
avatar
By Tax Dragon
06th Jan 2021 17:37

Got it, thanks.

Reference statistics, I thought all accountants loved them. So much so that I have heard them talk about groups of five (or fewer) in percentage terms... "our survey reveals that 60% of you think such and such" etc. And even "last year it was 50%, so it's heading the right way". You mean, you asked (or maybe employed?) one more person this year? Or last year there was an abstention?

Joke. I don't mean I'm making a joke, I mean it's a joke to convert minimal numbers to percentages. Like those TV ad surveys... 87% (small print: of 115 people who responded) said it's the best shampoo they've ever used, etc. Or 8 out of 10 cats... you asked only 10 cats?!... Or putting Vietnam ahead of the UK in that table.

Thanks (0)
Replying to whitevanman:
paddle steamer
By DJKL
06th Jan 2021 10:55

I doubt any of us has the data nor has enough knowledge of the sampling approaches undertaken to make any meaningful deduction,

"All experienced statisticians know that there are some circumstances in
which statistical calculations can be trusted, and others in which they can
only give answers of varying degrees of indefiniteness. The theory stated
in textbooks does not make it clear why this is. Deming’s distinction between
“enumerative studies”, and “analytic studies” lays the foundation
for a theory of using statistics, which makes it plain what other knowledge,
beyond probability theory, is needed to form a basis for action in
the real world."

Statistics and Reality (draft) Page 1, Nov 21, 1998
David and Sarah Kerridge

Thanks (0)
Psycho
By Wilson Philips
06th Jan 2021 11:42

At least we don’t have to deal with this kind of moronic behaviour:

https://apple.news/Ai9U70JR6Sgu_27f83e6QKQ

(For those without access I'll try and find another link)

https://twitter.com/therecount/status/1346540350941310976?s=21

Thanks (0)
Replying to Wilson Philips:
paddle steamer
By DJKL
06th Jan 2021 11:49

If you want to take a positive from it, maybe it is a good example of Darwin's natural selection in action.

Thanks (1)
Replying to DJKL:
avatar
By Tax Dragon
06th Jan 2021 17:38

In more ways than one. The way you mean might well be positive. The other way - that the more times the virus is transmitted (i.e. the more it spreads, e.g. at a super spreader event), the more (times) it will mutate - is profoundly negative.

Thanks (0)
avatar
By bernard michael
06th Jan 2021 12:03

Thanks for your input everyone my information has now arrived and I can work without thinking more B******s!

Thanks (0)