Urban75 Home About Offline BrixtonBuzz Contact

Coronavirus in the UK - news, lockdown and discussion

Things peaked then improved first in Scotland and then now gradually in England it seems. Yes the numbers are not constantly going down day by day now, but the general trend seems to be good. Did schools break up earlier there?

Personally I think combo of schools out, hot weather, greater herd immunity, vaccine protection, end of football, continuing social distancing, NHS pings/isolating are are all factors at play. Yes there is less testing but the decline in numbers in some areas cant be explained just by that .. numbers down 50% here in past 7 days for example, testing sure isnt.

numbers still high in 18-24s for example, where not all of these factors may apply so much.
 
But it appears to drop off in age groups older than school pupils, and it appears to drop off somewhat before end of term.

If school hols were the main factor (suddenly stopping transmission between under 20s) then you'd expect to see a drop off mainly in that age group first, and probably slightly after than the actual end of term. I think.

apart from the fact a million kids were isolating before term ended…
 
Tuesday = 23,340 new cases
Wednesday = 27,572 new cases
Today = 31,117 new cases

I don't like the look of those figures :(
It would have been incredibly unusual not to see such spikes in the positives by report date.

They have after all been a feature previously and their absence was perhaps an indicator of testing system issues being part of the decline shown. So in that sense its actually reassuring to see such spikes return.

However it is also too soon for me to confidently describe them as just being those spikes, depending on data in future days that might indicate something else at work.

Also I still prefer to look at cases by specimen date.

Either way it would be unusual if the extremely rapid rate of decline in the figures persisted. Things already fell so rapidly that people are scrambing for various explanations, and there was no reason to expect such a pace of decline to continue since that would quickly have resulted in an exceedingly low number of cases being achieved so rapidly as to lack credibility.

Positive cases for England by reporting date. Note the similar spikes/gaps in the previous wave and how we've seen unusually few of them in the downwards curve this time until recent days.

Screenshot 2021-07-29 at 16.49.06.png

England positives by specimen date shows something vaguely familiar and are indicative of anything from a typical blip bump to signs that things are levelling off or at least the rate of decline is slowing. But as usual the most recent data, shown here in grey, is incomplete so we just have to wait and see.

Screenshot 2021-07-29 at 16.48.38.png
 
Daily hospital admissions/diagnoses are currently compatible with a peak/plateau having really been acieved earlier this month. It is still early days though so I maintain a degree of uncertainty about this until I see an obvious, sustained decline in hospital admission figures.
 
It's certainly tempting to read a levelling-off into the England number of patients in hospital graph

Screenshot 2021-07-29 at 17.00.44.jpg

I'm not sure to what extent these figures tend to get added to retrospectively.
 
Obviously it's too soon to read anything into the rise in reported cases the last two days. Could be a blip, could be the start of something new. I've just liked looking at the rapidly falling new case numbers in the past week, whatever the causes. It is around now that we might expect to see the impact of the big reopening on the figures. Keep watching to see what happens next...
 
It's certainly tempting to read a levelling-off into the England number of patients in hospital graph



I'm not sure to what extent these figures tend to get added to retrospectively.
England does not retrospectively change such figures, at least not routinely. Other nations do at times retrospectively adjust certain hospital figures, sometimes continually.

Anyway it sounds like the data about why covid-positive people are in hospital which I've gone on about plenty, most recently because of a shit story in the Telegraph, has been made available. I have not looked at it yet, but here is the description from the page where it and other NHS data is made available in spreadsheet from:


A supplementary analysis of confirmed COVID-19 patients who are being treated primarily for COVID-19 can be found in the file below. Note that this covers acute providers only.

The majority of inpatients with Covid-19 are admitted as a result of the infection. A subset of those who contract Covid in the community and are asymptomatic, or exhibited relatively mild symptoms that on their own are unlikely to warrant admission to hospital, will then be admitted to hospital to be treated for something else and be identified through routine testing. However these patients still require their treatment in areas that are segregated from patients without Covid, and the presence of Covid can be a significant co-morbidity in many cases. Equally, while the admission may be due to another primary condition, in many instances this may have been as a result of contracting Covid in the community. For example research has shown that people with Covid are more likely to have a stroke (Stroke Association); in these cases people would be admitted for the stroke, classified as ‘with’ Covid despite having had a stroke as a result of having Covid.

The headline published numbers in publications to date have been “inpatients with confirmed Covid” without differentiating between those in hospital “for” Covid and those in hospital “with” Covid. Recognising the combination of high community infections rates, with the reduced likelihood of admission for those who contract Covid in the community and are fully vaccinated, the Covid SitRep was enhanced to add a requirement for providers to distinguish between those being primarily treated ‘for’ Covid and those ‘with’ Covid but for whom the primary reason for being in hospital was non-Covid related. In practice this distinction is not always clear at the point of admission when the patient’s record has not been fully clinically coded. In light of this trusts have been asked to provide this “for” and “with” split on a ‘best endeavours’ basis.
Remember some of the stuff they say there if shitty media try to make shitty use of this data to serve their shitty pandemic agendas. Some of what they've felt the need to point out does read like the sort of thing I'd like to shout at the fucking Telegraph.
 
OK I've quickly turned that brand new data into a graph. The proportions shown by this data dont reveal anything that would serve the Telegraphs awful agenda well, but I expect they will try anyway. Or they will cry about the 'best endeavours' basis of the data. Fuck the Telegraph. I dont think I will post this graph too often since to interpret it properly you absolutely have to pay attention to the text I quoted in my previous post, eg the stuff about people admitted for strokes.

Screenshot 2021-07-29 at 17.21.jpg
 
As well as a rebuttal of that shit from the Mail, the latest from Van-Tam also featured estimates of what vaccines have achieved.

Previously I was quite happy to point such figures out, and although I am still happy to do so, I must point out that we are quite far into a period where these figures become quite silly in places. This is because they are looking at what would have happened with no vaccines and no lockdown. And its fair to say that by now if we hadnt had vaccines, we would have had another lockdown or not relaxed restrictions so much in the first place.

So we end up with figures like 22 million infections prevented, when actually we'd have found another way to prevent rather a lot of those happening, or else we would have faced utter doom.

I assume this is why even the BBC stats man says that the 60,000 deaths figure is 'a bit of an exaggeration'.


Prof Van-Tam said: "I wish it were so. This is not 'all over bar the shouting'.

"I hope the worst is behind us but I think it's quite possible that we're going to have one or two bumpy periods in the autumn and in the winter, not only through Covid, but also through flu and other respiratory viruses as well."
 
Tuesday = 23,340 new cases
Wednesday = 27,572 new cases
Today = 31,117 new cases

I don't like the look of those figures :(
By the way when I replied to this earlier I suppose it may be possible for what I said to be misconstrued.

I dont know what will happen next at all, and my main point should have been that we usually need more than a couple of days worth of data in order to correctly identify meaningful trends. With a second point that positive cases picked up by testing were declining so quickly that such a rate of decline was pretty much impossible to sustain, eg we wouldnt expect cases to suddenly fall straight back to levels seem before this Delta wave got going, so they pretty much had to stop plummeting in the daily data soon.

If cases should start rising again, increasing in a sustained way beyond the levels they've fallen to recently, it will take a while before I can actually say that has happened with any confidence. 7 day rolling average comparisons that people may use have some lag in them by their very nature, and sometimes we can spot real trends in the raw daily data sooner than they show up in the 7 day average stuff. But its not that much quicker when doing it properly, I usually have low confidence in 2-3 days worth of data and then confidence starts to increase the more days beyond that I can add to the picture. Weekend reporting limitations and delays affect confidence timing a bit too, and are another reason people may choose to judge week on week instead of day by day.
 
Last edited:
I read an interesting theory today about how the pingdemic has essentially acted like a lockdown and network theory shows this would cut cases. But only temporarily.

No idea where i read it, so no link!
 
I read an interesting theory today about how the pingdemic has essentially acted like a lockdown and network theory shows this would cut cases. But only temporarily.

No idea where i read it, so no link!
Let us know if you find it or similar, cheers.

Its sort of doing my head in that I can blow my own trumpet about that aspect of this wave so much.

None of my points strike me as being novel or hard to come up with. So why are there certain moments in this pandemic when I manage to blurt out the right combination of details a bit earlier than other sources? I was going on about self-isolation etc being a sort of partial lockdown by stealth before the end of June. How come I decided to go on about that before others? I dont really get why these concepts dont arrive at the forefront of peoples minds at the same sort of time?

I suppose I'm moaning about how reactive as opposed to proactive so many of the sources that we may turn to for stories about whats going on are. Is it to much to ask that they might join the right dots ahead of time and share those expectations with us a bit earlier?

In some ways I suppose I've turned that into a coping strategy on the handful of occasions I've been able to do so in this pandemic so far. Feeling one or two steps ahead of the game is one way to come to terms with stuff and construct some order out of chaos. Quite a lot of the time Im only hours or a few days ahead, or I say something at the very last opportunity before it becomes invalid, which is always weird to look back on. And its actually quite frustrating on the few occasions where I've somehow got weeks ahead of where the newspapers etc are at in their appreciation of events.
 
Now in contrast to what I just said, various data and non-official modelling people on twitter tend to be quicker than me to start going on about trends seen in data. I only started following such people in recent months and at times I've found it a bit exhausting. I have copied quite a few of their techniques and the results can be interesting but Im often not confident to shout here about what they seem to be showing, at least not till more time has passed and data accrued.

A lot of them also make me feel I have to draw attention to details such as the values on an axis so that people dont misunderstand what the graphs are showing.

I suppose the following serves as a recent example. I post it here because the trend is interesting if sustained, and I will be gong on a lot more about this sort of age based rates of growth and decline in future if they contain interesting stories about where this wave is heading. But you need to be careful looking at these graphs and his use of language, ie note that so far he is talking about the rate of decline growing smaller, the figures highlighted in the first graph are not back into positive growth territory just yet. I suppose for convenience and strength of point reasons I'd probably have held off the 2nd point until the data actually showed that 20-24 age group back in positive growth territory.

 
I've finished my own look at positive cases by specimen date per age group for England.

To my eyes what the recent cases have done to the picture is make the declines more modest, increasing somewhat the plausibility of what is being shown in this data.

I really dont think the data as of today offers me a guide as to what happens next, it doesnt give me strong clues about whether that was the peak of a wave as we've seen in the past or something else like a euros spike where case growth will then resume after the fall from that spike.

There may be some regional stories emerging but I have to process all this data manually so I havent recently looked at any of those myself properly yet.
 
How plausible is it that the Euros could cause such a prominent spike? I'm not a football person myself...but would a few football matches really cause a large proportion of the population to act significantly differently to what they would have done otherwise? Are there many people whose nervousness about going to a pub is overcome by the desire to go there to watch a football game they could watch at home?
 
I'm using it as shorthand really, there could be other reasons for the same spike, or combinations of factors. I expect there will be attempts to analyse the role of the football eventually, but I dont know how long it will take before such things occur.

Certainly people have been very tempted to look for not very subtle clues in the case data.

Here are positive cases by specimen date (so most recent data is incomplete), broken down into a few age groups. Females on the left graph, males on the right.

A proper analysis would need to takke account of other factors such as differences in vaccination rates.

Screenshot 2021-07-29 at 22.51.jpgScreenshot 2021-07-29 at 22.51m.jpg
 
With a few exceptions such as Shaun, UK media are still failing to communicate the scale of current hospital pressure, although obviously there is quite some variation between different places.

 
By the way there are loads of caveats to those charts, and they should be used to monitor trends rather than take the numbers as totals. These forms of surveillance arent designed to count every case. And some forms of covid triage affect certain numbers.
 
Back
Top Bottom