Daily numbers, graphs, analysis thread 11

[BigChocFrenzy]

Welcome to thread 11 of the daily updates

Resource links:

Slides & data UK govt pressers
NHS England stats including breakdown by Hospital Trust
ONS UK statistics for CV related deaths outside hospitals, released weekly each Tuesday
Financial Times Daily updates and graphs
HSJ Coronavirus updates
Worldometer UK page
Covidly.com to filter graphs using selected data filters ONS statistics for CV related deaths outside hospitals, released weekly each Tuesday
Plot COVID Graphs Our World in Data

We welcome factual, data driven, and civil discussions from all contributors 💐

Currently, pillar 2 is >80% of new cases across most of the country

Particularly high Pillar 2 proportion in E. Midlands, Yorkshire and Humber

More FT charts;

I thought sage and independent sage were the same thing. There you go, you learn a new thing every day! And now I feel incredibly stupid!

@wintertravel1980

Last time I look at a piece of analysis from Independent Sage, they used IFR of 1% and a hypothetical (and absolutely unfounded) daily transmissions estimate of 8,000.

I do not think they can be calling themselves SAGE after that.

Aha, yes that would calculate to 20k deaths over the next 9 months then.

Fortunately all other stats we have suggest both of those numbers are bollocks.

They chose a v misleading name, imo

They should have called their group something else, to avoid confusion

I gather some of them belonged in the past to SAGE & associated sub-committees, but they showed a serious lack of originality re names

Far too much confusion from the government without scientists adding to it

"to 20k deaths over the next 9 months "

With inevitable local outbreaks of various sizes and winter to get through in that time,
I'd be relieved tbh if excess UK deaths only rise about 20k over that time to stay below 90k
It would be very manageable

829 cases and 176 deaths today. Slightly higher figures than last Wed but always difficult to say with deaths.

In terms of the last 7 days, the average cases is 892 so less than 1000 and 7 day average deaths is 118.

The 7 days prior to that the average cases was 1087 and deaths 132. So, overall, downward trend continues.

Out of the last ten days, only 2 days have seen cases above 1000 as well.

Data source: Worldometer and Gov.uk

Given that Tuesday can be when the weekend "catch up" appears I was a bit surprised to see only 689 new cases yesterday; I wonder if some of it's been carried over to produce today's 829?

With my cynical hat on, and remembering the coming relaxations on Saturday, I also wonder if we'll see some even larger numbers on the run-up as a warning to behave responsibly

Of course I could well be wrong and every figure may be an exact reflection of what's happening, but right now I'm finding it hard to have confidence in any of them

Worth also mentioning that from the NHS England stats (which obviously doesn't cover whole U.K. but a lot of the most infected areas), the number of deaths in hospitals is on average less than 50 per day now. This means that a lot of death is presumably happening in care homes and possibly at home too - I am not celebrating this, more noting that we have two epidemics that overlap to a degree.

The details about the 36 at risk areas includes places like Portsmouth.
Portsmouth is about 10th from bottom of that bottom 75 list

@Puzzledandpissedoff

My town is on that list and it is utter bullshit, they looked at the pillar one figures for two consecutive lists and listed the ones that had gone up by 100's of %, from 1-4 for example

Mine too, and the same thing jumped out at me. Admittedly it's a media piece and they're renowned for stirring up hysteria, but I simply don't get why the relatively small numbers are being used like this

Then again I don't get why they've been keeping the Pillar 2 numbers from local authorities either ... what the hell is going on here??

Derby pointed this out earlier with Suffolk. 50% increase in cases over a week.

2 one week and 3 the next 🤦‍♀️

I know obviously that deaths are announced later than they happen, but you can look back two weeks and most deaths will be captured within the two weeks if in hospital setting (where patients will likely have been tested, no post mortem etc).

itsgettingweird yes I saw that about Suffolk, but I think the report was based on the earlier, dubious graph which didn't reflect P2 results?

It may have been removed now but it's all a bit late since it's been plastered all over the net causing yet more confusion. We're lucky on this thread because we have some very clever posters who can explain it all properly, but not everyone's got access to that

@Puzzled
regarding the increase of positive tests compared with yesterday I'm guessing it's possible that more cases have been picked up just because mobile testing sites have been moved to Leicester etc. I know this is what happened in Anglesey with the recent outbreaks there they actively encouraged people to get tested instead of the usual policy of doing tests only if you're ill. If they are testing people in an area with lots of cases, given the high asymptomatic rates you would expect more positive tests to come out than otherwise. My theory anyway.

The latest MRC Biostats (Uni Camb) report is out, and says that infections are down to around 3k a day (which fits with the PHE serology report and Covid-Zoe, FWIW), and R is "very likely" to be below 1 across all of England.

www.mrc-bsu.cam.ac.uk/now-casting/

Be interesting to see the PHE report tomorrow. I wouldn't be surprised if they say they can't commit to seeing much change week on week as infections in their study are so low - they only found 14 last week I think. At that point numbers are really too low to infer much at all.

20,000 deaths over 9 months would be 500 deaths a day. They're claiming rates have to plateaued - which is debatable - but it would have to increase significantly if it was to get to 500. I think a group of scientists getting together and calling themselves something that is designed to confuse, then releasing sensational headline-grabbing claims is not helpful.

@Firefliess

20,000 deaths over 9 months would be 500 deaths a day. They're claiming rates have to plateaued - which is debatable - but it would have to increase significantly if it was to get to 500. I think a group of scientists getting together and calling themselves something that is designed to confuse, then releasing sensational headline-grabbing claims is not helpful.

Er, no. 20,000/9/30 = 75 deaths a day, They aren't that bonkers ;-)

Intteresting point, Frazzled; I don't know how long it takes for test results to find their way into the published figures, but increased testing sites - particularly in hot spots - would certainly explain some of it

Despite a few bumps at least it's a good thing that 7 day averages are still going down; we can only hope it continues

I think it's also missing the fact that a lot of the most vulnerable and elderly have already died, they can't die twice so it may be that things wouldn't ever get as bad again. We also are now much better at testing and keeping it out of care homes (I hope), so likely new outbreaks will be in working age people where we know it is less lethal. We need to ensure the NHS doesn't get overwhelmed and that hospitals and care homes stay covid-free as much as possible but otherwise I think we have to carry on.

I am concerned when I see the weekly surveillance that there are still outbreaks in care homes though, that's disappointing and I hope there is a plan to counteract it.

I thought we were going to start doing much more pro-active testing but that doesn't seem to be happening widely. We should be squeezing every drop from our testing capacity.

Did Norwich get anywhere with their plans to test the whole city? I'd like to see that happen and the surveillance study to enrol much greater numbers.

The real issue will be the time lag of deaths in the community. My friend died last Monday from suspected COVID-19 but they are doing a post mortem due to the fact she had a negative test earlier on. The coroner has said it will be at least 2 weeks before it's done ...

@patricia - Whoops, yes you are right I've had a long day today helping my son move house. Clearly used up all my brain capacity already today! So it's a plausible number if we assume deaths are trending towards around 75 a day - which looks to be around the upper end of what looks likely. I still think they're trying to grab headlines in an unhelpful manner

@Puzzledandpissedoff I was also a little surprised yesterday that new cases were in the 600s so I also wonder if some are 'carried forward'. It's possible.

Also possibly increased testing in Leicester may be picking up more cases.

Sorry to hear about your friend @Threesocks

Can I join you all?

I’m struggling to make sense of the continued high number of deaths reported on days like today. It appears that hospital deaths across the UK are now around 50 to 60 a day (or less) and the ONS survey is showing that over recent weeks hospitals have accounted for a little over 60% of all Covid deaths, so would expect to now be at less than 100 a day. Is the rest due to late reporting, or something else?

On numbers testing positive, I think there is good evidence the number of infections is genuinely coming down. The way the official statistics are reported is pretty dire for Pillar 2, so impossible to make sense of. For pillar 1 though we at least have the number of people tested and number testing positive. Total numbers being tested didn’t change much over June, but by my calculations the positive rate was 2% for the first 10 days, 1.7% for the next 10 days, and around 1.2% for the last 10; This also suggests the decrease in numbers has actually accelerated over the last week or so, despite the relaxations, which is good news. Maybe test and trace is finally starting to have an impact?

@aprillady - the other reported deaths are from care homes and other settings (such as homes). Firstly, the data reported is reported quite erratically due to few settings reporting over the weekend, so then there is a catch up over the next couple of days, which lead to big swings. Deaths in all settings are slowly going down, but they do appear to be going down in care settings more slowly than in hospitals.

And yes, all signs are that infections are coming down -when you look at the tests, the PHE reports, the Cambridge review out today, the hospital data and NHS triage data. Overall, trends at present genuinely seem to be down, although of course there will be local outbreaks and areas where things take longer to decline.

Do we have data on the number of people tested? I thought data was at best the number of tests posted out/carried out in testing centres. To my knowledge, there is no data on how many individuals have received a test of any kind, many people (front line NHS/care workers) have had several tests. Say a nurse is tested once a week for 10 weeks, then tests positive on the 10th week, that counts as 1/10 positive tests, but one individual.
Also, swabs and blood tests taken on the same person at the same time are counted as two tests.
I'd love to know how many individuals have been tested, and how many of those have ever tested positive. That would be useful data.