Daily numbers, graphs, analysis thread 6

[Barracker]

Welcome to thread 6 of the daily updates.

Resource links:
Worldometer UK page
Financial Times Daily updates and graphs
HSJ Coronavirus updates
Johns Hopkins Coronavirus Resource Centre
NHS England stats, including breakdown by Hospital Trust
Covidly.com to filter graphs using selected data filters
ONS statistics for CV related deaths outside hospitals, released weekly each Tuesday

Thank you to all contributors for their factual, data driven, and civil discussions.

If replicated across the country, it suggests our true death toll could be yet higher.

Pushing 40,000!

I can't see any factors specific to the Uk that would cause UK care home deaths to be such a different %

Why would UK care homes be so much safer ?
Are there particularly good quarantine arrangement, separation of infected residents etc ?

In Germany, once a care home resident has been discharged from hospital,
they are not sent back to care homes if they are positive for COVID

Instead, the state has rented hotels to quarantine them,
staffed and serviced of course, by people with PPE.

I think those found positive with only mild symptoms, who so far don't need hospital, are also put in these hotels

Despite this, there are still many deaths in care homes here,
due to infection by staff or (in the states they are allowed) visitors, who do not yet exhibit symptoms

More graphs, Yorkshire, E/W Midlands

E, SE, SW

That's a very impressive amount of work shoots

Do you have a graph for the north west please?

If some of those are illegible, there are zoomable versions here.

imgur.com/a/IYVpyDZ

The figures from County Durham don't match the ONS stats, which show for deaths occurring in weeks 14 + 15 (28 March to 10 April) 93 in hospital, 26 in care home, 2 at home.

It doesn't seem possible that the number of deaths in hospital for 25 days would be lower than ONS data for 14.

Cherryghost: i.imgur.com/zREmZP3.png

Salford much more than Manchester, for some reason

Interesting to note commuter belt places like Woking, Guildford, Sevenoaks, having relatively low death rates. Norfolk essentially totally unscathed.

Thanks for the work on these, Shoots. Just noticed Suffolk has accidentally been included with SW rather than E. Salford and Manchester are virtually the same city. I would guess the Salford hospital takes many from Manchester. Unless it’s by home address?

It's by home address

www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/methodologies/userguidetomortalitystatisticsjuly2017

SHoots That's odd, because BBC Newsnight are running that carehome story, with chart for Durham
With video here

https://mobile.twitter.com/bbcnewsnight/status/1253807288974008331?s=21

BBC Newsnightt@BBCNewsnight*

“If this were to be replicated across the rest of the country… then the real death toll could be far (greater)"

Policy Editor @lewis_goodall has seen data from Durham County Council, showing nearly 50% of Covid-19 deaths were coming from care homes.

and many thanks for the Zoomable versions of the graphs, much clearer

Coronavirus: England's 'excess deaths' among the highest in Europe

and it seems particularly England, within the UK

https://news.sky.com/story/coronavirus-englands-excess-deaths-among-the-highest-in-europe-11977394

New mortality figures suggest England has been as badly affected as Spain and Italy by the COVID-19 outbreak.
....
Further analysis of the EuroMOMO figures shows that the three countries with the biggest peaks in excess deaths are Spain, Italy and England.

England recorded the highest number of excess deaths anywhere in Europe for two consecutive weeks (week 14 and 15).
.....
England has also had more excess deaths than expected compared with Scotland, Wales and Northern Ireland.

If it’s by home address that’s strange re Salford and Manchester. Must have been an outbreak in Salford. My DS lives in Salford, me in Cheltenham!

Here's the volcano.
You can see the true peak on April 8th - currently standing at 841 deaths on that day.

A small note. Looking at the data behind the graph, it becomes visible that most deaths are fully caught up by the time two or three weeks have elapsed. However, in the last week (today's data, and the data on the 21st especially) rather a lot of older deaths are unexpectedly swept into these stats, when it had previously seemed that the book could be closed on those older dates. For example; today 587 England deaths reported. However, 32 of those were from March, even though it had begun to look like there were no more March deaths outstanding. It's not a huge proportion, to be sure, however it does imply that there are still potential backlogs in the system that appear in a batch from time to time.

It continues to bother me that there is the administrative glitch on March 29/30/31. These are after all, date-stamped deaths. It should be impossible for them to end up allocated in the wrong day, yet they clearly have been, across every trust in the country, and even as I watch each report add more data I can still see that March 31 continues to receive fewer than it should. I don't like it because there is no explanation that is satisfactory for why this might be. If 'Mr Smith' died on March 31, that date should be timestamped correctly, there should be no mechanism whereby his date of death is reported in error. These figures are meant to be the validated date of death numbers.

Thank you for the updated volcano Barrcker.

I rarely watch television, so don't see the daily briefings (and apologies if this has been asked before) but is this issue ever discussed at the briefings? Or is it possible that the government do not want the public to think that the actual peak may have passed in case this causes more people to break lockdown rules?

Also (and I accept I am being ridiculously optimistic here) could it be possible that the high excess death figures that we are currently seeing will be compensated for later in the year with potentially fewer than expected?

In other words, while I am not suggesting that those who have died from covid 19 are any less of a loss whether they were towards the end of their natural life with or without this pandemic, could the changing behaviour of the population at large lead to fewer excess deaths once this immediate threat is over, making the year on year figure not look as horrendous as it does at present?

Could the deaths that are reported late, and in batches, be ones that have had to be referred to a Coroner, or another checking process prior to being included in the numbers?

It’s reassuring to see the downward slope of the volcano. At least for now.

Personally, I am finding the location statistics some of the most interesting data, it’s salutary to note how the death rates compare with local demographics. I’ll look later on at some of the PHE local profiles to see how closely they match.

Thank you for this thread. It's about the only 1 I read atm.

Just a personal note on reporting in England.

Fil died on 16th April. He had tested positive for CV on 1st April.

Discharged (expected to recover) to a care home on 12th April. Died 16th April.

Coroner did a throat swab that came back negative last week. Because no doctor saw him at carehome and because hospital doctors expected him to recover there will now be a post mortem. Earliest date is 5th May.

Coroner expects cause of death to be recorded as CV. She is pretty sure he had a a cardiac failure caused by the virus.

So although fil died on 16th April he still doesn't have a death certificate so still isn't recorded anywhere. And wont be until after 5th may. So 3 weeks almost. I suspect that his aren't the only figures not yet on the graphs.

I find that fact interesting that Stevenage faired really well compared to Harlow and Watford which are very similar towns

I wonder if it's because Stevenage had a testing Center very early as I took my dd there at end feb. As she was ill on return from Italy ski trip.

With Watford I wonder if it's due to Harry Potter being there with 7000 visitors passing through Watford to get there from all over the world.

And that 5000-7000 visitors a day to Harry Potter !!

A very tentative and possibly foolish suggestion that someone somewhere forgot that there are 31 days in March. (I teach adult learners and know that a surprising number of apparently intelligent and well educated people don’t remember basic facts like this.)

The late reporting I believe is also down to the administration side of each trust, the ones that go back 3/4 weeks each day are often from the same trust, I think it was in Wednesdays figures that out of the hundreds of deaths pennine trust had declared 88 of them going back many weeks.

This thread is fascinating and so helpful so thank you for taking the time to extrapolate and share all this data!

I have a question, new to the thread so apologies if it’s been covered and I missed it, but is there any evidence to show the virus was in the UK before March? I know this is a thread for data but anecdotally many report the symptoms from Jan/Feb. It would be good to know if there’s anything in this. It’s not an impossible theory, given global travel.

Thanks!

@Moveonthecards, once the deaths started, it was here. Easiest way to figure it out

Thanks sofato5miles. But I guess not much in the way of knowing the extent in terms of ‘less serious’ cases (ie being considered at the time flu or another ‘winter illness’) or asymptomatic carriers.