Daily numbers, graphs, analysis thread 8

[Barracker]

Welcome to thread 8 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.

Thanks Bigchoc. I couldn't read the footnote on your image, it is tiny and fuzzy on my device!

OK, from the firdst paragraph of the hidden periods article:

If current trends continue, the end of the pandemic should be expected no earlier than March 2021, the global number of cases will exceed 5 million.

We are already on over 4.5 million cases, so that is a bit of a no brainer, isn't it? Or do they mean the death rate?

tbh, I'm wondering what on earth that Manchester Uni group think they are doing,
publishing something that could have so much effect on people's behaviour
... when their estimates are 5 x those of most other people

• I've seen a few estimates of 4-5% for the UK and now we have actual damn tests from Spain around that

it would be different if Man Uni had tests, but they just seem to be just using theory and extrapolation

There must be UK national antibody tests soon, so unless they've had a reassuring heads-up from these,
they certainly are being "bold"

btw, I've also read reports claiming that Spain and France have found early COVID cases too
and it's unlikely that could be making such a difference anyway, 5% vs 25%

I dunno BCF.

Maybe there's something in the water in Manchester! (Btw did you know Manchester was twinned with Wuhan? I only found that out last week).

Manchester group seems to have an agenda and they have been irresponsible both in the way they have framed it and by courting tabloid publicity, imo. If your results are 1. Different from everybody else's 2. Would be dangerous if everybody believes you and you are wrong, you need to present your results with more caution and perhaps humility.

From that hidden periods paper:

"real epidemic outbreaks in Germany, USA and the Republic of Korea probably occurred in November-December 2019, in UK – in early 2020"

That study of Germany's epicentre found 15% infected there and estimated 3% average in the country
(compared to Imperial's estimate of below 1% for Germany)

If Germany's outbreak started earlier, I'd expect the UK to have a lower % infection if those outbreaks actually did spread the infection earlier

• and somehow without any death peak during winter months that had lower total deaths than previous years

Actually, I can't believe that "hidden periods" article was only posted yesterday. They predict that the UK part of the pandemic will end on 20 Aug 2021, and will result in a total of 214,899 cases. Even with our limited testing we are already on 233,151 cases, so I think their estimates may be a bit out...

I find it so frustrating that there are so many pieces of "research" published by "respected" scientists and shouted from the media headlines when a decent read of the paper by even a layperson (ie me) can see glaring holes/ errors or simply dangerous assumptions.

Then there is the 21% found infected in actual blood tests in NYC, the hotspot in the USA

Why on earth would the entire country of the UK - which includes rural areas, small towns, 4 nations etc -
have anywhere near the % infection level of an international city that is one of the world's epicentre ?

Actually, I can't believe that "hidden periods" article was only posted yesterday. They predict that the UK part of the pandemic will end on 20 Aug 2021, and will result in a total of 214,899 cases. Even with our limited testing we are already on 233,151 cases, so I think their estimates may be a bit out...

I do think it HIGHLY unlikely that it was in Germany and Italy over a full month before the UK.

"Manchester was twinned with Wuhan?*

Thanks, red That is interesting
Maybe Manchester Uni was infected by the same strange infection that made Wuhan's figures so dodgy ?

I don't see any sign of the consequent excess deaths in France, Germany, UK, Spain over winter,

• and I'd expect some rise if there were any more than a few isolated cases of COVID brought in from Wuhan that didn't spread.

Total deaths in Germany over the last winter were unusually low - see the red 2020 curve compared to the grey curves from previous years

You can clearly see the high grey curve for 2017, which was a severe flu year

Similarly for other countries mentioned - where is the earlier increase in deaths ?

There is the claim - without evidence - that this early version was milder
However, scientists in the field have said that COVID is not the type of virus to have mutations that change much over time wert how the illness presents in patients
They say e.g. that the mutations to date won't affect being able to use just 1 vaccine for all

Even with our limited testing we are already on 233,151 cases

Though DH has made the point that we've had 233,151 positive tests rather than cases (people can test positive multiple times).

Maybe Manchester Uni was infected by the same strange infection that made Wuhan's figures so dodgy ?

Well when I found out it made me sit up.

The relationship and close ties between the council and business goes back to 1986 I beleive.

I am surprised we haven't seen at least some evidence of the outbreak early on in Manchester for this reason.

Lol big!

I'm supposed to be working. Random google...

assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/545779/gs-16-14-future-of-cities-twinning-contact-britain-china.pdf

@RedToothBrush

I'm not certain your husband is right.

The wording from the DHSC is

"As of 9am 14 May, there have been 2,219,281 tests, with 126,064 tests on 13 May.

1,593,902 people have been tested of which 233,151 tested positive.

As of 5pm on 13 May, of those tested positive for coronavirus, across all settings, 33,614 have sadly died."

Which does infer that the 233151 are on unique people, not repeated tests on the same person.

With this country and their attempt to keep the statistics looking good, who can really tell, though.... Though I would expect them to be trying to understate the number of cases, rather than overstate (by reporting the same people twice).

Ugh that Manchester paper. Been through peer review and still drops clangers like in the attached image.

I find it so frustrating that there are so many pieces of "research" published by "respected" scientists and shouted from the media headlines when a decent read of the paper by even a layperson (ie me) can see glaring holes/ errors or simply dangerous assumptions.

Agreed. I’m now enormously more sceptical of reports of scientific research than i was before this began and of “experts” in general.

On the off-chance the Manchester study is correct - that at least 25% of us have had it, that would do something quite stunning to our death statistics. It seems so unlikely that other countries quoting 5% max infection with very similar trajectories to ours have so many more deaths. How are they getting away with printing this stuff, let alone it getting on the front of newspapers?

Keep losing this thread.

Nick Stripe @NickStripe_ONS

THREAD – my first take on the COVID mortality reports we released today

1.Deaths occurring in April
2.Deaths in the care sector

Our data sources are:
-death registrations
-death notifications to CQC & CIW

1/n
The number of COVID-related deaths increased from 7% of all deaths in March to 37% of all deaths in April

Of these deaths (approx. 34k), COVID was the underlying cause of death in 95% of cases (approx. 32k)

i.e. 95% died from COVID not just with COVID

2/n
In just 2 months we have seen as many deaths due to COVID as the 3rd most common cause of death in the whole of 2018 - our last full annual analysis

Annualised age standardised mortality rates (ASMR) are c.70% higher than normal in Eng this Apr and c.40% higher in Wales

3/n
In 90% of COVID deaths, the deceased had at least 1 pre-existing condition. On av they had 2.3 conditions, down from 2.7 in Mar

Main conditions incl – Dementia & Alzheimer; Ischaemic heart disease; flu & pneumonia; chronic lower respiratory; diabetes; hypertensive diseases

4/n
COVID was by far the leading cause of death in April – ASMR 587 per 100k (Eng); 481 (Wal)

2nd was Dementia & Alzheimer Disease – ASMR 209 per 100k (Eng); 168 (Wal)

3rd was Ischaemic heart disease – ASMR 85 per 100k (Eng); 86 (Wal)

5/n
We already know that approx. 25% of excess deaths above 5 yr averages are not explained by COVID according to what has been recorded on death certificates

Today’s analysis starts to bring some insight into what might be going on.
Further work is underway

6/n
I have stated before that there are two likely reasons for this:

-Disruption to normal care pathways

-COVID being present and exacerbating other conditions, but not obvious to the certifying medical practitioner in the absence of a test

7/n
ASMRs for Dementia & Alzheimer have increased very significantly in April which may provide some evidence for the second reason.
Up 83% Eng & 54% Wal

Although less prevalent, ASMRs for “ill-defined conditions”, more frequently noted for the very old,
are also up c.100%

8/n
There have been c.23k more deaths of care home residents than we would expect so far this year, most in April

c.12.5k are COVID-related, with c.9k occurring in care home settings and c.3.5k in hospital settings

Care home residents now account for c.35% of all COVID deaths

9/n
The daily number of care home residents that died with COVID in hospital settings levelled off during April – to approx. 75-130 per day

But they increased in care home settings to reach a peak of 425 deaths on 17 April

10/n
A lower proportion of death certificates of care homes’ oldest residents in Mar & Apr mention COVID:

27% of all care home residents’ death certificates mention COVID

Whilst 33% of those aged 65-69 mention COVID

This falls to 24% of those aged 90+

11/n
From death notifications to CQC in Eng we can also assess at a headline level the number of deaths of recipients of domiciliary care

From 10 Apr (when CQC indication of COVID was first captured) to 8 May:

• 3,161 recipient deaths were notified, 593 (18%) indicated COVID

12/n
Looking at CQC notifications for the same time period from the previous 3 years, this is:

• c.1,600 more deaths of recipients of domiciliary care than were notified in 2019

• c.2,000 more than the 3 yr average

13/n
When using aggregate statistics, please remember that behind every number is someone who has died,
leaving family, friends and communities grieving in very difficult circumstances

We process & code every death registration/cert - it remains very real for us

Take care🙏

Office for National Statistics (ONS))@ONS*

There were 33,841 deaths involving the #coronavirus (COVID-19) that occurred between 1 March and 30 April 2020 (registered by 5 May 2020) in England and Wales.

Of these, 32,143 (95.0%) had #COVID199* assigned as the underlying cause of death

http://ow.ly/plIr50zGYxX

Compared to the 5-year average
the rate of deaths due to Dementia and Alzheimer disease was significantly higher in April 2020,

we are currently investigating the increase in non-COVID deaths and will publish more in the upcoming weeks

http://ow.ly/OjgL30qGgkL

"Manchester was twinned with Wuhan?

Thanks, red That is interesting
Maybe Manchester Uni was infected by the same strange infection that made Wuhan's figures so dodgy ?

I know our number of cases doesn't appear to be decreasing dramatically but considering the drop in hospital admission over a week and number of critical care beds can anyone work out statistically if it's more tests and picking up more cases due to any symptoms or continuous percentage of positives correlates with number of admissions and continuing death rate and increased testing is having no bearing on the figures

Thanks bigchoc I suggested above it would be really useful to know what the other excess deaths were assigned a cause of. I would suspect there is an increase in DV, suicide and possibly heart attack/stroke when the hospitals weren't overwhelmed but you couldn't get through on 111 or 999 to get there.
But I'd also be interested in comparing the stats on Murder (things like knife crime etc) and RTAs and compare how much these have dropped to see the bigger picture of how this pandemic has affected us nationally as world as internationally.

The results of Manchester study may only be credible if you assume that only a fraction of C19 patients produce antibodies.

Most of the antibodies studies I have seen so far focused on hospitalised or older patients. It has been suggested that younger and healthier group may fight C19 with T cells and may not produce standard antibodies (IgG).

As far as I know, most of the approved antibody tests assess IgG levels so they may miss people who have had milder cases.

Of course, there is a question whether there is such a thing as T-cell immunity and if people who got C19 mildly and did not produce antibodies may get reinfected. I am sure many scientists are trying to find answers to those questions but for now we do not seem to have sufficient data.