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 💐

I agree @ big choc, for me the lockdown date was key to uk high death rate. I watched in horror from France at the uk late lockdown. One week later in France would have been devastatingly worse and i cant imagine..I also have to stop myself reading some threads about the uk over counting!

Hugely grateful to all posting such great analysis on these threads. Kept me sane, I have to say. As well as frustrated at the quality (or otherwise) of our journalism.
Agree. First question today from bbc journo was a rewrite of the same question they always ask.
Do they think we are opening up too quickly?

I thought we'd had that answered frequently? And it was answered again. It had already been explained how it would happen in the opening speech.

I do think some of the scare mongering and not enough stats has had an effect on some people. I live in an urban town in a very mixed socio economic area. I see people who seem to have never heard of Covid and others who walk and drive around at all times wearing gloves and face mask. They aren't even near anyone!

I love this thread for the stats and I've been able to make what I think is an informed decision one my every step. I've tracked cases in my town, neighbouring towns and my county.
I also will make different judgements if I see cases rise.
But at the moment we are at 1/100k population .

Impact of coronavirus in care homes in England: 26 May to 19 June 2020

First results from the Vivaldi study,
a large scale survey which looked at coronavirus (COVID-19) infections in 9,081 care homes providing care for dementia patients and the elderly in England.

https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/articles/impactofcoronavirusincarehomesinenglandvivaldi/26mayto19june2020

Factors affecting the risk of infection in care homes
.....
care homes using bank or agency nurses or carers most days or every day are more likely to have more cases in residents
(odds ratio 1.58, 95% confidence interval: 1.50 to 1.65),
compared with those care homes who never use bank or agency staff
....
care homes in which staff receive sick pay are less likely to have cases of coronavirus in residents
(odds ratio 0.82 to 0.93, 95% confidence interval: 7% to 18%),
compared with those care homes where staff do not receive sick pay.

Our media never shine a spotlight on the good stuff. We are world leading at research and vaccines but it never gets a mention.
Nz have done well at eliminating the virus but they are now stuck waiting for other countries ( maybe the uk) to save them with a vaccine. They can't make their own.

Totally agree with above from another poster.

That's really interesting @Bigchoc. Especially the bit about sick pay, as that applies also every type of workplace. There's been so much focus on the "science" - by which read epidemiology and mathematical modelling - and not enough at all of the psychology/sociology of human behaviour. And mandating full pay for anyone who tests positive, awaiting test results or self isolating would be such a simple thing to pass legislation on. I'd also like to see airlines forced to rebook for free any passengers who have to cancel flights due to symptoms. So great to see some statistics showing there are things you can do to influence whether people stay home when sick or not

outbreaks ?

PHE have stated that their "other settings" for outbreaks includes workplaces

An "outbreak" is defined as when two or more lab-confirmed cases of Covid-19 have been linked to a particular setting
e.g. a care home, a church, a meat plant

However, I wish statements about the number of outbreaks also always included the total number of cases
as they are likely to be very different in the above examples

• which matters when assessing the significance

There are articles that care home outbreaks have reduced from 112 to 58 in the last week,
but that workplace outbreaks have increased by 100% to 36 outbreaks,
with the attached PHE graphic

e.g.
https://www.theguardian.com/society/2020/jul/02/suspected-covid-19-outbreaks-in-english-workplaces-double-in-a-week?CMP=sharebtntw

However, how many total cases are there in those 36 outbreaks in the workplace ?
We shouldn't have to dig this out each time
(Germany's RKI report normally gives both)

Clearly the number of outbreaks is still much smaller than those in care homes, but the total number of cases in each is important too
(obviously that can change quickly - one outbreak at a large meat plant has completely dominated infection figures for all Germany for a couple of weeks)

@itsgettingweird that’s so impressive, can you pleas explain how you managed to track cases to your town? I’m trying to understand the data and I’m a long way off that level of understanding

"Our media never shine a spotlight on the good stuff. We are world leading at research and vaccines but it never gets a mention."

There have been several articles on this, but there is a limit to how often people will read about a vaccine that may arrive sometime in the future,
maybe from the UK, also maybe from the USA or France.

There have been 67,000 excess deaths in the UK, which is bound to fill far more articles

• especially as many could have been avoided by better decisions and systems.

The media do report BJ's claims of "world-beating" contact tracking, App etc
which hype may cause the genuinely world class (not world-beating) UK research to be overlooked.

I would rate the ONS as genuinely world-beating
No other country I know of has such superb statistics, extensive reports & analysis and so up to date.

It's a shame statistics aren't sexy enough to be shouted about by politicians.

Just seen this. They seem to be implying that severity is NOT an age issue. I know @ShootsFruitAndLeaves will settle this. www.google.com/amp/s/m.huffingtonpost.co.uk/amp/entry/eight-in-ten-elderly-care-home-residents-with-coronavirus-asymptomatic-new-study-finds_uk_5eff84dbc5b6ca97091e349a/

The first drug to cut mortality was actually found by British researchers - and is cheap and readily available. This is currently more important than a vaccine, although we have a couple of decent candidates there too. Our scientific research is world leading, unlike some of the governments claims for e.g. the track and trace system.

We are also pretty good at muddling through - so our government is totally incompetent but people sew scrubs and make visors on school 3-D printers to help provide PPE.

France were at one stage planning to use full face snorkel masks to supply oxygen to patients - incidentally using a valve designed in Britain. Anyone know if this actually happened? There was a worldwide attempt to repurpose CPAP machines to supply oxygen, again I'm unsure of the extent to which that happened but certainly the focus changed from using ventilators to other means of delivering respiratory support.

I'm surprised that people claim to know the actual details of the contracts with the private testing laboratories - who really knows what data was requested as part of the contract. What is clear is that the processing of tests was extremely poor. e.g. www.independent.co.uk/news/health/coronavirus-testing-lighthouse-labs-nhs-deaths-delay-a9589381.html and www.hsj.co.uk/coronavirus/exclusive-test-data-from-commercial-labs-going-into-black-hole/7027619.article

Note that one of these suggests that area information was supposed to be provided.

France were at one stage planning to use full face snorkel masks to supply oxygen to patients - incidentally using a valve designed in Britain. Anyone know if this actually happened? There was a worldwide attempt to repurpose CPAP machines to supply oxygen, again I'm unsure of the extent to which that happened but certainly the focus changed from using ventilators to other means of delivering respiratory support.

A hospital in Warrington was in all the papers as they cut their death rates hugely by doing this I think. I don't know if that became widely used but I think hospitals generally realised that ventilators weren't always the best solution.

@oldbagface

A couple of points... The study states that 80% weren’t symptomatic “at the time of testing” rather than they were and remained assymptomatic. That said, it seems a reasonable proportion of older people asymptomatic. However, the evidence is overwhelming that age is an extremely significant factor in severity and morbidity for those that are symptomatic.

Thanks, BigChoc and Patricia for the clarification on tests; I'd got it in my mind that something was different / had changed between the ONS and daily figures, so it's good to have the facts

"The first drug to cut mortality was actually found by British researchers ... This is currently more important than a vaccine"

The first Coronavirus drug was remdesivir, which shortens the length of illness by about 30% and slightly reduces mortality.
It was judged so important by US Public Health that Trump has bought up the next 3 months global production though !

Dexamethasone is more important, but it still only reduced deaths for some of the most serious cases -
and would actually worsen those not on O2

The researchers say it would have saved about 5,000 people
==> A vaccine is much much more important - but just is not yet available

We also need treatments to stop mild cases becoming serious

I have everything crossed for the oxford vaccine to get us all out of this mess. It would be such an added bonus for it to come out of the UK!

I don't care where it comes from, or where any drug comes from,
just that it is safe and comes quickly - and isn't bought up by Trump

HuffPo is completely blocked by my content blocker - seems too entwined with ads and I cba to adjust settings

Is there a paper referenced, or authors ?

Yes my only concern is if it is made somewhere else and then we can’t get hold of it because of Trump (or anyone else) being selfish....

@OldBagFace

study is here www.gov.uk/government/publications/vivaldi-1-coronavirus-covid-19-care-homes-study-report/vivaldi-1-covid-19-care-homes-study-report

It's really quite shit and poorly described.

Best I can see, care homes do have more staff than residents, because they aren't all working at the same time. So the 210,620 staff and 172,066 residents sounds about right. But there should be 300k+ residents in 15,000 or so homes, but they have only 9000 homes here.

Also it says there might be duplicates here?

And it's not clear how they calculated the prevalence, apparently since March (?), of 10.5% of residents infected. Perhaps they did a good job with valid methodology, but who knows because nobody could be bothered to explain that there.

At any rate they are saying 3.9% tested positive May 11 - June 7, but not clear if that's unique residents or how many times tested or anything.

They are saying 80% asymptomatic but who knows because this 'study' is absolute shit.

The ONS report is now quite good

www.ons.gov.uk/file?uri=%2fpeoplepopulationandcommunity%2fbirthsdeathsandmarriages%2fdeaths%2fdatasets%2fdeathsinvolvingcovid19inthecaresectorenglandandwales%2fcurrent/julydeathsinvolvingcovid19inthecaresectordataset02072020155122.xlsx

If we look at Table 5 & 6, we can see that there were very large excess deaths during April but they disappeared during May

i.imgur.com/Hw7JUj2.png

For example at the beginning of March there were no covid deaths, and hospital deaths were 40-60 daily, while care home deaths were 300-380

On 26 March, 496 deaths were recorded in care homes, and 73 in hospitals (of care home residents), with 18 and 26 covid-19 deaths for each respective cohort.

It's likely however, the true total of covid-19 deaths in care homes was 100+, since the total number of deaths was more than 100 over any day to mid March.

The hospital deaths, meanwhile would have been much more likely to be tested, and indeed while 160 care home residents died in hospital on 20th April, the worst day, 133 of these were with covid-19, so net was only 27.

In fact if we consider the 2nd to 8th of March, when no care home patients died with covid-19 in hospital, 365 people died in hospital, or 52 per day. At no time after 22 March has the number of non-covid daily care home resident deaths in hospital exceeded 50.

I.e. there were no excess (beyond covid) deaths at all in hospital.

There were however thousands of excess deaths in care homes. These occurred from day 1 more-or-less, so it's highly improbable that these are lonely Alzheimers patients, but much more likely undiagnosed covid-19.

So sure, maybe lots of these old people with covid were asymptomatic.

But it seems that didn't stop it killing them!

Covid-19 killed lots of old people without doctors noticing that they had been killed by covid-19.

The 10.5% total prevalence seems like a wild under-statement to me, as we've not that far off that number of total dead during the pandemic.

oh beg pardon I see ONS have a fuller description of the study as linked to by bigchoc above.

www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/articles/impactofcoronavirusincarehomesinenglandvivaldi/26mayto19june2020

I think the impact of funding cuts and councils absolving themselves of social care to disabled people In their home due to covid is really worrying.

This BBC article this morning says that two thirds of all people who died from covid were disabled. This seems extraordinarily high; what classification are they using for ‘disabled’?
Is it - those who need care (ie mainly Care home deaths - mainly those 75+). Or anyone who had a pre-existing condition?
Coronavirus: Why disabled people are calling for a Covid-19 inquiry www.bbc.co.uk/news/uk-53221435

I'm dreaming about a day when we hear "no new cases in the UK". I'm not sure it'll ever even happen. At least we have reached a point where number of deaths are below the 5 year average for this time of year. When it comes to new cases, do you think that day will come, and if so, predictions of how long it'll take to get there?

@shoots Thanks for the ONS link to the care home study that you posted. It looks as if the reason they focused on 9000 odd care homes rather than 15,000 is that they were only looking at homes for over 65s or dementia patients (who would be largely over 65,) So excluding homes for disabled younger adults I assume?

The numbers so look rather low - though, as you say, the study is a bit scant on detail. One possibility is that they may have asked them only about current residents - even without Covid the average time someone spends in a nursing home is about 6 months (people typically enter when they're already very frail) - so they will have typically lost about half their residents who were there in the peak of the epidemic (plus those who died from it, obviously) It's not clear that this is what they've done, but they say they use the figures to estimate the total proportion of staff and residents infected, and to do this you'd want to focus on the current population not anyone who's lived/worked there in the last 5 months. It would be poor if they have done this though.

Another possibility that they don't discuss is non response bias - only 56% of care homes responded to their phone calls. Responding is very quick and easy for those who simply have to tell them they've had no cases. It's more complex for those with cases who may have been unsure of the figures, or anxious about sharing personal data or reputation. A smaller sample with 100% response rate (or close to) would be more convincing.

It may also be low because most of the excess deaths (and so we assume most of the cases) happened when testing rates were low and they are being asked only for confirmed cases.

ONS: Coronavirus (COVID-19) related deaths by disability status, England and Wales: 2 March to 15 May 2020

Statistical weighting is vital to properly assess the magnitude of risk, such as disability
This ONS report illustrates again that socio-economic factors significantly worsen risks.

Once the stats re disability are fully normalised wrt such factors,
the risks due to disability are reduced from the alarming headline rate, but are still significant.

The hazard ratio for disability is higher for females:

https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/coronaviruscovid19relateddeathsbydisabilitystatusenglandandwales/2marchto15may2020

The fully adjusted model illustrates the results after further adjusting for region, population density, area deprivation, household composition, socio-economic position, highest qualification held, household tenure, multigenerational household and occupation (including key workers and exposure to others) in 2011.

Therefore, the fully adjusted results show relative differences in risk between non-disabled and disabled groups that are specific to disability status
and are not associated with any of the factors listed by which members of the groups might differ.

Adjusting for these factors substantially reduces the estimated risk of death involving COVID-19 for disabled people relative to the non-disabled group.

Lurker someone on here linked to a huge table and it showed the UTLA as well as LTLA.
I had to scroll through for a longtime but I managed to find each day for my town!