Daily numbers, graphs, analysis thread 14

[BigChocFrenzy]

Welcome to thread 14 of the daily updates

Resource links:

Uk dashboard deaths, cases, hospitals, tests - 4 nations, LAs, English regions
Slides & data UK govt pressers
[[https://www.gov.uk/guidance/coronavi
rus-covid-19-information-for-the-public UK stats]] list of reports added daily by PHE & DHSC
PHE Surveillance report infections & deaths released every Thursday with sep. infographic
ONS England infection surveillance report ONS UK statistics for CV related deaths, released weekly each Tuesday
Daily ECDC report UK & EEA
Worldometer UK page
Plot FT graphs compare countries deaths, cases, raw / million pop
Covidly.com world summary & graphs
Plot COVID Graphs Our World in Data additional data

We welcome factual, data driven, and civil discussions from all contributors 📈 📉 📊 👍

There will surely be other cases like that, care homes were not in a bubble.

Yes, it is true for idiosyncratic cases but, again, it does not seem to be the primary vector. Again, from SAGE:

assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/897497/S0343_Care_Homes_Analysis.pdf

The means of introduction of new disease cases to care homes is multiple: linked to the connection between care homes and hospitals, and the connection between staff and community (medium confidence). We cannot say which is more influential at this stage.
• We expect (high confidence low evidence) that hospital discharges and visits may have been an important source of introduction of C-19 infection to care homes.

I believe the facts that we know now (e.g. genome analysis of London care homes and PHE stats on outbreaks) seem to indicate that the hospital discharges have been "more influential" than transmissions from care home staff. I fully accept the analysis is not entirely conclusive.

I will see if I can dig out more info on asymptomatic transmissions in care homes. I seem to recall I have seen an interesting report on this.

Adding my thanks to @littleowl1 - really helpful to have the emails (for two areas - mine and my parents') arrive so promptly this morning.

The and visits above is the bit that I meant; people visiting their loved ones is a vector that is really hard to quantify, but provides a more likely explanation for low case areas like the South West than hospitals. To be clear, I was not suggesting you missed anything in your interpretation of the report, I'm criticising the report itself as having gaps. It started with assumptions and looked to prove them, which is ok, but unfortunately they didn't include all other possible factors to rule them out and didn't do any genome sampling to get firm proof. They also refer to the London genome sampling that implicates staff, despite putting it down as a lower probability assumption. And they don't supply all the figures I would like to see to be able to do effective analysis. So I'm not happy they have done enough with this report.

And another thing - WTF is up with Liverpool care home workers (pg 14 onward) that they were off sick with respiratory diseases at the end of April and hadn't gone for tests? Hancock was begging daily on the TV for people to take tests to reach his magic 100k number, they would with vulnerable people but - nah, just take a few days off. No wonder we can't get a lid on this virus, people are ridiculous.

following on from @Firefliess yesterday evening

An initial view of London boroughs colour coded by week. Six groupings based on the weekly cases per 100,000 across all English LAs (excl upper tier county councils, as these are all covered by their constituent district councils). Essentially a quartile split plus highest and lowest 5%, so for example the top "quartile" reflects the 75th to 95th percentile - given the disproportionate case per 100,000 rate at the upper levels this probably needs splitting further, perhaps to 85th to 95th and 75th to 85th, given the much narrower spread using the second, third and fourth quartiles seems to work well. Splitting out the lowest 5%, as currently included, may indeed be unnecessary.

The coding number (eg 6 = in top 5%; 1 = in bottom 5%) is probably superfluous, and if anything beyond the colour to show the grouping it might be better to show the actual cases per 100,000 fig for each LA. Just need to fix the application of a multiple If statement plus application of conditional formatting into a single formula.

A starter for ten though - and available in this format for every LA currently ordered by region. Have used London as the example because of the striking movement - very high initial cases, then a quite rapid decline to quite low percentile, and now a bit of a relative increase again.

Actual data included at the moment to just re-inforce that levels are hugely lower today than back in March, April , May.

A starter for 10 though :)

13 hospital deaths today @ nhs england
0 deaths @ Wales, Scotland, NI

@wintertravel1980 Firstly, the research seems to indicate that contrary to the intuitive explanation that temporary staff spread out infection across care homes, genetic analysis did not find any overlaps across different facilities. In other words, each of the reviewed care homes had to deal with multiple COVID introductions at approximately the same time.

I just wanted to add that not all care homes use temporary staff, the home I work in hasn't used temporary staff in the 5 years I've been there, the home I worked in previously hadn't used temporary staff in many many years. In both cases we found that if we had higher staffing levels we could cover sicknes/holidays and it didn't work out any more expensive than paying high fees to agencies and provided many benefits.

This is worth a different thread to avoid disruption here of one topic, but I would like to invite you all to my soon-to-be-posted thread about what is an acceptable rate/100,000 people that the UK can cope with for now.
It might be useful then when we look at figures to internally say "ah, we keep on top of anything under 2/5/10/20 per 100,000, so that level is ok." I'd be interested in getting everyone's thoughts please.

@MRex

This is worth a different thread to avoid disruption here of one topic, but I would like to invite you all to my soon-to-be-posted thread about what is an acceptable rate/100,000 people that the UK can cope with for now. It might be useful then when we look at figures to internally say "ah, we keep on top of anything under 2/5/10/20 per 100,000, so that level is ok." I'd be interested in getting everyone's thoughts please.

@MRex very opportune

I was just playing around with my percentile charts. Fixed - "bad", "better" "good" etc thresholds are a simple visual option me being lazy so views and discussion on that would be very useful.

I've currently gone for a visual starter with under 5 per 100,000 as the max to head into "green territory", with anything below 2 better again, and zero of course being the ultimate standard. I'd probably rather set a lower number than 5 per 100,000 as a potentially acceptable level.

PHE twitter feed

"Owing to technical difficulties with data processing, the England and UK cases and tests data is not expected to be updated today.

The #COVID19 data dashboard is available here: coronavirus.data.gov.uk"

the daily updates have been pretty shambolic for several days now; missing counties, missing data, no data at all, and same again today.

Given they've been publishing daily data for getting on for six months now it does make you wonder:

• single point of failure ; the one person who knows what to do has finally been allowed to take some leave;
• softening us all up for less regular reporting?

I saw a good point made on Twitter which asked how the gov dashboard copes with private tests, given that 220 (ish) of the Northampton factory positive tests were carried out privately. Might be a reason for the delay?

@HoldingTight
is indeed a good point. As they don't fall into the 'official' pillars it wouldn't surprise me if the government failed to count them.

Love the coloured tables @boys3 (I only leaned how to do conditional formatting on Excel about a year ago and currently use it for everything!)

Re problems reporting - I suspect it's related to increased numbers of private providers undertaking Pillar 2 tests who've not yet set up the systems necessary to share data with government. I'm sure the aim is to get it daily and keep it that way - because they do need people to keep an eye on things and act accordingly.

Care homes were being forced to take patients with Covid in order to keep funding payments.

No. They had to take patients in order to receive extra funding. So in order to get extra money to deal with the outbreak they had to participate in dealing with the outbreak.

wintertravel some off the cuff reactions before I look at the studies further.

London care homes include high levels of BAME staff, noted in one of the reports. BAME groups in London had high levels of community transmission.

Movement of staff between homes possibly varies between areas. Do you have any data on size of home by region? In the rural south west I'd expect movement of staff to be more significant.

The South West experienced one of the first outbreaks of Covid-19 when someone returning from a ski trip attended a Torbay church service. I would expect that to have spread to care homes.

It would not be surprising if some of the patients moved out of hospital to free up beds turned out to be asymptomatic or presymptomatic - but biased studies dont prove that.

I am surprised how few people seem to know that in order to have enough beds for Covid patients NHS hospitals were converting theatres to ICUs (leaving some for emergencies) / learning how to make one machine cope with two patients/ having your surgeons become ICU doctors and your junior doctors nurses. Does no-one remember the plans to prioritise patients because the NHS anticipated that capacity would be exceeded? Or that one London hospital did have to declare a critical incident? www.bbc.co.uk/news/uk-england-london-52812457 (and was apparently followed by a couple more).

No. They had to take patients in order to receive extra funding. So in order to get extra money to deal with the outbreak they had to participate in dealing with the outbreak. No some have contracts for x number of beds. They don't get to pick and choose, if they are being paid for 10 beds and they currently have 9 residents their contract means they take the next person allocated. This is how it works in my area, we considered doing it a couple of years ago as it gives the home a guaranteed base income but we decided against it.

I have no idea if that is national.

Haven't read the last couple of pages yet but what's going on with the figures, patients admitted to hospital especially?
The last week or so, Wales seem to have had more (days where) patients admitted than England, which sounds very surprising to me. The last couple of days of date especially seem to put Wales much higher...
Am I right in thinking this is an error / data issue rather than true?

@ancientgran

No. They had to take patients in order to receive extra funding. So in order to get extra money to deal with the outbreak they had to participate in dealing with the outbreak. No some have contracts for x number of beds. They don't get to pick and choose, if they are being paid for 10 beds and they currently have 9 residents their contract means they take the next person allocated. This is how it works in my area, we considered doing it a couple of years ago as it gives the home a guaranteed base income but we decided against it.

I have no idea if that is national.

The post I was referring to specifically said that they were threatened with the loss of funding if they didn’t take Covid patients, blackmailed essentially, but linked to a Sky News story about access to extra funding to support dealing with the pandemic. What you said may well be true but not the same thing.

Oh I am so so glad you like it so much! What an amazing endorsement. If anyone else would like to sign up you can do so for free at www.covidmessenger.com.

@Shitfuckoh

Wales count differently. They count anyone who presents that may have covid. England only counts those that test positive. That's why the Wales number is so high, because it's always going to be massively out.

@alreadytaken The South West experienced one of the first outbreaks of Covid-19 when someone returning from a ski trip attended a Torbay church service. I would expect that to have spread to care homes. I live in Devon,, the main issue seemed to be the Torbay school he attended which was closed, I think for over a week and when it reopened some students were still isolating, presumably the ones who had been in most contact with him. The skiing trip was in half term so late February but I think it was the 1st or 2nd March when it was reported in the press.

Although the local church closed for deep cleaning I don't think any outbreak was connected to the church although a local surgery and other schools closed. I don't know if anyone at the surgery was confirmed with covid, if it was a GP or nurse I imagine they would be more likely route into care homes.

The outbreak was in Torbay, they have had 28 deaths from covid in care homes and under 25% of homes have had outbreaks. I think both are lower than national average but I'm sure one of the statistics experts on here could be more certain of that. Not sure how it compares to the rest of Devon or indeed the wider South West.

The home I work in isn't in Torbay so I don't know anymore details.

@PrayingandHoping Ah! Thank you, it makes sense now but what a strange way of doing it.

@Witchend

I’m writing from my mobile so I can’t seem to see how to directly reply to each. I’ve been offline this afternoon for a much needed break from screens.

The feedback above is brilliant - I can’t tell you how helpful it is to hear what works/what doesn’t. and it gives me ideas on what to change/build next!

I will add a note to make it clear about the absolute number vs per head of population. I felt absolute numbers keep it clearer for some - Even if I personally prefer per head of pop. I see it with my mum - her eyes glaze over when I talk about per head of pop!

I love the scrolling idea for those who have signed up for multiple areas. Will definitely add that to the to-do list.

On the stats - I’m not adding them manually - I wrote a program and data model to extract the data from the governments daily data publication and insert it into emails. It’s still a pretty time consuming process though. Hopefully if we have enough subscribers and get some sponsorship the first thing I’m doing is getting a programmer to automate it more so I don’t have to get up at 5am everyday!!!!!

So all I can ask is that you share, share, share with family friends and colleagues.

Forward the email to people so they can sign up. Follow us on Facebook.

If we have enough subscribers i feel confident we will get sponsorship and then I can keep the thing going for the duration of the pandemic. The 5am starts are beginning to bite. I’m not as young as I used to be! So I’m hoping we get sponsored sooner rather than later.

For anyone else who wants to sign up it’s www.covidmessenger.com

@AugustBreeze @MRex @Pebble21uk @Frazzled2207

@littleowl1

I'm a bit late to the party (I'm catching up on the last couple of days) but I've signed up & eagerly await my email tomorrow morning 👍🏻

I've shared it on Facebook as well, I know of a few people who I think would be interested.

That is one epic outbreak in Northampton.