Daily numbers, graphs, analysis thread 9

[Barracker]

Welcome to thread 9 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.

T cell immunity is part of the immune response leading to production of antibodies. Depending on which T cell type. Immune response is a process so until virus fully beaten there will be some replication and hence shedding. Whether measurable and of a large enough dose to give infection are also different issues. It may also be that the T cell response is part of the cytokines storm that causes significant problems in some people.

Sorry posted this on wrong forum.

Neurotrash
And if you want to understand the Republican mindset that is causing the USA numbers, have aread of this
eu.ydr.com/story/news/2020/05/28/pa-rep-brian-sims-calls-out-gop-members-hiding-coronavirus-cases/5273786002/

Interesting theory about vascular links

https://elemental.medium.com/coronavirus-may-be-a-blood-vessel-disease-which-explains-everything-2c4032481ab2

Listening, wow. Horrific.

What are the numbers today?

2445 cases, 215 deaths

Fascinating thread on Twitter which explains T cells, antibodies etc:
twitter.com/askeladdentx/status/1266479519789191168?s=21

Late lockdown seems mostly what has caused the UK's unusually high level of
excess deaths = total deaths - historical average over the same period

The UK is also unusual in having had hotspots across the country

• another sign of lockdown being too late - and probably why it is taking so long for deaths & cases to fall to the level of comparable European countries.

https://www.ft.com/content/6b4c784e-c259-4ca4-9a82-648ffde71bf0

Unlike other hard-hit countries, there was a large rise in the excess death rate across all parts of the UK,
with London reporting by far the biggest jump.

By contrast, in Italy, the epidemic was concentrated in the northern region of Lombardy,

and there were two hotspots in France:
one around Paris and the other around the eastern city of Mulhouse, near the German and Swiss borders.
....
Although the exact number of infections can only be estimated at the moment of lockdown,
the statistical relationship between that and excess deaths is strong.

Given we were originally estimated to be 2 weeks behind italy is it true that our new cases are falling lower than italy's 2 week ago. I looked at news reports for mid may suggesting cases around 1 to 1.5k per day but am not sure how their testing compares with ours.

Also not seen any figures on how many cases our soon to be launched track and trace can deal with. I know the ons estimated 8000 new cases a day. This seems to many to contact trace

Also surprised on briefing today we have no idea about what demographics and professions are being infected. Syrely it would be possible to generate some rudimentary information such as job role postcode etc as part of the online portal?

Oops, just seen my weird App typo at 10:37 am

"we'd need to compare [COVID + excess deaths] for each country,"

Sorry, obviously I mean excess deaths !

From those graphs @ big are we seeing a very slight increase in excess deaths? It looks like it is going up again ever so slighty in some areas of the uk, and thought is saw the same on the euromomo graphs.

walksen We need to compare like with like and I haven't seen what the Italian version of ONS estimates for their real number of cases,
as distinct from cases they confirmed by test

To avoid differences in testing - which would distort case figures, I would suggest comparing the UK hospital COVID deaths to Italy deaths 2 weeks ago
(because Italy does not include care home deaths)

Italy on 16 May had 153 deaths
To avoid possible data blips, I'd also compare the rolling 7-day average of each

@humphriescushion If you look at the original source document for the FT graphs (www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/deathsregisteredweeklyinenglandandwalesprovisional/weekending15may2020#deaths-by-region-in-england-and-wales)

it shows excess deaths with an uptick, BUT makes the statement that COVID deaths (which they break out and the FT don't) are decreasing - so the uptick is not due to an increase in COVID deaths.

I believe it is something to do with a backlog of reporting non-COVID deaths, as the data is on day death was registered, not when it occurred. So it appears that, in England and Wales, there was a backlog of non COVID deaths reported about 2 weeks ago - deaths that are likely to actually have occurred up to about 6 weeks earlier. Which results in a strange looking uptick on the graph, but doesn't actually mean that week saw more deaths.

funnily enough the slides havent appeared on the website yet - seems to me that deaths are now levelling off rather than falling and hospital admissions are showing a very slight rise while still being below the level of a week ago, confirmed cases up too.

Ah ok thanks @ patricia, i will have a look. Looked a bit strange.
Yes i have been looking for the daily briefing slides for today as well @ already.

A lot of the "excess" deaths are unconfirmed COVID though, or with COVID a factor, especially in care homes

(which is why ministers & govt advisers keep saying that excess deaths is the most reliable thing to take)

Excess deaths are those above the historical average for the period, so we need to consider what else would be causing deaths to rise, especially in care homes

We can assume that deaths on any day are a cumulation of deaths occurring up to several weeks ago.
Was there some specific holdup that suddenly caused a lot of deaths to be released at once ?

I can't find the footnote, as iPad is refusing to zoom atm

According to the ONS, the first May bank holiday affected registration - see figure 3 on the link I posted in the previous slide.

And their explanatory text -

"The number of death registrations in Week 20 was impacted by the early May Bank Holiday, which took place on Friday 8 May 2020 (in Week 19). The number of deaths registered on the early May Bank Holiday fell to 88 deaths compared with 2,950 deaths registered on the previous Friday (Friday 1 May 2020). Trends seen in Week 19 and Week 20 should therefore be interpreted with caution, as deaths not registered on the early May Bank Holiday were likely registered in the following week (Week 20).
Early May Bank Holiday normally takes place on a Monday, however, to celebrate the 75th anniversary of VE day, the bank holiday took place on Friday 8 May. In this analysis a week is defined as Saturday to Friday. Having the bank holiday on the Friday rather than the Monday results in a larger effect on registration delays as the death would have to be registered in the following week. On a normal bank holiday, deaths that cannot be registered on a Monday could still be registered on the Tuesday to Friday of the same week, reducing the impact of the delay on the weekly figures.
When looking across the last two weeks and looking at a rolling seven-day period, which smoothed some of the effect of the Friday bank holiday, we see that the number of deaths overall has continued to decrease. It is important to note that we expect the number of deaths to return to a normal level two weeks after a week with a bank holiday."

Ah was just about to copy that. So appears to be bank holiday blip.

I hypothesized earlier in these threads that one of the key reasons for our high death rate was our multiple seed points exacerbated by half term travel into mainland Europe, esp Northern Italy, and then by the time we locked down 3 weeks later those "seats" having begin multiple points of exponential growth. Anecdotally a friend travelled from USA to Germany on 29 Feb, to an area only at that time declaring 60 cases, and came home a week later then tested positive to covid 19 on 9th March. She thinks she got it from a woman coughing on a train. What are the chances? But it happened.

I do wonder if by contrast we will be relatively more spared than some other countries should a second wave arrive?

Is there any data anywhere on people who have contracted COVID whilst observing a strict lockdown? By strict I mean maybe leaving the house for nothing other than a supermarket shop once a week or less?

Whatsnext that's a really fascinating thread!

FWIW I deal with burial authority records in my work
last winter was low numbers - mild
this spring has been heavy
but still in line with the normal rolling 12 month average
(this covers ashes, interments and second burials)

I do think that the only true comparison will be the excess over rolling averages
AND
that will rely on good data collection and publishing
eg Tanzania records 3% of births and 20% of deaths