Meet the Other Phone. Only the apps you allow.

Meet the Other Phone.
Only the apps you allow.

Buy now

Please or to access all these features

Covid

Mumsnet doesn't verify the qualifications of users. If you have medical concerns, please consult a healthcare professional.

Daily numbers, graphs, analysis thread 7

981 replies

Barracker · 28/04/2020 12:53

Welcome to thread 7 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.Flowers

OP posts:
Thread gallery
127
ShootsFruitAndLeaves · 05/05/2020 21:40

That apple.news thing is really quite obnoxious. Why do they think they have the right to stick their branding in front of other people's content/URL/brand name? At least Google's amp crap still includes the original link in the link.

ShootsFruitAndLeaves · 05/05/2020 22:25

@BigChocFrenzy that looks like 6596 excess deaths to 19 April. Not convinced the later date is completely up-to-date. We had 30,500 excess deaths to 17 April, and around 37,500 to 24 April, so roughly 32,500 to 19 April.

Hence 550 per million E&W, 575 per million Belgium.

Certainly close!

BigChocFrenzy · 05/05/2020 22:59

To reinforce the debunk upthread,
the very knowledgeable @TheMship (a scientist working in the field) posted this on Westministenders:

"If you see people panicking over what I think is a completely irresponsible media story about a "more contagious" form of the virus, please can you drop them this very informative link?"

https://twitter.com/BillHanage/status/1256422856436613126

whatsnext2 · 06/05/2020 10:08

Both @MillicentMartha and myself posted links about different strains yesterday (Tuesday at about 1300).

Anecdotally and from the Covid tracker app it might appear that the lesser strain was circulating much earlier but as the IFR was lower wasn't picked up.

I would have thought equally it might be good - if there is any herd immunity from first strain or it might also be bad - if all models used R and IFR from first strain but second strain has higher values of each?

Is there a possibility that subsequent infections might in a small proportion of the populations have a more extreme inflammatory response?

BigChocFrenzy · 06/05/2020 12:25

"Anecdotally"

but we don't have evidence of a weaker strain that was circulating widely

Scientists in the field say that the mutations of COVID, because of its structure, vary far less in their effects than say flu
Best explanation I've found is this:

https://www.city-journal.org/coronavirus-vaccine
Every virus has a genome composed of genetic material (either RNA or DNA) that encodes instructions for replicating the virus.

When a virus infects a cell, it accesses machinery for making copies of its genomic instructions and follows those instructions to make viral proteins that assemble, with copies of the instructions, to form more viruses
(which then pop out of the cell to infect new cells, either in the same host or in someone new).

There is a critical difference between coronaviruses and flu.

The novel coronavirus genome is made of one long strand of genetic code.
This makes it an “unsegmented” virus—like a set of instructions that fit on a single page.

The flu virus has eight genomic segments, so its code fits on eight “pages.”

That’s not common for viruses, and it gives the flu a special ability.
Because the major parts of the flu virus are described on separate pages (segments) of its genome, when two different flu viruses infect the same cell, they can swap pages.

Imagine two people with eight-page reports fighting over a copy machine.
In the tussle, some copies might turn out to have a mix of pages from two different reports.
This page-swapping process, where viruses exchange parts of their genome, is called reassortment.

The flu can change rapidly when multiple strains pass through the same host.
But coronavirus, as a one-page report, tends to stay together,
and while coronaviruses can swap sections

  • in a process known as recombination - it is difficult to achieve and thus rare. (Imagine two pages ripping in the same way and swapping pieces that get glued together again.)

Coronavirus does mutate.
.....
A vaccine is like a description of a wanted criminal:
it tells your immune cells whom to look out for.

So long as the suspect’s appearance doesn’t change too much, then the vaccine works.
....
What we don’t have to worry about is the virus rapidly mutating away from our vaccines as fast as flu can,
because owing to its simplicity, it can’t pull off the flu’s face-swapping tricks.

3luckystars · 06/05/2020 12:30

Are they including care homes/nursing homes now?

BigChocFrenzy · 06/05/2020 12:31

We need to wait for the results of the population studies with antibody tests

That will give us an idea about whether all these hopes for a hidden mass immunity from hidden previous infections are justified

The problem is that on MN everyone and his dog had a nasty bug in January,
or even going back to September, that they are sure was COVID

Since these posters are almost all young or middle-aged MNers and they post about how horrible the bug was,
then surely the 85+ group would have been dying in significant numbers much earlier.

However, the total deaths statistics show that they were not dying.

whatsnext2 · 06/05/2020 13:37

@BigChocFrenzy the problem is that the mutation alters the spike or ‘face’ of the virus so it may have a bearing on vaccine effectiveness.

Also in addition to ‘anecdotally’ there was the case from December in France. As I said it may have had a lower IFR. Surely if there was a nasty bug out there earlier in the year,COVID or whatever, with significant death rate it would have shown in the figures; so either everyone imagined it or it had low IFR.

Jrobhatch29 · 06/05/2020 13:54

Or it was fluHmm the antibody tests that other places have already done are not showing that huge amounts of the population have had it. Why would it be causing chaos in wuhan in january but quietly circulating as a mild illness here without anyone noticing? My facebook is now full of "told you i had it at christmas" because theyve found one early case in france. I would love it to be true. Also with how much people mix at christmas surely it would have exploded?

Jrobhatch29 · 06/05/2020 14:12

Also, and this is anacdotel as well, im a primary school teacher and schools were wiped out by norovirus before christmas across the country. It was allover news. Schools had to put in additional hygeine measures and lots of schools had to close due to high levels of staff absense. Surely if this was noticed and picked up by media, a new respitatory illness would have been noticed? There would have been lots of absense in all jobs

MillicentMartha · 06/05/2020 15:01

Ha ha, anecdotally I had the most weird flu type thing in January. Terrible headache, fever, cough. No one seemed to catch it from me though... And I work in a school. There are viruses around all the time. The excess deaths with this one are what makes it different.

The different strains on the link I copied previously I thought were interesting but I’m not trying to scaremonger. I just like the stats.

ShootsFruitAndLeaves · 06/05/2020 15:12

Here's what happens when you use data which are not of the same type to draw conclusions (re: IFS ethnicity study above)

See attached graph. What I have done here is compare the death counts from the three cohorts:

  1. NHS data - has half of the data truncated by excluding care homes (very old, very white), and deaths at home (presumably old, white)
  1. ONS data - includes care home/home deaths, but only 30% of these aren't certificated with covid-19, whereas 100% of NHS deaths are.
  1. True cohort - excess deaths for older age groups, covid-19 counts for younger people, who are not at risk (for sane values of 'risk', and outside medical contexts) at all from covid-19. We don't see excess deaths not captured in the covid-19 death counts till age 55+, where it's about 10%, but for 80+ it's 50% extra deaths, which is a HUGE number considering that 80+ is already the largest group in the NHS cohort

I have multiplied the 2 smaller cohorts by a constant number so that the sum of deaths in each equals the third. Here we can see that the NHS cohort is laughably unrepresentative and it just dangles pathetically in the 80+ age group, as these people die at home and in care homes.

Second graph shows overrepresentation by age group within the NHS data. Ignore the 0-19 group as it's such a tiny tiny risk, and we see that there's 40% overrepresentation of ages 20-60, with less and less overrpresentation as we get older. The exact point where the NHS cohort becomes UNDER-represented isn't clear because of their silly 20-year boundaries.

Suffice to say that if you use a cohort which is incredibly front-loaded with younger people, disproportionately minorities, and then you compare this with the age profile for a second cohort which is not nearly so front-loaded, and the real cohort is older again, then you are going to draw faulty conclusions, considering that your model shows an exponentially increasing risk in cohort 2 that's not even properly there in cohort 1, because it's so horribly truncated on the right-hand side of the curve, and your ethnic population is given, based on cohort 2, a much lower risk of death because it is MUCH younger than your white British/Irish population, but then your real death count is something else.

In other words if you toss away half the deaths, much older and therefore whiter, by ignoring the excess deaths and only analysing the ethnicity of the necessarily younger population in hospitals, then you won't get valid results at all.

Daily numbers, graphs, analysis thread 7
Daily numbers, graphs, analysis thread 7
NewAccountForCorona · 06/05/2020 15:48

That can be illustrated nicely with the Irish data (sorry, the link is fucking impossible to hide, I've lost it again).

Basically something like 57% of cases are in women, 43% in men, but it's the opposite way around with the death rate. Which means that men are much more likely to die from it.

When you add in that over 70% of the over-80s are women, and the median age for death in Ireland is 81, then men are even more likely to die.

If you just looked at cases you'd say "oh, women are more susceptible, look how many of them are being infected".

The HSE has fantastic infection rate figures, broken down nicely by sex and age and everything else you can think of. But they are hiding the equivalent for deaths somewhere arseways. About once a week I find it, bookmark it, and lose it again Hmm

ChazsBrilliantAttitude · 06/05/2020 15:52

The case in France isn’t anecdotal is it? I thought they had retested samples from someone who had pneumonia in December.

whatsnext2 · 06/05/2020 16:15

@ChazsBrilliantAttitude no French case not anecdotal, my bad grammar/ punctuation. Tim Spector’s COVID symptom tracker analysis , suggests there was a bad respiratory infection earlier in year.

Elmerrrrrrrr · 06/05/2020 16:22

I thought the French case was most likely a false positive

ChazsBrilliantAttitude · 06/05/2020 16:24

It is true there was something horrible around Christmas time. DS2 was really I’ll with a very high temperature for days. I then got a milder version. Who knows? In a normal year I would have just labelled it “one of those things”

ChazsBrilliantAttitude · 06/05/2020 16:24

They retested the sample for the French case several times.

Elmerrrrrrrr · 06/05/2020 16:26

If you look at Dr Gaetan Burgio's twitter feed he explains why its likely that French sample wasn't covid

azaleanth90 · 06/05/2020 18:17

Sorry to leap back in after pages, but can anyone say if the new figures we get daily include care homes?

dairyfairies · 06/05/2020 18:18

we have more than 6k of new infections as of today. I don't understand this. The numbers of new cases seem to go up even though we have been in lockdown . should they not go down? and why is the government making noises about easing restrictions from Monday onwards when our cases are in fact increasing instead of coming down. Or are they not going up and we are just detecting more because testing increased?

there seem to be some pretty knowledge people on this thread. can anyone help me out here?

BirdieFriendReturns · 06/05/2020 18:25

Surely if they are testing more people, cases will go up?

There could be another hundred thousand people or more at home with symptoms who don’t need to seek help.

Elmerrrrrrrr · 06/05/2020 18:26

I assume cases are going up because they're testing more.

dairyfairies · 06/05/2020 18:28

I assume cases are going up because they're testing more.

but is it that simple?