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.

Yes, SHoots When it comes to comparisons with previous years, we should really stick to comparing for the same weeks
COVID may have caused 2020 to have all the top 8 weeks for the mid-March - mid-May period

I was hoping also that we would soon have the date of death, which is much more useful, not just of registration but from your last post this could take several months ?

If you look at the ONS release by local authority we have weekly counts for deaths by date of death. However I don't think this was published in the past, and it's not something that is aggregated particularly.

Actually we don't need to wait a year - there are some deaths not registered for many many months due to inquests, etc. ,but they are rare. It's more that a week out we definitely DON"T have all the deaths, but we can estimate roughly how many there will be because we know what % are roughly registered after 2, 3, 4, 5 etc weeks.

The issue is more that while we have reasonably accurate data for weekly deaths by occurrence NOW, the previous weekly records were caused by old people dying of flu combined with post-NY registration delays, but nobody really cared, so the ONS never bothered to publish those figures by week of death, because it was like we had 14,000 registrations in a week rather than 12,000, the true number of deaths might have been 13,000 and it wasn't really something anyone was interested in discussing....

Yes that spanish stuff shows it could be another 6-12x as many deaths yet as have happemed in each country. Though antibodies are 1month behind so it will be slightly higher.
But this is why people who keep saying the numbers of dead under 40 is so low 3k or something. Well it's not when you account for herd immunity. The chance of you dying may well be low but the total of dead people is high.
Once you account for the facy teachers are only teaching 2% of stufents then you get 65 educational staffx3 which is higher than the drs and nurses which is also over 100.
I dont think it really matters what is on the DC as long as the person wasnt murdered. The ons clearly shows the year on year high spike.
But over time the deaths will stop being in nursing homes or over 70s and become fewer and under 68.

I agree we couldnt have been sweden whether that is vit d or density or general heath/state of our healthcare making us unfit/nhs loding more. If we have the same % as spain and italy and similar deaths we are similarly unhealthy.
I want to see the deaths in 40-60yo who are normal weight. Because they said rhe levels of obesity in the icu are similsr to uk population but comparing countries with less deaths who are slimmer.

It really is a nightmare not least as so many countries have an uneven spread so when holiday season comes the locals will have to hide. But if london flares up will they just lock that down? Because that will make immunity difference larger.
I expect we will have 4-6w of release then not a lockdown but a dont go to work etc. With 160k infected it wont take long to spike up as we are not trying to prevent it by trimming off the infected.
Tbh i would be annoyed if i were a hcp theyve just got over a peak and now racing to another one. There was a fb video of a dr saying exactly the same and that people shouldnt be rushing off back to work/school. It is certain they think the school holidays will be a natural break like with swine flu.
Tbh i can see why people are suicidal its not the lockdown its the stress of it. And thought of all these waves. Could do with a dignitas over here.
Im hoping they get testing thge hcp as we can only hope they have much higher immunity making them and patients in for other things slightly safer.
If the gov admitted the lack of testing for antibodies was about money they should say and refer us to the best independent one there id for people to pay. All these 70yo in fear they ciuld have had it. Still need to be cautious in case you can get another strain etc but at least the mental strain would largely go.
There was an outbreak in a daycare near montreal where i think 17 kids got it and a few staff. Im interested in though if the parents caught it off the kids.

But this is why people who keep saying the numbers of dead under 40 is so low 3k or something. Well it's not when you account for herd immunity. The chance of you dying may well be low but the total of dead people is high.

No, read more carefully.

That is the total under 60.

The total under 40 is 227 for England and Wales, and we absolutely would have been completely unequivocally wrong to have locked anything down or changed our behaviour in any way whatsoever if it was only about the deaths under 40.

The risk is EXCEEDINGLY SMALL if you are not at a very minimum middle-aged.

Once you account for the facy teachers are only teaching 2% of stufents then you get 65 educational staffx3 which is higher than the drs and nurses which is also over 100.

Just to clarify doctors and nurses aren't particularly at risk of dying from what I can see, it's just that their deaths are highlighted. If you take the age profile and large population of teachers then you'd expect some dozens of deaths even if they were all jobless since January. It doesn't follow for a moment that they've been exposed since schools closed. It's likely many of the dead died when they were teaching 100% of students, and of the remainder we don't have any evidence that they were infected at work.

Shoots, I understand that. The graph just highlighted that for everyone over 45, Covid was having a fairly similar effect in increasing deaths. You’d expect less of a proportion of deaths at 45 than 90, obviously, but the increase is broadly similar for that range. So perhaps it isn’t inherently much worse for older people than any other disease as they are more likely to die of any illness. It is more noticeable that the young - between 15-45 are seemingly less affected, though that might be partially due to less deaths from accidents and crime. Children under 1 always have a high death rate, sadly, due to congenital disabilities, birth trauma, cot death and domestic violence etc.

No, it's much worse in that if you for example have a 1 in 10,000 risk of death and that doubles to 1 in 5,000 then that's much better than going from 1 in 10 to 1 in 5, for example.

It's not a 15-45 thing particularly, there's a fairly strict exponential increase in growth, so 15 will be a statistical zero, and 40-45 won't be at all.

Proportionately you’d expect a greater risk of death from any illness with age, though. If the risk doubles, from whatever base line, it still doubles.

I only mentioned 15-45 as that’s the range of that set of data on the graph.

Yes it’s a pain that the ONS 2020 data is bunched into 15-44, it’s such a huge band and doesn’t enable us to make generalisations about “schoolchildren”. Prior years data is in 5 year bands.

Ah yes i thimk ive picked up 3k from another thread.
I did mean with the teachers that theyve caught it before lockdown so thats why if they had worked through they would have had 3x the deaths but agree we dont know when they died or whether it was caught elsewhere.
The fact its more ta could be that they are older and sit closer.
I do find it odd that nursi g would be relatively low. Even with masks it should be worst affected job role. But if it is the masks bringing it dwn then shows where uk need to focus as all desths in excess are avoidable.
I think looking at nhs statistics the 40-59 deaths are 1905 so still working age and also parents and coukd demonstrate my point that once you x thst up to allow for reaching her that is a very big number.

The nhs is only hospital deaths and i think there were several at peak where they died at home probably at least 4 just in the papers.
So i assume there are slightly more than are logged on nhs but wont be realised on the year on data as you may have been saying because so many safely at home not injuring themself.

not sure which data you are looking at, but the ons certainly publishes 2020 in 5 year bands.

The site posted earlier looking at Covid death reporting also has a very interesting article about the GDP6 gene and GDP6 deficiency, the contraindications for treatment with hydroxychloroquine and BAME deaths here.

Gene frequency is between 5% and 25% in tropical Africa, the Middle East, tropical and subtropical Asia, some areas of the Mediterranean, and Papua New Guinea
ref here.

GDP6 deficiency seems to also be related to severity of anaemia in relation to sickle cell disease and trait and can occur comorbidly ref here.

ONS in 20-year bands:

Only 12 dead from COVID in 0-19 group

www.england.nhs.uk/statistics/wp-content/uploads/sites/2/2020/05/COVID-19-total-announced-deaths-13-May-2020.xlsx

Interesting article in The Times today about the different r(t) rates in different regions.

London is apparently 0.4 which some are postulating represents a degree of herd immunity. My only personal bet would be localised herd immunity in those working in the highest risk settings.

I am guessing that, as others go out more, we will see more cases again.

What does everyone think?

apple.news/AHhzGnGPYRdCYWu78_0qXOQ

Yes I'd imagine it's entirely possible to have e more herd effect in London by now Interesting to see if that then filters out to commuter towns too.

The north was affected later than London and south so is obvious the rate will still be a bit higher there

I can't see that link Larry, is there a regional breakdown?

The difference between R0 in London and other areas is quite striking.

It can be partially explained by the fact that London has got relatively few care homes so the generic transmission rate is not inflated by the care home spread (which is pushing R0 up in other regions).

Re: herd immunity - I still think it may be a factor even though preliminary antibody testing seems to indicate that only ~10% of the London population has got antibodies. However, we do not know what percentage of the population has got innate resistance to C19 (i.e. they might not get infected even if they get exposed to mild to moderate viral load). If this number is meaningful, extra 10-15% of acquired immunity might start making a difference.

Also weren't the antibody tests from a 2-3 weeks ago? There can be a lag of 3-4 weeks before antibodies are developed, so you'd expect the prevalence to be higher now.

One london factor could be bcg rates...
I would think too a lot more of london were afraid by initial high numbers. Whereas people overconfident elsewhere.
More londoners in masks?
It would be amazing if more were unable to catch it.
I wonder if they found that with animal trials though?
Even if its 10% it would be more due to time lag on antibodies.
I do think the rich elderly stuck at home is factoring in the decision making (the Q) once the plebs have had it they can all come out.

Of course more younger people probably smoke ?

Larry London has a much lower death rate / million than NYC iirc

and that NYC study found 21% has antibodies

So,15% for London ?
(same as the hotspot in Gangelt, Germany)

The % required for herd immunity varies according to the R0
The estimates I've read from scientists are in the 60% - 85%

However, presumably some of the "low-hanging fruit" have caught it
and hence the available population to infect is no longer 8 million, but say only 6 - 7 million ?

Has anybody seen covid19-phenomics.org/PrototypeOurRiskCoV.html this risk calculator?

More importantly after lockdown is relaxed:

There will be some social distancing, but mostly the types that have most effect,
with blanket lockdowns only in very local hotspots

No country seems to be returning to the old "normal"
with e.g. max public events, millions flying off to cheap holidays

Even if governments relax all rules, many shielded and quite a few vulnerable people will be social distancing
So many of those most at risk will be aware of this and changing their lifestyle

I hope masks become widely accepted - it is a comparatively painless precaution,
which may be key to how places like Singapore, Hong Kong, S Korea have been so successful in squashing COVID

Yes, my calculations are very similar to yours, BigChocFrenzy.

The latest NYC number I have seen is 25% so I am also assuming 15% for London. Non-infected population is 6-7 million so the question is how easy or difficult it is for C19 to spread out outside of "low hanging fruit".

R0 of 0.4 (if it is accurate) seems to indicate that further spread may be more difficult (e.g. some people may need higher viral load, others may be naturally immune). There is a reasonable argument that when UK starts relaxing measures, they can be more slightly more flexible with London (which would sound completely counterintuitive to people not familiar with the data).

What is a concern is what level of immunity - if any - will be present in those who had only a very mild illness

Also, some people have a theory that there is a large immune population who had COVID without any symptoms at all

• are they indeed immune, or just lucky the 1st time ?

We really need mass sampling around the country with antibody testing, to get some idea of % infected so far

• and also studies on the effectiveness of different antibody levels.

Guardian has just mentioned this article:

onlinelibrary.wiley.com/doi/abs/10.1111/ijcp.13528

It is a peer reviewed paper (which is extremely rare for C19 related publications) that suggests up to 29% of the UK population might have had COVID.

The number is higher than what gets picked up through antibody testing since some people appear to fight C19 with T-cells without developing antibodies.

Of course, it does not answer the "first time lucky" question but it is still an interesting data point.

Infection passed on twice as often in the northeast

www.thetimes.co.uk/article/coronavirus-infection-passed-on-twice-as-often-in-the-northeast-j7g67bw6t

Does that work, I think I used a ‘share token’?

Bigchoc,

I am not sure whether deaths vs immunity (nyc vs London) is such a linear relationship due to differential demographics, obesity and deprivation rates.

What I do think is those locked down have very low rates of immunity but also very little exposure. However those working may have a high level of immunity.

If the r(t) is half the rest of the country, I doubt it is all explained by Londoners being more compliant.