Daily numbers, graphs, analysis thread 6

[Barracker]

Welcome to thread 6 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.

@ShootsFruitAndLeaves Thank you for your graphs (I've been behind with this thread) - There does look to be a clear increase in the numbers, spreading out from London.

In terms of numbers, I will also be interested in the months and years ahead to see how many older people will die prematurely due to frailty exacerbated by the lockdown. I suspect we may never really know.

Sometimes scientists have to use the most reliable calculation method there is, but accept it is not perfect
In my field of Maths / Physics modelling, this was my normal.

I don't know of any better method of calculation than used in this paper, other than to use averages for each comorbidity,
because the condition of each patient would not be given in more detail in the anonymised medical records they had available
It would also not include intangibles like
e.g. the will to live of a patient, whether their doctors were the best in the hospital or the worst

All countries seem to show a 2-3 x higher risk for males of all ages
provided we normalise wrt the greater number of females in the higher age bands

It seems that the people in the study were being compared against the most appropriate like for like group. To have broken them down into individual long term conditions, and controlled for inter-dependencies between these conditions was quite rigorous. To have a study that not only looked at the relations between conditions, but also between varying degrees of those conditions seems almost impossible in the time frame the study was put together over. Although the authors point out that caveat, it strikes me as more a suggestion of a manner of improvement, rather than as an admission of shortcoming. They do not even discuss whether such information is available anywhere. It would certainly be a difficult feat to draw out severity from medical records. How would this be recorded? To my knowledge, doctors do not score conditions. They might prescribe different dosages of drugs, but this would also be dependent on other factors. There may be clues, but it would be a research task in and of itself to draw out case severity from medical notes.

So, this is the best information we currently have, and it leads to a conclusion of a longer amount of years of life lost than previously suggested. It seems to me that if this study is to be dismissed, then the burden of evidence lies on those dismissing it. Without this addition detailed information, surely the strongest position that someone disagreeing with this study can take is that average life expectancies are of little use when dealing with such a biased sample and that therefore years lost to life is actually unknowable? And if years lost to life is actually unknowable, then you cannot then simultaneously take the position that these people were all liable to die very soon anyway.

uk
So if its 0.7% fatal
18k deaths is that 2.6m or so immune?
4% of population?

Assuming that people once reovered cannot catch it again for a year or so

That would probably have been better on the same chart....

Ah thanks, shoots
Overlaying makes the comparison very clear

Shoots have you updated your day of death graphs for today? I was wondering what the peak/plateau was looking like.

Yes I posted a few pages back

Thank you. This thread has become so fast moving I missed it.

There is a continuing narrative that the dead would have died anyway within a few months

If that was genuinely the case, it would mean the death rate would fall to compensate for those who died a few months prematurely... I very much doubt it will turn out like that!

I've been following these threads for a while now.

I understand that your primary focus is data and graphs but I was intrigued to read about this man's recovery from Covid-19. In addition to double pneumonia in his lungs, his bowel went into failure www.rte.ie/news/coronavirus/2020/0423/1134406-miraculous-recovery/

John Burn-Murdoch@jburnmurdoch (FT)

Why do I say UK daily deaths may have peaked?

Here’s week-on-week change in daily deaths.

This gets rid of weekly reporting patterns and asks, are more people dying than at same point last week?

In UK, blue bars mean we’re now seeing fewer deaths than same day last week.

Hospitals in almost every region now have fewer covid patients than same time last week

• Suggests UK is at or near peak for new confirmed infections,

though UK testing still lagging,
and care homes of course absent from this view

Also deaths, cumulative deaths,

daily deaths in regional hotspots internationally
London & NYC have peaked, some US states still increasing

A wee point re figures and negative tests, sorry if already pointed out or obvious.

It was pointed out to me that nhs workers who've had it but didn't get a test are being tested to make sure they're clear of it before returning to work. The person who mentioned it had their test over a month after initial symptoms and was neg, though classic symptoms. I was also told of many cases where tests were going to be taken and then people were told they weren't going to bother etc.

Until a really reliable antibody test is done and everyone is tested, especially those who think they've had it, true fugues are out of our reach. But I think everyone knows that.

Thanks for those graphs BigChoc. I find those the easiest to understand.

There's no way trump can wiggle out of those! He's certainly the "most best" at Coronavirus figures!

Though we still don't trust China?

Interesting article from the NY Times on the antibody tests. A random sample of 3,000 people was tested at grocery stores and 21.7% of those tested in NYC had antibodies to the virus. The test was calibrated to err towards producing false negatives, so there is confidence that this is a true reflection, if not an underestimate of the true rate.

This suggests a death rate of around 0.5% of those infected which is consistent with the Pasteur Institute and that NYC was on the way to herd immunity.

www.nytimes.com/2020/04/23/nyregion/coronavirus-antibodies-test-ny.html#commentsContainer

Interesting. i'd like to see some more results from UK population sampling- antibody (if we can get a reliable test for larger numbers) or antigen. Maybe that would be one way of hitting the 100k target next week!

Germany has established a reliable antibody test and has started a largescale program of testing samples from bloodbanks and also from around the country

The later groups will be tested repeatedly over time to check how immunity is spreading

Interim report in May

COVID-19 case cluster study (Gangelt Municipality)
Lead - Prof. Hendrik Streeck (Inst. Virology, Uni Bonn)

This area is an early epicentre with the highest % cases in Germany, spread after a carnival

The study used questionnaires, took throat swabs and tested blood for the presence of antibodies (IgG, IgA).

Work is ongoing, but the recent interim report found that 15% of the population are immune

Infection rate would be much lower over the country as a whole

https://www.land.nrw/sites/default/files/asset/document/zwischenergebniscovid199casestudyygangelt_0.pdf

Interesting interview in Nature with the epidemiologist behind Sweden’s approach

www.nature.com/articles/d41586-020-01098-x?utm_source=twitter&utm_medium=social&utm_content=organic&utm_campaign=NGMT_USG_JC01_GL_Nature

15-20% immunity would decrease the amount of distancing measures required to bring R0 below 1. And, generally, these higher immunity rates will be found in centres of high population density (London, for instance).

Thank you for this thread, so interesting and informative.

Neurotrashwarrior I am hearing a lot about the likely high rate of false negative test results as well. One example, friend’s 70+ MIL was in ICU on a ventilator for over a week (now recovered thankfully). Doctors were ‘in no doubt whatsoever’ that she had cv but 5 (!) tests proved negative. The medics told family that the tests are hopeless, they reckoned picking up 30% of cases. Also the theory behind ‘reinfection’ seems to be faulty test results. It does make you wonder about genuine infection figures.

On another point did anyone see the article on the use of sleep apnea machines to treat patients - looks v promising. news.sky.com/video/coronavirus-patients-astonishing-progress-with-black-box-treatment-11977910

The UK govt initially tried this approach of asking people to take similar social distancing measures
The public didn't do so, hence with rocketing infection levels. lockdown was inevitable.

In Sweden, there is much greater trust in the authorities, in experts too and much greater compliance with advice

Plus the much lower infection and death rate in the entire cluster of Scandi countries compared to W&S Europe needs to be investigated in depth,
to see if there are feasible lifestyle /supplementation / diet changes that can be copied.

Also copying measures such as masks and phone tracking from Asian countries with low death rates
(I exclude China due to great unreliability of data)

I'm not sure how a greater sense of civic responsibility & trust can be copied though, which seems common to both those groups of Scandi and Asian countries