Chat about Novel Coronavirus - 2019-nCoV - COVID-19

I'll mainly be drinking gin.

They can also detect parkinsons or so i think ive read.

I just looked in to it quickly seems very early stages. Interesting though.

https://clinicaltrials.gov/ct2/show/NCT04324606#studydesign

https://amp.ft.com/content/67e6a4ee-3d05-43bc-ba03-e239799fa6ab#click=https://t.co/YPPhk6f4ms

I'll mainly be drinking gin.

1. villa-ru
   

https://www.bbc.co.uk/news/world-52373888

Coronavirus: World risks 'biblical' famines due to pandemic - UN

And now this too, war torn countries are going to be hit again :(

Yeah, it happens, if you are very unlucky, but all evidence points towards this being a pretty rare occurrence especially for cycle-fit mid-30s.

Ok, but going by Italy and Spain UK could have 10-20k dead under 50's plus a huge question mark over the indiscriminate percentage of people left with potentially permanent lung damage.

Any government embracing that outcome is led by a shitheel.

going by Italy and Spain UK could have 10-20k dead under 50's

What data from Italy and Spain are you looking at to reach that conclusion?

Italy c. 1% of deaths are 50s or less, goes up to c. 2.5% for Spain.

This may be in part due to Spain conducting 500000 fewer tests compared to Italy, while Spain shows more confirmed cases?

https://www.statista.com/statistics/1106372/coronavirus-death-rate-by-age-group-italy/

Maybe. Or their health system was overwhelmed at more died.
Edit, actually, maybe not. They must have been tested to be in Covid death stats and you'd think that the ratio of age would remain true, unless old people just not being tested/recorded as Covid mortality, which is likewise possible.

Would be great if everyone didn't take everything here as a submission to a peer reviewed journal, it's a discussion thread and we're all relying on the stats and data that's out there, which as we all know is dependent on testing policies etc. Having to repeat the same shit all the time is very wearing.

FT proposing that UK deaths actually double that reported for instance, so we're potentially massively underreporting just as China did. If you want to wait for reliable, watertight data before discussing then it's going to be a long long time...

1 Attachment

That's the snag isn't it?

We don't have great data, but decisions will have to be made sooner rather than later.

Hopefully SK is a good model with decent data.

This is 4 or 5 days old so probably has been posted before, but in case it hasn't:

"The inside story of the Ventilator Challenge - the muddled thinking, the wasted time the political egoism." - Twitter thread by FT public policy editor Peter Foster.

And the corresponding FT article "Muddled thinking punctures plan for British ventilator".

I think the best quotes from the twitter thread are:

The government insists that this was all done with the advice of both top doctors AND the regulators. But is that true? When I showed that spec to @AlisonPittard last week, she was amazed. Yes docs were prepared to have less-than Rolls-Royce features, they could live with a Ford, not a Ferrari. But this is a spec, basically, for a Trabant.

And:

The 'cut-the-crap' 'how-hard-can-it-be?' attitudes that leads to headless decision making. It's embarrassing. [...] Expert people TEARING their hair out at the willful numbskullery of the people at the top.

Basically, this entire 'ventilator' effort was, to resurrect that handy old phrase, an omnishambles.

FT link:
Yes. There's problems with death data.
Simplistically:
when you register a death, there are four sections on the death certificate.
1 (a)
1 (b)
1 (c)
2

1 is the cause, b and c are contributory.
2 is underlying.
There's free text. The data is cleaned and processed by ONS.

The best thing to look at is the ONS file, it will be behind. There will be bumps. The excess deaths is probably the right thing to look at, to get an idea of what is happening. BUT. It depends on how the death certificate is being interpreted in the data.
Are the FT counting the 1(a) COVID-19, are they including those people with metatstatic cancer listed in 1 (b) or 1 (c) or even II?

Rather than throwing numbers around and making calls about "the numbers are up/down/spiking/flattening" and then trying to compare to different countries, what should happen is:
these are the UK deaths
these are the UK deaths we think (using our definitions) directly attributable to COVID (ie. no underlying condition really)
these are the deaths of people who would have died, but died sooner because of COVID

I would like to know what the strategy around testing is going to be, or what other countries are doing. Rather than everyone just shouting out "but italy, but spain, but germany does xyz tests why aren't we?"

I think, not sure, but this thread is to try and get a grip on what's going on around us. If you're going to throw things around, it's probably a good idea to put links to everything/something. Vague "i read" is as bad as "HIV MADE IN THE LAB TRUMP WILL WIN" to many people.

David Spieglhalter - https://twitter.com/d_spiegel
is always ALWAYS worth a read.

Yeah, there and Germany likely to give the best impression of the real infection and mortality rates. Until then I think it would be callous of the Government not to err on the side of caution.

I watched the interview with Sweden's advisor posted a couple of days ago and what he says regarding 'total number of deaths will be the same over time, so maybe some people here die sooner but our figures will equal other countries eventually' makes sense to a point but it's also a bit of a gamble, and he admitted that their strategy to protect vulnerable people but let it pass through the younger population has failed straight off the bat (supposedly because of immigrants working in care homes). So yeah, all just sounded a bit cynical.

There was a tweet yesterday with a table containing the dates of death and the dates that they were reported. (This for deaths in UK hospitals). It's the basis for the updated charts that are being produced with backdated deaths. Interesting to see just how long some deaths take to come through the system, possibly due to awaiting test results (possibly even post mortem testing) or just general bureaucracy / manual processing at the ONS.

Aha, found it: https://twitter.com/RP131/status/1252586603580731393?s=20

I think you / we/ I have 5 days from death to register a death. And then it'll go the registrar's office. And you might die in hospital in Southwark, but actually be from somewhere else. And then, so who gets the death? etc and so on.

Could we not simply look at the ave deaths over the last 10 years, extrapolate that imaginary line through now, and everything above it can be considered COVID19?

Very basic i know as we probably had an ave drop on non COVID as we are all forced to stay at home and not be stupid or exposed, but you get what i mean?

Or will that only be available when we have more information, as is the caviat always.

If you want maths then look here: https://twitter.com/cheianov/status/1252657345831882760

Could we not simply look at the ave deaths over the last 10 years, extrapolate that imaginary line through now, and everything above it can be considered COVID19?

Kind of. The problem is someone who has a stroke and doesn't get to hospital in time due to infection fear is an excess death caused by the crisis but isn't helpful to include when you are trying to work out if you can relax restrictions.

Cheers, that's actually the first time I've seen someone explain a little bit how some of the modelling works.