Donate to Remove ads

Got a credit card? use our Credit Card & Finance Calculators

Thanks to johnstevens77,Bhoddhisatva,scotia,Anonymous,Cornytiv34, for Donating to support the site

Coronavirus - Modelling Aspects Only

The home for all non-political Coronavirus (Covid-19) discussions on The Lemon Fool
Forum rules
This is the home for all non-political Coronavirus (Covid-19) discussions on The Lemon Fool
dealtn
Lemon Half
Posts: 6072
Joined: November 21st, 2016, 4:26 pm
Has thanked: 441 times
Been thanked: 2324 times

Re: Coronavirus - Modelling Aspects Only

#312383

Postby dealtn » May 26th, 2020, 4:00 pm

I had heard, but haven't investigated myself yet, that we are now running at "negative excess deaths". Presumably with fewer Covid deaths, and associated even fewer non-reported but likely attributable to Covid deaths, but with the country not yet back to normal, there are few additional deaths, yet offsets from fewer road accidents, work accidents etc. aswell.

XFool
The full Lemon
Posts: 12636
Joined: November 8th, 2016, 7:21 pm
Been thanked: 2608 times

Re: Coronavirus - Modelling Aspects Only

#312390

Postby XFool » May 26th, 2020, 4:17 pm

Mike4 wrote:
dealtn wrote:
Mike4 wrote:
No.

For a start they are only looking at hospitalised cases, by which time it's too late to be administering HXQ. It needs giving early when symptoms first appear, not as a last gasp attempt to save a dying patient. Secondly, there is no mention of zinc being given at the same time, without which we know it's effectiveness is compromised.

So why do twice as many die as the control? I thought you asked whether there were fatalities, not whether it worked if given earlier and/or with zinc.

I only tried to answer the question you asked.

I asked if anyone could cite a fatality attributed to HXQ. You haven't cited one, have you? Just put forward some circumstantial evidence suggesting it happens, no actual case(s) cited. You haven't even found one attributed to HXQ being taken for malaria, and there have been billions of doses taken!

The Lancet study is a collection of observational studies as I understand it, not a 'gold standard' placebo controlled double blind trial. Observational studies are helpful but have massive limitstions. Selection of the cohort to study is one major failing. How were the subjects to study selected by the Lancet? For a start, only people ill enough to arrive in hospital, probably. HXQ is most effective when used as a prophylactic or when first infected. There is an Indian study of 10,000 police who were offered HXQ as a prophylactic. Approx half elected to take it, and half declined. Nine deaths from C-19 recorded so far in the cohort declining it, and none in the cohort taking it. why did the Lancet study choose not to look at that I wonder.

We are still waiting for a placebo-controlled double-blind trial to nail this either way, once and for all. Until we get that, all these anecdotal studies are worth Jack.

WHO halts hydroxychloroquine trial for coronavirus amid safety fears

The Guardian

Malaria drug taken by Trump could raise risk of death and heart problems, study shows

"The World Health Organization has said it will temporarily drop hydroxychloroquine — the malaria drug Donald Trump said he is taking as a precaution — from its global study into experimental coronavirus treatments after safety concerns."

That Lancet paper is cited as the spur for the WHO action.

GrahamPlatt
Lemon Quarter
Posts: 2059
Joined: November 4th, 2016, 9:40 am
Has thanked: 1032 times
Been thanked: 824 times

Re: Coronavirus - Modelling Aspects Only

#312474

Postby GrahamPlatt » May 26th, 2020, 8:24 pm

https://www.npr.org/sections/coronaviru ... 0520650826

The WHO's chief scientist, Dr. Soumya Swaminathan, says the review was prompted by the article in The Lancet, which was not a randomized control trial but still large. .... She says the WHO hasn't yet seen data that showed a problem with hydroxychloroquine in its own study, but The Lancet article raised questions among many of the investigators involved in the WHO's trial.

Jury’s still out.

Now POTUS is not on my Christmas list, but I do object to the press failing to differentiate between treatment & prophylaxis.

And yes, this should be on t’other thread. Sorry.

PS there’s another study here https://www.nih.gov/news-events/news-re ... t-covid-19 where they’re enrolling what they anticipate to be a different demography to the WHO - over 60s with co-morbidities. Now that’ll be interesting.

servodude
Lemon Half
Posts: 8271
Joined: November 8th, 2016, 5:56 am
Has thanked: 4435 times
Been thanked: 3564 times

Re: Coronavirus - Modelling Aspects Only

#312515

Postby servodude » May 27th, 2020, 12:44 am

dealtn wrote:I had heard, but haven't investigated myself yet, that we are now running at "negative excess deaths". Presumably with fewer Covid deaths, and associated even fewer non-reported but likely attributable to Covid deaths, but with the country not yet back to normal, there are few additional deaths, yet offsets from fewer road accidents, work accidents etc. aswell.


I've been trying to find these figures myself also; I'm interested in comparing across nations
This is about the closest I've found: https://ourworldindata.org/excess-mortality-covid

There was an effort from some quarters when the excess deaths from the UK included swathes outwith the COVID attributed figures to claim that they were caused by the "lockdown" (it's not really a good term for what it is but I'll use it)
- some from the same quarters were simultaneously claiming that reports of COVID deaths were being inflated in the UK

I think that now we're a bit further from that period it's probably worth a revist/review?

As far as I'm aware Denmark has had no excess deaths during the lockdown period (https://www.thelocal.dk/20200417/death- ... statistics)
- this implies that the COVID deaths replaced lives saved by lockdown
- we now appear to be in a situation where the UK looks similar - and I think for the same reasons (people avoiding death because of the lockdown)

So why did the UK figures look so different back then?
- under attribution to COVID seems like the most believable reason
- otherwise what? lack of Hygge? orders more domestic violence in the UK? a population on the edge of life so much that staying in doors is fatal?

I think we have to assume that for the purposes of working out what's happening in the population regards this virus that all the excess deaths be attributed to COVID (any that weren't being compensated for by those saved in lockdown)
- which I do think most suggestions in this thread already did

and that if you meet people whose first assumption on seeing the reports was GOTO StopLockdown that you should consider trying to refactor their logic processes to avoid GOTO ;)

- sd
(Nothing inherently wrong with GOTO if it's scoped properly - but it probably deserves it's bad rap)

Alaric
Lemon Half
Posts: 6035
Joined: November 5th, 2016, 9:05 am
Has thanked: 20 times
Been thanked: 1400 times

Re: Coronavirus - Modelling Aspects Only

#312517

Postby Alaric » May 27th, 2020, 12:59 am

servodude wrote:I think we have to assume that for the purposes of working out what's happening in the population regards this virus that all the excess deaths be attributed to COVID (any that weren't being compensated for by those saved in lockdown)
-


The statistic least open to manipulation is the week by week count of people dying. The graph of this is periodically published. Up to almost the end of March it's not that different to an average of the last five years. In April it was around twice as many as normal. Since then it's dropped back a bit, with the observation it's currently (in May) running at about the same level as a "bad" January, one where there's something of a flu outbreak.

The count of excess deaths against recent averages is broadly around the same as the count of those certified as COVID-19 related.

Those who work in this field, the jargon being "Impaired Lives Annuities", observe that the life expectancy even for those potentially terminally ill is still measured in a handful of years rather than months.

Mike4
Lemon Half
Posts: 7086
Joined: November 24th, 2016, 3:29 am
Has thanked: 1637 times
Been thanked: 3794 times

Re: Coronavirus - Modelling Aspects Only

#312530

Postby Mike4 » May 27th, 2020, 6:41 am

I'm surprised to see Malaysia missing from the list. A country of some 30m people and only 4 deaths per million.

They appear to be doing something right.

Edit to add: Oops I wss replying to the list posted by Zico, in post 312366.

Mike4
Lemon Half
Posts: 7086
Joined: November 24th, 2016, 3:29 am
Has thanked: 1637 times
Been thanked: 3794 times

Re: Coronavirus - Modelling Aspects Only

#312531

Postby Mike4 » May 27th, 2020, 6:57 am

XFool wrote:WHO halts hydroxychloroquine trial for coronavirus amid safety fears

The Guardian

Malaria drug taken by Trump could raise risk of death and heart problems, study shows

"The World Health Organization has said it will temporarily drop hydroxychloroquine — the malaria drug Donald Trump said he is taking as a precaution — from its global study into experimental coronavirus treatments after safety concerns."

That Lancet paper is cited as the spur for the WHO action.


Dr Didier Raoult questions the validity of the data used for the Lancet meta-study which the W.H.O are relying on in taking this action. He points out in a tweet that the homogeneity of the data across five continents is not credible, suggesting the data is corrupted or even frigged.

https://twitter.com/raoult_didier/statu ... 7866764288

johnhemming
Lemon Quarter
Posts: 3858
Joined: November 8th, 2016, 7:13 pm
Has thanked: 9 times
Been thanked: 609 times

Re: Coronavirus - Modelling Aspects Only

#312533

Postby johnhemming » May 27th, 2020, 7:05 am

Mike4 wrote:suggesting the data is corrupted or even frigged.

I wonder how one would identify the latter.

servodude
Lemon Half
Posts: 8271
Joined: November 8th, 2016, 5:56 am
Has thanked: 4435 times
Been thanked: 3564 times

Re: Coronavirus - Modelling Aspects Only

#312568

Postby servodude » May 27th, 2020, 9:34 am

johnhemming wrote:
Mike4 wrote:suggesting the data is corrupted or even frigged.

I wonder how one would identify the latter.


For some stuff it means it's out by more than a couple of bawhairs!

I remember as a young servodude a couple of decades ago a lovely girl we worked with had accidentally dropped "frigged" in to an external report as a failure category (I think it was a Sony project we'd been contracted to do); not quite realising it was an "internal" classification rather than a technical term.

-sd

Mike4
Lemon Half
Posts: 7086
Joined: November 24th, 2016, 3:29 am
Has thanked: 1637 times
Been thanked: 3794 times

Re: Coronavirus - Modelling Aspects Only

#312571

Postby Mike4 » May 27th, 2020, 9:53 am

johnhemming wrote:
Mike4 wrote:suggesting the data is corrupted or even frigged.

I wonder how one would identify the latter.


Look for the numbers written over tippex, I guess....

dspp
Lemon Half
Posts: 5884
Joined: November 4th, 2016, 10:53 am
Has thanked: 5825 times
Been thanked: 2127 times

Re: Coronavirus - Modelling Aspects Only

#312577

Postby dspp » May 27th, 2020, 10:17 am

servodude wrote:
johnhemming wrote:
Mike4 wrote:suggesting the data is corrupted or even frigged.

I wonder how one would identify the latter.


For some stuff it means it's out by more than a couple of bawhairs!

I remember as a young servodude a couple of decades ago a lovely girl we worked with had accidentally dropped "frigged" in to an external report as a failure category (I think it was a Sony project we'd been contracted to do); not quite realising it was an "internal" classification rather than a technical term.

-sd


Frigged is a technical term in every industry I've worked in :)

As many of us have noted, there seems to be a lot of transmission going on undetected, that is really messing up the modelling. Only now that sufficiently reliable & widespread testing is taking place is that hypothesis becoming firmer.

regards, dspp

Mike4
Lemon Half
Posts: 7086
Joined: November 24th, 2016, 3:29 am
Has thanked: 1637 times
Been thanked: 3794 times

Re: Coronavirus - Modelling Aspects Only

#312583

Postby Mike4 » May 27th, 2020, 10:40 am

dspp wrote:Frigged is a technical term in every industry I've worked in :)


Have some here really not not encountered the term do you think? Like you, I've been using it since I was at skool. Probably learned it from my dad.

Frigging is to fiddle and adjust things to suit your needs.

So to frig the data means to deliberately change it or leave important bits out to support your argument. Dr Didier is suggesting The Lancet set out with the preconceived notion that HXQ doesn't work, and when the data didn't support this notion, someone frigged the data until it did. What other explanations might exist for the percentage of patients with hypertension to be near identical across all five continents. Same for smoking when the percentage of smokers in the general populations varies widely.

He is suggesting the data used by Lancet 'smells' wrong. Here is the data chart he is questioning, from his Twitter post. Not that I understand most of it.

Image

dealtn
Lemon Half
Posts: 6072
Joined: November 21st, 2016, 4:26 pm
Has thanked: 441 times
Been thanked: 2324 times

Re: Coronavirus - Modelling Aspects Only

#312592

Postby dealtn » May 27th, 2020, 11:02 am

Mike4 wrote:He is suggesting the data used by Lancet 'smells' wrong. Here is the data chart he is questioning, from his Twitter post.


Well if they only presented the data for USA and Europe that would still be pretty damning. A sample size of nearly 80k, representing 83% of the total.

I guess a proper large size double blind is the way to go.

johnhemming
Lemon Quarter
Posts: 3858
Joined: November 8th, 2016, 7:13 pm
Has thanked: 9 times
Been thanked: 609 times

Re: Coronavirus - Modelling Aspects Only

#312901

Postby johnhemming » May 28th, 2020, 8:49 am

The CDC's current best estimate of a symptomatic fatality rate is 0.4%
https://www.cdc.gov/coronavirus/2019-nc ... arios.html

Scenario 5: Current Best Estimate


They estimate that 35% are infectious, but not symptomatic.

That would imply an IFR of 0.26%.

With a higher proportion that are asymptomatic one would expect a lower IFR.

This looks at the proportion of infected people that are asymptomatic.
https://www.cebm.net/covid-19/covid-19- ... mptomatic/

zico
Lemon Quarter
Posts: 2139
Joined: November 4th, 2016, 12:12 pm
Has thanked: 1074 times
Been thanked: 1086 times

Re: Coronavirus - Modelling Aspects Only

#312963

Postby zico » May 28th, 2020, 10:42 am

FT has published a country comparison based on excess deaths - generally accepted as a more accurate measure than the Covid deaths table I posted the other day. UK comes out significantly worse than any other country by this measure (approx 900 deaths/million, Italy next worst approx 800 deaths/million)

The UK has suffered the highest rate of deaths from the coronavirus pandemic among countries that produce comparable data, according to excess mortality figures.

The FT has made these comparisons for the first time because the level of excess deaths in other hard-hit European countries, such as Italy and Spain, has returned close to normal levels. This means death rates in those countries are unlikely to overtake the UK unless they suffer a second wave of infections.


Two main reasons for this, which are probably inter-related.
First, UK was slow to lockdown (as measured by number of infections before lockdown).
Secondly, UK was affected in all regions where Italy and France had hotspots, but also regions unaffected by Covid-19. There's an interesting chart in the FT article showing just how many regions in Italy and France were virtually untouched.

UK government is suggesting results need to adjusted for age of population, but FT article says
The FT analysis shows that the UK's excess deaths figure remains the highest whether younger people are excluded or the analysis is limited to pensioners."


Natalie Dean, assistant professor of biostatistics at the University of Florida, said some countries such as Italy had “bad luck” as they were caught by the virus early on. This gave other countries, such as the UK, time to learn lessons, she added.

“I was very surprised by the delayed response in the UK. Given what we were observing in Italy at the time and that the UK was on the exact same trajectory, had the same very steep rise, I was surprised to see discussion about waiting. There was an immediate need to stop what was happening,” she said.

“For London, in particular, it was clear that there was a steep rise, so it’s reasonable to think earlier intervention would have saved lives.”


Link to article is below (no paywall).

https://www.ft.com/content/6b4c784e-c25 ... 8ffde71bf0

spasmodicus
Lemon Slice
Posts: 259
Joined: November 6th, 2016, 9:35 am
Has thanked: 65 times
Been thanked: 117 times

Re: Coronavirus - Modelling Aspects Only

#312974

Postby spasmodicus » May 28th, 2020, 11:04 am

johnhemming wrote:The CDC's current best estimate of a symptomatic fatality rate is 0.4%
https://www.cdc.gov/coronavirus/2019-nc ... arios.html

Scenario 5: Current Best Estimate


They estimate that 35% are infectious, but not symptomatic.

That would imply an IFR of 0.26%.

With a higher proportion that are asymptomatic one would expect a lower IFR.

This looks at the proportion of infected people that are asymptomatic.
https://www.cebm.net/covid-19/covid-19- ... mptomatic/


yes, many have long suspected that the IFR is below 0.4%

regarding "They estimate that 35% are infectious, but not symptomatic", I wonder how this impacts the effectiveness of the UK govts "track and trace" scheme. This relies on infected people reporting themselves i.e. "But from now on, everyone with symptoms should ask for a test online or phone to arrange a test by calling 119". But if 35% fail to report, because they don't have symptoms and an unknown proportion prefer not to for privacy or other reasons, or even do not realise that they have to report, how effective is this system likely to be?

It's hard to make sense of all the statistics being bandied around at present. The "zero cases in London" report from the mayor a few days ago also seems strange to me. I have been monitoring cases on the BBC website for my area, Kingston (KT2 postcode) which is within Greater London

https://www.bbc.co.uk/news/uk-51768274

Total confirmed cases in KT2 (population 177,000, or about 2% of Greater London) over the last 8 days increased from 509 on May 20th to 520 reported today, i.e. between 1 and 2 new cases per day.

S
(somewhat mystified)

johnhemming
Lemon Quarter
Posts: 3858
Joined: November 8th, 2016, 7:13 pm
Has thanked: 9 times
Been thanked: 609 times

Re: Coronavirus - Modelling Aspects Only

#312992

Postby johnhemming » May 28th, 2020, 11:47 am

A lot of what the government are doing is not really thought through. One benefit of Cumming's eye test is that it will be harder for them to keep the lockdown going which in my view is a good thing as we need to bring it to an end (as we have done for some time).

Mike4
Lemon Half
Posts: 7086
Joined: November 24th, 2016, 3:29 am
Has thanked: 1637 times
Been thanked: 3794 times

Re: Coronavirus - Modelling Aspects Only

#312997

Postby Mike4 » May 28th, 2020, 11:57 am

johnhemming wrote:A lot of what the government are doing is not really thought through. One benefit of Cumming's eye test is that it will be harder for them to keep the lockdown going which in my view is a good thing as we need to bring it to an end (as we have done for some time).



One sometimes wonders if the whole thing was a "cunning plan", to break the lockdown without officially lifting it.

The point being, if it causes a second wave of infection the guvverment can say nothing to do with us, we dun our best, the people brought it on themselves by getting all bolshie.

johnhemming
Lemon Quarter
Posts: 3858
Joined: November 8th, 2016, 7:13 pm
Has thanked: 9 times
Been thanked: 609 times

Re: Coronavirus - Modelling Aspects Only

#313007

Postby johnhemming » May 28th, 2020, 12:21 pm

In a battle between cockup and conspiracy in government cockup normally wins hands down.#

In any event we will need cooler weather or air conditioning for infections to grow until autumn.

ReformedCharacter
Lemon Quarter
Posts: 3120
Joined: November 4th, 2016, 11:12 am
Has thanked: 3591 times
Been thanked: 1509 times

Re: Coronavirus - Modelling Aspects Only

#313016

Postby ReformedCharacter » May 28th, 2020, 12:39 pm

johnhemming wrote:
In any event we will need cooler weather or air conditioning for infections to grow until autumn.

Is that right? I thought that (contrary to what I would have expected) infections spread even in quite dry and hot conditions. Happy to be corrected if that isn't the case.

RC


Return to “Coronavirus Discussions”

Who is online

Users browsing this forum: No registered users and 5 guests