Difficult Covid Preprint to debunk

Get your science fix here: research, quackery, activism and all the rest
Post Reply
User avatar
jimbob
Light of Blast
Posts: 5276
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Difficult Covid Preprint to debunk

Post by jimbob » Tue Sep 22, 2020 8:14 pm

Due to all the diagrams and tables being placeholders:

Press release:

https://www.lboro.ac.uk/news-events/new ... id-deaths/
Research from Loughborough and Sheffield Universities suggests new approach to recording COVID deaths 3 July 2020
A new study by academics at the Universities of Loughborough and Sheffield, and economic consultants at Economic Insight, seeks to provide a more statistically robust approach to the question: “How many deaths in England and Wales are due to COVID-19?”.

Currently COVID-associated deaths or excess deaths are used to track the impact of the virus. However these figures may be distorted as to record a COVID-associated death you require only weak evidence that COVID ‘may’ have contributed to the death, and counting excess deaths assumes that any variation in weekly mortality relative to a five year average represents ‘excess’ deaths due to COVID without taking into account other drivers of mortality.

The new study, co-authored by Karligash Glass, Professor of Financial Economics and a respected economic data scientist within Loughborough University’s School of Business and Economics, Professor Anthony Glass (Professor of Managerial Economics, the University of Sheffield), Sam Williams (founder and Director of Economic Insight) and Alasdair Crookes (Consultant at Economic Insight) proposes an enhanced approach to crunching the COVID numbers.

The authors have used statistical techniques to better identify deaths due to COVID, applying the excess death framework more robustly and controlling for other factors that affect mortality. The key findings are as follows:

Actual deaths due to COVID are some 54% or 63% lower than implied by the standard excess deaths measure, and reported excess deaths likely include a significant number of non-COVID deaths.
While it is well known that COVID deaths are concentrated in the elderly, the study finds them to be particularly acute in the very elderly (75-84 and 85+ years old).
Over the lockdown period as a whole Government policy has increased mortality rather than reduced it.
This final finding may be alarming to some, but the authors point out that the lockdown initiative was always intended to ‘flatten the curve’ (delaying the spread to avoid overburdening the NHS) and not necessarily to lower mortality rates. They also assert that the overall increase in mortality is a result of significant unintended consequences of the lockdown, for example, reduced A and E attendances and reduced cancer and cardiac treatments.

Speaking about the new study, Loughborough’s Professor Karligash Glass said: “In addition to providing a new way of measuring the impact of COVID-19, this study raises questions not simply about the efficacy of the blanket lockdown response, but also whether Government communications to encourage public compliance have inadvertently driven other, more harmful, behaviours.”

Professor Karligash Glass and Professor Anthony Glass are also academic associates of Economic Insight, a leading UK economics consultancy.

The paper, ‘An Improved Measure of Deaths due to COVID-19 in England and Wales’, is due to be a published as a working paper on SSRN: https://papers.ssrn.com/sol3/papers.cfm ... id=3635548
Let's look at the paper:

https://poseidon01.ssrn.com/delivery.ph ... 29&EXT=pdf
Screenshot 2020-09-22 211425.png
Screenshot 2020-09-22 211425.png (360.32 KiB) Viewed 3089 times
Have you considered stupidity as an explanation

User avatar
bob sterman
Dorkwood
Posts: 1123
Joined: Mon Nov 11, 2019 10:25 pm
Location: Location Location

Re: Difficult Covid Preprint to debunk

Post by bob sterman » Tue Sep 22, 2020 8:53 pm

From "Economic Insight" (https://www.economic-insight.com/) the first author Sam Williams...
Sam Williams is a leading economist known for helping clients solve key strategic problems relating to commercial decision-making, regulation and competition policy. He is particularly adept at translating complex analyses into clear, straightforward messages that help clients make informed choices. He is a trusted advisor to major corporates, regulators and government departments at the most senior level – and routinely provides board level briefings. Sam has a track record of demonstrably delivering enhanced performance for his clients, from designing value-adding pricing strategies, through to influencing regulators and competition authorities on business-critical issues.
The second author Alasdair Crookes...
Alasdair has experience in providing expert economics support for regulated companies throughout price control processes (including econometric techniques), assessing the market impacts of competition policies and commercial arrangements, and developing incentive frameworks to effectively fulfil industry targets.
It's wonderful to see people from so many walks of life mucking in to help with the battle against COVID-19.

When the next banking crisis comes around maybe all the clinicians, epidemiologists, virologists etc can reciprocate by writing papers on the appropriateness of bailouts for the wealthy and austerity for the rest of us.

User avatar
Bird on a Fire
Princess POW
Posts: 10137
Joined: Fri Oct 11, 2019 5:05 pm
Location: Portugal

Re: Difficult Covid Preprint to debunk

Post by Bird on a Fire » Tue Sep 22, 2020 11:08 pm

The tables and figures are at the end of the pdf. That's quite common with preprints as journals require manuscripts to be formatted that way - table and figure placement is about the only service they provide authors in exchange for trying to charge $35 to access content they got for free.

Figures start page 28; Tables start page 35
We have the right to a clean, healthy, sustainable environment.

User avatar
sTeamTraen
After Pie
Posts: 2551
Joined: Mon Nov 11, 2019 4:24 pm
Location: Palma de Mallorca, Spain

Re: Difficult Covid Preprint to debunk

Post by sTeamTraen » Wed Sep 23, 2020 12:05 am

They seem to have two main arguments.

One, a lot of the people who died were old/in nursing homes and might have died within a few months anyway. I have some sympathy for this in principle: In August 2003 there was a panicked reaction across France because (allegedly) thousands of people were dying in nursing homes in a heatwave. Every night, whichever doctor was not on vacation and prepared to say the largest number of dead went on national TV. I think the claimed casualty rate peaked at 14,000. It was decreed that this must Never Happen Again, and the government earmarked billions to install air-conditioning in every nursing home --- to be paid by everyone working for free on a public holiday (which one was not defined; the effects of this hasty f.ckup are still felt in France today). Meanwhile Spain, Belgium, Luxembourg, etc reported zero extra deaths and there was no visible run on the funeral parlours. However, subsequent analysis showed that there had been a modest bump in the number of deaths in August (people normally die less in summer), compensated by a small drop by the end of the year. So the heatwave may have finished off a few people who were at death's door. That said, the numbers with COVID seem too huge for this to be a plausible explanation.

Two, the idea that the excess deaths observed were not due to COVID. Given that only two major non-random things happened in 2020 compared to the average of the previous five years --- i.e., COVID and the lockdown --- they are not shy about suggesting that a lot of the excess deaths might be due to the lockdown itself. I find this highly implausible, especially with the big numbers that they are wheeling out. Yes, a lot of people's cancer appointments were put back a while, and this may have impacts over the next few years. But people were not keeling over in huge numbers in A&E of heart attacks, or early-stage cancers that suddenly grew 1000x bigger between March and May.

So in both cases we will find out in due course; the "would have died by Christmas" ones by looking at the overall excess mortality for the year, and the "lockdown killed people" one from comparing the numbers for the main causes of death over time.

In the meantime, I think this paper is mostly speculation (by economists, not epidemiologists), not to say innuendo. Where were all these excess people dying of not-COVID? At home or in hospital? The latter is where they take you when you have a heart attack or a sudden huge cancer. :roll: Why did we not hear about this? Were cardiac surgeons rushed off their feet trying to save people with emergency bypasses but unable to operate because the COVID squad had commandeered all the surgical masks? Were ambulance crews regularly turning up to find cold, lifeless patients, or having them die in the van in the car park? Anecdotes are not evidence, but generally speaking when tens of thousands of people are dying unexpectedly, the anecdotes come along anyway.
Something something hammer something something nail

User avatar
jimbob
Light of Blast
Posts: 5276
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: Difficult Covid Preprint to debunk

Post by jimbob » Wed Sep 23, 2020 6:28 am

Bird on a Fire wrote:
Tue Sep 22, 2020 11:08 pm
The tables and figures are at the end of the pdf. That's quite common with preprints as journals require manuscripts to be formatted that way - table and figure placement is about the only service they provide authors in exchange for trying to charge $35 to access content they got for free.

Figures start page 28; Tables start page 35
Thanks. I was looked at the tabls and decided t wasn't worth looking further quickly. The very few papers I've participated in, we basically did all the formatting and table placement.
Have you considered stupidity as an explanation

User avatar
jimbob
Light of Blast
Posts: 5276
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: Difficult Covid Preprint to debunk

Post by jimbob » Wed Sep 23, 2020 6:30 am

sTeamTraen wrote:
Wed Sep 23, 2020 12:05 am
They seem to have two main arguments.

One, a lot of the people who died were old/in nursing homes and might have died within a few months anyway. I have some sympathy for this in principle: In August 2003 there was a panicked reaction across France because (allegedly) thousands of people were dying in nursing homes in a heatwave. Every night, whichever doctor was not on vacation and prepared to say the largest number of dead went on national TV. I think the claimed casualty rate peaked at 14,000. It was decreed that this must Never Happen Again, and the government earmarked billions to install air-conditioning in every nursing home --- to be paid by everyone working for free on a public holiday (which one was not defined; the effects of this hasty f.ckup are still felt in France today). Meanwhile Spain, Belgium, Luxembourg, etc reported zero extra deaths and there was no visible run on the funeral parlours. However, subsequent analysis showed that there had been a modest bump in the number of deaths in August (people normally die less in summer), compensated by a small drop by the end of the year. So the heatwave may have finished off a few people who were at death's door. That said, the numbers with COVID seem too huge for this to be a plausible explanation.

Two, the idea that the excess deaths observed were not due to COVID. Given that only two major non-random things happened in 2020 compared to the average of the previous five years --- i.e., COVID and the lockdown --- they are not shy about suggesting that a lot of the excess deaths might be due to the lockdown itself. I find this highly implausible, especially with the big numbers that they are wheeling out. Yes, a lot of people's cancer appointments were put back a while, and this may have impacts over the next few years. But people were not keeling over in huge numbers in A&E of heart attacks, or early-stage cancers that suddenly grew 1000x bigger between March and May.

So in both cases we will find out in due course; the "would have died by Christmas" ones by looking at the overall excess mortality for the year, and the "lockdown killed people" one from comparing the numbers for the main causes of death over time.

In the meantime, I think this paper is mostly speculation (by economists, not epidemiologists), not to say innuendo. Where were all these excess people dying of not-COVID? At home or in hospital? The latter is where they take you when you have a heart attack or a sudden huge cancer. :roll: Why did we not hear about this? Were cardiac surgeons rushed off their feet trying to save people with emergency bypasses but unable to operate because the COVID squad had commandeered all the surgical masks? Were ambulance crews regularly turning up to find cold, lifeless patients, or having them die in the van in the car park? Anecdotes are not evidence, but generally speaking when tens of thousands of people are dying unexpectedly, the anecdotes come along anyway.
Accurate phrase there:

" I think this paper is mostly speculation (by economists, not epidemiologists), not to say innuendo"
Have you considered stupidity as an explanation

User avatar
jimbob
Light of Blast
Posts: 5276
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: Difficult Covid Preprint to debunk

Post by jimbob » Wed Sep 23, 2020 7:21 am

Hmm
]A limitation of the current excess deaths metric (a comparison of deaths
to a 5 year average for the same week), is that it attributes the entirety of the variation in mortality
to COVID-19. This likely means that the metric is overstated because there are a range of other
drivers of mortality. We address this by estimating novel empirical Poisson models for all-cause
deaths (in totality; by age category; for males; and females) that account for other drivers including
the lockdown Government policy response
I am not sure that a Poisson model makes sense for a pandemic. Yes a classic example was deaths from horse kicks in the Prussian Army, but that was something considered random and with a constant rate.

I *am* only looking at the abstract, but can't see what one would do to a Poisson model to make it fit.
Have you considered stupidity as an explanation

User avatar
Bird on a Fire
Princess POW
Posts: 10137
Joined: Fri Oct 11, 2019 5:05 pm
Location: Portugal

Re: Difficult Covid Preprint to debunk

Post by Bird on a Fire » Wed Sep 23, 2020 10:41 am

Poisson distribution is the standard way of modelling number of events in a population, and is apparently considered a "basic model" in epidemiology https://pubmed.ncbi.nlm.nih.gov/7797329/

I'm not an epidemiologist, but they're also widely used for population models in ecology. The processes modelled don't have to be random for Poisson to be a good fit.
We have the right to a clean, healthy, sustainable environment.

User avatar
jimbob
Light of Blast
Posts: 5276
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: Difficult Covid Preprint to debunk

Post by jimbob » Wed Sep 23, 2020 11:46 am

Bird on a Fire wrote:
Wed Sep 23, 2020 10:41 am
Poisson distribution is the standard way of modelling number of events in a population, and is apparently considered a "basic model" in epidemiology https://pubmed.ncbi.nlm.nih.gov/7797329/

I'm not an epidemiologist, but they're also widely used for population models in ecology. The processes modelled don't have to be random for Poisson to be a good fit.
Thanks - I can only read the abstract, but it does touch on my basic issue with its use in this situation:
Results and conclusions: We illustrate that the three distributions, although superficially different, often lead to similar results. We argue that epidemiologists should often obtain similar results regardless of which distribution they use. We also point out that application of all three distributions can be inappropriate if assumptions of independence or homogeneity of risks fail to hold. Finally, we briefly review how these basic distributions can be used to justify use of other distributions, such as the Gaussian distribution, for studying disease-exposure associations.
I'd say that uniformity and independence were both invalid assumptions in this case.
Have you considered stupidity as an explanation

User avatar
Bird on a Fire
Princess POW
Posts: 10137
Joined: Fri Oct 11, 2019 5:05 pm
Location: Portugal

Re: Difficult Covid Preprint to debunk

Post by Bird on a Fire » Wed Sep 23, 2020 12:22 pm

Full paper is here: http://citeseerx.ist.psu.edu/viewdoc/do ... 1&type=pdf

Here's the paragraph on Poisson assumptions:
Epidemiologists frequently use the Poisson distributionto analyse data from follow-up studies when the sum-mary data involve counts of cases. Typical applicationsinclude studies of cancer, cardiovascular disease, otherchronic diseases, and mortality. Data required for analy-sis consist of the total number of cases and the person-time of follow-up for each subgroup of interest. Use ofthe Poisson distribution to model variability in countsof cases is probably reasonable if we can assume thatthe disease occurs independently in different peopleand in the same person at different points in time,that the likelihood that a new case will occur in a shortperiod is proportional to the number of people, andthat disease risks are homogeneous across people andtime. These assumptions have a more formal statement,calle
Obviously the occurrence of covid isn't independent between cases, so there is a question there as to whether that makes enough difference at a population level to affect the validity of the model. I;m not sure about the homogeneity of risks thing - that doesn't apply to covid, but nor would it to cancer, cardiovascular disease etc. As is normally the case with analysing real-world data the issue will be whether the deviation from assumptions is significant, rather than just theoretically present.

If I'm understanding their description of the model correctly, though, there's a much bigger problem (p13-14), which is that they seem to be modelling covid deaths using a general regression model, ie treating each week of data as an independent point, when of course the underlying process (covid infection) is highly serially dependent, and they don't seem to be accounting for temporal autocorrelation in any way. If that's the case then the model is all kinds of f.cked, and would be whatever distribution they chose for the response.
We have the right to a clean, healthy, sustainable environment.

User avatar
Bird on a Fire
Princess POW
Posts: 10137
Joined: Fri Oct 11, 2019 5:05 pm
Location: Portugal

Re: Difficult Covid Preprint to debunk

Post by Bird on a Fire » Wed Sep 23, 2020 12:35 pm

It's a kind of model I'm not familiar with, so I might be out of my depth a bit here.

Basically, they are using the total excess death figures, which will include a portion due to covid and a portion that isn't. For the covid portion, the most important explanatory variable for the number of cases (and therefore deaths) in a given week is the number of cases last week, which AFAICT isn't explicitly included.

They could be doing something a bit more sophisticated, for instance using a Markov chain and mixture distributions, seeing as their hypothesis is that the excess deaths are a mixture of deaths, some caused by an infectious disease and others caused by other stuff (which it might be fairer to model as a Poisson process).

I note that most of the examples they cite are of modelling mortality from non-infectious agents like train crashes.

Anyway I'd better get back to work ;)
We have the right to a clean, healthy, sustainable environment.

User avatar
jimbob
Light of Blast
Posts: 5276
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: Difficult Covid Preprint to debunk

Post by jimbob » Wed Sep 23, 2020 2:46 pm

Bird on a Fire wrote:
Wed Sep 23, 2020 12:22 pm
Full paper is here: http://citeseerx.ist.psu.edu/viewdoc/do ... 1&type=pdf



If I'm understanding their description of the model correctly, though, there's a much bigger problem (p13-14), which is that they seem to be modelling covid deaths using a general regression model, ie treating each week of data as an independent point, when of course the underlying process (covid infection) is highly serially dependent, and they don't seem to be accounting for temporal autocorrelation in any way. If that's the case then the model is all kinds of f.cked, and would be whatever distribution they chose for the response.
Yes, that bit is what I was meaning about the Poisson analysis. And from a quick look at their data - they don't have enough datapoints for them to apply a Poisson distribution to anything other than each week's numbers. Which as you say, is stupid.
Have you considered stupidity as an explanation

User avatar
jimbob
Light of Blast
Posts: 5276
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: Difficult Covid Preprint to debunk

Post by jimbob » Wed Sep 23, 2020 2:52 pm

sTeamTraen wrote:
Wed Sep 23, 2020 12:05 am
They seem to have two main arguments.

One, a lot of the people who died were old/in nursing homes and might have died within a few months anyway. I have some sympathy for this in principle: In August 2003 there was a panicked reaction across France because (allegedly) thousands of people were dying in nursing homes in a heatwave. Every night, whichever doctor was not on vacation and prepared to say the largest number of dead went on national TV. I think the claimed casualty rate peaked at 14,000. It was decreed that this must Never Happen Again, and the government earmarked billions to install air-conditioning in every nursing home --- to be paid by everyone working for free on a public holiday (which one was not defined; the effects of this hasty f.ckup are still felt in France today). Meanwhile Spain, Belgium, Luxembourg, etc reported zero extra deaths and there was no visible run on the funeral parlours. However, subsequent analysis showed that there had been a modest bump in the number of deaths in August (people normally die less in summer), compensated by a small drop by the end of the year. So the heatwave may have finished off a few people who were at death's door. That said, the numbers with COVID seem too huge for this to be a plausible explanation.

Two, the idea that the excess deaths observed were not due to COVID. Given that only two major non-random things happened in 2020 compared to the average of the previous five years --- i.e., COVID and the lockdown --- they are not shy about suggesting that a lot of the excess deaths might be due to the lockdown itself. I find this highly implausible, especially with the big numbers that they are wheeling out. Yes, a lot of people's cancer appointments were put back a while, and this may have impacts over the next few years. But people were not keeling over in huge numbers in A&E of heart attacks, or early-stage cancers that suddenly grew 1000x bigger between March and May.

So in both cases we will find out in due course; the "would have died by Christmas" ones by looking at the overall excess mortality for the year, and the "lockdown killed people" one from comparing the numbers for the main causes of death over time.

In the meantime, I think this paper is mostly speculation (by economists, not epidemiologists), not to say innuendo. Where were all these excess people dying of not-COVID? At home or in hospital? The latter is where they take you when you have a heart attack or a sudden huge cancer. :roll: Why did we not hear about this? Were cardiac surgeons rushed off their feet trying to save people with emergency bypasses but unable to operate because the COVID squad had commandeered all the surgical masks? Were ambulance crews regularly turning up to find cold, lifeless patients, or having them die in the van in the car park? Anecdotes are not evidence, but generally speaking when tens of thousands of people are dying unexpectedly, the anecdotes come along anyway.

And of course, when you look at the timing of this - you see the number of excess deaths less COVID deaths rising and falling with COVID deaths and falling during lockdown.

If these excess deaths were due to lockdown, one would expect them to get worse as lockdown progressed.
Image
Image
Have you considered stupidity as an explanation

OneOffDave
Clardic Fug
Posts: 211
Joined: Tue Nov 12, 2019 9:26 am

Re: Difficult Covid Preprint to debunk

Post by OneOffDave » Wed Sep 23, 2020 5:56 pm

Generally over the lockdown period, infectious disease outbreaks from things that weren't Covid fell through the floor. Things like Norovirus were running at around 10% of the level we'd expect to see for the time of year. Extra hand washing, closed schools and reduced mixing had an impact on transmission. Food poisoning outbreaks were smaller and rarer too, possibly because fewer mass catered events, large bbqs etc.

It could be argued that those people who put off going to hospital with chest pain who then had fatal heart attacks my not have been saveable anyway so was just a timing issue rather than an impact of lockdown.

User avatar
sTeamTraen
After Pie
Posts: 2551
Joined: Mon Nov 11, 2019 4:24 pm
Location: Palma de Mallorca, Spain

Re: Difficult Covid Preprint to debunk

Post by sTeamTraen » Wed Sep 23, 2020 6:34 pm

jimbob wrote:
Wed Sep 23, 2020 2:52 pm
And of course, when you look at the timing of this - you see the number of excess deaths less COVID deaths rising and falling with COVID deaths and falling during lockdown.

If these excess deaths were due to lockdown, one would expect them to get worse as lockdown progressed.
Not necessarily, if it were true (*) that lockdown was somehow causing immediate increases in deaths due to the awful acute stress of the level of lockdowniness. In fact that curve sort of follows the strictness of the lockdown measures, IIRC.

Of course it fails utterly because of the fact that most deaths, even premature ones, aren't "heart attack, roll around a bit, die in 15 minutes". Some people will die in six months from now because they couldn't start their chemo in March (not so much because of there being no hospital capacity as because it's not a good idea to trash someone's immune system when there's a nasty virus going around). Some people will die in 7 years time because their breast cancer screen was cancelled in March and they were away in July when the replacement was scheduled, so their incipient tumour had more than half a year to take hold. But this is all drops in buckets.
Something something hammer something something nail

User avatar
jimbob
Light of Blast
Posts: 5276
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: Difficult Covid Preprint to debunk

Post by jimbob » Wed Sep 23, 2020 6:41 pm

The black line was the full lockdown
Have you considered stupidity as an explanation

User avatar
Woodchopper
Princess POW
Posts: 7057
Joined: Sat Oct 12, 2019 9:05 am

Re: Difficult Covid Preprint to debunk

Post by Woodchopper » Thu Sep 24, 2020 9:38 am

There's no need to speculate on causes of deaths during the lock down which weren't attributed to COVID-19.

You can look at the data here: https://www.ons.gov.uk/peoplepopulation ... 10july2020

The most common cause of death was from Altzheimers or Dementia. Its difficult to see how those deaths could have been caused by the lock down itsself. The deceased would very likely have already been in care homes (yes I know some people with those diseases live at home, but if they are so severe that they are a cause of death then living in a care home is more likely).

I recall that when the data was released someone from the ONS suggested that the most likely reason for the increase in deaths due to Altzheimers or Dementia was that they were actually undiagnosed COVID-19 cases. These are people who probably can't complain about shortness of breath.

Another explanation could be a dramatic fall in the quality of care due to staff shortages (due to staff being sick or in quarantine). If that occurred then the deaths would be an indirect effect of the pandemic rather than something caused by the lock down.

User avatar
sTeamTraen
After Pie
Posts: 2551
Joined: Mon Nov 11, 2019 4:24 pm
Location: Palma de Mallorca, Spain

Re: Difficult Covid Preprint to debunk

Post by sTeamTraen » Thu Sep 24, 2020 6:51 pm

Woodchopper wrote:
Thu Sep 24, 2020 9:38 am
There's no need to speculate on causes of deaths during the lock down which weren't attributed to COVID-19.

You can look at the data here: https://www.ons.gov.uk/peoplepopulation ... 10july2020
Thanks. I remember being frustrated in late March that we weren't going to get those numbers for a meaningful amount of excess deaths until the end of May, and then I forgot about it.
Something something hammer something something nail

PeteB
Clardic Fug
Posts: 205
Joined: Wed Nov 13, 2019 1:02 pm

Re: Difficult Covid Preprint to debunk

Post by PeteB » Fri Sep 25, 2020 3:41 pm

Yes, the ONS even did a study on it and examined the different possible explanations

https://www.ft.com/content/2e0a7116-593 ... a5bd554113

User avatar
bob sterman
Dorkwood
Posts: 1123
Joined: Mon Nov 11, 2019 10:25 pm
Location: Location Location

Re: Difficult Covid Preprint to debunk

Post by bob sterman » Fri Sep 25, 2020 6:04 pm

Some commentary on this preprint in this article...

https://bylinetimes.com/2020/09/23/scam ... in-the-uk/

Post Reply