PNE Online
Welcome to PNE-Online. Why not register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox! You can also join up as a forum Patron to help support in the running costs of the forum.

Coronavirus science and statistics(no politics)

This certainly seems to go against the narrative.

 
The red lines are 14 days after each of your lockdown dates. What's your reasoning that a lockdown on the 5th November with falling rates didn't combine to produce a complete drop in hospital patients? Purely because it was a variant (that then responded within 14 days of the next lockdown in exactly the same way with hospital figures?)

View attachment 3784

All three peaks clearly show the rate of increase starting to slow down BEFORE lockdown. The "inflection point" is that point where an exponential upward curve starts to flatten.. And this occurs before each peak. If lockdowns caused the decline you would expect an inverted V shape with rates continuing to climb exponentially AFTER lockdown.. Because it would take at least two weeks after lockdown for patients in hospital/infections to start falling. If it was lockdown that caused the decline why did the rate of increase start to decline BEFORE lockdown ?
 
All three peaks clearly show the rate of increase starting to slow down BEFORE lockdown. The "inflection point" is that point where an exponential upward curve starts to flatten.. And this occurs before each peak. If lockdowns caused the decline you would expect an inverted V shape with rates continuing to climb exponentially AFTER lockdown.. Because it would take at least two weeks after lockdown for patients in hospital/infections to start falling. If it was lockdown that caused the decline why did the rate of increase start to decline BEFORE lockdown ?
Welcome back. According to Snicky the dates of lockdown I’ve added are in green. Mine and your definitions of “flatten” would appear to differ as I’d personally say the lines look much flatter 2 weeks after lockdown than in the first week after a lockdown. I’d also go do far as saying rates continued to increase exponentially after lockdown looking at those charts. What is your definition of “flatten” at the point the green line intersects?
I assume the stats are on the dashboard too that produce the image2CC2DA7E-CD3D-4059-B2B3-D2A6D726D25B.jpeg
 
Welcome back. According to Snicky the dates of lockdown I’ve added are in green. Mine and your definitions of “flatten” would appear to differ as I’d personally say the lines look much flatter 2 weeks after lockdown than in the first week after a lockdown. I assume the stats are on the dashboard too that produce the imageView attachment 3798
The article did mention infections, not hospital admissions. There is a lag of a couple of weeks I think.
 
The article did mention infections, not hospital admissions. There is a lag of a couple of weeks I think.
Also just to quote sliper “Because it would take at least two weeks after lockdown for patients in hospital/infections to start falling.”

the green and red line are two weeks apart. So according to that “two weeks after lockdown, patients in hospital started falling”

so I’m kind of lost what point sliper has argued against?
 
But to be honest Me and my kids have met my parents and family for the first time in 18 months today.

If people still want to argue that lockdowns didn’t have a desired impact on hospital numbers then knock yourselves out.

The December/January wave and government wavering showed exactly what happens when lockdowns were on and off and if people really want to pretend they contributed nothing then knock yourselves out.

They might not have been the right answer or they might not be the long term answer but they made a difference to what was happening with infections and hospital usage, that much is clear
 
But to be honest Me and my kids have met my parents and family for the first time in 18 months today.

If people still want to argue that lockdowns didn’t have a desired impact on hospital numbers then knock yourselves out.

The December/January wave and government wavering showed exactly what happens when lockdowns were on and off and if people really want to pretend they contributed nothing then knock yourselves out.

They might not have been the right answer or they might not be the long term answer but they made a difference to what was happening with infections and hospital usage, that much is clear
I’m not sure anyone is arguing that Jake, not on here anyway. More like it’s not as cut and dried as it seems and that the chances of them happening again in the same way are low. There may be a better way.
 
I’m not sure anyone is arguing that Jake, not on here anyway. More like it’s not as cut and dried as it seems and that the chances of them happening again in the same way are low. There may be a better way.

I think you’ve made a giant leap there. We’ve (including myself)already stated that the long term impacts and comparative results are going to take some time to establish.

“there may be a better way” No doubt. That’s what research, time and hindsight give you along with all the other analysis and again, I’ve not disputed that in any of my last posts.

All I’ve pointed out based on the dates you provided earlier is that the statistics provided by the NHS show some form of correlation to those dates. It isn’t perfect, nor is it definitive, but any form of analysis can show that especially with the second and third lockdown imposed, and the shape of the graph it shows it had an evident impact when/if a lockdown was in place.

25205CE1-0FCB-46F8-B644-49AF7FCDF436.jpeg

If that chart was put in a GCSE paper last week and the question was give one feature of the trend/line between the green and red line pairs I can bet your bottom dollar that the Mark scheme would say more along the lines “increasing, rising, many more people in hospital by the second line”. I bet you wouldn’t get a Mark for saying “flat, already flattened, not increasing as much”, yet apparently that’s what it shows according to some
 
Also just to quote sliper “Because it would take at least two weeks after lockdown for patients in hospital/infections to start falling.”

the green and red line are two weeks apart. So according to that “two weeks after lockdown, patients in hospital started falling”

so I’m kind of lost what point sliper has argued against?

I'll concede on the point...that graph doesn't show that. I assumed the red lines were lockdown.

From the Spectator..

"In a peer-reviewed paper now published in Biometrics, I find that, in all three cases, Covid-19 levels were probably falling before lockdown. A separate paper, with colleague Ernst Wit, comes to the same conclusion for the first two lockdowns, by the alternative approach of re-doing Imperial College’s major modelling study of the epidemic in 2020. In light of this, the Imperial College claim that new infections were surging right up until lockdown one — causing about 20,000 avoidable deaths — seems rather questionable."

I had in mind a similar analysis to the peer reviewed data I had seen earlier and linked to in the quote.. and assumed wrongly that the red lines were the dates of lockdown.

 
Last edited:
Interesting story of a young athlete who rushed back into training after COVID. Poor lad missed out on the last Olympics because of injury, and will probably be too old for the next ones. A lesson worth bearing in mind for anyone who picks up the virus.

Scottish 800m runner Guy Learmonth has given up on qualifying for the Tokyo Olympics because of a respiratory problem after revealing he contracted Covid-19 in March.

The 29-year-old was due to compete at the British Championships, which double as Olympic trials, this weekend.

But he is now taking time out after belatedly seeking medical help.

Learmonth says missing a potential Olympics debut is the "hardest and most brutal decision of my life".

"I didn't want to admit anything was wrong," he added on Instagram.

"I was one of many that got Covid at the Euro Champs in March and after initially getting through it, I just cracked on with training almost immediately, raced a lot across America and Europe and put so much stress and strain on my body without the proper recovery.

"I was reluctant to accept that Covid had anything to do with it but as the races went on and my performances started going backwards, I knew deep down something was wrong with my body.

"I finally took the plunge and after a series of tests and respiratory scans this week we've gotten to the bottom of things instantly.

"The worst thing of all is it is a relatively quick recovery process and if I just admitted that something was wrong then I would have had this sorted in March and not two days before the Olympic trials. My own hubris got the better of me."

Learmonth competed at the 2014 and 2018 Commonwealth Games, but missed out on possible selection for the Olympics in Rio five years ago because of injury.

He secured a third British indoor title with victory in Glasgow in February last year.

"It's a bitter pill to swallow," he said. "These last 18 months I've genuinely changed my entire life to gear for Tokyo and to fight for my place but I won't be able to make that walk this time around.

"It has been the hardest and most brutal decision of my life. But I will be taking a small step back to focus on my recovery, some rehab and all going well over the next four weeks I hope to be back racing towards the end of the season in August/September time."
 

"Robust T Cell immunity in unexposed individuals."
 
They did say that they had limited confidence in their assertions and I read several prominent scientists disputing the findings earlier this week. There seemed to be surprise that Henaghan co authored a study but wasn’t listed in the contributors list. Something about his links to collateral global.
I’ve long thought that the term mask can be badly used, with there being a huge difference between a bit of cloth worn over and over again and a N95 (or similar) mask. With covid being almost exclusively an airborne spreading disease, common sense surely says properly covering mouth and nose should protect the wearer. Apparently mask wearing wasn’t particularly strict in the studies.

I’m also dubious about them saying hand gel works but proper masks don’t with it being airborne - especially bearing in mind the studies done previously were they failed to find covid on surfaces from rooms were many infected people had been.


That's an interesting counter argument, thanks, with some important questions.

My first observation is that they, validly, point out the apples and oranges argument between community and healthcare. However, they then try to use healthcare settings to question community settings. In other words, trying to use the orange to disprove the apple. Healthcare, where people are trained to wear them correctly, is different from the community, where most are not - and many don't. The same for the type of masks - the vast majority of mask mandates included surgical masks which, by their own words, are not designed to stop a respiratory virus. Those of us pointing this out a couple of years ago were labelled dangerous subversives! :D :p

They do quote one study as showing that masks reduce the transmission of influenza - when you actually look at the paper they cite, it does not state that. It points to a single study in healthcare settings that *may* show a reduction in transmission - and specifically states that there is a gap in the literature - little evidence, in other words - to show that there is any reduction in the real world. If they are going to pick apart the analysis of others, then this is very disingenuous. To be honest, I get the impression that they only read the abstract rather than the full article.


That's before we look at the papers they omitted - most of which state 'little evidence' or focus only on healthcare settings. In the same way, they contradict themselves by stating that the N95 has to be worn at all times to be effective or people can be infected anywhere in the building - and then claim that there is a type of herd protection against transmission if only some people wear masks in a healthcare setting (the paper they quoted was not definitive, btw, and accepted experimental flaws - I again think they have been abstract skimming rather than reading). Can both of these be true at the same time?

My other issue is that they argue that the study assesses the impact of mask advice rather than the efficacy of the mask itself. However, they then go on to quote the Bangladesh study as proof of mask efficacy. A study that, of course, set out to test the efficacy of mask advice, not the masks themselves.

I have many concerns with the Bangladesh study - especially in how it is extrapolated to the wider population. Firstly, there was very little attempt to filter out any confounding variables - such as changes in behaviour. Were people more likely to isolate or social distance, for example?

Secondly, people trying to extrapolate it to show that masks work rarely account for time and place - the villages are rural and sparsely populated. Would the relatively small effect found still apply in a more crowded city where the amount of covid carriers you encounter will swamp that? Indeed, that is one of the contradictions in their reply - they argue about the build up of particles in crowded settings, but the slight effect found in the Bangladesh study is unlikely to make much difference. Seropositivity is also very crude - people could have caught the virus before the study.

In terms of time, it was an eight week window, a snapshot. If there was any slight reduction in seropositive individuals, would this be the same in three months, six months, a year? Too many people think of this virus only in the short term - when it keeps circulating, you are eventually going to get it. In addition, the Bangladesh study attracted a lot of criticism for its statistical analysis that overemphasised the findings, although determining whether that is correct is way outside my paygrade!

Finally, they quote a study suggesting that the use of *any* face mask reduces transmission. Yet, they earlier claimed that surgical and cloth masks do not stop respiratory viruses. Again, can both of these be true, or are there problems with the 'well-designed real world study.' (There were some questions about the methodology, btw).

Overall, I thought this article was a bit rushed and didn't dive particularly deeply into the research. Good to see a debate, but this has not changed my view that there is little evidence that masks are effective - certainly in the community rather than healthcare.
 
For the initial critique, I didn't look at the authors and instead played the ball. For the second part, I looked into the background of the people contributing to the Conversation article (I am already aware of Trish Greenhalgh) and it seems that they have pushed masks all the way through the pandemic. Indeed, that would explain the issues with the article and suggests why they appeared to look for research that strengthened their narrative - that isn't necessarily a problem as science needs debate and a plethora of views. However, it does not excuse the sloppiness, cursory research, and misrepresentation of what conclusions the papers actually draw - call it professional pride, but I find that much harder to forgive :D

Very disappointing - I give them a D-
 
Top