Seasonal effect?

[bob sterman]

Carl Heneghan is frequently in the news telling us that COVID-19 is not as big a threat as we think.

But what on earth is he going on about here?

https://twitter.com/carlheneghan/status ... 1142833152

He writes "This is what it looks like in Oldham - shows restrictive measures didn't affect the outcome as the infection is currently increasing in line with a seasonal effect" with this figure...



For this July-August seasonal increase - does he know Oldham is in the Northern Hemisphere and neither month is known for its particularly high levels of respiratory viruses?

Looks much more like cases declining after something happened at the end of March, and then increasing again after something happened in early July. Wonder what that could be???

[nezumi]

It must be that he thinks that because it's a different season on the calendar it must be because of that. Because, as we all know, nature always obeys our arbitrary calendar and we never ever get snow in spring or a warm day in December. Because science. Because his statement bears no relation to any understanding of epidemiology at all. Please tell me this man is not an epidemiologist?!

[bob sterman]

It must be that he thinks that because it's a different season on the calendar it must be because of that. Because, as we all know, nature always obeys our arbitrary calendar and we never ever get snow in spring or a warm day in December. Because science. Because his statement bears no relation to any understanding of epidemiology at all. Please tell me this man is not an epidemiologist?!

He's not just a Clinical Epidemiologist - he is Director of the Centre for Evidence-Based Medicine (CEBM) at Oxford...

https://www.phc.ox.ac.uk/team/carl-heneghan

He's generally pretty vocal in his opposition to most "lockdown" stuff - and doesn't tend to acknowledge when "lockdowns" might have worked - as in this tweet...

https://twitter.com/carlheneghan/status ... 0591808512

"Out of 213 countries with Covid, China is 201st when it comes to cases per popn. - anyone have a good explanation for this anomaly?"

Err....just perhaps this could have something to do with vast quarantine centres and the most brutal and violently enforced lockdown in the world with some people having doors to their apartments welded shut? (not saying we should follow these examples).

[nezumi]

He's not just a Clinical Epidemiologist - he is Director of the Centre for Evidence-Based Medicine (CEBM) at Oxford...

Then I clearly do not understand as much epidemiology as I thought I did. Better get back to studying

[OneOffDave]

We're not even 3 weeks into meteorological autumn and given the case data lag it's far too early to see a seasonal effect. It'd be interesting to know if he's adjusted that data for testing rates as that will have an impact on case numbers reported.

[sTeamTraen]

What is the seasonal explanation for the curves of, say, the United States or Australia?

[bob sterman]

It just boggles the mind that the Director of the "Centre for Evidence-Based Medicine" could put this out as "evidence" of anything.

Stamped with a "Trust the Evidence" logo.

[badger]

Trying and failing to get my head around this.

Is he saying that "restrictive measures didn't affect the outcome" because to him, 'affecting the outcome' means reducing the case numbers in a much more pronounced way. Like to almost zero after lockdown? And that therefore the small effect on numbers (as he sees it) therefore must be down to something else, because if lockdown had worked, it would have had a much bigger effect, so it must be.... seasons, yes, that fits. It's seasons, not lockdown? Wtf?

It's easier for me to believe his Twitter is a parody account, or the organisation which he represents isn't as kosher as it sounds, than believe that anyone vaguely numerate could make the argument in that tweet and feel it was well reasoned and publicly defendable.

Please help me understand.

[PeteB]

He's just sh.t and doesn't admit he's wrong

A clearer cut example

https://twitter.com/carlheneghan/status ... 9668413440

[PeteB]

Or as James Annan says

https://twitter.com/jamesannan/status/1 ... 1050054656

“Seasonal effect” had me ROFLing.

I wonder if there’s anyone left who still hasn’t worked him out yet?

[bob sterman]

Heneghan is in the news pretty much every day - talking about new evidence that supposedly shows that COVID-19 is not a major threat any more, that we shoudn't introduce particular lockdown measures, or require people to wear masks etc etc. A lot of the evidence is published in the form of "rapid reviews" in the CEBM blog. Rather than in peer-reviewed papers or on pre-print servers where others could comment (they don't allow comments on the CEBM reports)...

https://www.cebm.net/oxford-covid-19-evidence-service/

What they call the "Oxford COVID-19 Evidence Service" mixes science with advocacy (e.g. reports titled "Does the resurgence of COVID-19 in Spain exhibit exponential growth?" are mixed with others such as "Bin the rule of six").

Heneghan himself authors a new report on the blog 2-3 times a week - describing new evidence that is apparently so important that it urgently requires a big media engagement drive and pubic dissemination campaign. Not many scientists can even dream of producing work 2-3 times a week that is of such rigour and significance that it demands the attention of the nation's media and policymakers on each occasion.

I guess these engagement activities keep him very busy - as although he has authored hundreds of publications in peer-reviewed journals - over 20 this year alone - none of his work on COVID-19 appears to have yet appeared in the form an a peer-reviewed empirical report in a journal (at least I can't find any on Scopus / Pubmed yet).

[sTeamTraen]

I guess these engagement activities keep him very busy - as although he has authored hundreds of publications in peer-reviewed journals - over 20 this year alone - none of his work on COVID-19 appears to have yet appeared in the form an a peer-reviewed empirical report in a journal (at least I can't find any on Scopus / Pubmed yet).

Google Scholar (which picks up a lot beyond peer-reviewed articles) lists 47 publications for him this year.

11 of those mention COVID or SARS-CoV-2 in the title, but they seem to be mostly working papers on the CEBM web site (of which he is a director), with a couple of preprints that might never see the light of day in a peer-reviewed journal. One of these investigates whether COVID-19 can be [PDF] transmitted by the orofecal route, so if you don't like this guy's research, I guess you can eat sh.t.

[dyqik]

What is the seasonal explanation for the curves of, say, the United States or Australia?

The US at least has people moving indoors more in summer as it's too f.cking hot outside, and moving inside in winter, as it's too f.cking cold outside. And moving inside in fall, because the forest fires have made it so smoky that you can't breathe outside. Spring in most regions lasts for two weeks.

[MartinDurkin]

He's not just a Clinical Epidemiologist - he is Director of the Centre for Evidence-Based Medicine (CEBM) at Oxford...

Isn't this where Ben Goldacre works? Heneghan's stuff seems like the sort of thing he would have written about in his old Guardian column.

[badger]

He's not just a Clinical Epidemiologist - he is Director of the Centre for Evidence-Based Medicine (CEBM) at Oxford...

Isn't this where Ben Goldacre works? Heneghan's stuff seems like the sort of thing he would have written about in his old Guardian column.

They do indeed work at the same place.

Professor of EBM at the University of Oxford. Director of the Nuffield's CEBM. It seems like there's plenty of people who haven't worked him out yet...

[PeteB]

https://twitter.com/davidnhamilton/stat ... 6299676678

[PeteB]

I notice Henneghan and Gupta were invited to the No 10 summit to overrule the scaredy cats Whitty and Vallance

[bob sterman]

I notice Henneghan and Gupta were invited to the No 10 summit to overrule the scaredy cats Whitty and Vallance

Getting the "superforecasters" in???

Would this be the same Professor Gupta who told us in March that by that point COVID-19 had infected almost 50% of the popultion - only to be proved wrong when serology studies showed that it was closer to 5% even by May. And who pretty much every week since then has been telling us that we've almost reached "herd immunity"???

[badger]

Gupta, Hennegan and Sikora authored the letter mentioned in the article, and they're all pushing the herd immunity agenda pretty hard from what I can see on big platforms.

Sikora has form (though a large following) but Gupta and Hennegan present as reasonable authorities and are indeed "leading academics".

Am not sure how their call for "more targeted measures" would actually work, maybe move all the over 65s to the isle of man so the rest of us can carry on as normal? But they are getting a lot of traction from people whose trust in the Govt has been worn down and are desperate not to go into lockdown again.

I see a lot of people asking why most of the places hit hard first time around are doing better now, and therefore it must be some kind of immunity, and therefore Gupta etc are onto something (T cells/cross reactivity has already given us much higher immunity in some places, or much more than we think). And not just KBF types, it's much broader than that.

[bob sterman]

I see a lot of people asking why most of the places hit hard first time around are doing better now, and therefore it must be some kind of immunity, and therefore Gupta etc are onto something (T cells/cross reactivity has already given us much higher immunity in some places, or much more than we think). And not just KBF types, it's much broader than that.

Of course - that will happen eventually. It's just that Gupta has been banging on about us being on the cusp of "herd immunity" since March. Even suggesting that the threshold for "herd immunity" for this virus might be as low as 20% of the population exposed. Conveniently ignoring data such as what happend on the Charles de Gaulle aircraft carrier (60% of crew managed to get infected). Serology has shown London is around 20% and the city does not appear to be unaffected by the recent rises ("seasonal effect"???).

Eventually spread will slow down - and she'll finally be right. Stopped clock and all that.

[dyqik]

What is the seasonal explanation for the curves of, say, the United States or Australia?

The US at least has people moving indoors more in summer as it's too f.cking hot outside, and moving inside in winter, as it's too f.cking cold outside. And moving inside in fall, because the forest fires have made it so smoky that you can't breathe outside. Spring in most regions lasts for two weeks.

Of course, the most powerful season in the US is election season. Or maybe college football season.

[badger]

I see a lot of people asking why most of the places hit hard first time around are doing better now, and therefore it must be some kind of immunity, and therefore Gupta etc are onto something (T cells/cross reactivity has already given us much higher immunity in some places, or much more than we think). And not just KBF types, it's much broader than that.

Of course - that will happen eventually. It's just that Gupta has been banging on about us being on the cusp of "herd immunity" since March. Even suggesting that the threshold for "herd immunity" for this virus might be as low as 20% of the population exposed. Conveniently ignoring data such as what happend on the Charles de Gaulle aircraft carrier (60% of crew managed to get infected). Serology has shown London is around 20% and the city does not appear to be unaffected by the recent rises ("seasonal effect"???).

Eventually spread will slow down - and she'll finally be right. Stopped clock and all that.

Hmmm, I get the stopped clock thing and am not saying Gupta is right, but using aircraft carrier data when discussing herd immunity threshold calculations *can* be ignored, can't they? The R number would have been very different and we are not calculating for a population sharing each other's space on a tin can with recycled air. It's the sort of argument that plays into Gupta's hands.

London does look to be about 20% of antibody prevalence, but again if pre-existing immunity exists, or other kinds after infection, we need to acknowledge it's possibility, even if we don't yet know the percentage.

"Not unaffected"? Most London boroughs positive numbers have fallen this week. Some by as much as half. Probably due to testing farce, but numbers have been relatively low for sometime, even after back to work, back to pub, holidays, beaches, protests, raves and now schools have been back for over two weeks. I have some sympathy for those looking at Whitty & Vallance yesterday and not feeling like they were getting enough of an explanation.

[AMS]

Can't find where I saw it now, but saw recently a good explanation of how pre-existing herd immunity (e.g. vaccinated population) works a bit differently to how herd immunity develops naturally, where the fraction that are immune is not constant over time. The upshot is more people become infected by the disease being spread - e.g. you might stop the virus spreading if (say) 60% are pre-immune, but in a naive population, 75% might catch it before the HI stops it. Importantly, the higher the R number, the bigger this overshoot will be, so slowing the spread of the virus also decreases the endpoint in the total number who'll actually catch it.

[badger]

Can't find where I saw it now, but saw recently a good explanation of how pre-existing herd immunity (e.g. vaccinated population) works a bit differently to how herd immunity develops naturally, where the fraction that are immune is not constant over time. The upshot is more people become infected by the disease being spread - e.g. you might stop the virus spreading if (say) 60% are pre-immune, but in a naive population, 75% might catch it before the HI stops it. Importantly, the higher the R number, the bigger this overshoot will be, so slowing the spread of the virus also decreases the endpoint in the total number who'll actually catch it.

Interesting, also in the light of this pre-print of herd immunity in Brazil https://www.medrxiv.org/content/10.1101 ... 20194787v1

Just tweeted by @AdamJKucharski

Some people claiming that this supports theory of HIT of 40%, but not sure how...

[PeteB]

I was listening to Neil Ferguson on the Life Scientific this morning and he said "It's not having an appreciable effect as yet, The most heavily effected area of the UK is London and there a few London boroughs where there is a hint of herd immunity having a effect but it's really too early to tell"

I would have thought 20% infected would have quite an effect if R is just over 1

R = R0x

So it would reduce R =R0*0.8

so for the same level of transmission comparing an area with X=.95 and x= 0.8 you would expect to see a significant difference

Doesn't mean it's safe, by any means but more wiggle room