Cochrane Survey: Provide feedback on Cochrane’s Policy on Conflicts of Interest

Rules are utterly useless if they are not universally applied. This changes nothing when arbitrary exemptions are given for ideological and political reasons.

There's a saying that enforcement is 9/10 of the law. With BPS stuff there is no enforcement at all, requirements are optional and violations are retroactively cleared. More rules means nothing given that.

 

This arbitrary distinction between financial and non-financial CoI seems incredibly naive and short-sited. It's about potential motivations to mislead, and there is no shortage of very powerful non-financial motivations in that regard. If an interest has a non-trivial potential to motivate deception of any sort, then it is a valid CoI; whether it is financial or not should be irrelevant. To treat them differently seems a deception in itself.

I think people sometimes just use the phrase Conflict of Interest when they really just mean Interest, without thinking about what they are saying. It then becomes easier to conflate the two and argue less importance.

 

This policy does nothing to address problems of the past. It continues to ignore the fact that Cochrane reviews uncritically include studies such as PACE (and also ludicruously underpowered studies that even NICE ignore) in its reviews and ignores methodological weaknesses which may or may not be caused by the triallists wanting the result to go a certain way. It is nonsense to say that Cochrane is better than all other journals because of this new policy. They are a journal that only publishes systematic reviews. With a review there is absolutely no need for ANYONE involved in it to have anything to gain or lose depending on which way the results go. So why do they let any clinicians who have developed a treatment, or is a proponent of a treatment, get involved in a review trials of that treatment in any way at all? And/or have studies of theirs included in the review. And because they comprise 50% or 33% of the authorship, and don't do the risk of bias assessment, that's OK? It's absolutely ridiculous. This is why NICE ignores Cochrane and does its own reviews with independent reviewers doing the technical work and a supposedly independent panel drawing up the guideline. Although an unbiased panel is not guaranteed either. To judge whether a trial has been done well or not is a technical task. However, you do need to be able to judge whether decisions about which outcomes to measure and report are valid. And also be ever so slightly suspicious when the triallists don't declare their COIs to participants, in breach of the Declaration of Helsinki, and spend thousands of taxpayers money trying to hide the data. What is this nonsense of going from 50% no conflict to 66% no conflict? It's a complete joke. I thought Lisa Bero would think it was a complete joke too. Rant over.

 

Yes, I agree with all of that.

I thought, though, that Bero might have a point in that the word 'interest' is now being manipulated to make it possible to exclude people who might actually be unbiased - the people that Ilora Finlay described as 'interested but disinterested'.

 

Which aligns with the notion that people conflate "Interest" with with "Conflict of Interest", and from that seem to argue that a (Conflict of) Interest is not necessarily such a big deal. Sometimes maybe confused thinking, and sometimes I'm sure deliberate misdirection. How people can argue that non-financial CoI (real CoI) are less significant than financial ones is beyond me, but it does seem to fit with some of the weird logic that some of the BPS proponents favour.

I agree entirely that an Interest should not be assumed, by default, to be a CoI, but it should be assumed that anything which might conceivably be a CoI should should be examined to see if it really is one. But many interests will not of course be CoI.

 

Which is all a pretty silly discussion because this is an important concept in the law where it has been discussed and defined rather well in a clear process already. Being interested in the question is not an issue, having a declared interest in a specific outcome is. It's still not an easy problem but here it's like the entire discussion is happening in a vacuum.

On this it goes beyond direct financial interest. Several PACE researchers have been advising those treatments for decades, they built their reputation as "top researchers", nevermind how cringeworthy that is, entirely on those treatments. Even without the specific conflicts of interest of being paid advisers to medical insurers, they had a very strong incentive in proving themselves right, as admitting the uselessness of their treatments, which they found in PACE as in every prior trial, would have essentially ruined their career, made it clear they had given bad medical advice, in a paid capacity, for years. They had interest in the questions being asked, no issue here, but more importantly were deeply vested in a particular outcome that would have affected them personally. It's scandalous that they were even allowed to grade their own work, having decided on the test questions themselves in the first place. To further having taken part in the Cochrane review is just a complete breakdown of the entire process.

It's just weird that there should be so much discussion over something another field has largely settled already. Conflicts of interest are generic, not even particular to research or medicine. Recusal is not a penalty and it's not as if there aren't enough people out there with sufficient skills and no stake in any outcome. The 66% rule is particularly absurd.

 

NICE are also clear on this.

https://www.nice.org.uk/Media/Defau...rocedures/declaration-of-interests-policy.pdf

[my bold]

So to my mind Cochrane are just a*** licking the BPS crowd.

 

We may be talking at cross purpose but this is not what I was meaning.

The word interest has two quite different meanings in English, just as 'plain' and 'bank' do.
One meaning can be a synonym for hobby. The other is a synonym for stake.

So I am interested in research into ME. Research into ME is an interest of mine. But I do not have an interest in research into ME. There are situations where being interested might be expected to be linked to having a stake but that is not generally true. You can be interested in South American politics but have no reason to be thought to have an interest in South American politics.

The problem is that 'non-financial interest' is being manipulated with ambiguous language to include hobbies as well as stakes. That means that someone who is knowledgeable about a subject and shown interest in it can be disqualified from influencing a decision despite having no stake in the issue and there being no reason to think they have.

 

Thank you. Yes, that is an in important distinction.

I recall hearing a similar manipulation of meaning relating to interest a good few years ago by a Sun reporter (no surprises there). When quizzed whether he thought his actions were truly in the public interest, he responded by saying that there was considerable public interest in what he had reported. I don't think he understood the distinction.

 

I'm sure he would have done. This seems like a stock response, probably much rehearsed, to a well known problem.

 

Only skimmed all this and first thought the Bero reference could still be a misleading one, only quoted by the autors because the paper's title fitted. Still not sure about this, not able to read it all.

Didn't read the following either, but saw the diagram and thought it could be helpful to adjust it to what was going on with:

1) the PACE investigators and the Science Media Centre plus one of the SMC's main figures downplaying his role in the PACE trial

2) some funny things about the SMC, journalists and prizes (suggesting each other and awarding each outher prizes)
https://www.s4me.info/threads/trial...-an-upcoming-article.7944/page-10#post-141323

3) a Reuters journalist and her links to both the SMC and the lead author of the Cochrane review assessing the evidence of GET as a treatment allegedly investigated, among others, in the PACE trial

4) the role of a co-author of the SMILE trial in Cochrane's assessment tool for risks of bias
(https://www.s4me.info/threads/trial-by-error-lead-author-of-cochrane’s-new-bias-guideline-is-lp-study-co-author.11055/,
https://www.s4me.info/threads/rob-2...in-randomised-trials-2019-sterne-et-al.11025/)

5) Paul Glasziou's role as both a buddy of the PACE trial investigators (author of the later withdrawn Cochrane review protocol on GET individual patients data ) and a HANDI working group author [even co-founder/ head (?)] using the Cochrane review on GET to promote exercise as a safe and effective treatment
https://www.s4me.info/threads/“graded-exercise-therapy-chronic-fatigue-syndrome”-by-the-handi-working-group-2019.10469/#post-185083

6) BMJ and The Lancet peer review process of papers dealing with ME/CFS and how the editors defended papers against valid criticism




from:
Moynihan Ray, Macdonald Helen, Heneghan Carl, Bero Lisa, Godlee Fiona. Commercial interests, transparency, and independence: a call for submissions BMJ 2019; 365 :l1706, https://www.bmj.com/content/365/bmj.l1706

(Ah, this is the BMJ, so no surprise to meet Fiona here. Of course, some of the other PACE-SMC-REUTER-Cochrane-SMILE links could also just be normal professional collaboration.)

 

Dollars for Profs
Dig Into University Researchers' Outside Income and Conflicts of Interest

https://projects.propublica.org/dollars-for-profs/

eta: could do with one of these for the UK

eta2:
https://twitter.com/user/status/1204360911303172096