RoB 2: a revised tool for assessing risk of bias in randomised trials (2019) Sterne et al.

This is certainly true, given that the PACE authors acknowledged a significant mistake in the Lancet correspondence about how they characterized the population of a database they used to developed their bogus "normal range." The point has never been corrected in The Lancet, and Dr Horton has ignored the request to do so.

 

And PubMed have screwed up the links to the comments too. Part 1 (April 1989) is here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1711821/?page=1
Part 2 (May 1989) is here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1712003/?page=1

Letters are easy to bury, and hard to find, even when properly cited.

 

Webinar: RoB 2 Domain 3: Bias due to missing outcome data

Presented by Higgins and Sterne.

https://training.cochrane.org/rob-2-domain-3-bias-due-missing-outcome-data

 

https://twitter.com/user/status/1291690302567264256

 

I haven't watched this yet https://training.cochrane.org/resource/cochrane-new-conflict-interest-policy-non-financial-interests

 

But I contacted the presenter on Twitter ;-)
https://twitter.com/user/status/1293150705923117056

 

Grundy studied under the mentorship of Lisa Bero http://www.quinngrundy.com/ Lisa Bero, the new Cochrane research integrity tsar, co-wrote a paper with Grundy saying that non-financial conflict of interest isn't conflict of interest at all. https://www.ncbi.nlm.nih.gov/pmc/ar... or institution is,not a conflict of interest.)

 

https://twitter.com/user/status/1311762673810448408


https://twitter.com/user/status/1311763350070734849

Code:

https://twitter.com/Healthy_Control/status/1311762673810448408

See also forum thread: https://www.s4me.info/threads/new-d...ture-over-medication.16182/page-6#post-288998

And: https://www.s4me.info/threads/new-d...ture-over-medication.16182/page-8#post-292505

 

Interesting that Lisa Bero chooses to interpret CoI that way.

I'm pretty sure that this special interpretation differs from CoI guidelines in business and law for example.

 

Yes, I looked up a while back and it certainly does not have to be financial.

 

Bero says...."Focusing on interests such as personal beliefs, experience, or intellectual commitments can divert attention from financial conflicts of interest, which have the potential for widespread influence. The result is an erosion of the evidence base and confidence in science, making it vulnerable to competing groups’ claims [12], as we are seeing with issues as diverse as childhood vaccination and climate change."

Oh dear, well we'd better stop diverting attention then. Selfishly causing the erosion of trust in the nice people of Cochrane and the crappy data from crappy studies they polish up and put in their reviews.

 

But those are not interests in the sense that is relevant. An interest is some potential benefit that may accrue, not a belief or experience. Nobody is focusing on those. This resounds with rationalisation with deliberate confusion of term usage. Again, the message in the background is that big Pharma is bad but equally biased thinking of a touchy feely primary care sort is always good.

 

yes, you're right. She doesn't mention researcher allegiance at all. Which is relevant.

https://www.nationalelfservice.net/...ce-the-achilles-heel-of-psychotherapy-trials/
https://bmjopen.bmj.com/content/5/6/e007206

 

I see that as an example of some really poor quality reasoning. Turning a reasoned thoughtful consideration of what is a conflict into a polemic of competing opinion is unhelpful. Again and again, I see this unhelpful attitude of it's all so complicated let's simplify it all. (An example within this group but on a different topic would be the ubiquitous fatigue -- let's not tease out differences let's turn everything into generic fatigue)

Yes, thinking can be hard work. Good thing they get paid to do it. So no excuses then.

 

2020 Bias Methods Group Open Meeting October 27 & 29

https://methods.cochrane.org/bias/news/2020-bias-methods-group-open-meeting

 

https://twitter.com/user/status/1415934562761330696

 

https://twitter.com/user/status/1415824667286274050


how indeed.

 

Just to highlight the inconsistency here:








Clearly a lack of joined up thinking - or maybe just thinking - here.

 

The bigger question is how is it that this is common practice?

Wait, I know, it's because it is common practice, in the sense that people commonly do this. None of this is accidental, it's built-in to EBM, an alternative to the scientific method. How people expect that using a lesser methodology would yield good results is a real puzzle. No one ever checks these things.

That's easy to fix. Except the problem is people, that people are using this method to make their job easy, don't even to bother doing any real work. In hindsight this is bad but that was obvious from the start.

In computer science we respect GIGO. Why medicine can't do this I don't understand but if we can do this, they sure can, they just choose not to.