Trial By Error: Lead Author of Cochrane’s New Bias Guideline is LP Study Co-Author

No amount of randomisation impacts on the high risk of bias when using subjective measures in unblinded trials.

No matter how good or sophisticated your statistical analysis, if your data is subject to bias your results are worthless. Not sure this language is acceptable, but there is no way of getting around ‘Shit in, means shit out’.

This is nothing but a cynical attempt to lend spurious credibility to bad research.
 
Is this because that dude has first-hand experience with how bias works and is therefore an expert?

Does Cochrane know that LP literally entails standing on paper and shouting at it?

Is all this a realityTVshow and the writers are just experimenting with how much they can get away with until their audience has a problem with their suspension of disbelief?

...can someone please break the 4th wall now?
 
Peter Trewhitt said:
No amount of randomisation impacts on the high risk of bias when using subjective measures in unblinded trials.

No matter how good or sophisticated your statistical analysis, if your data is subject to bias your results are worthless. Not sure this language is acceptable, but there is no way of getting around ‘Shit in, means shit out’.

This is nothing but a cynical attempt to lend spurious credibility to bad research.
Click to expand...
As I understand it randomisation is to avoid selection bias. Other forms of bias need other strategies to deal with them. Randomisation is just one important tool amongst a good many important tools needed for properly run trials; the integrity of a trial is presumably only as strong as the weakest link in the toolchain.
 
Sean said:
They seem to have left out the 'controlled' bit of RCT.
Click to expand...
There is a very good reason for this I'm sure, as this link will help to clarify:

'PACE was not a "randomised, controlled trial" as stated by Chalder et al in their rapid response'

https://www.bmj.com/content/350/bmj.h227/rr-26
For example, Robert Courtney points out:
“Although it was a large and expensive government-funded trial, the PACE trial, as with most cognitive-behavioural research, was open-label and failed to control for placebo effects and biases such as response bias
Click to expand...
[my bold]

Which is why, as I understand, the 2011 PACE paper did not describe it as controlled:

"Comparison of adaptive pacing therapy, cognitive behaviour therapy, graded exercise therapy, and specialist medical care for chronic fatigue syndrome (PACE): a randomised trial"
 
Last edited:
Barry said:
Which is why, as I understand, the 2011 PACE paper did not describe it as controlled:

"Comparison of adaptive pacing therapy, cognitive behaviour therapy, graded exercise therapy, and specialist medical care for chronic fatigue syndrome (PACE): a randomised trial"
Click to expand...
It gets better:

PACE was originally described as 'controlled' in the PACE protocol paper (2007), but not in the main paper (2011). Which means that either the PACE authors decided not to claim controlled status for the final publication, or it was denied to them by the reviewers or journal editors.

Either way I seriously doubt they relinquished that status willingly. Clearly they wanted that status for PACE.

The word 'controlled' is mentioned only once in the main text, in relation to the PACE team's search strategy of the literature, and a handful of times in the reference section in the titles of some of the studies they cited.

PACE was most certainly not controlled.

Anybody claiming any kind of expert authority status who makes that claim has simple demonstrated that they are incompetent, or lazy (didn't do their due diligence), or a liar (relying on their audience not knowing any better and not doing their due diligence).
 
Sean said:
PACE was originally described as 'controlled' in the PACE protocol paper (2007), but not in the main paper (2011).
Click to expand...
Yes, I didn't want to laden my post with too much. I suspect there must have been rumblings between 2007 and 2011 that brought them up short. I've a vague recollection a reviewer maybe brought the issue up.

Which makes you think: If a £5M trial was funded on the basis of its protocol saying it would be a controlled trial, but was in fact implemented without being controlled, is that not in itself a gross misuse of public funds? The funding was for a controlled trial. Would PACE have got the funding if they had described it more accurately from the outset, as a "randomised trial" rather than a "controlled randomised trial"? It sounds perilously close to gaining financial advantage by deception.

@dave30th: Could there be any mileage in this latter point at all?
 
You must log in or register to reply here.
Back
Top Bottom