James C Coyne blogpost on how serious he considers nondisclosure of COIs incl re nonpharms tx

[Tom Kindlon]

James C Coyne is an influential psychologist who is not afraid to criticise other psychological researchers.

Here's his latest blogpost (it's quite long)

BMC Medicine gets caught up in Triple P Parenting promoters’ war on critics and null findings
By James Coyne PhD
Posted:March 16, 2015

Undeclared conflicts of interest constitute scientific misconduct.

Why we should be as concerned about conflicts of interest of interest in evaluations of nonpharmacological treatments, like psychotherapy.

http://blogs.plos.org/mindthebrain/...g-promoters-war-on-critics-and-null-findings/

I'll post a few extracts.

 

[Tom Kindlon]

Until recently, promoters of the multimillion-dollar enterprise controlled perception of their brand of treatment. They authored most reports of implementations and also systematic reviews and meta-analyses. They did not report financial conflicts of interest and denied any conflict when explicitly queried.

Some of us are concerned that some similar issues could arise with non-pharmacological interventions like CBT and GET (although generally there is no patent with these so it's not exactly the same).

There were reports in social media of how studies with null findings have been previously sandbagged in anonymous peer review or how authors were pressured by peer reviewers to spin results. Evidence surfaced of 3P founder Matt Sanders attempting to influence the reporting of a supposedly independently conducted evaluation. It is unclear how frequently this occurs, but represents a weakening of the important distinction between independent evaluations and those with conflicts of interest.

Are we witnessing the decline effect in the evaluation of 3P? Applied to intervention studies, the term refers to the recurring pattern when weaker results accumulate from larger, more sophisticated studies not conducted by promoters of the intervention who initially had produced glowing reports of efficacy and effectiveness.

And spousal relationships are reportable conflicts of interest.

Why get upset about conflict of interests in evaluations of nonpharmacological trials and reviews?

My colleagues and I played a role in improving the tracking of conflicts of interest going from industry-supported clinical trials to inclusion in meta-analyses.Our criticism prompted Cochrane Collaboration to close a loophole in investigator conflict of interest not having been identified as a formal risk of bias. Prior to the change, results of an industry sponsored pharmacological trial could be entered into a meta-analysis where the origins were no longer apparent. The collaboration awarded us the Bill Silverman Awardfor pointing out the problem.

It’s no longer controversial that in the evaluation of pharmacological interventions involving financial conflicts of interest are associated with inflated claims for efficacy. But the issue is ignored in evaluating nonpharmacological interventions, like psychotherapies or social programs like 3P.

Undeclared conflicts of interest in nonpharmacological trials threaten the trustworthiness of the psychological literature.

Readers are almost never informed about conflicts of interest in the trials evaluating psychotherapy evaluations and their integration in meta-analyses. Yet, “investigator allegiance” a.k.a. undeclared conflict of interest is one of the most robust predictors of effect size. Indeed, knowing the allegiance of an investigator more reliably predicts the direction of results than the particular psychotherapy being evaluated.

Politicians will accept ineffective social programs if they are in the service of being able to claim that they are not just doing nothing, they are offering solutions. And the ineffective social programs are particularly attractive when they cost less than a serious effort to address the social problems.

If journals have a policy of commentaries not undergoing peer review, it should be indicated at each and every commentary that is the case. That would be consistent with COPE recommendations concerning non-peer-reviewed papers in journals identifying themselves as peer-reviewed.

 

[OverTheHills]

Politicians will accept ineffective social programs if they are in the service of being able to claim that they are not just doing nothing, they are offering solutions. And the ineffective social programs are particularly attractive when they cost less than a serious effort to address the social problems..

Sounds very familiar

 

[eafw]

Indeed, knowing the allegiance of an investigator more reliably predicts the direction of results than the particular psychotherapy being evaluated.

This is quite a significant statement, any further info on this (had a quick look and can't find anything) ?

Assuming it's correct it's another reason why people with a vested interested in promoting CBT should not be running, reviewing or propogating studies on it.

 

[Tom Kindlon]

Indeed, knowing the allegiance of an investigator more reliably predicts the direction of results than the particular psychotherapy being evaluated.

This is quite a significant statement, any further info on this (had a quick look and can't find anything) ?

Assuming it's correct it's another reason why people with a vested interested in promoting CBT should not be running, reviewing or propogating studies on it.

Yes, it's interesting.
There have been threads on this e.g.

Some papers on researcher bias/allegiance in psychiatry
http://forums.phoenixrising.me/inde...s-allegiance-in-psychiatry.22424/#post-418226