Saturday, September 29, 2007

Conflicting views about disclosure

Medical journals are becoming increasingly strict in their disclosure requirements concerning authors’ financial conflicts of interest. Dr. Jerome Kassirer addresses the problem in a recent MedGenMed Webcast Video Editorial. He correctly points out that disclosure, while necessary, does not “fix” the problem of potential bias in practice guidelines and journal reviews. But the focus of this editorial is inappropriately narrow, singling out financial ties with drug companies while ignoring multiple other important conflicts.

Concerning disclosure, Kassirer states:

Disclosure may alert readers of possible bias, but it requires them to become mind readers. Did the author pen an unvarnished manuscript? Were his opinions subtly influenced by consulting or speakers' fees? Did he intentionally bias the material to satisfy the company whose fees help to pay for his daughter's college tuition? Interpreting the opinions and recommendations of financially conflicted authors in editorials and review articles is more akin to reading a mystery novel than reading a scientific paper.

Mind readers? Absurd as that seems some of the harshest cynics seem to think they can read authors’ minds. For the rest of us, disclosure alerts to possible bias and suggests the direction of such bias. While this doesn’t require mind reading it prompts discerning readers to look critically at the authors’ recommendations and to check primary sources. Those are skills doctors learned, or should have learned, in training.

Kassirer concludes with a call to go beyond disclosure and eliminate conflicts by finding “unconflicted experts”. But isn’t that an oxymoron? Anyone with expertise in a field is likely to have a personal stake. As Thomas Stossel recently pointed out, excluding all experts with potential personal or commercial interest risks leaving us with “the second best and not so bright”. The New England Journal of Medicine realized this in 2002 when editors found they had to relax their conflict of interest policy in order to find qualified experts.

The focus on drug company ties represents selective outrage in the debate over conflicts of interest. Drug company influence is only one of many conflicts, financial and otherwise, which may bias reviews and practice guidelines. A fair and intellectually honest approach to the issue should question all conflicts of interest that might impact conclusions and recommendations drawn from medical research. Should gastroenterologists, for example, who might gain financially from doing procedures, write guidelines on the indications for endoscopy and who should perform it? Should invasive cardiologists participate in the writing of guidelines for cardiac catheterization and stent implantation? As this list goes on and on one has to wonder how many “unconflicted experts” even exist.

1 comment:

Paige Hatcher said...

I can't find the citation (I heard the stat at a conference), maybe you can, but an interesting recent study actually claimed that disclosing a conflict of interest made authors/speakers seem MORE credible.