Blog | Wednesday, November 12, 2008

Can you trust the medical literature?

Just went to a session on this topic led by Joseph Ross, MD, the author of an April 2008 JAMA study on ghostwriting. He said research indicates that editors prefer studies with positive findings, that studies from U.S. and/or "prestigious" institutions are more likely to be accepted, and that industry-sponsored trials are accepted more often by journals.

In cardiovascular research, studies are 37% more likely to favor newer treatments than existing treatments. In general research, published studies are two times more likely to be associated with pro- than anti- drug industry findings, he said. How does this happen? Some of the culprits include:
--Industry designs studies in such a way as to bias them toward favorable findings
--Industry analyzes results in favorable ways, and displays/states results misleadingly
--Industry suppresses unfavorable findings, choosing to submit positive trials for publication
--Industry supports ghostwriting, in which a reputable person is often brought in at the last minute to edit a study rather than participate in it or its analysis, yet is listed as the lead author.

How to fix the problem? Dr. Ross suggests:
--Prospective public registration of all clinical trials
--Public reporting of all safety/efficacy outcomes
--Improve integrity of study authorship. For one, the principal investigator should be listed as the lead author on studies.