News

Drug Companies Seem to Get the Results They Want in Trials, Study Finds

June 05, 2007

The results of studies that compare two competing drugs depend largely on who sponsored each study, according to an analysis of 192 clinical trials of statin drugs appearing in the June issue of PLoS Medicine.

Previous analyses of industry-financed drug trials had found that they were more likely than those with no industry ties to favor experimental drugs over placebos. The new paper was one of the first to look at trials that compared two drugs head-on.

Lisa A. Bero, a professor of clinical pharmacy and health-policy studies at the University of California at San Francisco, led a team that looked at all the randomized, controlled trials published from 1999 to 2005 that compared two drugs, one or both of which were statins, the cholesterol-lowering drugs that form a market worth more than $20-billion. (Of those 192 studies, 136 were led by academic researchers.)

Dr. Bero’s team found that the results and conclusions of the trials favored, by wide margins, the statin made by the sponsor of the trial. Sponsorship by the company whose drug was being tested meant results 20 times as likely, and conclusions 35 times as likely, to favor the test drug. “This bias in drug-drug comparison trials should be considered when making health-policy decisions regarding drug choice, such as drug formulary decisions,” the researchers write. —Lila Guterman