Bias in Drug-Company Studies?
Industry-Funded Drug Studies More Likely to Find Drugs Work
June 2, 2003 -- When a drug company pays for a study of its product, it's four times more likely to be happy with the findings than when the study has independent funding.
That's the main finding from an analysis of 30 studies that compared drug-company-funded trials with trials with other sources of funding. Led by Joel Lexchin, MD, of York University in Toronto, it appears in the May 31 issue of the British Medical Journal. This special issue carries several articles and editorials alleging that drug companies have too much influence on medical research and doctors' practices.
Do they? Drug companies pay for more than half of all clinical research. They give free samples of their drugs to doctors and encourage them to prescribe them. They pay doctors handsome speakers fees; fly them to free seminars at company-sponsored educational meetings; offer doctors gifts, free meals, and more. They also sponsor seemingly independent patient groups that advocate use of the sponsors' products.
There's good and bad in all of these things. The strongest charge, however, is that drug-company research is somehow tainted. The Lexchin study does not show this, but it does raise some disturbing points.
"We found that if drug companies fund a trial of a product they make, it is four times more likely the trial will turn out to be favorable to that product than if the trial were funded by someone else," Lexchin tells WebMD. "That conclusion applies to a wide range of different drugs: Medicines for arthritis, psychiatric drugs, birth control pills, and others. And it applies across a wide range of diseases."
Fraud? No. There was no hint that drug-company-sponsored trials were poorly done.
"We actually were surprised that the methodology of drug-company-funded research was as good -- or in some cases better -- than trials getting other sources of funding," Lexchin says.
The problem, he suggests, is two-fold. Drug-company-funded studies with negative results are less likely to be published. And industry-funded trials are more likely than other trials to compare the product with placebo, with another less-active drug -- one not usually used as a first-line treatment -- or with the wrong dose of an active drug.