Registering Clinical Trials Doesn't Go Far Enough

The New York Times on Saturday editorialized in favor of mandatory registration of clinical trials that test drugs. The industry has agreed to a voluntary registration system in the wake of revelations that drug makers failed to reveal that most antidepressants A) didn't work in kids; and B) led a few kids to attempt suicide.

This week, a Food and Drug Administration advisory committee will consider whether the U.S. should ban the use of most antidepressants in kids, like they've done in Great Britain. I'm guessing they will make a half-hearted gesture in the direction of greater scrutiny -- perhaps adding warning to physicians to keep close tabs on their teenage patients for the first few weeks after prescribing drugs like Paxil and Zoloft.

This weak response to the potential dangers of these marginally effective drugs (the trials revealed that the difference in the recovery rate between depressed kids on placebo and depressed kids on drugs is virtually indistinguishable) shows why clinical trial registration -- the reform du jour -- doesn't get at the root problem of the drug testing system in the U.S.

As long as the drug industry is sponsoring and conducting clinical trials, it's not likely that the public will ever get straight information about the usefulness of drugs. Before the 1980s, independent academic investigators conducted most clinical trials. Today, industry funds nearly two-thirds of all trials. The physicians who enroll patients in these trials not only get financial support for the tests, but often wind up landing lucrative consulting and speaking gigs with the firms when the trials turn out positive.

It's a structural bias that skews the information supplied to the public, an effect that has been well documented in the medical literature. Trials funded by industry are much more likely to turn up positive results than trials funded by independent investigators.

I tested this theory on my own last February. A colleague at the Center for Science in the Public Interest and I conducted a study that pulled the two dozen placebo-controlled clinical trials on the use of antidepressants in kids that had been published in the past decade in the medical literature. We separated the studies into those funded by industry and those funded by independent investigators and compared the results.

Better than 90 percent of the industry-funded trials showed the drugs were "safe and effective" in children. Only five of nine trials conducted by independent researchers came to the same conclusion.

And when one looked at the dozen trials submitted to the FDA by industry (but never published) to get patent extensions for these drugs, only three of 12 had positive results and at least one of those positives was in fact an after-the-fact combination of two failed trials into a single data set. That enabled the industry-funded researcher to statistically show marginal effectiveness.

The FDA and industry kept those results hidden from the public for several years. Only after Britain banned the use of the drugs in kids and the FDA began considering the question here did that information finally become available. A registry would have made that information known much sooner.

But it wouldn't have changed the fact that industry-funded researchers still conducted most of the trials. These researchers got to pick the patients; they got to judge the effectiveness of the drugs (for antidepressants, this can be quite subjective); they got to interpret the data. Even in situations where the clinicians were double-blinded (they don't know which patients are on drug or placebo nor do the patients), the subtle bias generated by their desire to see their sponsor's drug work could have influenced all those decisions and ultimately the conclusions drawn from the trial.

The only way to eliminate this bias entirely is to have totally independent researchers conduct clinical trials. Former New England Journal of Medicine editor Marcia Angell calls for this reform in her new book, The Truth about the Drug Companies, which is getting widespread attention now. It's a reform that is long overdue. Only then will we get a truly objective view of the value of new drugs.

Posted by gooznews at September 12, 2004 05:03 PM