(Page 2 of 2)

Some critics, however, have argued that medical journals themselves have been a part of the problem. A growing number of studies in recent years have shown that journals publish more trials with positive results than those with negative or inconclusive ones. And critics say the journals have moved too slowly to address such issues.

Dr. Catherine D. DeAngelis, the editor of The Journal of the American Medical Association, said the idea of requiring trial registration had been kicking around among editors for about a decade. She said the issue came up again during a discussion at a meeting earlier this month of the International Committee of Medical Journal Editors, partly out of frustration.

"We have tried editorials," said Dr. DeAngelis. "We tried getting the pharmaceutical companies to do it. We tried talking to leaders in government. But it hasn't happened."

While Dr. Drazen and Dr. DeAngelis said their group was likely to decide over the coming months what course to follow, it is not clear how the drug industry will react. Last week, Merck said it would support the idea of a government-run test registry. And GlaxoSmithKline said it would soon begin posting on its company Web site the trial results of all its drugs on the market, including tests for potential new uses of them.

Some other companies and the drug industry's trade group, the Pharmaceutical Research and Manufacturers Association of America, said last week that they could not comment because they had not seen specific registry proposals. But one official of the trade group raised concerns that registries could release company trade secrets or present data in ways confusing to doctors and the public.

Whatever the case, the example of the little-known test of Celexa in adolescents shows how medical journals can now miss information about a major trial of a drug that is the subject of an article.

Dr. Nancy C. Andreasen, the editor of The American Journal of Psychiatry, which is the flagship publication of American Psychiatric Association, said it was the responsibility of a study's authors to provide a scholarly overview of the published articles discussed in their paper. She said that her publication did not specifically ask authors or companies that sponsor trials about unpublished studies.

"We didn't have a checklist that includes that question," Dr. Andreasen said. She added, though, that the publication regularly reviews its policies.

The Celexa trial in question was run in Europe from 1996 to 2002 and was sponsored by H. Lundbeck, the Danish company that developed the drug.

Forest Laboratories sells the drug, which is generically known as citalopram, in this country under a license with Lundbeck.

A spokesman for Lundbeck said the company reported the trial results to Forest, although he could not say when. Forest executives did not respond to written inquiries from The Times seeking that information.

But Forest executives apparently had an opportunity to know about the European test before the publication of the positive trial's results this month in The American Journal of Psychiatry. Forest executives said they presented safety data concerning potential suicide risk from both the positive study and the European trial last fall at a medical conference. It was around that time that regulators in Britain and this country expressed concerns that several antidepressants might cause some depressed teenagers to consider suicide; the issue is still under study.

The Lundbeck spokesman said that an abstract about the European trial had been presented in April at a Swedish medical meeting, and both companies said that an article about that trial was being prepared for publication. Both companies also said that they did not promote the drug's use in children because regulators had not approved it for pediatric use. (Doctors can legally prescribe a drug for any use, once it has been approved for at least one purpose.)

Dr. Andreasen and other journal editors interviewed said that a single failed trial of a drug did not mean that the treatment was ineffective, because the study's design might have been flawed. By the same token, of course, a single positive test of a drug does not necessarily mean that it works.

In a Lancet article in April, British researchers sought to compare the benefits and risks that widely used antidepressants pose for children and adolescents, based on published and unpublished data. They reported that their analysis of the pooled results from two unpublished Celexa trials - the one since published in The American Journal of Psychiatry and the European study cited in the Danish textbook - suggested that citalopram was unlikely to produce a "clinically important reduction in depressive symptoms."

"With no good evidence for efficacy and the potential for increasing the risk for suicide, the risk-benefit balance is unfavorable," the researchers reported.

Dr. Karen Dineen Wagner of the University of Texas Medical Branch in Galveston, who was the lead outside investigator on the study published in The American Journal of Psychiatry, did not respond to interview requests through a hospital spokeswoman. The two other outside researchers involved, however, both said that Celexa worked well in their test and that the young patients did not experience increased suicidal thoughts.

"I don't know what the raw data looks like from the European study," said one of them, Dr. Adelaide S. Robb of the Children's National Medical Center in Washington.

She said that she was informed by Forest executives in 1999 that the European study was under way but that she was never told that it had been completed.

Download the Print Edition of The New York Times to Your PC.

<<Previous | 1 | 2