Delays That Can Cause Death

By Anastasia Toufexis.

Dr. Anne Willoughby had a dilemma. She is the director of a study begun in 1988 to see if infusions of immunoglobulin (IG) in children infected with the AIDS virus could boost their weak immune systems and help stave off illness. An independent panel that had been monitoring the investigation informed Willoughby early this month that children receiving monthly doses of IG, a protein produced by the body's cells, were faring significantly better than those given a placebo. The youngsters had fewer bacterial infections and fewer hospitalizations. News of the advance in treatment could have enormous consequences for up to 20,000 children in the U.S. who carry the virus and the 1,700 to 2,000 HIV-infected babies who are born each year. But how and when should the information be released?

That question is at the heart of a growing debate among scientists, 200 of whom gathered this month at the National Institutes of Health in Bethesda, Md., to air the issue. Traditionally, researchers have published their findings in medical journals, like the New England Journal of Medicine or the Journal of the American Medical Association. Supporters insist that this approach helps ensure credibility by allowing time to weed out sloppy science. Editors send submitted papers to outside scientists for comment, a system known as peer review. But the process can be lengthy. Between the time a paper is first offered and its publication, months or even more than a year can pass. Critics charge that such lapses can cause critical delays in the application of new treatments and may even cost patients their lives. They favor speedier release of medical news, through mailed bulletins to physicians or general press conferences.

Defenders of the publish-first route maintain that the peer-review system, introduced in England in 1665, is a crucial quality-control mechanism. "It avoids mistakes," declares Dr. Arnold Relman, editor of the N.E.J.M. "It minimizes the hype, exaggeration and biases and prevents the dissemination of premature and unwarranted conclusions. Short-circuiting that process is not a good idea."

Many doctors agree. They complain that patients sometimes hear about developments through the press and come into medical offices waving a news clipping and demanding the latest treatment. Physicians say that without seeing a complete scientific report, they have no way of assessing the information or knowing whether it is appropriate for a particular patient.

Opponents, however, contend that major studies today receive intense independent scrutiny while they are under way, and thus the journal peer- review system may not be as necessary as before. Embargoing medical news until publication, they say, may actually put patients in greater jeopardy than allowing information to be released quickly. Though journal editors insist that the review process can be speeded up for urgent papers, they concede that sometimes an important report has gone unrecognized. For example, Yale University researchers sent a paper to the N.E.J.M. in October 1989 noting that patients with spinal-cord injuries who were treated with high doses of steroids while in the emergency room had much better recoveries than patients who did not receive the drugs. The impact of the discovery is enormous: about 10,000 people a year suffer such spinal-cord injuries. But the paper was not published until seven months later.

Moreover, some critics charge that medical editors' defense of the status quo is aimed more at safeguarding journal profits than at assuring the spread of accurate information. The N.E.J.M. is notorious for refusing to publish papers if the findings have already received publicity in the press. Editor Relman defends the practice: "We share every publisher's desire to be interesting and original. What's wrong with that?" What's wrong is that the policy impedes the free flow of scientific information by intimidating researchers into not talking about their work before publication. Having a paper appear in the N.E.J.M. can enhance chances of winning new grants or help push a faculty member toward gaining tenure.

Still, scientists increasingly appear to be finding it difficult to justify withholding important information. Three years ago, researchers from the National Cancer Institute discovered that women with a certain type of breast cancer who received aggressive drug therapy had less chance of having a quick relapse than those who received standard treatment. Dr. Vincent DeVita, then director of the NCI, decided not to wait for publication. He issued a clinical alert. Says DeVita: "By the time that article was published, 50,000 women could have passed through that window of opportunity."

For Willoughby too, the choice in the end proved simple. After consultation with colleagues, officials at the NIH and the Food and Drug Administration, a decision was made to release the results of the study of immunoglobulin in children -- at a press conference and in overnight letters sent to physicians. Says Willoughby: "Once we were certain that the results were sound, there really was no alternative."

With reporting by Dick Thompson/Washington