Dartmouth Medical School

For Release: March 17, 2006
Contact: MedNews Office

Researchers Advocate Caution when Disseminating and Adopting Treatment Based on Preliminary Medical Research

HANOVER, NH—As more media and practicing physicians are reporting and acting on preliminary medical results presented at scientific meetings, researchers at Dartmouth Medical School and the VA Hospital are urging the media and medical community to treat this data with caution. Citing a case in which patients were harmed due to the presumption that newer treatments are better than older ones, the article outlines ways to improve the dissemination of these preliminary medical results and ensure that patient safety remains the top priority.

Writing in an editorial in the March 15 issue of the Journal of the National Cancer Institute, Drs. Steven Woloshin and Lisa Schwartz, members of the VA Outcomes Group in White River Junction, Vermont, describe the details surrounding the approval of a lung cancer drug called Iressa based on promising preliminary results from a small, uncontrolled study. Despite concerns of fatal side effects and the fact that a similar tested drug was already on the market, the FDA approved its use through an accelerated process. The FDA later pulled the drug after Iressa did not improve survival in larger randomized trial of 1700 participants. But in the two years the questionable drug was on the market, it was prescribed to more than 200,000 patients who could have been using a safer, more effective alternatives.

"There is a strong push to get the results of research out to the public and medical community as quickly as possible and we think the question that needs to be asked is "what is the rush?" said Woloshin.

The authors acknowledge that when there is no comparable treatment on the market and patients have run out of alternatives, speed has its virtues. But when preliminary findings turn out to be false, "patients get hurt: they are exposed to ineffective or harmful treatments, or they forego good alternatives," they write.

These problems have led Woloshin and Schwartz, associate professors of medicine at Dartmouth Medical School, to offer suggestions on how to improve the dissemination of preliminary results. Since preliminary results are often presented at scientific meetings and conferences, the authors believe that meeting organizers could help doctors make more informed decisions by posting entire presentations on meetings' websites, not just the abstracts. Organizers should also establish the expectation that major presentations be published at (or near) the time of the meeting, write the authors, who point out that new treatments are not fully investigated until published in a peer-reviewed journal, sometimes years later. And there are cases when the preliminary data is refuted and never published at all, add Woloshin and Schwartz.

They also believe that researchers presenting early data should also include a caution or "limitation" header to remind reporters that findings presented at scientific meetings may change substantially by the time they are published in a peer-reviewed journal. "By encouraging journalists to insert more cautions about the preliminary nature of the research and the weakness of the science, this preliminary results would be publicized more responsibly," said Schwartz.

Regardless what happens on the dissemination side, physicians have a responsibility to their patients to approach preliminary research cautiously, write Woloshin and Schwartz. This would entail making sure that benefits outweigh possible harms and that safer, possibly more effective alternatives do not already exist for their patients.

"The underlying message of our editorial is that it should be a more considered decision about whether to broadly disseminate or adopt new research findings when they are often in preliminary stages and have not been properly vetted to ensure there an actual improvement over standard of care," said Schwartz.

The editorial was written in response to a study in the same issue of JNCI in which Sharon Giordano, et al investigated the substantially increased use of chemotherapy drugs called taxanes after preliminary data was shared at a scientific meeting 5 years before final, peer-reviewed research was available.