Industry conflicts common in cancer studies
■ Journal articles published by authors with financial conflicts are more likely to have positive outcomes, a new study shows.
- WITH THIS STORY:
- » Conflicts in cancer research
Nearly a third of oncology studies published in influential journals are authored by researchers with financial conflicts of interest, according to a May 11 study published online by the journal Cancer.
Researchers examined more than 1,500 studies published in 2006 in high-impact journals such as Cancer, The Lancet Oncology, and The New England Journal of Medicine, noting when authors disclosed conflicts of interest.
The review found some kind of financial conflict, such as industry funding of research, consulting income or stock ownership, in nearly half the articles presenting the results of prospective clinical trials.
These conflicts were associated with industry-friendly findings, the study said. Randomized-trial studies with author conflicts were twice as likely to report statistically significant positive overall survival outcomes from the treatment under study.
"The frequency of conflicts of interest we observed is concerning, especially because the presence of these conflicts appears to influence study outcomes," said Reshma Jagsi, MD, PhD, the study's lead author and assistant professor of radiation oncology at the University of Michigan Medical School.
"Disclosing conflicts of interest is an important first step, but it is probably not sufficient," Dr. Jagsi said. "We need to direct our efforts toward disentangling research from industry ties. We can reduce the reliance of researchers on industry ties by providing other sources of funding."
A shift of that sort appears to be under way, with the National Institutes of Health -- the government's leading research grant-making agency -- receiving $10 billion in extra funding in the federal stimulus package.
Oncology not alone on conflicts
The study's results are similar to earlier findings not limited to oncology. For example, a Jan. 22, 2003, article in The Journal of the American Medical Association systematically reviewed 1,140 studies and found that about 25% of biomedical researchers had industry ties.
"The fact that there are extensive conflicts of interest -- this is pretty much the case in most disciplines," said Jerome P. Kassirer, MD, professor of medicine at Tufts University School of Medicine in Massachusetts and former NEJM editor-in-chief. "It just hasn't been studied particularly as extensively as it has [in this study]."
Dr. Jagsi said her study likely underestimates the frequency of conflicts because her research team largely relied on authors' disclosures and did not do additional investigation to uncover undisclosed industry ties.
Now that medical journals routinely ask authors to disclose any potential conflicts, Dr. Jagsi said, editors should apply extra scrutiny to studies with conflicts, perhaps requiring an additional review of figures by a staff biostatistician.
That was an idea that Dr. Kassirer, a fierce critic of cozy physician-industry relationships, rejected.
"You ought to be scrutinizing every study the same way," he said. "You ought to look at the statistics, look at study design, see how many patients were dropped from the study. You have to do the hard work as an editor and evaluate every single paper."
Another expert derided the study's findings. Simply tallying the percentage of cancer studies with conflicts disclosed without regard to the effect on patients is unhelpful, said Thomas P. Stossel, MD, professor of medicine at Harvard Medical School in Boston.
"Without patient outcomes data, the Cancer article is meaningless -- as are most papers of this sort," said Dr. Stossel, who directs the division of translational medicine at Brigham and Women's Hospital, also in Boston.
Dr. Stossel said pursuing conflicts of interest is a "waste of time" because the evidence that they harm patients is overblown. If there is a problem with a particular study's design, it should be disputed on the merits, not by pointing to industry ties, he said.
"If you set these things into the context of the number of interactions between industry and physicians and industry and clinical investigators, the relationship is overwhelmingly positive," Dr. Stossel said.