| NEW YORK
NEW YORK (Reuters Health) - Economic ties that could bias drug trials and patient care might remain hidden due to tangled disclosure rules at medical journals, a new study reveals.
Researchers found that of 131 cancer journals, only 112 had policies requiring researchers to state conflicts of interest, such as drugmaker stock ownership or speaker fees. And among journals that did have such policies, the rules were all over the map.
"Journals can't even agree on what a conflict of interest means," said Dr. Aaron S. Kesselheim of the Harvard Medical School in Boston. "It is certainly confusing to authors and to readers."
Scores of studies have shown that when researchers have a financial stake in their work, their reports are more likely to promote drugs and downplay side effects.
Medical journals are the main line of communication between doctors and form the basis of treatment guidelines published by the government and medical groups.
Ultimately, the research trickles down to patients, dictating the drugs or medical devices they get from their health provider.
"Physicians -- like other professionals -- are influenced by incentives, especially financial incentives," said Kesselheim. "Conflicts of interest and financial relationships can have an impact on the research process and on the reporting of research, and they can also have an impact on physician behavior."
But the fact that a study involves a financial conflict of interest doesn't mean it shouldn't be published, because the research can be solid and important, he added.
"The solution that a lot of journals are embracing now is this idea of transparency," explained Kesselheim, whose findings appear in the journal Cancer. "What a lot of journals are saying is, we will publish this research, but we will publish it with a disclosure."
That will put readers on their toes when the read the report, and alert them to potential biases.
But researchers looking at journals in cardiology and pediatrics have found that disclosure policies are anything but standardized, as has one recent investigation by Reuters Health (see reut.rs/khoB5a).
So Kesselheim and his colleagues surveyed all cancer journals, noting the publications' disclosure policies as of late 2009.
Nearly one in eight of the 131 they found didn't have disclosure requirements, and half of those that did had no clear definition of conflict of interest.
Mostly, journals defined conflicts of interest as ties to a company that could benefit from the research findings, or relationships that could bias scientists' conclusions, influence their integrity or embarrass them.
Kesselheim and his colleagues also went through more than 1,700 reports from high-impact journals. Of the 27 publications that included editorials, which are particularly vulnerable to bias, just 14 had disclosures.
Some journals promised they would keep disclosures confidential, while other said they would penalize researchers who failed to reveal conflicts of interest.
"There is no consistency in the oncology literature about how conflicts of interest are defined and how they are reported," said Kesselheim. "It is certainly a strange way of trying to achieve transparency."
Still, compared with just a decade ago, the field has come a long way, Kesselheim added.
"There has been a lot of progress," he said, "but there is still a ways to go."
SOURCE: bit.ly/pu0nOW Cancer, online June 29, 2011.