Conflict Rules at Journals Introduce Bias

Related articles

Two of the most difficult issues we as consumers of health information confront are who to trust and how much to trust them. There is a lot of misinformation on the Web and even reporters in the mainstream media sometimes get the story wrong.

The few sources of health news that have not received much criticism are respected medical journals, such as The New England Journal of Medicine and The Journal of the American Medical Association. These journals do everything they can to protect their reputations. But in doing so, might they go too far, to their own detriment?

The editors of The New England Journal of Medicine cried "mea culpa" last week, asking forgiveness for what top editor Dr. Marcia Angell enigmatically referred to as "the most serious mistake for which we have had to apologize."

The transgression? The editors claimed they had violated their own publication's policy 19 times over the past three years by permitting researchers with potential financial conflicts of interest to write review articles on the safety and efficacy of pharmaceuticals.

The apology and reaffirmation of standards was apparently meant to keep science free of bias and improve the quality and credibility of the research published in one of the nation's most prestigious medical journals. Unfortunately, however, this "confession" only highlighted an ill-advised policy that consigns research to being judged not on its scientific merits, but rather on whether a particular type of purported conflict of interest contaminated the quality of the work.

The Journal's policy requires it to select for its pharmaceutical review columns only authors who have no recent history of research grants, consultant fees or any other financial exchange with companies who manufacture the drugs being reviewed. The policy applies not only to the funding of the researcher in question (here the Journal adhered to its policy), but also to the funding of the institution that employs the researcher. This was not enforced and was the focus of the apology.

Thus the Journal is chastising itself for failing to disqualify authors whose department or medical school, as opposed to the researcher himself or herself, had at some point accepted corporate research money. In fact, requiring such financial disclosure does not promote good science nor does it serve to eliminate all or most sources of bias.

First, there is the question of financial conflict of interest at the New England Journal of Medicine itself. The vast majority of the institution's estimated $20 million annual budget comes frompharmaceutical company advertisements. How, without being "holier than thou," can the editors disqualify authors on the basis of financial ties when their own salaries are paid by the same companies?

Second, the term "conflict of interest" encompasses far more than the potential bias linked to corporate-sponsored research. If Researcher A is required to disclose that he has consulted for Pfizer and is thus disqualified as an author, why should Researcher B, whose stock profile may be 80 percent Pfizer, be considered unbiased and acceptable? Is the next step in "author disclosure" going to be a tell-all of personal and family members' investment portfolios?

Where does the attempted exorcism of conflict of interest for researchers stop? And why is its scope restricted to financially driven bias? In a l993 commentary titled "Conflict of Interest: The New McCarthyism in Science" published in JAMA, Dr. Kenneth Rothman presented a stinging critique of the financial disclosure requirements of scientific journals, noting that many potential biases other than financial ones abound.

Should authors who write about homosexuality and AIDS be required to reveal their sexual preference? This stance was once discussed in the mainstream. Should researchers assessing the health benefits and risks of alcohol be required to divulge if they are Mormons? Should researchers funded by ideologically fueled foundations that are hostile to profit-making enterprises of all kinds be considered suitable as authors for assessing the value of pharmaceuticals? As Rothman wisely notes, "Like McCarthyism in the 1950's, this frenzy of credential checking is apt to get way out of hand."

Third, policies that exclude as authors any and all who have even the remotest form of funding from industry leave a highly self-selected group of potential authors in the pool. Perhaps those who remain eligible because they had no financial ties with drug companies were not selected as consultants because of an inferior academic record. Or perhaps they were passed over because they had an obvious antipathy to what they view as obscene profits of the pharmaceutical industry. Are these scientists free of conflict of interest?

What is ultimately most insulting and demeaning about the conflict of interest rules of financial disclosure is that they, at least subtly, discredit the purged authors without giving them a chance at self-defense. They also disparage the integrity of the time-honored scientific peer review process in which a work is submitted to a panel of peers who judge the submission -- usually anonymously -- not by its funding, but by its merits.

If medical research is funded by Merck, Lily or Zeneca and is approved in an independent peer review process at the New England Journal of Medicine or any other journal, how might any information about potential funding relationship help the reader analyze the study on its merits?

Bias stems from many sources other than financial ties. The integrity of science relies on the rigors of the scientific process. The idealistic but impractical purge of corporate-funded research is no substitute for scientific scrutiny. If only it were that easy.