Wednesday, April 15, 2009

The JAMA Two contd. - Douglas Melnick and Adriane Fugh-Berman write:

Editing Ethics: JAMA’s New Conflict of Interest Policy(Media) Permanent link

The role of a medical journal should be to publish relevant, worthy, peer-reviewed articles, and the appraisal of what is relevant and worthy must include determining authors’ conflicts of interest. Two editors at JAMA, however, appear to think their role includes the repression of public information and the silencing of concerned readers who expose conflicts of interest.

Our story, to date: A study published in JAMA found that Lexapro, compared to placebo, lowered the rate of depression in stroke patients. Robert Robinson, the lead author, was quoted widely in the media extolling the virtues of antidepressants for stroke victims. In a letter to JAMA, researchers Jonathan Leo and Jeffrey Lacasse noted that Robinson’s study also found psychotherapy superior to placebo but that the paper contained no direct statistical comparison between Lexapro and therapy. In a response, Robinson admitted that the drug and psychotherapy were equivalent. Subsequently, in an unimpressive bit of investigative reporting consisting of a cursory Google search, Leo and Lacasse discovered that Robinson had been on the speaker’s bureau of Forest, the manufacturer of Lexapro, and had not disclosed this. Leo informed JAMA of this undisclosed conflict of interest in October 2008. JAMA stated that it would investigate, and did not communicate further with Leo.

Leo and Lacasse then wrote a notably dispassionate recounting of the story, which was accepted for publication by the BMJ. Leo called and e-mailed JAMA editors to let them know about the impending publication. JAMA ignored him. Months later, in March 2009, the commentary was published in the BMJ’s Rapid Response section, triggering JAMA’s own rapid response. JAMA editors demanded a retraction of the BMJ article and apparently attempted to intimidate both Leo and Leo’s dean, Ray Stowers, who told the Wall Street Journal that DeAngelis threatened to “ruin the reputation of our medical school.”

Days after the BMJ publication, JAMA ran a letter and correction clarifying Robinson’s and a coauthor’s conflicts of interest. In a move that must have press officers everywhere cringing, Catherine DeAngelis, JAMA’s editor in chief, apparently told an enquiring WSJ reporter that Leo was “a nobody and a nothing.” DeAngelis has since denied that quote, but her feelings are clear in the denial, which concedes, grudgingly, that Leo is a “somebody doing something very important.” Even if the WSJ got the original quote wrong, which we doubt, the paper trail left by the editors is damning. The Wall Street Journal cited an e-mail from DeAngelis to Stowers that reads: “As I’ve already expressed to you, I don’t want to make trouble for your school, but I cannot allow Jonathan Leo to continue to seek media coverage without my responding. I trust you have already or soon will speak with him and alert me to what I should expect.”

DeAngelis and fellow editor Phil Fontanarosa then wrote an editorial that provides a startlingly detailed description of unacceptable behavior. Under the flag of protecting the reputation of the journal, the editors admit to trying to dissuade Leo from sharing his information. They admit to trying to get Leo in trouble with his dean. They admit to threatening to ban Leo from the pages of JAMA. They even admit to being "strong and emphatic" in their bullying efforts.

Revealing exactly what happened in excruciating detail makes little sense when one is completely in the wrong. Instead of apologizing, the editors go on the attack, accusing Leo of a “serious ethical breach of confidentiality” for going public with public information. Even more astoundingly, the editorial announced a new policy at JAMA that seems more suited to a totalitarian country than to a journal published in a country where free speech rights are part of the Constitution. The new policy would “require that the individual bringing the allegations provide a written detailed explanation of the unreported conflicts of interest and provide documentation to support the allegation. The person bringing the allegation will be specifically informed that he/she should not reveal this information to third parties or the media while the investigation is under way…”

The editors’ idiosyncratic perceptions of their proper role risks destroying rather than salvaging the journal's reputation. It’s a relief that JAMA appears to be operating in a unique universe; editors of other journals have politely distanced themselves. In interviews with the Wall Street Journal, Fiona Godlee, editor in chief of the BMJ, said, “No one group or organization should have a monopoly on investigating a piece of work”, and Jerome Kassirer, former editor of the New England Journal of Medicine, said that Dr. Leo “did nothing wrong.”

The weighty public responsibility borne by medical editors does not include silencing physicians who have facts to share that the editors prefer remain out of the public eye. The underlying arrogance reflected in JAMA’s edict is disturbing. Influential medical journals shape physicians’ perceptions--and the very practice of medicine--by what they publish. Medical journals should provide a vigorous free forum for discussion, and medical editors should view themselves as the servants rather than the tyrants of the medical community.

It is as unethical for a medical journal editor as for a pharmaceutical company executive to attempt to prevent an independent researcher from revealing facts that affect the interpretation of medical information, and ultimately patient care. If the journal cannot be bothered to perform the simplest fact checking on the conflicts of its authors, it should be grateful that the public is willing to perform this role. Certainly authors who submit manuscripts for publication are responsible for stating their own conflicts, but journals should take a more journalistic approach: Trust, but verify. Participation in a drug manufacturer's advisory board or speaker’s bureau is easy to find. And authors shouldn't be coddled if they fail to state a significant conflict of interest. It’s high time for journals to get serious about punishing authors who lie.

Douglas Melnick, M.D., is a preventive medicine physician in the Los Angeles County Department of Public Health. Adriane Fugh-Berman, M.D., is an associate professor at Georgetown University Medical Center.
Source

If you are into this, there's more here!

No comments: