Three weeks ago, the editor of the prestigious Journal of the American Medical Association called Dr. Joseph B. Martin, dean of Harvard Medical School, and said, as she describes it, ``Joe, do something with your kids!"
Dr. Catherine D. DeAngelis, the editor, told the dean that in violation of the journal's policies, Harvard authors of three recent articles had failed to disclose relevant financial ties with drug companies.
Responding to DeAngelis' s concerns, Martin told her that he plans to send a letter laying out the conflict-of-interest disclosure requirements for JAMA and another top medical journal to all 8,000 members of the medical school's faculty, which encompasses most doctors practicing at more than a dozen Boston hospitals and research centers.
The Harvard cases and others like them are fueling mounting concern about potential conflicts of interest in medical research.
At issue is the danger that researchers who receive money from for-profit companies -- whether for speaking fees, consulting, or conducting drug trials -- may, consciously or unconsciously, be biased by that money.
While some critics say that scientists should accept no money from for-profit companies, DeAngelis and others contend that, at the very least, researchers must reveal any funding they have received that might conceivably influence their findings or recommendations, so that colleagues and patients can judge the authors' credibility.
JAMA requires authors to sign a statement listing all financial interests that might be perceived as influencing their article and has published these disclosures since 1990.
DeAngelis has undertaken a campaign to enforce her journal's disclosure requirements and, as part of that effort, posted an editorial online yesterday in which she said she has begun to publicize authors' failures to follow JAMA's policies.
In the editorial, DeAngelis singled out Harvard Medical School, saying that it was ``somewhat surprising" that three consecutive cases of nondisclosure brought to the journal's attention by authors or readers involved Harvard authors.
DeAngelis said in a phone interview yesterday that she believes that the Harvard authors' omissions stemmed from ignorance of the journal's policies, rather than malice.
She said Harvard simply has so many faculty members and hospitals that it is hard to get a message across to everyone.
Martin declined through a spokesman to comment, but confirmed that he planned to send out the 8,000 letters describing the disclosure policies of JAMA and the New England Journal of Medicine. He also planned to discuss with faculty whether Harvard Medical School needs to improve its rules on financial ties.
The conflict-of-interest issue comes up periodically, spokesman Don Gibbons said, and a large committee finished reexamining it a couple of years ago.
In the three-page editorial yesterday, DeAngelis points out that it is impossible for medical journals to check the financial ties of each of the thousands of authors whose papers are published every year, and there is no central, up-to-date database on the topic.
JAMA has clarified its rules on disclosure and, when omissions are pointed out, corrects them in print, she wrote. But different journals have different rules, so some confusion is likely, she said.
One of the JAMA papers that DeAngelis cited described a major five-year study that found that when women went off antidepressants during pregnancy, their depression was highly likely to recur.
The study, published in February, was funded by the federal government and included no drug company funding.
But a Tufts obstetrician, Dr. Adam C. Urato, wrote to JAMA to complain that its authors had received money from antidepressant manufacturers in the past. His letter, published last month, stated that because the study dealt in part with the question of stopping antidepressants during pregnancy, readers should be aware of the potential for prodrug bias.
The lead author, Dr. Lee S. Cohen of Massachusetts General Hospital, responded in the journal that the study did not address the question of whether ``antidepressants as a whole or a particular antidepressant" should be prescribed. He and his colleagues saw no potential conflict of interest.
But in retrospect, he wrote, they should have opted for ``utmost transparency," and they support the idea of full disclosure. He reported having received grant money, consulting fees, or speaking fees from a half-dozen drug makers, including
DeAngelis said she ``wouldn't blink twice" if Cohen and his colleagues submitted another article to JAMA: ``They are very good researchers; they're good people," she said.
A second paper, published in May, found that heart-disease research funded by for-profit companies was more likely to favor new treatments over the old standard of care than research funded by other sources.
The lead author, Dr. Paul Ridker of Brigham and Women's Hospital, later apologized, saying that while he had received no funding for the study, he had neglected to mention past consulting fees and grant money from a half-dozen drug companies.
The third Harvard-connected paper involved an apparent link between women with a particular kind of migraine and the risk of heart attack and stroke.
The study, led by Dr. Tobias Kurth of Brigham and Women's Hospital and published July 19, found that migraines with aura, visual disturbances like flashing and blind spots, seemed more closely linked to stroke and heart attack than typical migraines.
The study was funded by federal money and a foundation and did not look at treatment options, so the Kurth team reported no potential conflicts.
But DeAngelis wrote in JAMA that all past and present ties with the makers of drugs used to treat migraine or heart disease needed to be disclosed.
In a letter posted on the JAMA website, Kurth reported receiving research grants from three companies that make pain medication.
The lesson, Kurth said in a phone interview yesterday, is: ``It is best to report any relation to the [journal] editors and let the editors decide whether it's relevant or not."
Liz Kowalczyk of the Globe staff contributed to this report. Carey Goldberg can be reached at firstname.lastname@example.org.