Sorry, there are no polls available at the moment.
  • BIOETHICS FORUM ESSAY

The JAMA Imbroglio

It’s hard to know how to describe the current imbroglio over the actions of the JAMA. What started out looking like “please don’t shoot the messenger” has come to resemble “David and Goliath.”

The saga began with an academic scientist’s concern that researchers’ biases – born of their connections to pharmaceutical manufacturers–can skew their interpretation of clinical trial results in favor of drugs. Yet it has progressed to a highly publicized controversy and a request by the American Medical Association that JAMA’s oversight committee investigate threats allegedly made against the scientist by the senior editors of its flagship journal.

The incident has not only endangered the reputation that the journal has cultivated with its highly touted conflict-of-interest rules but opens it to criticism both for apparent abuse of power and for the folly of its newly announced policy, which forbids anyone who reports potential ethical breaches to JAMA from making such allegations public. This “whistleblowers’ gag rule” lacks ethical justification and will, if enforced, either discourage the reports of COI violations, belying JAMA’s professed interest in uncovering such problems, or more likely will drive them all into the general press in the first instance, which the journal most certainly doesn’t want.

Like many debacles, this one had a simple beginning. Last May, Robert G. Robinson, M.D., of Carver College of Medicine at the University of Iowa, and a number of colleagues published in JAMA the results of a randomized controlled trial aimed at determining whether, over the first year following acute stroke, treatment with escitalopram (Lexapro), a selective serotonin reuptake inhibitor approved for management of depression and anxiety, decreases the frequency of depression developing among stroke patients, compared with problem-solving therapy or a placebo. The authors concluded that the use of both active therapies resulted in a significantly lower incidence of depression over 12 months of treatment compared with a placebo, but that problem-solving therapy did not achieve significant results over placebo using the intention-to-treat conservative method of analysis. In press coverage of the article, Robinson made clear his conclusion that the off-label preventative use of antidepressants was advisable in all stroke patients who can tolerate the drugs.

In mid-October, JAMA published a number of letters about the article along with a reply from the authors. One letter, from Jeffrey Lacasse, Ph.D., of Arizona State University in Phoenix, and Jonathan Leo, Ph.D., of the DeBusk College of Osteopathic Medicine, a new school at Lincoln Memorial University in Harrogate, Tenn., pointed out that there was selective reporting of the results, in that the effects of problem-solving treatment and Lexapro had not been directly compared. Robinson and his coauthors acknowledged this failure and said their analysis showed no statistical difference between the drug and problem-solving therapy in preventing depression.

Given his general interest in COI in medicine, Leo was disturbed to find that Robinson had not declared that he had previously received funding from Forest Laboratories, the maker of Lexapro. As he later wrote, “The financial relationship with Forest Laboratories was well-documented and easily discoverable via a Google search, as evidenced by Dr. Robinson’s previous self-disclosures in varied sources.” Therefore, a day after JAMA had published his letter and Robinson’s reply, Leo reported the apparent COI to JAMA, which responded that it would investigate and correct the record as appropriate.

On Feb. 5, 2009, Fiona Godlee, the editor in chief of BMJ, a leading medical journal, published an introduction to five commentaries it had commissioned to coincide with the release of a report by the Royal College of Physicians on relations between industry, academia, and the National Health Service, in which Godlee invited readers to submit their own reactions. Leo and Lacasse submitted a commentary in which Robinson’s undisclosed conflict and apparent bias for drugs over equally effective psychotherapy was Exhibit A.

Rather than directly attack JAMA for failing to have uncovered the conflict, they focused on how researchers’ relations with drug companies “complicate the decision making process” for patients with mental disorders and their physicians by skewing the evidence-base on which they are reliant. Furthermore, they pointed out that subsequent corrections – such as the admission by Robinson et al. in JAMA five months after the original article – are unlikely to receive much media coverage and hence the equal efficacy of a non-drug therapy “will continue to remain a mystery to most of the news reading public.” They concluded that it has become a sign of naiveté to take “the conclusions of clinical trials at face value.”

The problem is not limited to just a couple of isolated cases but involves the entire culture of medicine that has developed over the past ten years. Put bluntly, the scientific machinery is broken. There is no easy fix, but surely patients deserve better.

Before publishing the BMJ commentary, Leo says he had repeatedly contacted the JAMA editorial staff to give them an opportunity to correct any factual errors, but he received no response. However, when he notified them on March 5 that his commentary was being published that day – and told them that his “initial skepticism” that JAMA would “set the record straight” about the Lexapro study seemed to have been “well founded” since the journal “had done nothing in almost five months” – the reaction was dramatic. JAMA’s editors immediately demanded that he withdraw the BMJ letter, though they failed to describe any factual errors that would justify such action.

Instead of reasons, JAMA’s executive deputy editor is alleged to have told Leo that he was “banned from JAMA for life. You will be sorry. Your school will be sorry.  Your students will be sorry.” (The editor later asserted that he informed Leo simply that given “his apparent lack of confidence in and regard for JAMA, he certainly should not plan to submit future manuscripts or letters for publication.”)

To avoid any misunderstanding about the point of their letter, Leo and Lacasse requested that the BMJ append to it a statement, which was published the following day: “We are fully aware that JAMA is concerned about conflicts of interest and has taken a leading role in promoting policies to benefit the medical community. We are pleased to report that we learned at the end of business on Thursday (3/5/09) that the JAMA Editorial Staff has looked into this matter and will be discussing it in the forthcoming March 11 issue.”

That was not, however, enough to satisfy the senior editors at JAMA. Indeed, the editor in chief followed up with a call to Leo’s dean. In a recent editorial, she and the executive deputy editor described her purpose and actions thus:

[S]ince Leo apparently did not appreciate the serious implications of his actions, despite our attempts to explain, we felt an obligation to notify the dean of his institution about our concerns of how Leo’s actions were potentially damaging to JAMA’s reputation. We sought the dean’s assistance in resolving the issue involving a member of the faculty of his institution to assure there would be no need to publicly identify that faculty member. No dean wants his or her institution implicated in a publication reflecting improper behavior by a faculty member. We fully expected a professional and appropriate response and assistance with resolution, as has occurred when we have notified other deans about related issues in the past, such as in other cases involving undisclosed financial conflicts of interest and cases of duplicate publication.

Some bioethicists have described JAMA’s actions as “thuggery” or worse. Whistleblowers who cross the editors will be “publicly identif[ied]” and telling anyone other than JAMA about a COI is “improper behavior,” equivalent to the acts of clear wrongdoers, such as “cases involving undisclosed financial conflicts of interest and cases of duplicate publication.” The dean describes the call as a simple case of attempted intimidation; the JAMA editors themselves depict their “tone” as “strong and emphatic,” reflecting “just how seriously we take the responsibility to ensure a fair process of investigation and above all, to protect the integrity of science and the reputation of JAMA.” There can be little doubt that they thought the journal’s reputation was at stake, but unfortunately what they did drove a stake through that reputation.

What is surprising is not that the editors expected Leo to care about any harm to JAMA’s reputation caused by his truthful description of an article the journal had published (especially since Leo had actually gone out of his way to make clear that “JAMA is concerned about conflicts of interest”). Rather it’s amazing that they could claim for a minute that someone who reports a conflict-of-interest to the journal engages in “a serious ethical breach of confidentiality” by disclosing the COI to the media.

Leo not unreasonably regards this assertion as an attempt to stifle free speech. Beyond that, it is simply nonsense on stilts. He was under absolutely no duty of confidentiality simply because he had reported the problem to JAMA. The new policy, announced in the editorial in which the editors describe and defend their treatment of Leo and his dean, compounds the problem by generalizing the principle the editors think he breached, namely that people who complain of conflicts of interest in JAMA’s articles must “not reveal this information to third parties or the media while the [editors’] investigation is under way.”

The new policy also contains a grudging step in the right direction – one which might have avoided the problem had it been in place when Leo first told JAMA of his plans to mention the Robinson case in his letter to the BMJ – namely that the complainant “will be informed about progress of the investigation, upon request, as appropriate, and will be notified when the investigation is completed.”

In responding as they did to Leo’s factual complaint that Robinson had failed to provide an accurate description of his conflicting interests, the editors not only struck at the messenger but also inflicted a wound on their journal. More’s the pity that the editorial, which had the potential to heal that wound and restore JAMA’s reputation, simply made the editors appear more arrogant. It would have been refreshing if instead the editorial had been about one-eighth as long and had simply said that:

  1. JAMA took Leo’s allegations seriously and conducted a thorough investigation, the results of which were about to appear when his letter was published online in the BMJ,
  2. this experience has led the JAMA editors to institute a new policy of keeping the person who brings allegations forward apprized of the progress of the investigation, while asking in turn that the person hold off publicly discussing the matter until the pending investigation has been completed, and
  3. they were sorry if, in the heat of the moment and with the sting of feeling criticized unfairly (since they were about to publish the correction), they said some unkind and untrue things or seemed to threaten Leo and his medical school with retaliation.

A person in Leo’s position might well conclude that a thorough investigation can be more efficiently conducted if the editors are allowed to investigate quietly, and the new policy of keeping such complainants informed of the progress of the investigation may be enough to keep them from surmising a cover-up when – as in the Robinson case – it takes five (!) months to disclose something that is easily verified. But if these prudential reasons don’t persuade a complainant and if the complainant has other reasons to go public (for example, to warn physicians and patients not to rely on the findings of a study because of undisclosed bias that may have distorted the authors’ conclusions in ways that could pose imminent harm), the complainant is certainly under no duty to keep quiet.

Indeed, the only good thing to say about this aspect of JAMA’s newly announced policy is that is likely to be short-lived because of its harmful consequences for the journal: anyone with a serious complaint of undisclosed COI against a JAMA study will have to seriously consider whether he or she wouldn’t be better advised to go first to the press (or other outlets, like public interest groups, that are capable of getting prominent attention to such claims) and then let JAMA react after the fact, rather than go to the journal first and be told that one is now wearing a muzzle for as long as the editors take to conduct their investigation.

In sum, the acts of which the editors stand accused – failing to catch a lead author’s readily detectable conflict of interest concerning a major article, threatening the young assistant professor who brought this lapse to their attention, and then publishing a defensive editorial that contains an improbable version of the events, lodges ridiculous claims of “improper behavior” against the whistleblower, and promulgates an unjustifiable and counterproductive new policy on reporting conflicts of interest – seem to reflect the sort of hubris that has brought down the great and powerful since ancient times. Perhaps, if the journal oversight committee conducts a thorough inquiry and publishes the results, we will know more. But one has the sense that whatever prompted the editors to follow their self-destructive course may never be clear. At least for the moment, it is all very puzzling and rather sad.

Alexander M. Capron is a University Professor and holds the Scott H. Bice Chair in Healthcare Law, Policy and Ethics at the University of Southern California.

Published on: April 14, 2009
Published in: Conflicts of Interest in Research

Receive Forum Updates

Recent Content