January 26, 2010
Scrutiny emerges concerning conflicts of interest in veterinary literature
By: Jennifer Fiala
For The VIN News Service
The Journal of Veterinary Internal Medicine requires authors to fill out the above form to complete the online article submission process. The Journal of the American Veterinary Medical Association, American Journal of Veterinary Research and Journal of Small Animal Practice have no provision for declaring a conflict of interest during the online submission procedure.
Dr. George Glanzberg, like many veterinarians, reads human as well as veterinary medical literature.
So it has not escaped his attention that in human medicine, controversy has erupted over the apparent failure of a number of study authors to report conflicts of interest and the editors of some prestigious peer-reviewed journals to ferret out such relationships.
Critics contend that financial or professional ties can cloud objectivity, and in turn, skew research outcomes. Glanzberg, a small-animal practitioner in Vermont, surmises that if unreported conflicts are of concern in human medical literature, they're likely to be a problem in veterinary literature as well, even if the stakes are not as high.
It's a concern that's not deeply shared by many veterinary journal editors, according to a recent poll conducted by the VIN News Service (VNS) regarding whether unreported conflicts are rife in the profession's publications. Yet Glanzberg and others consider conflict of interest statements to be paramount to the transparency and openness that's needed to ensure the integrity of all scientific literature.
Putting it simply, he said, “I’d like to have the option of knowing what I am reading."
Glanzberg originally expressed his concerns in a 2006 Veterinary Information Network (VIN) discussion when he challenged a Journal of the American Veterinary Medical Association (JAVMA) policy to not print a conflicts of interest statement following the studies that it publishes. Such statements are common in publications like the New England Journal of Medicine.
While his concerns initially failed to attract much attention on the VIN message boards, the topic was reawakened last August when the New York Times revealed that Wyeth had paid ghostwriters to author 26 scientific papers backing the use of hormone replacement therapy in women. Those works, the Times article suggested, were then signed by physicians and published in 18 peer-reviewed medical journals, crafting a medical consensus that benefited sales of Wyeth's Prempro and Premarin.
VIN members also reacted to news that Merck and Elsevier created fake medical journals to market drugs without disclosing the drug company's sponsorship.
(Merck and Wyeth, now sold to Pfizer, each maintain animal health subsidiaries that make pharmaceuticals used in veterinary medicine.)
Then last September, the Journal of the American Medical Association published results of an in-house survey that showed ghostwriting is extremely common. Results from the online questionnaire crafted for the study showed that of the authors who responded, 8 percent admitted that they had published a paper that contained a significant contribution from an unacknowledged person.
The survey did not ask whether industry was involved in these situations, but that clearly was the implication.
Responding to these reports and a growing awareness that drug companies have become an even more significant driver of medical research and its publication, medical journal editors have drafted a standardized form for reporting financial conflicts of interest. They hope this form will be more comprehensive than others used in the past and make reporting less confusing for potential authors. The form was unveiled by the International Committee of Medical Journal Editors (ICMJE) in October.
The committee cannot force journals to use the form but urges that they do.
To determine what role, if any, potential conflicts of interest play in veterinary literature, the VNS recently contacted a handful of veterinary medical publications.
All the veterinary journal editors reached by the VNS said that they were aware of the new ICMJE form and the growing awareness of the conflicts problem in human medicine. They noted that there is no similar effort underway for veterinary journals.
Nor is there as much of a need, the majority of the editors stated.
Why? Because the large, Big Pharma-sponsored clinical trials that populate human medicine are uncommon in veterinary medicine, they said. In fact, most veterinary-related studies are funded by government entities like the National Institutes of Health, various foundations or not at all. Even when funding comes from grants or industry, the trials tend to be small with less at stake, they added.
What’s more, it is much harder to hide conflicts of interest in veterinary medicine than in human medicine, noted Dr. Kurt J. Matushek, JAVMA's editor-in-chief. Considering the veterinary research community is relatively small, reviewers are usually familiar with authors and their clinics or laboratories.
“It can be self-policing, up to a point,” he said.
Dr. Kent Lloyd, associate dean for research at the University of California, Davis’ (UC Davis) veterinary medicine program, said that, in his experience, veterinary journals ask authors if they have financial conflicts of interest. Moreover, he said that at most veterinary schools, there are additional integrity protections, which probably help the journals.
For example, UC Davis researchers are not allowed to accept grants or funds that come with strings attached or put restrictions on what the researcher reports or how they can report it.
“It’s dealt with upfront,” he said. “An investigator doesn’t take money from a company to conduct research using university facilities without dealing with the university.”
To grasp whether veterinary journal editors perceive a problem with unreported financial conflicts of interest and whether they took steps to prevent such conflicts from impugning their journals, the VNS attempted to survey them. Initially, a series of questions was e-mailed to 20 editors without much response. Then Dr. Mary Christopher, organizer of the International Association of Veterinary Editors (IAVE), published the survey on the group’s listserv, which reaches more than 150 veterinary editors worldwide.
The VNS asked the editors whether they had policies for reporting conflicts or perceived a problem with unreported conflicts. The editors also were queried about the separation between editorial content and advertising.
In all, six editors responded, including one from a German journal and one from a non-peer reviewed journal. One of the six responses came from an editor who had seen the survey but was too busy to act on it. As a generalization, the handful of editors who responded indicated that they routinely ask authors about conflicts and were unaware of any problem with unreported conflicts.
Dr. M.D. Salman, editor of Preventive Veterinary Medicine, said his journal requires a conflict of interest statement from every author.
The only exception to the general consensus came from Dr. Karen Overall, editor of the Journal of Veterinary Behavior, who said that the publication had become “fussy” about authors reporting the financial backers of studies. Overall reviews manuscripts for other journals and notes that there have been cases where it was thought that some funding was relevant, but it was not reported.
“There are many papers on — for example, pheromonal products — where the authors have not disclosed that they have been paid consultants for the company,” she said. “I only know this because I have later seen people’s CVs (curriculum vitae) in another context and it is listed.”
The IAVE's Christopher said that although there may not be a serious problem with conflicts of interest, reported or not, that does not mean editors take the situation lightly. She notes that at the group’s last annual meeting there was a detailed presentation on conflicts given by the American Veterinary Medical Association’s Dr. Craig Smith, associate editor with the publications division.
However, Dr. Paul Pion, a boarded cardiologist, published author and president of VIN (the parent company of VNS), said it is a “crock” to suggest that veterinary research is devoid of potential conflicts. The question, he said, “is not the size or nature of what triggers the potential conflict. What matters is that it exists at all. If there is a potential conflict, it must be disclosed explicitly, no exceptions.”
Pion believes that it is the responsibility of journal editors to more explicitly define a potential conflict of interest for their authors in order to help them recognize potential conflicts that they might not think to disclose.
It's "not because (authors) wish to hide them,” explained Pion, “but because they don’t recognize them as potential conflicts."
Pion continued: “We all believe we are immune from influence. Well, we are not. We are human. We all want to ‘do right’ by those who are nice to us. So, it is up to the profession and especially the journals to do more to help authors recognize potential conflicts of interest to protect authors from criticism for not revealing what others might consider potential conflicts of interest.”
Pion and his cardiologist colleague Dr. Mark Rishniw, a VIN consultant and Cornell University fellow, highlight a recent example that they believe illustrates how, when it comes to reporting conflicts, veterinary editors and journals are dropping the ball.
Pion and Rishniw point to an article in the July/August 2008 issue of the Journal of Veterinary Internal Medicine (JVIM) that prompted the cardiologists and two of their UC Davis colleagues to author a series of letters to the editor, expressing their concerns with a study on the effects of benazepril on survival and cardiac events in dogs with asymptomatic mitral valve disease.
Benazepril is an angiotensin converting enzyme inhibitor marketed by Novartis Animal Health — a company that, three years earlier, had gifted a $150,000 Vivid 7 ultrasound machine to the facility where the study's authors work .
While the authors apparently reported to JVIM that they had no conflicts of interest upon submitting their article, the ultrasound machine donation was revealed in the study's acknowledgments section.
Rishniw and Pion emphasize that their concerns do not lie with the authors — researchers who, they claim, likely did not intentionally misrepresent their conflicts or skew the study. Pion said he "believes in the integrity and intent of the authors of the paper” and is not attacking the ethical standards of the authors.
Rather, the situation illustrates a problem with the leadership of journals that do not go far enough to explore potential conflicts, outline their parameters to authors or take the issue seriously, he said.
“We are all human; most of us are honorable and nice people,” Pion explained. “But I also believe that we are all influenced by that mutual niceness and desire to find positive results (investigator bias) that can lead to the subconscious desire to not publish a bad outcome about a drug that we believe works and is produced by a company that has been good to us. This creates a very dangerous situation where we think we are immune from this form of potential conflict of interest and bias. The fact that our colleagues checked 'none' when asked about potential conflicts demonstrates that they don’t recognize the potential for a conflict of interest inherent in this type of gift from the maker of the drug under study.”
In response to a VNS query, the benazepril study's primary author, Dr. Jean Louis Pouchelon, took offense at the notion that his scientific integrity was in question.
“Our team has always declared potential conflicts of interest or sponsors involved in our studies, and we did so in the Acknowledgments section of the aforementioned paper,” he said.
Pouchelon explained that the echocardiography system that Novartis donated to his school was not tied to any person or project.
In response, Pion reiterated his belief that the problem lies with journal's editors: "The (JVIM) editor should have seen the acknowledgment and recognized it as discordant with checking the none box and ensured that when the article was published the gift was clearly declared as a potential conflict of interest."
JVIM editors did not respond to repeated requests from the VNS for comment.
Pion believes that, in general, veterinary medical journals are too cavalier about potential conflict reporting statements. He does not buy the argument that there is not enough money in veterinary research to bias scientists. “We’re not talking intentional cheating or fraud. We’re talking about investigator bias that we all need to work hard to defend against for ourselves and each other.
“We can’t pretend it isn’t a factor,” he added. “It doesn’t really matter whether the funding or donation was $5 or $1 million. What matters is how receiving a gift affects us. We are all full of potential conflicts, and we owe it to the profession to acknowledge that fact.”
That's what Vermont’s Glanzberg wants to happen. He knows that there are conflicts of interest in research, sometimes financial, and he doesn’t care how researchers characterize them.
He just wants to know about them and believes that it's time for veterinary journals to follow their counterparts in human medicine by making reporting requirements more rigorous.
“This is no longer penny-ante stuff that we needn’t be paying attention to,” Glanzberg said.
Freelance journalist Timothy Kirn contributed to this article.
Search VIN news
All news categories