Two weeks ago, Edmund Levin and George Stewart, members of the American Academy of Child and Adolescent Psychiatry, sent a letter to the editor of the Academy’s journal, requesting an explanation as to why a controversial study led by former Brown Professor Emeritus of Psychiatry and Human Behavior Martin Keller has not been retracted.
The paper — which details the findings of Study 329 and focuses on the effects of the drug Paxil on adolescent depression — has been continually criticized since its publication in 2001.
While Levin and Stewart have worked to get the paper retracted, Jon Jureidini, a professor at the University of Adelaide in Australia and a member of the nonprofit Healthy Skepticism, has been working with his team to reanalyze the original data and republish the results.
A controversial history
Since its publication, Keller’s paper, which suggests that Paxil is an effective treatment for adolescent depression, has been criticized for being ghostwritten by associates of GlaxoSmithKline — the drug company that makes Paxil.
In 2006, Keller publicly acknowledged that GSK had given him tens of thousands of dollars during and after the time the study was conducted. Keller did not respond to multiple requests for comment for this article.
A Senate investigation in response to a U.S. Department of Justice lawsuit confirmed the presence of ghostwriting in the paper, said Paul Thacker, a fellow at the Edmond J. Safra Center for Ethics at Harvard who participated in the Senate inquiry.
But Rachel Klein, one of the 22 cited authors on the Keller paper and a professor at New York University, said while she thinks GSK played a role in writing the paper, it was not ghostwritten.
Continued research in the years after the article’s publication has suggested Paxil is linked with an increase in suicidal ideation in adolescents, The Herald previously reported.
The study is continually cited in other papers as evidence of Paxil’s effectiveness, Jureidini said. That a fraudulent paper is still cited “really has to be addressed,” he added.
“I think the concern is legitimate,” Klein said. “But I think it is too bad because the paper presents everything. As long as you give the information, you’re not misleading.”
The University has declined to support any efforts to have the paper retracted, The Herald previously reported.
Leemon McHenry, a member of Healthy Skepticism, and Jureidini have written to the University administration to ask for support in having the article retracted, but administrators have refused, McHenry said.
“The University takes seriously any questions about the soundness of faculty-conducted research,” said Marisa Quinn, vice president for public affairs and University relations. “While the University cannot comment on individual personnel cases, it does take appropriate actions whenever such questions are raised… We have effective policies in place, and those policies are consistently applied, although they are confidential.”
Clinical Professor of Family Medicine David Egilman ’74 MD ’78 said though no institution has ever punished a faculty member due to ghostwriting, “It’s a social responsibility to make sure that fraudulent information is not published by faculty.”
Retracting the paper
Levin said he first became interested in Study 329 due to his association with AACAP.
Though 13 years have passed since the publication — and despite the Senate investigation finding the Keller article to be fraudulent — there “still has been no effective action in getting it retracted,” Levin said.
Levin and Stewart’s regional division of AACAP has sent two different letters to Andres Martin, the editor of the journal. In addition to the one sent two weeks ago, the group sent a first letter requesting the study’s retraction.
In response to the first letter, Martin wrote that the paper did not meet the criteria for retraction, Stewart wrote in an email to The Herald.
The group was also informed that AACAP leadership had instructed the ethics committee not to investigate the article, Stewart said, adding that he was not sure why the committee had been “muzzled.”
Martin has not yet responded to the group’s second letter, asking for an explanation, Stewart wrote.
Martin did not return The Herald’s requests for comment.
Reanalyzing the data
Jureidini, along with four team members, is currently reanalyzing the data from Study 329 to “write the paper as it should have been written,” he said.
The decision to do so came in the wake of the British Medical Journal’s Restoring Invisible and Abandoned Trials initiative, announced June 2013. The initiative “calls for third party authors to publish or republish unpublished and misreported clinical trials,” according to a document released by the BMJ.
In the document, the BMJ named Study 329 an “abandoned study” — a study no one is working on but which is “misreported” and has not been corrected or retracted.
Jureidini said he and his team have been reanalyzing the data since late last summer and hope to have a draft completed in the next month.
“What they have found is that the fraud and corruption in the Keller article is far worse than anyone ever expected,” McHenry, who has worked closely with Jureidini, said. Specifically, the paper is far worse in terms of safety and efficacy, he added.
Very few people, including Keller, ever looked at all the raw data, McHenry said, adding that most of this was done by statisticians associated with GSK.
It is “ironic” that his team has looked at the data more than any of the authors of the Keller article, Jureidini said.
Levin and Stewart both said they support the study’s reanalysis. It may “be unrecognizable when Jureidini publishes it,” Levin said.
Jureidini said he hopes the original Keller article will be retracted in the wake of the republished one.
But Klein said she does not think the new results will show the Keller article to be fraudulent.
“I’m all for the data being examined,” she said. “If it is true that the results are very different, that would be a very different situation, but I can’t imagine that that would be the case,” she said.
A clinical perspective
Despite the Keller paper’s suggestion of the efficacy of Paxil, many psychiatrists said they do not prescribe the drug to patients.
Multiple psychiatrists interviewed said Paxil has worse side effects and withdrawal symptoms than other antidepressants and should not be prescribed as a first-line drug.
Allegations of ghostwriting and the role GSK played in Study 329 raise issues related to the relationship between psychiatrists and drug companies.
Study 329 and the Keller article are “a very good soap box from which to talk about the atrociously dishonest and morally unethical things the pharmaceutical industry has done,” Levin said.
Thacker said studies like Study 329 show that psychiatry is not independent from drug companies, and pharmaceuticals have “captured and owned the field of psychiatry.”
“You can’t live with them and you can’t live without them,” said Louis Velazquez, a practicing child and adolescent psychiatrist, referencing drug companies.
Research psychiatrists also rely on pharmaceutical companies to fund their research, said John Fanton, a child and adolescent psychiatrist at Baystate Medical Center and an assistant professor of psychiatry at Tufts University School of Medicine, who was completing his residency at Brown when the allegations about Study 329 arose.
When there is an economic downturn, there is “a legitimate need to identify and access other finding streams,” he said, adding that pharmaceuticals often will provide the most money for research.
But there is an “inherent conflict” between researchers and pharmaceutical companies, because researchers are required to share both good and bad knowledge, and pharmaceutical companies may have different goals, Fanton said.
If the paper is retracted, it will create “ongoing problems” for GSK, Thacker said. “There’s a lot of egos and a lot of money at stake.”