Journal Faces Lawsuit Over Discredited Study
Internal documents later showed that GlaxoSmithKline hired a PR firm to ghostwrite the article, cherry-picking data and recruiting 20 co-authors to lend credibility
The company then used the paper to market Paxil to doctors.
Meanwhile, a lawsuit filed last month against the American Academy of Child & Adolescent Psychiatry (AACAP) and Elsevier, which publishes the organization’s journal, JAACAP, demands that the journal retract the article.
Attorney George W. Murgatroyd III, acting on behalf of the public, filed the suit in the Superior Court of the District of Columbia Civil Division.
Published in 2001, the article ignited decades of controversy. Critics say it overstated Paxil’s benefits and downplayed its risks, including increased suicide risk among teens.
Known as the “Keller article,” after lead author Dr. Martin Keller, then chair of psychiatry at Brown University, the paper reported partial results from Study 329, an investigation into whether paroxetine was safe and effective for treating depression in adolescents.
GlaxoSmithKline, now GSK, which manufactures Paxil, funded the study.
The article reported that “paroxetine is generally well tolerated and effective for major depression in adolescents” — a claim now widely known to have been based on distorted results.
The study actually found the drug was neither safe nor effective. Internal documents later showed that GSK hired a PR firm to ghostwrite the article, cherry-picking data and recruiting 20 co-authors to lend credibility.
The company then used the paper to market Paxil to doctors.
Peer reviewers raised concerns about the study’s data. As soon as the article went live, practitioners wrote multiple letters to the editor asking why statistically significant indications that the drug caused serious adverse events, including “suicidal gestures,” were dismissed in the clinical trials and not addressed in the paper.
According to the complaint, the Keller article became one of the most cited articles supporting the use of antidepressants in child and adolescent depression — even though evidence from two other GSK trials confirmed the drug was ineffective and risky.
Even though the U.S. Food and Drug Administration (FDA), which never approved the drug for use in children and adolescents, raised concerns about the study.
Calls for JAACAP to retract the article began in 2010, bolstered by:
- A 2015 British Medical Journal reanalysis confirming the drug’s dangers and data manipulation.
- Evidence from GSK’s internal files and depositions in other Paxil lawsuits.
- A 2012 U.S. Department of Justice case in which GSK paid $3 billion to settle criminal fraud charges related to Paxil and two other drugs.
Still, the JAACAP and Elsevier have so far refused to retract the article.
Murgatroyd has represented a dozen families whose children died by suicide or were severely injured in a suicide attempt, allegedly as a result of taking Paxil. His litigation team has deposed all the article’s authors and has combed through GSK’s internal documents.
Both JAACAP and Elsevier continue to profit from the article, according to court documents. It costs $41.50 to download from the JAACAP website, and $33.39 to download from Elsevier’s ScienceDirect website.
The complaint asks the court to “redress the knowing publication, distribution, and continued sale of a false and deceptive medical article that has misled physicians, consumers and institutions for over two decades and continues to endanger adolescent mental health and safety as well as public trust in scientific integrity.”
JAACAP last week published an “expression of concern” indicating that the Committee on Publication Ethics (COPE) will manage recommendations and guidance.
COPE is an international, nonprofit organization that provides guidance to journal editors on publication integrity. COPE does not investigate whether there is research or publication misconduct — it only examines whether the journals involved followed its code of conduct.
AACAP did not respond to The Defender’s request for comment. Elsevier responded only that it would need more time to respond to such a request.
Study 329 revealed serious safety risks, including suicidal behavior
Before publishing the Keller article, GSK funded three studies to test the safety and efficacy of paroxetine to treat depression in children and adolescents. The drug failed to demonstrate efficacy in all three trials.
The first study — Study 329, completed in 1998 — also revealed serious safety risks, including suicidal behavior. Later studies confirmed those risks. According to court documents, GSK knew the “disappointing” results of Study 329 would be a commercial disaster for the drug.
However, the study had some minimal positive results, which could indicate the possibility of efficacy. It met 15% of the outcomes the researchers had been hoping for to show that Paxil worked. GSK officials privately conceded these results were not sufficient to show efficacy.
Yet GSK planned to publish selective data from the study in a prestigious medical journal to claim the drug worked.
GSK hires PR firm to write first draft of JAACAP article
The drugmaker hired a private public relations company, Scientific Therapeutics Information Inc. (STI), to write the article. An employee drafted it and sent it to Keller, who was selected to be the lead author and finish the publication process.
STI’s role was not listed in the final draft submitted to JAACAP.
Later in 1998, GSK’s second study, number 377, also failed to show efficacy. The study also showed four times the number of serious adverse events related to suicidality as the placebo study, according to court documents.
That same year, although GSK by then knew of two studies showing the drug was ineffective, the drugmaker decided not to publish any data from Study 377 and instead paid “directly or indirectly” three prominent psychiatrists — Karen Wagner, M.D., Ph.D., Dr. Neal Ryan, and Keller, who had worked on Study 329 — to promote Paxil as a safe and effective treatment for adolescent depression at conventions, forums and at a meeting of the American Psychiatric Association.
The third study, 701, concluded in 2001, also failed to demonstrate efficacy six months before the Keller article was published — yet GSK and JAACAP went ahead with publication.
Twenty authors who were psychiatrists were added as authors of the Keller article. Two GSK employees, James P. McCafferty and Rosemary Oakes — the only authors without medical or doctoral degrees — were also added, although their affiliation with GSK wasn’t disclosed.
Testimonies from depositions in other trials indicated that 10 of the authors didn’t even comment on the paper, and none of them had access to raw clinical trial data — although they all said they did.
None of them disclosed conflicts of interest, and all of them signed off on the article as their work.
GSK used Keller paper to sell $1 billion worth of Paxil to teens
Once JAACAP published the article, GSK’s sales team began promoting Paxil as “safe and effective” for teens.
GSK sent the article to all of its 2,000 Paxil sales reps. In the three years that followed the article’s publication, it is estimated that the company made over $1 billion from sales of the drug to adolescents, the court documents state.
In the following few years, regulators in the U.K. and the European Union issued alerts about the dangers of paroxetine based on its link to suicide ideation in teens.
In 2003, the FDA issued a similar warning, saying there was “no evidence” the drug was effective.
In the following years, some of the authors began to discuss their concerns about suicidality internally, although they made no changes to the article, according to court documents.
Have JAACAP and Elsevier refused retraction to protect authors?
The complaint filed last month alleges that JAACAP editors and Elsevier refused to retract the Keller article “in an apparent attempt to shield at least five of the Keller article authors who are prominent members of the AACAP from possible ramifications of retraction.”
Conflicts of interest among the article’s authors are glaring. Keller, Wagner and Ryan all received money to promote Paxil as safe and effective in the years before publication, according to the complaint.
Two authors, McCafferty and Oakes worked for GSK, the complaint said, and did not disclose that in the article.
Several authors of the Keller article went on to hold influential positions within the AACAP. Wagner was president from 2017-2019.
Dr. Gabrielle A. Carlson also served as president of the organization from 2019-2021. Before that, she was chair of AACAP’s Program Committee from 2011-2014, and won AACAP’s Virginia Q. Anthony Outstanding Woman Leader Award.
Dr. Graham Emslie has served on the JAACAP Editorial Board. Dr. Boris Birmaher has chaired committees in the AACAP and published numerous articles, editorials and organizational practice parameters.
The current editor-in-chief of the journal, Dr. Douglas Novins, who was not an author on the article but holds final decision-making power over retraction, has worked closely with some of the authors — co-authoring editorials with both past presidents.
Dr. David Healy, co-author of the BMJ article that reanalyzed the data from Study 329, told The Defender that for years, he and others who had been investigating this issue assumed the journal had been duped by GSK, but later realized the journal “was not duped — it was complicit.”
Keller and then-editor Dr. Mina Dulcan were close friends, Healy said — a relationship revealed in transcripts of interviews Dulcan did for a set of BBC programs.
See more here childrenshealthdefense.org
Please Donate Below To Support Our Ongoing Work To Defend The Scientific Method
PRINCIPIA SCIENTIFIC INTERNATIONAL, legally registered in the UK as a company 
incorporated for charitable purposes. Head Office: 27 Old Gloucester Street, London WC1N 3AX.