Skip to Content
No Fees Unless We Win 855-948-5098

Misleading Paxil Study Exposed in New BMJ Study


Study 329, a Poster Child of tainted scientific research, has finally gotten its comeuppance.

Study 329 was published in 2001 by over a dozen well-credentialed academics whose names lent credence to the study’s conclusions, but unbeknownst to the medical community and public, the study was ghostwritten by a PR company hired by the manufacturer (SmithKline Beecham, aka GlaxoSmithKline or “GSK”), and the study’s reported findings are, according to a newly published study, contrary to the actual underlying data.

Although Paxil has never been FDA approved for use in any pediatric population, the study was used to encourage doctors to prescribe the drug to adolescents.  Over the past 10 years, the study has been hotly criticized by academics, the media, and in lawsuits, including ones brought by the Attorney General of the State of New York and the Department of Justice.  Even so, requests for retraction have repeatedly been rejected.

A group of dogged academic researchers decided that Study 329 would be the perfect candidate to undergo thorough reanalysis under the RIAT protocol (“Restoring Invisible and Abandoned Trials,”) a new project whose purpose is to “improve the integrity of the [scientific] process to determine drug efficacy and safety.”

The researchers were given full access to the same data the original “researchers” were given, but they looked at the actual data, not just summaries provided by the company.  The lead author of Study 329, Dr. Martin Keller, admitted in a deposition taken by Baum Hedlund that he did not look at the raw data, but only at summaries provided to him by GSK.  Not surprisingly, the RIAT researchers’ conclusions were opposite to that of the original authors.  Not only did the RIAT analysis show Paxil to be ineffective, but that the drug can cause suicidal behavior in a significant number of young people.

Through litigation over the past 15 years, Baum Hedlund was able to obtain access to hundreds of thousands of internal confidential documents and depose a number of the key players in the saga of Study 329.  The documents and testimony revealed, amongst other things, close financial ties between the “authors” of Study 329 and GSK, the lengths to which Study 329 was influenced by financial interests, and the fact it was ghostwritten by a company called Scientific Therapeutics Information (STI), which was paid for by GSK.

“I am proud of the role our litigation played regarding Paxil and Study 329 in helping to bring to light the manipulation of the efficacy and safety of Paxil in adolescents.  We have seen, first hand, the consequences this manipulation has taken in our representation of those families harmed by this conduct.  The true results of Study 329 were overshadowed by an opportunity to obtain enormous profits by spinning negative results into falsely positive ones. Study 329 was not an isolated, renegade event—it is simply one instance where the company got caught. Hopefully, this signals to academics, regulators, and company personnel that transparency and honesty is the better route, and the public and governmental agencies paying for these ’results’ should question their reliability.” Statement by Michael L. Baum, senior managing shareholder of Baum Hedlund Aristei & Goldman

We applaud the team of researchers who have spent the last 10 years uncovering the truth and conducting the comprehensive and arduous reanalysis of the original Paxil Study 329: Jon Jureidini, Ph.D., MB BS, FRANZCP, David Healy, M.D. FRCPsych, Joanna Le Noury, Ph.D., John M. (Mickey) Nardo, M.D., B.S., (Mathematics), Melissa Raven, Ph.D., Catalin Tufanaru, M.D., MPH, MClinSci (EBHC), Elia Abi-Jaoude, MD, FRCP(C); and Peter Doshi, who conceived the idea of a new standard and protocol for re-authoring studies that have not been published or have been misleadingly published.

For more information about Study 329:

Past articles about Study 329:


  • Please enter your first name.
  • Please enter your last name.
  • Please enter your phone number.
    This isn't a valid phone number.
  • Please enter your email address.
    This isn't a valid email address.
  • Please enter your city.
  • Please enter your state.
  • Please make a selection.
  • Please make a selection.
  • Please enter a message.