Independent researchers have recrunched a drug company's data for the antidepressant Paxil and reached the opposite conclusion about its safety and effectiveness in children and adolescents.

The first study — published in 2001, of 275 adolescents aged 12 to 18 with major depression — was paid for by the drug company. It was known as Study 329.

"The findings of this study provide evidence of the efficacy and safety of the SSRI, paroxetine, in the treatment of adolescent depression," the study concluded. 

Now a reanalysis of the original data, published Wednesday in BMJ, formerly the British Medical Journal, contradicts those findings.

There were safety concerns before but the extent wasn't realized until now, said Dr. Elia Abi-Jaoude, a psychiatrist at the Hospital for Sick Children in Toronto and one of the authors of the reanalysis.

Dr. Elia Abi-Jaoude

There were safety concerns about Paxil but the extent wasn’t realized until now, says Dr. Elia Abi-Jaoude. (CBC)

"This particular drug in that particular study was ineffective and unsafe," Abi-Jaoude said. "The more general point and, arguably, the more important point is that without access to the data and without access to the study protocol, you know very little about what happened in a study. It can be published in a top journal, peer reviewed [yet] you know very little about what happened."

One of the problems is that a small group with a clear interest in the outcome of the study have outlined the findings without anyone else having access to the raw data, say scientists campaigning for independent scrutiny of all clinical trial research.

Dr. David Henry of the Dalla Lana School of Public Health at the University of Toronto wrote a journal editorial accompanying the reanalysis, "Liberating the data from clinical trials."

"The actual benefits of some treatments are clearly being exaggerated and they're being promoted on the basis of these claims, and this is resulting in both over diagnosis and over treatment of some conditions, and we need to start to correct that balance. We can only start to correct that by getting access to the original data," Henry said.

The controversy over Paxil was part of a complex legal case against the drug company, now called GlaxoSmithKline. In 2012, GSK was fined $3 billion US in court settlements.

GlaxoSmithKline has never admitted wrongdoing. The journal that published Study 329 has not retracted the paper.

 Dr. Stan Kutcher, a professor of psychiatry at Dalhousie University in Halifax, was one of the authors of Study 329.

In an email to CBC News regarding the reanalysis, Kutcher said "this pejorative and biased paper was created by authors who, in contrast to acceptable scientific enquiry, had predetermined what they would report."

Kutcher said he and the other original authors are drafting a full response.

Exaggerated sense of effectiveness

In 2003, GlaxoSmithKline warned of serious side-effects of SSRI antidepressants when taken by children and teens.

In general, Abi-Jaoude advises people to consult their doctor before starting a new medication, including SSRIs, which he prescribes in some cases. The same advice applies when considering stopping a drug: Approach all drugs with a healthy dose of skepticism, he suggests.

In his commentary, Henry acknowledged data sharing is not without risks that need to be considered, such as threats to patient confidentiality and investigators' own agenda. It is also a time-consuming process and a major undertaking in terms of human and analytical resources.

Henry called original data from clinical trials "treasure troves of information" that can be used to not only reconfirm findings but also to generate valuable new ones. 

With files from CBC's Kelly Crowe