A study claiming an antidepressant was safe for adolescents did not disclose data showing a link to suicidal or self-harming behaviours: new analysis

An antidepressant was misrepresented as safe for adolescents, says new Australian-led research which has ramifications for drug companies and the reporting of trials.

The groundbreaking study re-analysed the original research findings on paroxetine, sold as Aropax, Paxil and Seroxat, a drug originally reported to be effective and safe for depression in adolescents.

The re-analysis of the controversial Study 329, published in 2001, used previously undisclosed data from the study that showed the drug was linked to serious adverse reactions, including suicidal or self-harming behaviours.

"Our study also revealed that paroxetine was no more effective at relieving the symptoms of depression than a placebo," said Professor Jon Jureidini from the University of Adelaide's Critical and Ethical Mental Health Research Group.

"This is highly concerning because prescribing this drug may have put young patients at unnecessary risk from a treatment that was supposed to help them."

The study, published in the medical journal The BMJ, is the first under its initiative called Restoring Invisible and Abandoned Trials (RIAT).

It aims to expose misreported trials, to have undisclosed outcomes of trials published and misleading publications corrected.

The BMJ's associate editor Peter Doshi says the original study, published in the Journal of American Academy of Child and Adolescent Psychiatry, should be retracted.

He said the new study puts "pressure on academic and professional institutions to publicly address the many allegations of wrongdoing".

The original manuscript was not written by any of the 22 named authors, but by an outside medical writer hired by the drug company which funded Study 329, he said.

The company, now called GlaxoSmithKline, was fined a record $US3 billion ($A4.17 billion) by American authorities in 2012, in part for fraudulently promoting paroxetine.

Prof Jureidini said the researchers learnt a lot about incorrect reporting and the considerable fallout that can be associated with distorted data.

"Regulatory research authorities should mandate that all data and protocols are accessible," he said.