A reanalysis of data from a controversial study published in 2001 that opened the doors for doctors to write millions of prescriptions of an antidepressant to children and teens shows that the drug is not only ineffective but potentially unsafe. The new study, published Wednesday in The BMJ, details a number of issues with the original analysis that made paroxetine, marketed as Paxil by GlaxoSmithKline, look more favorable.

Here's what you need to know about the new findings:

Tell me about the old study.

The original paper, published in the Journal of the American Academy of Child and Adolescent Psychiatry and known as Study 329, purportedly showed that the drug is safe and effective in adolescents with depression.

The study has been revealed to have been largely ghostwritten by a consultant on the drug maker’s payroll and has been investigated by both by both federal and state authorities. The Food and Drug Administration has required so-called "black box" warnings on the packaging for Paxil since 2004 after it received reports of links to suicidal thinking. The company has also faced a number of consumer lawsuits by parents of teens who said their children had serious reactions to the drug.

The paper’s lead author -- Martin Keller, a psychiatry professor at Brown University -- was revealed to have financial ties to drug companies in a 1999 investigation by the Boston Globe.

What's this new study everyone's talking about?

The team conducting the reanalysis -- led by Jon Jureidini, a professor at the University of Adelaide in Australia -- went back to the raw data and came to starkly different conclusions than the first paper. They found that the effect of the drug was no different than a placebo and that there was an increase in harms among study participants, the most alarming of which was an increase in suicidal thoughts and behavior.

Jureidini and his colleagues said that Study 329 contained numerous transcription errors and other problems that masked issues and violated the own company’s protocols for statistical analysis.

“It’s hard to think there wasn’t some mischief being done,” Jureidini said. He said it looked like there may have been "deliberate attempts to play down the adverse event profile."

Given all the investigations over the past 14 years, why is this BMJ paper significant?

The journal and its authors have stood by the research all these years. Some of the clinical guidelines for prescribing antidepressant medication to children and adolescents around the world still refer to and build on this old paper.

“A small group of us knew that the drug didn’t work all that well, but actually the wider world still thinks the drug works quite well,” David Healy, a co-author who is a researcher at the School of Medical Sciences at Bangor University in Wales, explained in a conference call with reporters a few days before the study was published.

What is GlaxoSmithKline's take?

The researchers said the sent the company the paper in March but did not receive any response. On Wednesday, when the study was published, the company released this statement:

We were able to help this team to carry out their reanalysis by providing access to the detailed data from the original trial. This reflects our commitment to data transparency – we publish the results of all our studies regardless of whether they are positive or negative and we are the only pharmaceutical company to be part of the AllTrials campaign. We have also led the way in giving external researchers access to the very detailed patient-level data behind our studies, granting access to more than 50 research teams around the world so they can independently use our data in their research.
Importantly, the findings from this team’s analysis appear to be in line with the longstanding view that there is an increased risk of suicidality in paediatric and adolescent patients given antidepressants like paroxetine. This is widely known and clear warnings have been in place on the product label for more than a decade. As such we don’t believe this reanalysis affects patient safety.
This particular trial, which was conducted in the 1990s, was included in detailed reviews carried out many years ago by regulators and by GSK which identified the increased risk. It’s something the medical community and regulators are aware of."

How common is it for researchers to go back and re-analyze the data for a study that has already been published?

This happens very rarely and part of the reason is that it's often difficult for outside researchers to get their hands on the raw data.

The study was part of an initiative called RIAT (Restoring Invisible and Abandoned Trials), which “encourages abandoned or misreported studies to be published or formally corrected to ensure doctors and patients have complete and accurate information to make treatment decisions.”

Fiona Godlee, editor in chief of The BMJ, said this week that the reanalysis highlights the “the extent to which drug regulation is failing us” and called for new legislation that would ensure that the results of all clinical trials are made fully available and that individual patient data is released for independent review.

Peter Doshi, an assistant professor at the University of Maryland School of Pharmacy and an associate editor of The BMJ, said the reanalysis is a “model for the future of publishing” in that it not only contains a 10-page analysis of the research but makes the data public.

“The data are there for anybody to go and do their own homework. You don’t have to take it on trust anymore,” Doshi said.

Read more: