Many supporters of AllTrials will be interested in a study published in The BMJ today, a reanalysis of previously hidden clinical trial data. The new research used data from a 1990s clinical trial of the GlaxoSmithKline (GSK) antidepressant drug paroxetine. Today’s findings contradict a 14-year-old analysis of the data referred to as Study 329, which found paroxetine to be safe and effective for treating adolescents with major depression.

The new research is the first reanalysis of a drug study under the RIAT (Restoring Invisible and Abandoned Trials) initiative, which calls on companies and academic funders to publish detailed trial information for independent scrutiny. The RIAT team was able to access the original clinical trial data using GSK’s patient-level data access portal, where researchers can request access to this information.

Tracey Brown, Director, Sense about Science and co-founder of AllTrials:

“When all trials are registered and results reported, it becomes possible for researchers to work out what data are available. GSK has gone further and made its patient level data available to researchers. It is disappointing that there are still so many companies not reporting trials. Researchers, doctors, patients and, in July, their shareholders have said they want transparency about trial results. This will confirm their views.”

Sir Iain Chalmers, coordinator of the James Lind Initiative and co-founder of AllTrials:

“Among pharmaceutical companies, GSK under its current management has led the way in promoting clinical trial transparency and provides a practical mechanism to make trial re-analyses possible. The reanalysis of Study 329 illustrates the knowledge dividends from the company’s new policies and contrasts strikingly with the scientific misconduct that characterised the company’s behaviour under previous management. Today’s GSK has shown moral and scientific leadership that puts to shame many in the academic community.”

Study 329 has been cited 572 times and profoundly influenced clinical practice at the time of publication, yet we now know that the analyses were misleading. “Liberating the data from clinical trials has the potential to benefit patients, prevent harm, and correct misleading research”, writes Professor David Henry at the University of Toronto in an accompanying editorial in the BMJ.

More coverage:

Bad medicine: Drug companies cannot be trusted to tell the truth about the efficacy of their products – outside scrutiny is crucialThe Independent

Seroxat study under-reported harmful effects on young people, say scientiststhe Guardian

Only full disclosure of drug trial results will maintain trustNew Scientist

Australian doctors lead push for better drug trials by big pharmaThe Australian Financial Review

Antidepressant Paxil Is Unsafe for Teenagers, New Analysis says – The New York Times

Landmark Analysis of an Infamous Medical Study Points Out the Challenges of Research Oversight The Chronicle on Higher Education