Clinical Trials: Show Us the Data
The Carlat Psychiatry Report, Volume 11, Number 11, November 2013
Uri Cohen, MD
Dr. Cohen has disclosed that he has no relevant relationships or financial interests in any commercial company pertaining to this educational activity.
“It is no longer possible to pretend that a report of a clinical trial in a medical journal is enough to allow full independent scrutiny ….” Fiona Godlee, Editor, British Medical Journal (BMJ)
Too often our literature presents an oversimplified picture driven by some agenda, usually commercial. We prescribers, or, more accurately, “research consumers,” need a more complete and accurate description of what’s actually observed in clinical trials. This would help us use the literature to inform the nuanced clinical judgments we face every day with each patient.
Several major campaigns are currently underway to promote greater transparency of clinical trial data. These include the AllTrials (www.alltrials.net) and the RIAT (Restoring Invisible and Abandoned Trials) initiatives. They are not aimed solely at researchers, regulators, or journal editors, nor do they follow the more familiar patterns of looking for potential conflicts of interest (COIs), limiting collaborations between academia and industry, or advancing particular ideological agendas. Instead, they’re predicated on the idea of educating and empowering research consumers like us around what we need for clinical decision making, through access to clinically relevant information from trials themselves.
What gets published in the psychiatric literature (and, indeed, all medical literature) are highly condensed proxies of original, primary, clinical trial information. The true “methods” and “results” sections of a trial are the clinical study report (CSR) and raw data sets. These documents can be quite large, and very often they’re seen in full only by the sponsor of the trial.
As clinicians we would be better served by complete and accurate access to all the trial data, to help us assess the true efficacy and side effects of an intervention.
One of the foremost illustrations of this is from the antidepressant literature. Study 329, was a multi-site trial of paroxetine (Paxil) in the treatment of adolescent depression sponsored by SmithKline Beecham (now GSK). It was summarized as supporting the conclusion that “paroxetine is generally well-tolerated and effective for major depression in adolescents” (Keller MB et al, J Am Acad Child Adolesc Psychiatry 2001;40(7):762–772).
In fact, original documents from the trial itself, later made public by court order (something that has been rarely done for any study), have raised serious issues not just with the conclusions but, much more importantly, with how closely the published paper represented the clinical study report and data sets. In significant ways that would not be apparent to even the most careful reader or peer reviewer without the original trial information, effectiveness and adverse events were not accurately represented (see, for instance, Jureidini JN et al, Int J Risk Safety Med 2008;20(1–2):73–81). As of now, the ongoing handling of Study 329 remains emblematic of the broader issue of how limited access to raw data from clinical trials may undermine the integrity of our existing literature.
That legacy is particularly relevant to newer treatments of depression, including those discussed in this month’s issue. We are moving away from an era of “blockbuster” medications towards one of more targeted and personalized care. As research consumers, whether or not we can reassure ourselves that we are being provided complete and accurate information from the trials on a particular intervention becomes even more important. A number of research sponsors are already engaging with clinical trial transparency initiatives, and it will be increasingly possible to differentiate between those researchers and companies that do—or do not—provide true transparency (see, for instance, Doshi P, BMJ 2013;347:f6754).
As information from trials and other sources grows, and related meta-analyses and “big data” syntheses expand, so too does the importance of access to the actual methods and data. Future issues of TCPR will examine these campaigns and initiatives in more detail. For now, I will close with this quote from Richard Lehman at BMJ: “Sharing of accurate and complete data is the basis for true evidence-based medicine …. This is where the rebirth of humane scientific medicine must begin.”