Let sun shine on clinical results

Shock waves rumbled through the medical community in February when the Food and Drug Administration convened a hearing to investigate the possibility that certain antidepressants did not help young patients and, in some cases, actually might have increased their risk of suicidal behavior. The previously unavailable clinical trial data that surfaced because of these deliberations changed the understanding of how the medications should be used -- especially among young patients.

It also was part of a series of events, including the agency's initial June 2003 caution regarding the antidepressant paroxetine and second expanded advisory issued in March, that highlighted the life-and-death nature of making clinical trial results more accessible. The unfolding story since has become a rallying cry for efforts to open up the clinical trial process.

The AMA has now added its voice by calling for a specific set of reforms. Concerns long have been simmering over the need for more transparency in the clinical trial process, and the AMA Council on Scientific Affairs put together a compelling case. In a June report, it outlined the potential for publication bias in pharmaceutical research and how such biases threaten to distort the medical literature, thereby affecting the validity and findings of subsequent reviews and analyses and even decisions by government agencies and health plans.

Ultimately, a comprehensive clinical trial registry is viewed as a mechanism that would correct these shortcomings by putting a wealth of trial information in a central place.

That's why delegates to the AMA Annual Meeting in Chicago last month urged the U.S. Dept. of Health and Human Services to establish such a registry for all clinical trials and require every trial to have a unique identifier.

The AMA is not alone in advancing this position. Its policy is consistent with a proposal being considered by the International Committee of Medical Journal Editors that would require all clinical trials to be listed in a registry as a requirement for publication.

The House of Medicine also recommended that all trial results be made available through publication or some kind of electronic repository. This point is critical. No repositories currently include unpublished findings from trials. But scientists estimate that as much as 50% of results are simply left to gather dust -- a fact that undermines good medicine. This information, if available, could have the potential to improve patient care or inform treatment decisions.

And currently, scientific progress might sometimes be slowed because researchers are unknowingly left to reinvent the wheel, revisiting research questions that others already have proven unsuccessful simply because there is no complete databank to check.

The availability of clinical trial findings, the openness of science and the avoidance of bias in medical journal publication are imperative to the AMA. They are connected to the very underpinnings of medical practice.

Physicians need access to information that would allow them to view the entire picture of how medicines work -- in general and in specific populations -- as well as how they compare to other therapeutic options and what choices could serve their patients best.

But according to the CSA, this ideal is not always the reality. It appears that positive findings are more likely to be seen in print than negative results, the report notes. Also, an association exists between pharmaceutical industry sponsorship of clinical research and publication of data that favor the sponsors' products. Additionally, the publication of negative results could be delayed, especially when compared with the timing of positive ones. And generally, the bulk of clinical trial information is unpublished at the time of a product's marketing.

Already, a multitude of smaller registries of trials exist, usually limited to specific illnesses or life-threatening conditions. Taking the next step, the development of a comprehensive registry, will require a cooperative venture among all participants -- government, industry and medicine. But it can be done, and physicians and patients alike will benefit from a better informed discussion of the best science and the best treatments.