You may not know that when embarking on a clinical trial, it is important to register the trial on an approved clinical trials registry, ideally before data is collected. While rules have tightened up to require this for many journals (including our own), some studies may still not be registered until after a trial is completed (especially when results are found to be positive) or may not be registered at all. Even then when a trial is registered, and the results are not what was hypothesized, is that information ever shared with the public through published studies? Are you aware that you have access to these clinical registries to see if other studies on a published topic are also there but with different results?
To help sort through the issues of registration, completion of pediatric clinical studies and their publication, Shamliyan et al. (doi: 10.1542/peds.2010-2847) looked at a large number of closed studies at ClinicalTrials.gov (the United States trial registry) to see if they were ever completed, and whether their results were ever published. Unfortunately, this story does not currently have the happiest of endings, in that more than half the registered trials for children do not have results available through the registry or in publication. As an editor, I can only peer-review the manuscripts we receive for clinical studies, but this article will make you wonder what we never receive because investigators opted not to share results publicly. That in turn, introduces bias in the body of clinical research shared through journals like ours. Hopefully this bias will be reduced as a result of articles such as this and our own requirement that all studies we consider be registered before data collection begins.
A great example of whether or not all results have been published can be found in an article we are releasing this week online by Carrasco et al. (doi: 10.1542/peds.2011-3285) on the efficacy of serotonin receptor inhibitors (SRIs) in treating repetitive behaviors in autism spectrum disorders. The authors reviewed all registered randomized double-blind placebo-controlled studies of these drugs in this setting, and found some trials were never published. The findings in this article are not dramatic but important nonetheless because the authors use their findings to challenge those who study the use of pharmacotherapy for autism to be sure that all trials and results have been published or reported publicly. Without that happening, the efficacy of these SRIs cannot be accurately assessed.
Finally, a commentary by Scott Denne (doi: 10.1542/peds.2012-0621), stresses the need for improvement in registration and reporting of results in clinical trials. Read the above two interesting articles and the accompanying commentary to learn more about a problem you may not have thought about — that not all clinical studies get registered, completed, or published!