A new research paper in JAMA looking at 1017 clinical trials has found that most trials stopped before the planned end remain unpublished, and overall 56% never published results. An international research group was given access to the records of all the clinical trials approved by six ethics committees in Switzerland, Germany and Canada between 2000 and 2003. They searched for journal articles written about these trials and found only 567 papers, so only 56% of the trials had results published in a journal.

The researchers found that 25% of the trials were ended earlier than planned, primarily due to problems recruiting volunteers. Around 60% of those trials had not been published. The researchers identified nine trials that were stopped because it became clear that the treatment had a beneficial effect – all of these trials were subsequently published as journal articles.

The researchers conclude: “The non-publication of results from discontinued—or from completed—[clinical trials] represents a waste of valid data that could contribute to systematic reviews and meta-analyses.”

Prevalence, Characteristics, and Publication of Discontinued Randomized Trials JAMA. 2014;311(10):1045-1052. doi:10.1001/jama.2014.1361 http://jama.jamanetwork.com/article.aspx?articleid=1840235