An investigation into unpublished stroke research data has shown that 19.6% of completed clinical trials, which could potentially influence patient care, are not published in full. Researchers writing in BioMed Central's open access journal Trials describe how these unpublished studies included more than 16,000 participants and tested 89 different interventions.
Peter Sandercock and his colleague Lorna Gibson worked with a team of researchers from the University of Edinburgh, UK, to search the Cochrane Stroke Group's Specialised Register of Trials for completed trials of pharmacological interventions for acute ischemic stroke, and to determine how many of these were ultimately published. He said, "Failure to publish trial data is to be deprecated as it sets aside the altruism of participants' consent to be exposed to the risks of experimental interventions, potentially biases the assessment of the effects of therapies, and may lead to premature discontinuation of research into promising treatments".
The researchers identified 940 trials, of which 125 were not published in full. The largest trial included 856 patients, while two unpublished trials included fewer than 10 patients each. According to Sandercock, "Several of the trials we identified may have been large enough to influence clinical practice and the findings of systematic reviews and meta-analyses".
Sandercock concludes, "Well designed clinical trials should be published because their results can benefit patients, justifying the risk to trial participants from experimental treatments. We found 22 unpublished trials that reported the number of deaths. In these trials, 636 people died, but no information was available on whether the experimental drug had contributed to any of those deaths".
551 different researchers were involved in conducting these studies and of these, 72 had been involved in more than one trial that remained unpublished.