The study, conducted by Yale university, incorporated 4347 interventional trials that took place between October 2007 and September 2010 in 51 academic medical centres. 2892, or 66%, disseminated results – either through publication in peer-reviewed journals, or by reporting them on clinical-trial registries.
Submitting results to clinicaltrials.gov, the national US registry created in 2000, has been mandatory since 2007, within one year of completion of trials on drugs subject to FDA regulation. But reporting rates remain worryingly low, with 1166, or 26.8%, of studies complying – and only 12.6% within two years.
“While there are statutory requirements to disseminate clinical-trial results, there aren’t any enforcement mechanisms,” said Yale assistant professor Nihar Desai, who worked on the study. “Investigators know that the requirements are requirements in name only.”
“If instead, funders announced that previous performance on results dissemination would be incorporated into funding decisions, or journals would incorporate this into the adjudication of manuscripts, or leaders of academic institutions incorporated it into the promotions process, I suspect it would lead to a sense of urgency to ensure timely dissemination,” he continued.
Across individual institutions, rates of publication ranged between 35% of trials conducted and 67.2%, and reporting to clinicaltrials.gov between 4.1% and 55.4%. Many of those who do release results only do so belatedly: 21.9% of the entire study sample took longer than two years to publish; 14.2% to report to clinicaltrials.gov.
“It beggars belief that the vital step of reporting results in a clinical-trial register continues to evade the research community,” said James Cockerill, co-founder of AllTrials, “Missing results are a disservice to medical science, and a disservice to the thousands of clinical-trial participants who take part in trials. These ‘leading’ academic institutions need to get their research house in order, immediately, without exception.”
The European Medicines Agency (EMA) launched its own clinical-trials results database, EudraCT, in January. Summaries of this information will be made available to the public on the EU Clinical Trials Register. Trials involving adults have to log results to EudraCT within one year; those involving children within six months. Penalties for non-compliance remain at the discretion of participating member states, but the pharmaceutical industry had so far been broadly “cooperative”, said an EMA spokeperson.
The EMA also last year introduced a policy of proactively publishing clinical reports submitted as part of applications to market medicines within the EU. The first reports are expected to be released later this year.
“Proactive publication enables public scrutiny that is expected to establish trust and confidence in EU regulatory decisions for medicinal products,” said the EMA spokesperson. “It enables application of new knowledge in future research and increases efficiency of medicine development. It also helps avoid clinical-trials duplication, limited unnecessary patient exposure to trials.”
A recent report published in Trials Journal criticised the existing process of release-on-request, citing the lengthy correspondence involved. But its authors praised the EMA’s wider drive towards greater transparency and public access. “The EMA is trying to do the right thing for science, for medicine and especially for the patients they serve. That is precisely why it is worth fighting to improve EMA practice, perhaps with more resources and regular open audits,” said the University of Maryland’s Peter Doshi.
Dr Desai also commended the International Committee of Medical Journal Editors’ January announcement, calling for members to demand that authors submit underlying data within six months, and for clinical-trials registers to ask for plans for data-sharing as part of the submissions process.
“There seems to be consensus around the need to disseminate, but we have yet to build the framework and apparatus required to ensure that it occurs,” said Dr Desai, “That is what is desperately needed. Announcements like this are creating a climate where dissemination and data-sharing more broadly are the norm.”