In this session, the speakers summarized the requirements of clinical-trial registry posting on the Web and discussed how they are related to publishing of clinicaltrial results in medical journals.

Tracy Beck provided insight into US legislation on clinical trials. She emphasized registration of clinical trials and disclosure of clinical-trial results as two main interests of science editors. Registration and results disclosure are regulated by the Food and Drug Administration Amendments Act (FDAAA) of 2007, Maine legislation, and the Food and Drug Administration Modernization Act, Section 113. Moreover, the World Health Organization (WHO) and the International Committee of Medical Journal Editors (ICMJE) have guidelines on clinical-trial registration and result disclosure. Beck said that as of 8 December 2008, Maine has recognized results only on clinicaltrials.gov as compliant with its regulations.

Beck said that according to the FDAAA, results of trials completed before a drug is initially approved must be posted not later than 30 days after approval. Results of trials completed after a drug is initially approved, however, must be posted not later than 1 year after the earlier of the estimated and actual completion dates. Beck noted that the results databank (clinicaltrials.gov) must be expanded not later than 27 September 2010 to include lay and technical summaries of results if these summaries can be written without being misleading or promotional.

Beck noted some key challenges posed by the regulations and guidelines. For example, whereas the FDAAA and Maine require clinical trials to be registered within 21 days of when the first patient is enrolled, WHO and ICMJE require registration before recruitment of the first patient. WHO has not addressed the timing of result disclosure, and ICMJE’s policy is still evolving. Beck said that it was not clear whether journals will consider lay and technical summaries as prepublication disclosure.

Trish Groves discussed current requirements for clinical-trial registration and used the BMJ to show how one journal relates to guidelines on clinical-trial registration and result disclosure. She gave an overview of the ICMJE rules on clinicaltrial registration, saying that in the case of trials in which “human participants are prospectively assigned to one or more health-related interventions, the effects on health outcomes need to be evaluated.” She explained that clinical trials begun before 1 July 2005 can be registered retrospectively but only before journal submission. She enumerated ICMJE criteria for suitable registries, including free public accessibility, openness to all prospective registrants, assurance of validity of registered data, and electronic searchability. Groves said that the ICMJE guidelines change often, so science editors ought to be alert.

Groves said that the BMJ enforces clinical-trial registration policies. First, research editors check registry entries regularly at the initial internal peer-review stage and reject trials that are eligible but unregistered. Second, research editors ask authors for registration details up front; if they are not supplied, the editors ask again before external peer review. Third, they ask for the full protocol and show it to reviewers. If a trial is registered, the research editors do not consider registration and results posting to be before publication, because the postings are not peer reviewed. On publication, the BMJ includes trial-registration details in each abstract.

Groves also briefed participants about WHO and European Union policies on pediatric trials. She said that since 2009 the European Medicines Agency has mandated trial registration and posting at eudract.emea.europa.eu with public access.

Groves suggested that editors of medical journals consider checking on Web sites such as clinicaltrials.gov and then contacting principal investigators about the possibility of publishing their studies.