Abstract
The editors of the world's leading medical journals profess to encouraging transparency in medical research, but the recent behaviour of the International Committee of Medical Journal Editors (ICMJE), a self-appointed group of many of the world's leading medical journals, suggests they have retreated into timid self-interest or been beaten into submission by their publishers.
The ICMJE was at the forefront of the move to make clinical trial registration compulsory. Its 2004 announcement and 2005 deadline for trial registration had a major impact on the number of trials being registered. 1, 2 The ICMJE's stance was more influential than the 1997 US Food & Drug Administration (FDA) Modernization Act, which was supposed to force the registration of trials in serious and life-threatening diseases but was largely ignored by drug companies.
Unlike the 1997 Modernization Act, last year's FDA Amendments Act (FDAAA) comes with much sharper teeth, and penalties of up to $10,000 per day for companies that fail to post results of studies on new drugs on ClinicalTrials.gov, the US government-funded trial register, within 12 months of the study ending. Even before this new act was passed, some drug companies had taken the welcome initiative of posting detailed summaries of their trials on their own websites. 3
The drug industry initially opposed trial registration, and sought to restrict the elements that had to be included or to delay public release of the details until new products were licensed. Companies' willingness to post trial details and results on their own websites was increased, or in some cases mandated, by lawsuits in the US against manufacturers who failed to reveal full information about their products. 3
Yet the bold editors who led the charge and demanded the registration of clinical trials have been far less welcoming of the logical next step of posting trial results summaries on publicly accessible websites. In fact, the near silence and restrictive position of the ICMJE may hinder public access to more comprehensive trial results summaries.
Ever since the FDAAA was passed in September 2007, clinical trials investigators and sponsors have requested clarification from editors about this new law's implications for publishing results in journals. 4 But it took the ICMJE until June 2008 to address this issue and, even when it did, only one of the member journals reported the decision. Unlike previous years, which have seen simultaneous announcements in most, if not all, of the ICMJE member journals, only the Croatian Medical Journal informed its readers of the committee's latest deliberations and the changes to the ICMJE Uniform Requirements relating to trial results posting. 5
The ICMJE realizes that it cannot stop companies from obeying the US law but, instead of welcoming increased openness, it has made it clear that if companies release more than the bare minimum required by the FDAAA, this could jeopardize publication in a peer-reviewed journal. According to the Uniform Requirements: ‘Editors of journals that follow the ICMJE recommendations may consider posting of more detailed descriptions of trial results beyond those included in ClinicalTrials.gov to be prior publication’. 6
Furthermore, it warns that: ‘ICMJE may need to amend these recommendations as additional agencies institute other mandates related to results reporting’. This uncertainty does not give companies or investigators confidence or encourage innovative methods for reporting results which might benefit patients.
Of course, journal editors, publishers and societies that make money from traditional peer-reviewed journals have a major competing interest in this debate. If trial results summaries that are posted on publicly accessible websites replace conventional reports of trials in journals, their revenues could dry up. But this vested interest should not stand in the way of an opportunity to reduce publication bias, prevent non-publication of unfavourable results, increase the reliability of guidelines based on meta-analyses and improve patient care.
Enlightened editors have argued that successful journals will be those that add value by providing synthesis, commentary and interpretation. 7 Busy clinicians will still, or perhaps increasingly, require guidance about findings relevant to their practice and the implications of new research. 8
Meanwhile, the ICMJE voices predictable concerns about ‘the potential problems associated with posting preliminary research results that have not yet undergone an independent peer review process’. Undoubtedly, we need a mechanism for correcting errors in posted results, but there is no evidence that journal peer review is good at detecting such errors, or even at detecting fraudulent data. 9 Indeed, the most recent studies suggest that peer reviewers fail to spot many problems in research reports and that many reports published in journals are inadequate. 10, 11
The value of peer review generally lies in increasing readability, removing ambiguity and toning down unsupported conclusions. 12 The review systems used by journals will continue to be useful for improving the quality of published research and ensuring that commentaries are balanced and helpful, but raising concerns about un-peer-reviewed data on websites seems like scaremongering.
Structured, public reporting of clinical trial results, linked to registers (where details about trial design must be posted before studies begin) offers exciting opportunities to improve the medical evidence base. Such systems, as well as increasing transparency and public access to information about medical treatments, should reduce selective reporting and publication bias. 8 Until now, journal editors have taken a lead in exposing such problems, 13 so it is disappointing that their courage seems to have failed them when faced with a possible solution which threatens their monopoly.
Editors, regulators, consumers and manufacturers should work together to ensure that web-based technology enhances our understanding of medical drugs and devices. Safeguards are needed to ensure that reporting of trial results is not hijacked to become an irresponsible form of direct-to-consumer marketing. It would also be preferable to see a truly international endeavour to optimize results reporting, not one led by any single legislature. 14 But defensive reactions and exercises in self-interest, such as the ICMJE's reaction to public posting of trial results, are a desertion of the moral high ground that editors and publishers love to occupy when it suits them. More importantly, it is a betrayal of patient care.
Footnotes
DECLARATIONS
Footnotes
Acknowledgements
None