Enhancing transparency and efficiency in reporting industry-sponsored clinical research: report from the Medical Publishing Insights and Practices initiative

Industry-sponsored clinical research has become more open and transparent in recent years (1), spurred by changes in policy, regulation and technology, and fostered by a general trend towards increased information access and sharing, which is also apparent across government and in other industrial sectors such as finance. Yet, despite advances in technology, publication of trial results in the form of conventional studies in peer-reviewed journals remains the gold standard and cornerstone of scientific disclosure. One expected consequence of increased transparency surrounding clinical trials is that journals will experience a significant growth in the number of manuscript submissions from industry-sponsored research, particularly manuscripts reporting the results of early-phase research. This increase could be substantial since, until recently, about 90% of Phase I studies and 60% of Phase II studies were never published (according to observations from one major company) as these studies tended to be considered of low interest or confirmatory in nature. The increase in the number of trials seeking publication comes on top of underlying, but more gradual, growth in the amount of research being carried out. Since 2005, the bibliographic database Medline has been growing at a rate of 2000–4000 references per day (2), and manuscript submissions to journals continue to increase according to many editors and publishers. In the next few years, the number of reports of clinical trials requiring publication is likely to increase still further. The US Food and Drug Administration Amendments Act (FDAAA) passed in 2007 requires summaries of the results to be posted on the ClinicalTrials.gov website within 12 months of the study end for most trials (3). As the ClinicalTrials.gov results posting format does not permit any textual introduction, interpretation or discussion of the findings, many funders are seeking to publish a traditional journal article at the same time as posting to provide more context. Increased trial registration and results posting may also increase the number of publications on compounds that do not get licensed. Another manifestation of growing transparency, arising from concerns about publication misconduct and poor reporting, is that editors require increasing amounts of information from authors at submission. For example, editors may request copies of related studies to check for overlap and of the study protocol to check that the study is reported responsibly. Journal submission systems also increasingly include questions about research ethics (such as IRB approval and patient consent) and competing interests. To ensure high standards of reporting, journals may also require evidence of compliance with guidelines, for example by submission of a completed CONSORT checklist with the manuscript. Thus, while increasing transparency of reporting may raise standards and bring benefits to readers, it comes at a cost of additional burdens on those reporting research. The evolution of reporting standards and guidelines is shown in Figure 1. Landmarks in Medical Publication Practice Guidelines. AMWA, American Medical Writers Association; CSE, Council of Science Editors; EMWA, European Medical Writers Association; FDA, Food and Drug Administration; GPP, Good Publication Practice for pharmaceutical companies; GPP2, Revised version of GPP (2009); ICMJE, International Committee of Medical Journal Editors; ISMPP, International Society for Medical Publication Professionals; MPIP, Medical Publishing Insights and Practices; PhRMA, Pharmaceutical and Research Manufacturers of America There have been suggestions that the publishing ‘ecosystem’ may not be ready to accommodate the expected flood of publications disclosing the results of all industry-sponsored trials which, when combined with increasingly stringent reporting requirements, is putting the system under stress. However, through close collaboration among all interested parties, it is possible to improve the efficiency of the publication process, thus reducing stress on journals, peer reviewers and authors, increasing publishing capacity and transparency, and accelerating public access to clinical trial results. Editors, publishers and pharmaceutical industry employees met recently to share their perspectives on the new disclosure paradigm for industry-sponsored clinical trials and to discuss opportunities to enhance the publication process. The Medical Publishing Insights and Practices (MPIP) initiative was started in 2008 by members of the pharmaceutical industry and the International Society for Medical Publication Professionals (ISMPP) to elevate trust, transparency and integrity in publishing industry-sponsored studies. The MPIP is a collaboration sponsored by Amgen, AstraZeneca, GlaxoSmithKline, Pfizer and ISMPP with the specific goals of: (i) gaining a better understanding of the issues and challenges faced by journals in publishing pharma-sponsored studies; and (ii) promoting collaboration between the medical publishing and pharmaceutical industries to develop and implement solutions to these challenges (Bristol-Myers Squibb also participated in phase one of this collaboration). MPIP has been a forum for journal–pharma engagement since its inception and aims to continue to promote cooperation between these industries. The first phase of this collaboration was focused on obtaining insights from journal editors and publishers on the issues and challenges perceived by journals to be barriers to transparency and trust. Results of this research were presented at the ISMPP Annual Meeting (April 2009). In the second phase of the initiative, which is ongoing, the focus is on brainstorming and implementing solutions to the issues raised in phase I. Most recently, MPIP organised a roundtable meeting to provide a forum for journal editors and publishers to discuss current issues with publication professionals from the pharmaceutical industry (see Acknowledgements for the list of participants). This article summarises the main discussion points and, in particular, outlines solutions that were proposed. The meeting identified a number of trends, which are considered in turn. Everybody involved in the publication of research is facing new challenges, but many of these can be lessened or resolved by collaboration and increased understanding. The recommendations from the meeting are summarised in Table 1. The publication of more clinical trial findings is likely to be beneficial, so venues need to be developed to accommodate the increased volume of studies that should be made publicly available. This might involve developing new models and alternatives to traditional peer-reviewed journals. As one participant noted ‘we need to find the most efficient way to find the right home for every paper’. The costs of electronic publishing are less volume-dependent than those for traditional paper-based journals, and therefore offer possibilities to escape from fixed page budgets. However, readers and advertisers still require printed copies, so some journals have responded by making the web and printed content different. For example, the BMJ now regards the web-based version of articles to be the definitive one, and allows authors almost unlimited space for this, including additional tables and multimedia items, while the print version of research reports has been shrunk to the ‘Pico’ format, which is essentially an extended structured abstract (4). The journal can thus publish more studies and include more non-research features in the printed journal without exceeding its page budget. Web publishing has also led to the development of alternative economic models for journals, in particular Open Access journals, which charge authors or funders, and new means of generating income for subscription-based journals such as pay-per-view facilities for individual articles. Such economic models for publishing are more flexible than the traditional system of funding journals mainly through annual subscriptions, and journals that can increase their revenue if they publish more studies are likely to be more responsive to the increased volume of submissions. The difficulty of obtaining peer reviews affects editors directly and authors indirectly, as reviewer refusals can cause delayed decisions. Funders and institutions might work with editors to seek ways to encourage suitably qualified individuals to act as reviewers. Publishers and editors might also develop methods for sharing reviewer comments among a wider pool of journals and therefore facilitating consideration by an alternative journal if rejected by the first target. Some journals suggest that authors should supply reviewer comments from other journals and find these can be useful, especially if the rejection was because of an article not fitting the scope of the first journal rather than the quality of the research. However, editors find authors are reluctant to admit that they have already been rejected by another journal. One editor suggested that authors consider a rejection like a ‘prior conviction’ and therefore do not want to admit to this by supplying reviewer comments from another journal. All those involved in publishing research could also work together to reduce editors’ difficulties in getting suitably qualified individuals to review articles. This might include initiatives such as identifying a wider pool of reviewers, increasing recognition for reviewers (for example by awarding CME credits for it) and providing reviewer training. Some journals ask authors to suggest reviewers, but this has had little effect on the problem. Data from one publisher suggested that reviewer refusal rates were increasing steadily year on year, but evidence from individual journals suggests there may be considerable variation, e.g. between disciplines, so some editors are already experiencing an acute problem. One participant expressed concern with the quality of reviews and wondered how this could be improved, when studies suggest that reviewer training has limited effect. An increase in cursory or superficial reviews may be a further indication of the low priority given to reviewing and the need to improve incentives for this activity. Publishers and editors might also seek to develop uniform standards and formats for submission to reduce the workload for authors. The ICMJE Uniform Requirements (5) were developed at a time when manuscripts had to be typed and one of their aims was to eliminate retyping and thus make it easier to submit to a different journal following rejection. With the development of word processing and bibliographic software, editors now seem to expect submissions to be formatted according to their individual requirements, so the goal of uniformity seems to have receded. Recognising this problem, and also perhaps to capitalise on having a series of journals, some publishers facilitate the transfer of manuscripts within their own systems. One model that has been successfully adopted by The Lancet and PLoS Medicine is to link a prestigious general journal to other titles (e.g., Lancet Oncology, Lancet Neurology, etc., and PLoS One). If a submission is rejected by the general journal, it can be passed on to another title (and, occasionally also in the opposite direction) if the authors agree. Other publishers of related journals have adopted more informal journal cascades or submission sharing but uptake by authors has been variable. In the field of neurology, editors of journals published by a number of organisations have formed a consortium allowing reviewers’ comments to be shared by different journals and thus increase the efficiency of the publication process. Further similar developments might benefit both editors (who find it hard to obtain reviews) and researchers (who would save time and trouble over re-submission). In addition, adopting uniform requirements for electronic submission and manuscript formats could reduce the workload of re-submission. Some journals encourage pre-submission inquiries and this could also save time if a preliminary decision could be given on a short-form submission such as an abstract or results posted on clinicaltrials.gov or on an unformatted manuscript. However, academic editors of smaller journals may not feel equipped to handle pre-submission inquiries. Journals could also reduce the number of unsuccessful submissions by providing more detailed information about their editorial scope and interests and posting this in a prominent location on their journal website. The BMJ has recently published a checklist for authors to determine whether their study is suited for the journal (6). Rejecting large numbers of studies causes an administrative burden to journals and is equally unpopular with authors. Journals could perhaps reduce inappropriate submissions by providing more information to authors and publication planners. One editor noted that a booklet for authors produced by his journal was very popular and in constant demand. Funders and editors might collaborate to educate researchers and planners to improve their journal selection tactics and avoid the over-ambitious targeting of inappropriate journals, which sometimes occurs. There is pressure to increase the speed of publication to ensure that full manuscripts are published by the time that results summaries must be posted (under FDAAA legislation). Some journals offer expedited review, at an additional charge to authors/funders, and achieve this by having a pool of trusted reviewers who are paid for delivering a rapid review. Initiatives described above to reduce inappropriate submissions that result in rejection and to facilitate re-submission following rejection should also increase the speed of publication. Commercial research sponsors, such as drug companies, need to restore confidence in their practices. This might be achieved by greater openness (e.g. publishing their publication policies – which some companies are now doing), explaining how they operate and describing their quality control mechanisms. One area of negative media focus has to do with the use of ‘ghost writers’. Although many editors recognise the contribution that publications professionals can bring to the process, media reports of ‘ghost writing’ (i.e. the unacknowledged use of writers), often linked to court cases describing actions that occurred several years ago, have fuelled misperceptions about legitimate use of medical writers and raised concerns among the general public and commentators (7). Media attention has focused on bad practice and irresponsible behaviour, and initiatives to raise standards and ensure ethical conduct have been much less widely reported, as good news rarely makes the headlines. The result has been increased concern about commercial companies’ publication practices and a loss of confidence in industry-sponsored research. Some commentators have questioned funders’ involvement at any stage of a publication (8), and there have been calls for competing interests to be disallowed rather than disclosed (9). Codes of conduct such as GPP (10,11) and the PhRMA policy (12) and specific guidelines such as those from EMWA (13), ISMPP (14), and WAME (15) may help establish the legitimate contributions that professional writers and publication planners can make, and also describe unacceptable behaviours. These societies can also help educate audiences on the distinction between ‘ghost writing’ and the legitimate use of professional medical writers whose contributions are disclosed and acknowledged appropriately. Increasing requirements for publications to conform with a variety of guidelines and journals’ individual styles may also require increased professional assistance from medical writers. Most journal editors recognise that the involvement of a professional writer can raise reporting standards and improve compliance with guidelines and overall editorial quality. The World Association of Medical Editors therefore states that ‘editors should make clear in their journal’s information for authors that medical writers can be legitimate contributors’ (15). Therefore, journal instructions and academic institutions’ policies need to define legitimate roles and actively encourage disclosure. To be effective, policies and codes of conduct need to be widely disseminated, endorsed by companies, with systems in place to monitor adherence. Concerns about redundant and overlapping publications and inappropriate publication strategies might be allayed by greater openness on the part of funders. Some journals request copies of related studies, and it might be helpful for funders to share details of their publication plans with editors to explain how many studies they plan to develop and how the studies relate to one another. Such information could be included in the cover letter. Submitting protocols alongside manuscripts and requiring registration for all trials may also help reduce selective and misleading reporting. Research funders are increasingly pro-active in publishing studies, due to new industry policies to increase transparency and pressure to ensure, where possible, journal publications are timed to coincide with the required posting of results on ClinicalTrials.gov, in order to provide appropriate context and expanded information. Coupled with pressure on academic and clinical researchers, who rarely have secretarial or administrative support from their own institutions, it is not surprising that companies are increasingly taking greater responsibility for publishing reports of their trials. Distrust and low credibility have become barriers to mutually beneficial relationships between big pharma and journal editors/publishers. As rebuilding trust begins with mutual understanding and insight, the overarching objective of the Medical Publishing Insights and Practices initiative is to obtain insights from journal editors/publishers that can lead to more effective partnership and solutions to increase trust and transparency in publishing of pharma-sponsored research. The MPIP roundtable meeting identified several areas in which the transparency and efficiency of publishing clinical research could be improved. We plan to continue to discuss solutions and encourage their realisation. We believe that only if editors, publishers, academic researchers and funders start to view each other as collaborators will we succeed in developing systems for publishing research that are sustainable and effective in our mutual goal of improving healthcare through the efficient and transparent transmission of knowledge. MPIP is a collaborative venture funded by Amgen, AstraZeneca, GlaxoSmithKline and Pfizer and coordinated by Leerink Swann consultants. We thank the participants at the roundtable meeting in Vancouver: Tad Campion (New England Journal of Medicine), Leighton Chipperfield (Elsevier), Leslie Citrome (International Journal of Clinical Practice), Robert Enck (American Journal of Hospice and Palliative Medicine), Lorna Fay (Pfizer), Chris Graf (Wiley-Blackwell), Trish Groves (BMJ), David Kent (Trials Journal/Journal of General Internal Medicine), Lionel Lewis (British Journal of Clinical Pharmacology), Jay Magrann (Current Medical Research & Opinion/Informa), Sheryl Martin (Obstetrics & Gynecology), Ann Murphy (The Oncologist), Jigisha Patel (BMC Medicine), John Shelton (Journal of Clinical Psychiatry), Gene Snyder (Envision Pharma), Philip Walson (Clinical Therapeutics), Maja Zecevic (The Lancet). Roundtable participants do not necessarily endorse the contents of this article. We also thank the staff from Leerink Swann who organised the meeting: Roland Andersson, Frank David, Eliza Kamenetsky. The authors are employees of the companies that are funding (or have funded) MPIP as shown by their individual affiliations, except for Elizabeth Wager who is a self-employed consultant who was paid by MPIP for her work on the meeting and preparing this report, and Al Weigel who is 2009–2010 President of ISMPP. All authors to the meeting in following the Elizabeth Wager the All authors the manuscript and to the manuscript.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.