Discussion
Given the growing demand for increased transparency in publications for clinical trials [
21]-[
24], the MPIP Initiative sought to identify authorship ambiguities encountered in industry-sponsored trials not well addressed by current guidelines and to develop an approach by which authorship decisions can be more clearly and easily determined. Investigation of the rationale used by the respondent groups in adjudicating challenging case scenarios identified a number of key gaps between authorship guidelines and how these scenarios are addressed. Most important among these, survey results highlighted the variability in responses both across and within peer groups despite some groups having high familiarity with and reliance on current authorship guidelines. Some cases also tested a number of areas where authorship criteria do not explicitly provide guidance, such as whether patient recruitment is considered a substantial contribution or the impact of timing on authorship criteria fulfillment [
20], which further contributed to response variability. These findings suggest the lack of clear authorship guidance on some scenarios frequently encountered in real life, and indicate that respondents acting on judgment contrary to guidance can contribute to lack of transparency. Second, despite this heterogeneity, respondents were uniformly confident in their own rationale for these authorship decisions. Since the survey did not specify or refer to any guidelines during the case scenario section, respondents likely believed their experience and judgment are a correct approach to resolving these scenarios. Finally, clinical investigators, who make up a significant percentage of authors on manuscripts from clinical trials, appear to be concerned with the importance or significance of the contribution rather than following external guidelines to determine authorship. This perspective is exemplified by clinical investigators ranking lowest among the groups surveyed for awareness of authorship criteria.
Through discussions with journal editors, clinical investigators and industry representatives on the gaps in authorship decision-making that emerged from the survey, we identified key themes for improvement. These themes included defining the criteria for substantial contributions early in the trial, ensuring all trial contributors agree to these criteria and documenting contributions to ensure these criteria are met. Expanding on these key insights, the Five-step Authorship Framework was developed to provide a clear and flexible process to facilitate more transparent and consistent authorship decision-making.
The success of the Five-Step Authorship Framework relies significantly on convening a working group early in the trial process to discuss which trial activities should be considered a substantial contribution in the context of current authorship guidelines. Given the wide diversity of trials undertaken and likely future changes in authorship guidelines, defining a comprehensive set of potential activities that meet the definition of `substantial’ for authorship is neither practical nor durable. Instead, utilizing the experience and judgment of a diverse group of core trial contributors in an open discussion format provides a transparent process for determining fair and comprehensive authorship criteria for a given trial. In addition, these authorship criteria should align with external guidelines, such as ICMJE and GPP2, to avoid potential bias in selection of criteria that exclude legitimate authorship. Documenting these definitions in authorship agreements ensures that all trial contributors are informed in advance and provides a basis for conflict resolution should a dispute arise. The flexibility of this framework permits authorship criteria specific to each trial, as some trial groups may define significant patient recruitment to be a substantial contribution while others may not (see Case 1). This adaptable format will help facilitate the continued movement away from patient recruitment as sole criteria for substantial contribution and place greater emphasis on intellectual contributions, thus better aligning with the evolving concept of authorship.
The authorship working group should also be responsible for determining who has fulfilled the criteria of substantial contribution to help draft the manuscript, ensuring the results are consistent with the aims of the trial. The initial author list should be determined through matching the documented contributions of potential pre-specified authors to the authorship criteria. In this way, the authorship working group would have a data-driven approach to determine which potential authors fulfill criteria 1 of the ICMJE authorship recommendations (see Cases 3, 5 to 7). In addition, the pre-specified author criteria provide an objective basis for evaluating whether to add or remove an author as contributions come to light during manuscript development (see Cases 2 and 4). Given the importance of its responsibilities, the authorship working group should be a stand-alone group and not subsumed into the trial Steering Committee. As outlined, the authorship working group will not require substantial additional burden and, hence, convening such a group should be feasible.
Our survey results identified divergent opinions both within and across respondent groups for how to adjudicate the cases, with varying awareness and reliance on authorship guidelines. Previous research has shown that authorship issues can be seen as either matters of defined rules or obligation [
25]. Decisions about issues using defined rules may differ by time or place, whereas for obligation, higher order rules define the right and wrong courses for action, and decisions are not contingent on specific social rules [
26]. Our research confirms that authorship decisions bring forth issues both based on rules (`formal criteria’) and obligation (`importance of contributions’), and participants try to compromise between these perspectives. Our finding that trial publication experience has little influence on the scenario adjudication suggests these opinions are not generated by clinical trial experience but are rather a characteristic inherent to the professionals taking part in clinical trials. Therefore, forming a diverse group early in the trial can facilitate this process by bringing together those who are more aware of guidelines and setting mutually agreed upon authorship criteria from both domains through broad discussion. These findings also independently align with the recent change in ICMJE authorship requirements, which added accountability for all aspects of work described in the manuscript as the fourth criterion [
2].
Our study has limitations related to its methodological approach, which was focused on collecting qualitative input from professional groups involved in industry-sponsored clinical trial publications rather than quantitative data, which can be difficult to collect via online surveys [
27],[
28]. The survey contained the possibility for self-selection bias, with respondents potentially having a special interest in authorship and, therefore, not being representative. However, the survey was not meant to collect representative input across all relevant characteristics from the four respondent groups but to learn about possible approaches to solving real-life authorship challenges in clinical research to inform potential solutions to these problems. In addition, the Five-step Authorship Framework will not resolve all authorship questions, especially as clinical trials continue to evolve in complexity. We see the Framework as a living guideline that will change together with clinical research.
The companies included in this publication will support implementation of this tool in their clinical trials to supplement best practice and would encourage other industry members to do so as well to ensure this framework is broadly embedded for maximum impact. Given the contribution from other stakeholders, such as journal editors, the authors will be reaching out to them and other organizations to ask for their help to foster awareness and uptake for the Five-step Framework.
Acknowledgements
The cases presented in this paper were based in part on the presentation ‘Overcoming Ethical & Compliance Concerns Surrounding Publication & Authorship’ by Chipperfield, L., Clegg, A., Gillon, J. et al. Elizabeth Wager at Sideview along with Kraig Schulz and Timothy Lee, both formerly of Leerink Swann Consulting, provided assistance in study design and survey development. The MPIP Initiative obtained a list of worldwide clinical investigators currently taking part in industry-sponsored clinical trials from the Adis Clinical Trials Insight database via Bryce McMurray, Dave Dustin and Adrian Beasley at Springer Healthcare. The following publishers assisted in disseminating the survey to editors: BioMedCentral, BMJ, Dovepress, Elsevier, Informa, Nature, SAGE, Springer, Wiley-Blackwell and Wolters Kluwer. The following organizations assisted in disseminating the survey to publication professionals and medical writers: AMWA, EMWA and ISMPP. Interim findings from the survey were discussed with 15 editors and one representative from medical journals at two independent roundtables organized by the MPIP Initiative on 3 December 2012 in New York City, USA and 9 April 2013 in Alderley Park, UK. Participants and their affiliations at the time of the two Roundtables were as follows: Sabina Alam, PhD (editor in chief, BMC Medicine), Patricia Baskin, MS (executive editor, Neurology), Wendy Battisti, PhD (director, scientific and medical publications, Janssen Research & Development, LLC), Ananya Bhattacharya, PhD (director, publication policy and education, Bristol-Myers Squibb), Matthew Cahill (associate vice president, global scientific and medical publications, Merck & Co), Edward Campion, MD (senior deputy editor and online editor, New England Journal of Medicine), Juli Clark, PharmD (director, global medical writing, Amgen), Thomas Clemens, PhD (editor in chief, Journal of Bone and Mineral Research), Finbarr Cotter, MD/PhD (editor in chief, British Journal of Hematology), Cynthia Dunbar, MD (editor in chief, Blood), Robert Enck (editor in chief, American Journal of Hospice and Palliative Medicine), Lorna Fay (senior director, publications management, Pfizer), Phil Fontanarosa (executive editor, Journal of the American Medical Association), Maureen Garrity (director, publication, Astellas), Thomas Gesell, PharmD (Envision Pharma, on behalf of ISMPP), John Gonzalez, PhD (publications director, AstraZeneca), Samantha Gothelf, PharmD (executive director, global medical publications, Bristol-Myers Squibb), Thomas Grant, PhD (director - publications, AstraZeneca), Trish Groves, MBBS MRCPsych (deputy editor, BMJ and editor in chief, BMJ Open), Daniel Haller, MD (editor in chief, Gastrointestinal Cancer Research), Charlotte Haug, MD/PhD/MSc (editor in chief, Norwegian Medical Journal), Carolyn Hustad, PhD (executive director, global scientific and medical publications, Merck& Co), Richard Irwin, MD (editor in chief, CHEST), Christine Laine, MD, MPH, FACP (editor in chief, Annals of Internal Medicine), Mary-Margaret Lannon (senior director, global publications, Takeda), Timothy Lee (analyst, Leerink Swann Consulting), Neil Lineberry, PhD (director, Leerink Swann Consulting - now with Navigant Consulting), Delong Liu, MD (editor in chief, Journal of Hematology and Oncology), Bernadette Mansi (head, publications and disclosure practices, GlaxoSmithKline), Ana Marušić, MD/PhD (University of Split), LaVerne Mooney, DrPH (director, publications management, Pfizer), Teresa Peña (global director, publications, AstraZeneca), Elin Reeves, MSc (head of publications, European Respiratory Journal), Kraig Schulz (managing director, Leerink Swann Consulting), Stuart Spencer, PhD (executive editor, The Lancet), and Ryan Stover (analyst, Leerink Swann Consulting). Quantitative survey results and a preliminary Five-Step Authorship Framework were discussed with six clinical researchers who took part in the survey and provided permission for follow-up: Michael Brown, MBBS, FRACP, FRCPA (Royal Adelaide Hospital, Australia), Glen Elliott, MD, PhD (Children’s Health Council, USA), Howard Katzenstein, MD (Monroe Carrell Jr. Children’s Hospital at Vanderbilt University, USA), Ludo Vanopdenbosch, MD (private practice, Belgium), Wolfgang Vanscheidt, MD (private practice, Germany), and Jonathan Wadsley, MB, BChir, MA, MRCP, FRCR (University of Sheffield, UK).
Funding
AM and DH were supported by a grant from the MPIP Initiative. NL is an external consultant employed by Navigant Consulting and was paid by the MPIP Initiative for his work.
Competing interests
TG is employed by Envision Pharma Group and served on the MPIP Initiative Steering Committee as a representative of the International Society for Medical Publication Professionals (ISMPP), neither of which provides funding support to the MPIP Initiative. BM, AB, MG, JC, SG, JG, CH, MML, LM, and TP are employees of companies sponsoring the MPIP Initiative, as shown by their individual affiliations. NL is employed by Navigant Consulting, a strategy consulting firm that advises life science and healthcare companies. The remaining authors declare that they have no competing interests.