Kevin Mayer Senor Editor Genetic Engineering & Biotechnology News
An association of pharma companies and publishing pros introduces a five-step plan for putting authorship guidelines into practice.
Perhaps well aware that what may be presented as “transparent” may instead be interpreted as a house of mirrors, the Medical Publishing Insights and Practices (MPIP) Initiative intends to close a credibility gap for what has become a kind of institution—industry-funded clinical research. The MPIP, an industry group founded by members of the pharmaceutical industry and the International Society for Medical Publication Professionals (ISMPP), is specifically interested in reporting issues.
Most recently, the MPIP has issued what it calls “The Five-Step Authorship Framework to Improve Transparency in Disclosing Contributors to Industry-Sponsored Clinical Trial Publications.” The framework was presented in October in BMC Medicine.
The framework emerged from a survey of 498 clinical investigators, journal editors, publication professionals, and medical writers. These participants were asked “how they would adjudicate challenging, real-world authorship case scenarios, determine the perceived frequency of each scenario, and rate their confidence in the responses provided.”
Such a survey may seem superfluous. After all, authorship guidelines have already been promulgated by groups such as the International Committee of Medical Journal Editors (ICMJE), the Good Publication Practice guidelines, the Committee on Publication Ethics, and the EQUATOR (Enhancing the QUAlity and Transparency Of health Resources) Network.
Yet the results of the MPIP survey suggest that authorship guidelines are often difficult to implement. The MPIP survey found that “low awareness, variable interpretation, and inconsistent application of authorship guidelines can lead to confusion and lack of transparency when recognizing those who merit authorship.”
For the multiple rounds of discussions it organized, the MPIP distilled the following “steps” to ensuring transparency in the authorship of clinical studies:
- Establish an authorship working group of core trial contributors as close as possible to trial commencement but prior to enrollment of first patient. For smaller trials where an authorship working group is not feasible, this committee may be a subcommittee of a larger Trial Steering Committee where one exists.
- Determine, in the context of the ICMJE authorship criteria and the specific trial, which authorship contributions will qualify as “substantial” and include these details in written publication agreements for all trial contributors.
- Implement a process to track and document in a concise format progress toward substantial contributions based on agreed-upon qualifications.
- Assess documented trial contributions for the specific study and invite those making a “substantial contribution” to participate as author(s).
- Ensure invited authors meet remaining ICMJE authorship criteria during content development and publication submission.
“As clinical trials continue to evolve in complexity, it is increasingly important for the designation of authorship for published research to be represented by those who provide significant intellectual contribution to its development and execution,” said Prof. Ana Marušić, M.D., Ph.D., department of research in biomedicine and health, University of Split School of Medicine, Split, Croatia, and lead author of the BMC Medicine article. “This is critical to ensure credit for the research, and more so, accountability and responsibility for accuracy and integrity of the work.”
Getting Clearer about Transparency
The comments in the preceding paragraph, which were made available in an MPIP press release, prompted GEN to follow up with Prof. Marušić directly.
GEN: The five-step authorship framework notes that authorship criteria can vary between clinical trials. Could such variability undermine consistency?
Prof. Marušić: It would be extremely difficult to develop a comprehensive list of criteria that accurately determine authorship for clinical trial publications both now and the in the future… [Our] conclusion was the authorship working group [should set] authorship criteria that match the particular trial activities while also following established guidelines and journal policies. By ensuring this group represents a diverse collection of trial contributors with experience from previous clinical trial work, you can provide a balanced perspective for what the authorship criteria should be.
GEN: The framework calls for establishing contribution criteria as early as possible, but allows that unanticipated authorship issues might arise afterwards. But must the “early authors” be unanimous before an authorship invitation is extended to a “late author”? (If so, why not just a majority?)
Prof. Marušić: Just as ICMJE guidelines should not be used as a selection or elimination tool, the “Five-Step Framework” is not meant to be a rigid set of rules to be voted on, but a tool to facilitate ongoing dialogue with the aim to bring the problems into awareness and to help objectively resolve authorship dilemmas that arise. Our position is that the “early authors” should be unanimous in extending authorship to ensure the process is fully transparent to all participants.
GEN: Would “holdouts” be identified?
Prof. Marušić: As mentioned above, awareness and constant discussion about these issues in view of accepted authorship guidelines is the key to resolve all predicted and unpredicted authorship dilemmas. Documenting relevant trial contributions in step 3 of the framework provides a data-driven approach to determine if further contributions warrant an invitation for authorship.
GEN: It appears the framework allows for the possibility that an entity may be allowed to make a substantial contribution to a paper, even though the entity refuses an invitation to be cited as an author. If contributors may remain ghosts, wouldn’t that leave in place one of the key “credibility gap” issues—the use of ghostwriters?
Prof. Marušić: This question probably refers to Case 4 from our study. [In this scenario, an investigator expresses disagreement with the data interpretation and conclusions as stated in the manuscript and does not want to be listed as an author or be acknowledged in the publication.] As shown in our study, most of the respondents offered solutions for this situation that involved being open and either continuing to list the contributor as an author or not listing but providing reasons for removing the author to the journal.
Similar to the ICMJE guidelines, the Five-Step Framework is about transparency of the authorship process. While it is not possible to compel anyone to serve as an author even if they meet the criteria set by the authorship working group, the goal is to make the overall process as transparent as possible.
Placing Transparency in Perspective
To round out our understanding of the MPIP Five-Step Framework, GEN confirmed that it seeks to address one of recommendations set forth in a 2012 article in the Mayo Clinic Proceedings (“Ten recommendations for closing the credibility gap in reporting industry-sponsored clinical research: a joint journal and pharmaceutical industry perspective”). It happens that the Five-Step Framework focuses on recommendation #5 (“Improve disclosure of authorship contributions and writing assistance and continue education on best publication practices to end ghostwriting and guest authorship”).
“In addition to authorship, MPIP is currently focused on recommendation #6 (“Report adverse event data more transparently and in a more clinically meaningful manner”), explained Bernadette Mansi, director of publications & disclosure practices at GlaxoSmithKline, MPIP Steering Committee Member, and a coauthor of the BMC Medicine study. “We identified key gaps for reporting adverse event data in clinical trial manuscripts through research with key stakeholders (trial sponsors, clinical investigators, and journal editors) and collaborated with journal editors to develop recommendations to address these persistent [adverse event] issues.
“Reporting negative or unfavorable results vs. positive results becomes a challenge as it relates to finding the appropriate venue for publishing data of this nature, since not all journals are interested in negative, unfavorable or confirmatory data,” continued Mansi. “From an authorship perspective, it is important to encourage researchers to disseminate all data irrespective of outcome in order to advance medical science and improve patient care.”