“The very act of writing can help clarify thinking” (1)
The effective presentation of new knowledge through published scholarship is essential to advancing the field of medicine, and the generation of outstanding scholarship is essential to advancing the careers of academic physicians. Indeed, academic reputation and standing remains largely based on the quantity and quality of publications across diverse career tracks in academic medical centers and universities (2). Academic writing is therefore important, but learning to write papers for publication is not easy. Moreover, learning to respond to the many complexities that arise when collaborating with colleagues in developing written scholarship can be very difficult—scientifically, interpersonally, professionally, and ethically. This is especially true in modern academic settings where expectations and standards for scholarly productivity are demanding, where rigorous intellectual work often derives from a shared process, and where interdisciplinary collaboration is highly valued.
Perhaps one of the most vexing issues encountered by resident physicians and early career faculty is how best to approach the assignment of authorship. There are both abstract and practical considerations involved, encompassing everything from how to deal fairly with latent power and hierarchy issues among authors to figuring out who will track down and type up the references in the right format. The aim of this article is to present a framework for thinking about authorship and to illustrate its value in working with different kinds of authorship issues and conflicts that commonly arise in the preparation of papers by more than one person.
Single authorship is in decline, and multiple authorship is on the rise (3, 4). In modern times, writing papers with others is the norm. During one’s career, academic physicians encounter opportunities to write papers with trainees, peers, senior colleagues, and other parties, such as pharmaceutical industries. Though each writing invitation offers potential for professional growth, whether by learning a new subject matter, improving writing skills, or refining interpersonal attributes such as cooperativeness, it also can provide the context for authorship issues and conflicts.
Over the last more than 30 years, the number of authors on papers, particularly papers in clinical medicine and the biosciences, have tripled to quadrupled (3, 4). This is largely due to the increasing complexity in clinical inquiries, necessitating the involvement of different specialists, and the increasing number of very large research projects, often conducted at multiple sites, sometimes in multiple countries. Some papers reporting results of large, multicentered clinical trials have more than 100 authors named in the byline. This increase in named authors on papers has led to the concern that coauthorship may, in certain circumstances, be granted to colleagues who have not rightfully fulfilled the criteria for true authorship. If colleagues who have not done the work are included as authors, it credits those who do not deserve to be credited and decreases credit to those who deserve greater credit. In addition to diluting work credit, unjustified multiple authorship also dilutes accountability for the work and greatly complicates allocation of responsibility (5). Where research mistakes or misconduct are discovered, both the liability that accrues and the responsibility for its correction is shared by all authors. This is the case even if the mistake or misconduct is isolated to a specific center or specific participant.
+
Example 1: Multiple Authorship
A department chairman asks six junior faculty members from different medical school departments to write up a research project. One or two of the six do the vast majority of the preparatory work and writing. The participation of the other four is nominal.
Gift or honorary authorship is a subcategory of multiple authorship in which senior researchers or senior academic physicians are included as authors though they have not rightfully met authorship criteria. Academic physicians of junior rank are often dependent on senior faculty, especially department chairs or section chiefs, for help in securing better jobs/salary, getting promoted, winning grants, and participating in future research projects. The power imbalance can create pressure, either explicit or implicit, on the junior authors to include senior colleagues as authors on what is the junior colleagues’ work (6). Apart from the inherent unfairness of taking undeserved credit, senior colleagues who insist upon or accept gift/honorary authorships model to junior colleagues that academic success is about maximizing the number of publications, even if actual contribution has been slim (7).
+
Example 2: Gift Authorship
A section chief demands first authorship of a paper written by a junior faculty member in his division. The paper was a product of a research project assigned by the senior colleague but one in which he had no subsequent involvement.
While multiple authorship may or may not be appropriate, “ghost authorship,” the denial of rightful authorship, is always improper. Ghost authorship occurs when a person contributes substantially to the research project or the writing of a manuscript, but is not listed as an author. Industry-paid authors, whether employees of the pharmaceutical or device industry or contracted medical media writers from a separate agency, often do the bulk of the writing of papers that report clinical research findings from studies sponsored by pharmaceutical or device industries (8). Yet these authors are usually not named in the byline.
In addition to the problem of who actually writes the manuscript, pharmaceutical sponsorship of research studies poses other authorship difficulties. Very frequently investigators, especially from nonacademic contract research organizations, have minimal if any input into the research design, no access to the raw data, and thus no ability to analyze and interpret the results of the data (9). Authors named on manuscripts are usually given the completed data analyses as generated by the sponsor. Less favorable results may or may not even be disclosed to would-be authors (10). Even when authors do write manuscripts themselves, they often are contractually obligated to obtain permission from the sponsor prior to submitting the manuscripts for publication. In the final review process, sponsors often edit the manuscript in significant ways to ensure that their product is reflected in the most favorable light (10).
Professional writers may also be hired to ghost write by parties other than pharmaceutical or device industries. For instance, some academic departments may retain professional writers to write on behalf of physician faculty members. This practice may be on the rise as clinical productivity demands increase while publication expectations remain unchanged.
+
Example 3: Ghost Authorship
A prominent psychiatric researcher is approached by a pharmaceutical company to coauthor the results of a large clinical drug trial conducted over multiple sites with other prominent senior academic colleagues. Other invited authors have agreed to participate. The researcher had no involvement with the design of the study, did not participate in the analysis of the data, and did not have the opportunity to offer revisions to the manuscript because of an impending, immediate deadline for submission.
+
Guidelines and Ethical Principles in Assigning Authorship
In an attempt to clarify authorship standards and reduce conflicts over authorship, especially within research teams, the International Committee of Medical Journal Editors (ICMJE) crafted a definition of authorship used by more than 300 medical journals throughout the world (11). It authoritatively specifies three necessary conditions for authorship.
Authorship credit should be based only on substantial contributions to (a) conception and design, or analysis and interpretation of data; to (b) drafting the article or revising it critically for important intellectual content; and on (c) final approval of the version to be published. These conditions must all be met (11).
The first author usually makes the greatest contribution to the design of the project, interpretation of the data, and writing the manuscript. The first author also assumes more credit and more responsibility than other authors and makes final authorship decisions, specifically the inclusion and ordering of other authors on the paper.
The last author is often the most senior and well-known member of the research team. In addition to fulfilling the ICMJE criteria, the senior academic physician often mentors the entire project and lends his or her name as last author to increase recognition of the papers produced by the research group.
This definition of authorship excludes mere participation in research, whether in the laboratory, hospital, or community. As such, active contributors to research enterprises, ranging from undergraduate laboratory research assistants to attending physicians serving as clinical investigators, are not considered authors. Contributions of research participants are usually noted in the contributors’ byline or the acknowledgments section.
+
Applying the Principles
Rightful authorship consists of two major components: personal effort and personal accountability. The ICMJE criteria mandate that each author participate sufficiently in the work so as to allow assumption of public responsibility for the finished product. The first two ICMJE prongs (a and b) are about doing the actual work. The third prong (c) ensures accountability for the work.
In the first example, in which multiple junior faculty are asked to participate in writing up a research project, authorship should be limited to include only the two colleagues who made a substantial contribution to the research project, to writing the manuscript, and to personally approving the final manuscript. Having been the ones directly involved, only these two authors are capable of taking responsibility for the work.
The second example, in which the section chief demands first authorship on a research paper written by a junior faculty member, illustrates how hierarchy dynamics can disadvantage junior participants. Responsibility resides chiefly with the senior faculty member not to abuse the power differential to his advantage. In this example, the section chief does not meet at least two of the three ICMJE criteria. He did not make a substantial contribution to the study design or analysis, and he did not draft or revise the paper for important intellectual content. As such, the section chief should withdraw his name entirely from the paper.
The third example, in which the senior psychiatric researcher is asked by a pharmaceutical industry to coauthor a paper, illustrates the result of common contractual arrangements between industry and academic physicians. As a byline author, the academic physician will likely be given a nearly completed manuscript which she did not write but to which she is expected to lend her name. Because the academic physician did not meet at least one of the ICMJE criteria, drafting or revising the paper for important content, she should withdraw her name from the paper.
As the above applications demonstrate, strict adherence to the ICMJE criteria safeguards the process of assigning rightful authorship. But are these solutions too idealistic? Can these outcomes be expected in the real world? Because ICMJE standards can be disregarded altogether or, more subtly, loopholes found or created, institutions would do well to draft policies about authorship assignment and have set mechanisms for mediating authorship disputes. In 2005, the Medical College of Wisconsin adopted a policy for authorship on scientific and scholarly publications. This policy adopted the ICMJE criteria as revised in 2003. In addition, it explicitly disqualifies honorary or courtesy authorships and authorships for research contributors. Finally, the policy sets forth a procedure to resolve disagreements about authorship. A concerned party first approaches the relevant department chair. If the dispute is not resolved by the chair or if the chair is involved in the dispute, the Research Affairs Committee is consulted. Finally, if the dispute is still not resolved, the dean of the medical school is approached to make a determination.
In addition to the dominant literature that promotes more rigorous application of the ICMJE criteria, other ways to think about authorship have also been advanced. Some alternatives to the conventional authorship model include contributor/guarantor and partial authorship. The contributor/guarantor model designates “contributors” instead of authors to manuscripts. At the beginning or end of the paper, each contributor is listed with reference to the specific contribution made to the project, such as literature review, project conception and design, statistical expertise, writing the first draft, and writing revised drafts (12). In addition, one contributor is designated the “guarantor.” This academic physician takes overall responsibility for the work (13). The chief advantage of this model is that it offers greater precision in authorship delineation.
The partial authorship model, in its most rudimentary form, proposes that for each published manuscript, credit for authorship is divided equally among all the listed authors. For instance, if four authors are listed for a particular paper, each author is accorded one-fourth (25%) partial authorship credit. An academic physician’s overall scholarship productivity is then judged by the sum of all partial credits earned. More sophisticated variations on this model include assigning differential partial credits to authors. Here, the first and last authors may be ascribed greater partial credit than other authors. One of the main advantages of this model is that it would discourage authorship inclusion of undeserving colleagues (14).
Authorship issues, conflicts, and occasionally misconduct arise in several forms. They can lead to disruptions, sometimes irreparable, in the working relationship of colleagues and can erode the overall integrity of the scholarship endeavor. Residents and early career physicians need to be educated about authorship rules, commonly encountered authorship problems, and equitable resolutions. Finally, junior colleagues are invited to consider alternative, perhaps superior, ways to credit authorship.