We recently conducted a survey of the Deputy Editors of The Journal of Bone and Joint Surgery which demonstrated that the most common method employed by journals to convey contributions and determine accountability and credit, author order and who is listed as the corresponding author, fails (1). Based upon the authorship byline, Editorial members had variable perceptions of authors’ roles (first, middle and last authors). Credit for research contributions was most often given to the first, second and last authors on the authorship byline.

It may be argued that the current system of reporting authorship without listing specific contributions of each individual author is only a problem if readers reach incorrect conclusions about who did what on a paper. Top medical journals such as JAMA, British Medical Journal, The Lancet, and Annals of Internal Medicine currently publish the contributions of each author. No surgical journal has yet to adopt this approach. Thus, to explore this issue further, we conducted a survey of the corresponding authors of all scientific papers published in The Journal of Bone and Joint Surgery from January 2003 through June 2003 with the following aims: 1) to identify each authors’ actual contributions to the published research study, 2) to determine how authorship order was determined, and 3) to determine rates of inappropriate authorship.

For each scientific article published in The Journal of Bone and Joint Surgery from January 2003-June 2003, we developed a single page questionnaire with the actual title and authorship byline as published. Each author was assigned a number (range 1-10 authors). Below the study title and author byline, respondents were asked to check any (or all) authors they believed to be involved with the following roles: study conception and design, data acquisition, analysis and interpretation of data, drafting the manuscript, manuscript revisions, statistical analysis, obtaining study funding, providing administrative support, and overall supervision of the study. We administered the final survey by email to all corresponding authors of scientific articles appearing in The Journal of Bone and Joint Surgery- American Volume between January 2003 through June 2003. We identified the corresponding author as the individual with full contact information and email address listed on the published paper. As a comparison group, we utilized responses from experienced readers about the contributions of the first, second, third and last author from our previous study (1).

Of 101 corresponding authors surveyed among 108 published scientific articles (7 authors published more than one article over the 6 month period), 82 completed the questionnaire (response rate=81%). The majority of studies focused upon joint reconstruction (41.6%) and lower extremity (18%).

Authors’ actual contributions to their research papers differed in all categories from readers’ perceptions about their roles (Table 1). Across all categories, readers were most likely to underestimate the roles of all authors; however, the greatest differences between actual contributions and perceptions were noted for the second and third authors.

Decisions about authorship order were most often based upon the amount of work done (59.6%), manuscript writing (11.2%), and study conception and design (12.4%). Respondents further added that 1) co-authors on their papers could not provide sufficient detail of the study to others (36% of respondents), 2) individuals involved in writing the manuscript (or parts thereof) were not included as authors (16.9% of respondents), and 3) some co-authors did not deservingly meet criteria for authorship ( 13.5% of respondents).

We confirm our hypothesis that the order of authorship alone is open to misinterpretation. More importantly, experienced readers, such as those in the Journal’s Editorial Board, did draw false conclusions about appropriate author credit and accountability based solely upon authorship order. It was concerning that the contributions of the second and third authors were highly under-estimated by readers which further highlights the problems with order of authorship as the sole means to draw conclusions about an individual's contributions.

The argument against publishing the contributions of each author in The Journal of Bone and Joint Surgery is primarily one of readers’ interest. Do the readers of The Journal care about who did what on a published paper? Is it the Journal’s responsibility to make authors’ contributions transparent to readers?

Listing contributions is important to those who publish papers in The Journal and to those whose promotions are based upon their academic output. Clearly, in this situation, accountability and credit for an author’s contribution are paramount.

Readers who do not rely on publications for academic promotion may still be interested in knowing who did what on a research paper. One benefit of such knowledge may be to enable a reader to address a specific question to the most appropriate author. For example, a question about the statistical analysis of a paper may best be addressed by the authors who was responsible for the statistical analysis. Ultimately, the question of whether listing who did what on papers may not be a decision for Journal editors, but rather one decided by readers.

Flanagin and colleagues surveyed 1169 corresponding authors across six medical journals to identify the prevalence of authors gift and ghost authorship(2). Ghost authors were designated as those individuals who contributed to the writing of the manuscript but did not receive authorship and gift authors were those individuals included as authors who did not deservingly meet criteria for authorship. Among 809 respondents (69% response rate), 11% admitted to ghost authors in their studies (range 11%-16%). Even more prevalent in the medical literature was the practice of gift authorship, which occurred 19% of the time (range: 11-25%). Ultimately, 1 in 4 authors misapplied authorship criteria. Our results of a 13.5% gift authorship rate are consistent with Flanagin’s study.

Inappropriate authorship is incompatible with the principles and ethics of scientific publication. Used appropriately, authorship establishes accountability and credit for scientific investigation (3). The Vancouver Group definition of authorship emphasizes the idea of responsibility. In the uniform requirements for manuscripts submitted to biomedical journals of 1993, the definition states “ Each author should have participated sufficiently in the work to take public responsibility for the content. Authorship credit should be based only on substantial contributions to conception and design, analysis and interpretation of data, drafting or revising the manuscript for critically important intellectual content, and final approval of the version to be published. Editors may require authors to justify the assignment of authorship” (3). To improve clarity and transparency of authors’ roles in a published scientific paper, surgical journals may adopt the strategy of “contributorship”. Simply put, the contributions of each author to study concept and design, data acquisition, data analysis, writing the manuscript, critical revision of the manuscript, administrative/technical support, and study supervision are explicitly reported in the published paper. However, not all contributors will be authors. For example, individuals who assisted in administrative aspects of a study but did not contribute substantially to a study should be listed in the contributors section but not as an author. Ultimately, explicit authorship criteria and publication of authors’ contributions may limit the degree of inappropriate authorship. REFERENCES:

1. Bhandari M, Einhorn T, Swiontkowski MF, Heckman JD. Who did What? (Mis) Perceptions About Authors’ Contributions Based Upon Authorship Order J Bone Joint Surg Am. 2003;85-A:1605-9.

2. Flanagin A, Carey L, Fontanarosa P, Phillips S, Pace B, Lundberg G, Rennie D. Prevalence of articles with honorary and ghost authors in peer-reviewed medical journals. JAMA 1998;280:222-224.

3. Rennie D, Flanagin A, Yank V. The contributions of authors. JAMA 2000; 284:89-91.