INTRODUCTION: The International Committee of Medical Journal Editors (ICMJE) does not recommend guidelines on authorship position, despite its important role in academic promotion. To identify perceptions of authorship position, the authors performed the first study of authorship position using only open-ended questioning. METHODS: Articles with at least 3 authors were selected from the top 10 cited and 20 other medical journals. The first, last, and 1 author in-between were selected and asked to identify contributions of first and last authors listed in medical journal articles. The responses were analyzed for differences between perceptions of first versus last authorship. RESULTS: Respondents were more likely to report ICMJE criteria being fulfilled by first authors; however, only 1.1% identified all 3 ICMJE conditions of authorship for either author. ICMJE criteria were no more likely identified by authors of high-impact versus low-impact journals. Significant differences existed between the understandings of appropriate roles for first- versus last-listed authors. First-listed authors were viewed at least 7 times more likely to be involved in study conduct, manuscript writing, have a major study contribution and perform the majority of the work involved. Last-listed authors were at least 7 times more likely to be viewed as having a minor or no contribution to the study, provide funding, be a laboratory head/mentor, hold a senior position and supervise/oversee the study. CONCLUSIONS: ICMJE authorship criteria were poorly identified by respondents in this open-ended, international sampling of medical journal authors. Although ICMJE criteria are ideally met by all authors, this does not seem to be well understood or accepted.
INTRODUCTION: The International Committee of Medical Journal Editors (ICMJE) does not recommend guidelines on authorship position, despite its important role in academic promotion. To identify perceptions of authorship position, the authors performed the first study of authorship position using only open-ended questioning. METHODS: Articles with at least 3 authors were selected from the top 10 cited and 20 other medical journals. The first, last, and 1 author in-between were selected and asked to identify contributions of first and last authors listed in medical journal articles. The responses were analyzed for differences between perceptions of first versus last authorship. RESULTS: Respondents were more likely to report ICMJE criteria being fulfilled by first authors; however, only 1.1% identified all 3 ICMJE conditions of authorship for either author. ICMJE criteria were no more likely identified by authors of high-impact versus low-impact journals. Significant differences existed between the understandings of appropriate roles for first- versus last-listed authors. First-listed authors were viewed at least 7 times more likely to be involved in study conduct, manuscript writing, have a major study contribution and perform the majority of the work involved. Last-listed authors were at least 7 times more likely to be viewed as having a minor or no contribution to the study, provide funding, be a laboratory head/mentor, hold a senior position and supervise/oversee the study. CONCLUSIONS: ICMJE authorship criteria were poorly identified by respondents in this open-ended, international sampling of medical journal authors. Although ICMJE criteria are ideally met by all authors, this does not seem to be well understood or accepted.