Variations in Instructed vs. Published Word Counts in Top Five Medical Journals

INTRODUCTION

Medical journals set word limits for research papers. After editorial and peer review, manuscripts are revised for additional detail and clarity. Assuming that a published article reflects the best possible version of a given paper, we assessed the degree to which the word count in each published article deviated from that in the journal’s instructions for authors.

METHODS

We included all original full research articles published in 2015 in the top five general medical journals (The New England Journal of Medicine [NEJM], The Lancet, JAMA, The BMJ and Annals of Internal Medicine), excluding meta-analyses and systematic reviews. The word count was calculated for each article’s abstract and main text, excluding figures, tables, headers and footers. We then calculated the absolute mean difference (95% confidence interval [CI]) in word count between the maximum number stated in the journal instructions and the actual number of words in the published articles within each journal, for the entire year of 2015.

RESULTS

A total of 719 articles were included (file available upon request). The abstract word limit for each journal varied from 250 (NEJM) to 400 (BMJ). Most published abstracts exceeded the allowable word count (Fig. 1, upper). Overall, Lancet had the largest absolute excess, 134 words (95% CI 119 to 149), between published and instructed abstract length, equivalent to a relative difference of + 45%; followed by NEJM (62 words over [95% CI 56 to 68], + 25% relative difference); JAMA (51 words over [95% CI 44 to 58], + 15% relative difference); and Annals of Internal Medicine (22 words over [95% CI 16 to 28], + 8% relative difference). In contrast, The BMJ, which had the highest abstract word allowance of 400 words, was 54 words (95% CI −65 to −43) below that limit, a relative difference of −14% (Fig. 1, upper).

Figure 1.

Instructed vs. published word counts for research articles in 2015, appearing in The New England Journal of Medicine (NEJM), JAMA: The Journal of the American Medical Association (JAMA), The Lancet, The BMJ and Annals of Internal Medicine (Ann Intern Med). The top panel presents the mean word counts for study abstracts, and the lower panel the word counts for the main text. Shown are the instructed word counts (red squares) and published word counts (black diamonds).

For the main text, the word limits ranged from 2700 (NEJM) to 4400 (BMJ). The prescribed limits were exceeded in NEJM (+ 25% relative difference), JAMA (+ 21% relative difference) and Lancet (+ 10% relative difference), while Annals of Internal Medicine (−1% relative difference) and BMJ (−7% relative difference) were within their limits (Fig. 1, lower).

DISCUSSION

The International Committee of Medical Journal Editors (ICMJE) puts forth uniform requirements designed to help authors and editors produce accurate, ethical and clear biomedical research articles.¹ Among the top five medical journals, there is no consensus regarding the optimal word limit for research papers. Clearly, some journals are flexible in allowing articles to exceed the stated word maximum. This may be a healthy reflection of the peer and editorial review process, with the final published version being the “highest standard.” However, if this is so, then the word length set by journals could be detrimental, requiring authors to unnecessarily omit important details at the time of submission. Among the top five medical journals, in additional to journal impact factor, article length was shown to predict the number of times an article was cited,² as was the case for astronomy journals.³ The ICMJE might find it helpful to engage in an open debate about the optimal abstract and text word count for research articles. Additionally, authors should be aware that the stated word count set by medical journals may be open to negotiation.

Compliance with Ethical Standards

Conflict of Interest

The authors declare that they have no conflict of interest.