- Anton Pottegård, research fellow1,
- Maija Bruun Haastrup, medical doctor2,
- Tore Bjerregaard Stage, PhD student1,
- Morten Rix Hansen, PhD student1,
- Kasper Søltoft Larsen, PhD student1,
- Peter Martin Meegaard, pharmacist3,
- Line Haugaard Vrdlovec Meegaard, pharmacist4,
- Henrik Horneberg, editorial manager1,
- Charlotte Gils, medical doctor2,
- Dorthe Dideriksen, pharmacist2,
- Lise Aagaard, professor1,
- Anna Birna Almarsdottir, professor12,
- Jesper Hallas, professor12,
- Per Damkier, consultant, assistant professor12
- 1Clinical Pharmacology, Institute of Public Health, University of Southern Denmark, DK-5000 Odense C, Denmark
- 2Department of Clinical Chemistry and Pharmacology, Odense University Hospital, Odense C, Denmark
- 3Capital Region Pharmacy, Clinical pharmaceutical Services, Rigshospitalet, Copenhagen, Denmark
- 4Capital Region Pharmacy, Clinical pharmaceutical Services, Nordsjællands Hospital, Hillerød, Denmark
- Correspondence to: A Pottegård apottegaard{at}health.sdu.dk
- Accepted 7 November 2014
Abstract
Objectives To describe the development of acronym use across five major medical specialties and to evaluate the technical and aesthetic quality of the acronyms.
Design Acronyms obtained through a literature search of Pubmed.gov followed by a standardised assessment of acronym quality (BEAUTY and CHEATING criteria).
Participants Randomised controlled trials within psychiatry, rheumatology, pulmonary medicine, endocrinology, and cardiology published between 2000 and 2012.
Main outcome measures Prevalence proportion of acronyms and composite quality score for acronyms over time.
Results 14 965 publications were identified, of which 18.3% (n=2737) contained an acronym in the title. Acronym use was more common among cardiological studies than among the other four medical specialties (40% v 8-15% in 2012, P<0.001). Except for within cardiology, the prevalence of acronyms increased over time, with the average prevalence proportion among the remaining four specialties increasing from 4.0% to 12.4% from 2000 to 2012 (P<0.001). The median combined acronym quality score decreased significantly over the study period (P<0.001), from a median 9.25 in 2000 to 5.50 in 2012.
Conclusion From 2000 to 2012 the prevalence of acronyms in trial reports increased, coinciding with a substantial decrease in the technical and aesthetic quality of the acronyms. Strict enforcement of current guidelines on acronym construction by journal editors is necessary to ensure the proper use of acronyms in the future.
Introduction
Acronyms—abbreviations formed from the initial components of a phrase or word1—improve the perception of complex, written information.2 3 Within the health sciences, researchers’ use of acronyms holds a long tradition, with the likely intention of branding their work into the minds of fellow researchers, clinicians, editors, or lay people.4
The use of acronyms in health sciences has been subject to intense debate.5 Authors have advocated against such use as they claim it has turned into MMMMM—a major malady of modern medical miscommunication6—and asserted that positive sounding acronyms are misused in clinical trials with negative outcomes.7 8 It has been suggested that editors should insist on eliminating the use of positive sounding acronyms9 or even bring a HALT (help acronyms leave (medical) trials) to the use of acronyms altogether.10
This heated controversy seems to be based on opinion rather than founded on rigorous scientific research. Few quantitative studies of this important topic exist, and to our knowledge studies on the technical and aesthetic quality of acronyms are virtually absent. We describe the extent and quality of acronym use within different medical specialties.
Methods
We included five major medical specialties in the analysis: cardiology, endocrinology, rheumatology, pulmonary medicine, and psychiatry. For each specialty we selected a disease that was central to the discipline and identified the most appropriate MeSH term for that disease. Using these MeSH terms, we searched PubMed for studies containing acronyms in their title that did not refer to a method (for example, randomised controlled trial). We restricted the search to randomised controlled trials in humans, reported in English, and published during 2000-12.
Acronym identification
In the included studies we looked for the meaning of the acronym in several sources in the order of title, abstract, full text, and trial registration (if any). AP, MBH, and MRH performed the initial search, further aided by CG, TBS, KSL, PMM, LHVM, and DD in identifying acronyms. In case of any uncertainty by the single reviewer, the information was double checked by both MBH and MRH.
Acronym evaluation
The evaluation consisted of both positive (BEAUTY, Boosting Elegant Acronyms Using a Tally Yardstick) and negative (CHEATING, obsCure and awkHward usE of lettArs Trying to spell somethING) criteria (box). We used a two step Delphi method to agree on these criteria.11 The final score assigned to each acronym was obtained by adding the BEAUTY and CHEATING score.
Criteria used for evaluation of acronyms
Positive criteria
BEAUTY—Boosting Elegant Acronyms Using a Tally Yardstick
Scores calculated:
1.5 points for each letter of acronym correctly used—that is, letters in the acronym that corresponded to the first letter in a word of the title
5 points if acronym was a real word
2 points if acronym related to the specialty of study
Negative criteria
CHEATING—obsCure and awkHward usE of lettArs Trying to spell somethING
Scores calculated:
−2 points for each letter incorrectly used—that is, not the first in a word
−1 point for each letter that was almost correctly used—that is, followed a correctly used letter
−1 point for each word in the full title not accounted for in the acronym (not counting prepositions and adverbs)
−2 points for each letter in the acronym that could not be attributed to a word in the full title
To assess the inter-rater reliability of the combined score we rescored 100 randomly selected acronyms.12 13 We also subjectively evaluated whether the acronym could be considered as “cool” (for example, had a witty cultural reference) or pretentious, or the quality of the language of the full title had suffered in a strained attempt to make the acronym fit better. We did not include these subjective measures in the overall score.
Finally, we identified a list of honourable and dishonourable mentions that for some reason did not obtain a particularly high or low score but still deserve to be highlighted.
Analysis
We reported the proportion of acronym use and the median quality score of acronyms over time. We reported the 25 highest and lowest scoring acronyms and the honourable and dishonourable mentions selected by the reviewers. One way analysis of variance was used to compare overall scores between different medical specialties. To determine if the prevalence of acronyms in cardiology was higher than that in the other specialties, we performed a χ2 test. The change in quality of acronyms over time was assessed using a Spearman’s rank correlation. For the top and bottom 25 acronyms, we identified the impact factor of the publishing journal in the year of publication, total number of citations, and average yearly citations.14 We compared the 25 highest and lowest scoring acronyms using an unpaired Student’s t test after log transformation.
Results
A total of 14 965 publications were identified, most of which were within the disciplines of cardiology (n=5063) and endocrinology (n=4994). Overall, 18.3% (n=2737) of the publications contained a total of 1149 unique acronyms (table 1⇓). The prevalence proportion of acronyms increased over time for all specialties, except for cardiology (P<0.01, fig 1⇓).
Basic search algorithm and results
Fig 1 Prevalence proportion of acronyms over time
Excluding 197 acronyms where we could not identify the full meaning, 952 acronyms underwent further evaluation. The median quality score was 6.5, with scores ranging from −18 to 22 (interquartile range 3.0-10.5). One way analysis of variance showed that the correlation between score and medical specialty was not statistically significant. Tables 2⇓ and 3⇓ present the 25 highest and lowest scoring acronyms. Over the study period the acronym quality score declined significantly (P<0.01, fig 2⇓). The honourable and dishonourable mentions are listed in tables 4⇓ and 5⇓.
25 best acronyms according to composite BEAUTY and CHEATING criteria (see box for details of scoring)
25 worst acronyms according to composite BEAUTY and CHEATING criteria (see box for details of scoring)
Honourable mentions
Dishonourable mentions
The intraclass correlation coefficient of the combined score was 0.91 (95% confidence interval 0.86 to 0.94), indicating almost perfect agreement.
Overall, 4.4% (n=42) of the acronyms contained poor language in an attempt to improve on the acronym, 11.5% (n=109) were designated as “cool,” with cardiology and pulmonary medicine in the lead with 12.9% and 10.7%, respectively, and psychiatry, rheumatology, and endocrinology following with 2.8%, 5.8% and 9.8%, respectively. Although 12.8% (n=122) of all acronyms were classified as excessively pretentious, this proportion varied between specialties: from psychiatry (19.4%), rheumatology (15.4%), pulmonary medicine (14.3%), endocrinology (13.9%), to, lastly, cardiology (11.8%).
The top 25 acronyms were published in journals with a median impact factor of 10.2 (interquartile range 6.8-28.9), whereas the bottom 25 had a median impact factor of 6.1 (3.3-11.4). This difference failed to reach significance (P=0.05). The top 25 acronyms had more total citations (median 69 v 29, P=0.02), whereas citations per year did not differ significantly (median 14 v 7, P=0.09).
Discussion
This quantitative and qualitative systematic study showed an increasing use of acronyms in the manuscript titles of four major medical specialties coinciding with a noticeable decline in the quality of the acronyms over time.
Cardiologists’ obsession with acronyms is well documented and has been the subject of in-depth analysis.6 8 15 16 17 18 Although the “10 commandments of acronymology” was suggested in 2003,6 these were never formally adopted by any cardiological society. No biologically plausible reason explains the apparent obsession with acronyms in cardiology. It may be hypothesised that fierce academic competition spurred the origin of such use, and that new researchers have been subject to peer pressure and assigned acronyms at all cost to avoid academic marginalisation and ridicule. Another hypothesis is a reversal of the process: cardiologists may first concoct a clever acronym and then design a trial to fit that acronym.
Between the top 25 and bottom 25 acronyms, studies with good acronyms had more citations than studies with poor acronyms. For manuscript titles with good acronyms we observed a non-significant trend towards publication in journals with a higher impact factor. Bibliometric assessment of academic production is closely associated with successful funding,19 20 as well as personal satisfaction, pride, and peer prestige of researchers.21 22 23 In line with our findings, a study found that using an acronym was associated with a twofold increase in annual citation rate.24 Furthermore, the length of a manuscript’s title has been identified as an independent predictor of citation rate.25 In that study, however, the authors failed to account for acronymisation in their regression model. This possibly represents a strong confounder, and we are confident that adjusting for acronym use would eliminate the apparent signal from title length.25 A causal relation cannot be inferred from our results though, and the issue of reverse causality remains a concern. We cannot exclude that well chosen and aesthetically satisfying acronyms increase the impact factor of the journals publishing them. However, we find it reassuring that acronyms that are technically correct and aesthetically satisfying are seemingly appropriately rewarded.
The Tolstoy manoeuvre
We observed several examples of what we designate the Tolstoy manoeuvre: if the title appears to quote extensive passages from War and Peace (>1400 pages), authors can fit any desired acronym by cherry picking letters. A striking example is ADJUST (Abatacept study to Determine the effectiveness in preventing the development of rheumatoid arthritis in patients with Undifferentiated inflammatory arthritis and to evaluate Safety and Tolerability, table 3). Incidentally, this represents a failed Tolstoy manoeuvre, as the “J” is not accounted for.
The good
Good acronyms are thoughtful, well designed, orthographically correct, and aesthetically satisfying. Acronyms such as CHARISMA, PREDICTIVE, and CAPTIVATE (table 3) are excellent examples and all likely to serve the purpose of the acronymisation to a meaningful extent. For pure inventiveness and imagination, some very good acronyms were included on the honourable mentions list, such as HI-5, DESSERT, and RATPAC (table 4).
The bad
The RATIONAL, RECOVER, and EXAMINE (table 3) acronyms may at first glance appear quite reasonable. On further examination, however, these acronyms reveal themselves to be poorly constructed. Consider the completely wonderful RATIONAL acronym, derived from “aspiRin stAtins or boTh for the reductIon of thrOmbin geNeration in diAbetic peopLe.” Orthographically, a worse acronym than this is literally impossible to construct. Although the acronym signifies that the study presents rational, clinically important data, as in “rational pharmacotherapy” or “rational allocation of resources,” such connotations seem disproportionate to the findings of the study.26
The ugly
We identified several acronyms that were seemingly randomly put together at the authors’ discretion and did not remotely resemble a recognisable word or phrase. Prominent examples include POLMIDES, ARMYDA-5, and METGO (table 3). The dishonourable mentions list includes abominations such as SU.FOL.OM3 and P-No SOS (table 5), leaving acronymologists around the world wondering why the authors bothered in the first place.
We conclude that the prevalence of acronyms in reports on clinical trials is increasing at the expense of their semantic and aesthetic quality. Given the academic importance of acronyms, we are surprised by the lack of effort dedicated to their construction. The growth of acronym use, especially those of poor quality, should be resisted.27 We believe that strict governance of current guidelines by journal editors will result in an aesthetic improvement and better use of acronyms.
What is already known on this topic
The use of acronyms by medical researchers to brand their studies in the minds of clinicians and fellow researchers is subject to controversy
The use of acronyms may be associated with a higher annual citation rate
What this study adds
The proportion of trials within major disease entities in rheumatology, endocrinology, pulmonary medicine, and psychiatry that uses acronyms is increasing
The technical and aesthetic quality of acronyms is decreasing
Notes
Cite this as: BMJ 2014;349:g7092
Footnotes
Contributors: AP, JH, and PD were responsible for the overall planning of the study. AP and TBS performed the statistical analyses and data management. All authors made major contributions to the planning of the study, data collection, and subsequent reporting of the work. PD, AP, and JH primarily drafted the manuscript. All authors revised the manuscript for important intellectual content and approved the final version. AP is the guarantor. The study design; collection, analysis, and interpretation of data; writing of the article; and decision to submit for publication were independent of any funding body. All researchers had access to all the data.
Funding: No specific funding.
Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; AP, JH, KSL, MRH, and PD have participated in studies using acronyms.
Ethical approval: Not required.
Data sharing: Statistical code and datasets are available from the corresponding author at apottegaard{at}health.sdu.dk.
Transparency: The corresponding author (AP) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies are disclosed.
This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.