It is taken for granted that the main aim of publishing abstracts submitted to the Mexican Congress of Rheumatology is to aid posterior interpretation. The abstracts are a type of academic text which must prioritise content over form – through a methodological framework. This demands the use of a language that may be understood over time without falling into ambiguities and that the contemporary reader is left with no doubts as to the author’s intentions1,2.

Under the premise that highly schooled individuals, such as rheumatologists should write well, we assessed the linguistic strengths and weaknesses of the abstracts submitted to the XLIX Mexican Congress of Rheumatology.

To this effect, all abstracts submitted to this congress were analysed. The authors’ names, memberships and identification data were removed. Linguistic assessment of each abstract was made by an expert who was unaware of the primary study objective. A structured tool was used with six domains defined a priori: reader comprehension (following the submission instructions), orthography (accents, full stops), syntax 1 (basic structure of the written piece), syntax 2 (indication provision errors), semantics 1 (ambiguity) and semantics 2 (inappropriate connotations). Scoring for each domain was classified as optimum (≤1 errors), substandard (2–4 errors) and insufficient (≥5 errors).

Three hundred and twenty-two (100%) abstracts were analysed and a total of 2,805 errors were reported, with an average of 8.7 errors per abstract (95% CI of 8–9, range of 0–45). The distribution of error frequencies showed that 3.5% of abstracts were classified as optimum, 24.5% as substandard and the remainder (72%) as insufficient. Within the subgroup of insufficient, 104 abstracts had more than 10 errors each. Table 1 shows the errors by domain. Correlation was found between the domains of syntax 1 and semantics 1 (r = .62). No correlation was found between other domains, in either the number of errors or the classification given by the abstract reviewers.

Our considerations from these data are as follows: the abstracts studied had one or more authors who were members of one or more institutions, and therefore these results did not evaluate rheumatologists who did not send abstracts or other types of written reports. The analysis of quality, the potential use and reproducibility of a scientific and/or academic communication depends on whether it is understood. We noted the errors by number and classification, so that the reader could define the “number of tolerable errors”, bearing in mind the formal minimum schooling of a rheumatologist (or resident rheumatologists is that of a 21-year-old. The abstracts already form part of the public domain and could be read in other countries, by physicians of other specialities and by the general public. The perception of the quality of the abstracts could have an impact on our sector.

This study is proof of concept, and explanations are therefore not permitted. However, in future studies it would be of interest to assess whether there is discordance between formal schooling and functional schooling, an underestimation of the importance of the Congress or in the shared responsibility of the corresponding author and co-authors in the final content of the abstract. It would also be of interest to assess whether these findings occur in other areas. Meanwhile the logical action to follow is the discussion of these data by academic groups of the speciality1,2.