tipology:letter

numero rivista e pagine: HSR Proceedings in Intensive Care and Cardiovascular Anesthesia 2011; 3(1): 77-78
PDF version

A is for Abstract

Authors: M. John*1

Head of Medical Humanities International MD Program, Professor of Biomedical Communication Skills, Faculty of Medicine, Vita-Salute San Raffaele University, Milan, Italy

Corresponding author: * Corresponding author:
Prof. Michael John
UniversitÓ Vita-Salute San Raffaele, Milan, Italy
Via Olgettina, 48 - 20132 Milan, Italy

E-mail: michael.john@hsr.it

Everyone involved in the world of biomedical science knows, or at least should know, what an abstract is. It is interesting to consider that originally the ‘abstract’ was found at the end of a paper. It was the concluding paragraph that ended the work with a bang and allowed us to remember what it was all about. It has since been moved to the front of the paper where its job is to attract reader interest and, vulgarly speaking, sell the rest of the article to the peer population. Therefore, the abstract is fundamentally important. It might just be the only part of your paper that most people will read, providing the title is sufficiently stimulating of course. However, that is another story.
Coming back to the abstract, PubMed gives free access to abstracts to allow readers to decide if they want to invest in reading the whole paper. Writers should therefore think of the abstract as a form of advertising that contains the message of the paper and transmits it in a clear, concise and attractive manner. It’s a bit like the back cover of a book or the teaser to an action movie.
The Vancouver Group says that abstracts should state the purposes of the study or investigation together with basic procedures, main findings and the principal conclusions, while emphasizing new and important aspects that qualify the work.
The good bits of the full paper should therefore be highlighted and the audience just can’t wait to see the rest. However, an abstract must never include data that are inconsistent with or absent from the main body of the paper. This might seem obvious, but Pitkin et al. showed that this rule is often, quite amazingly, broken [1].
It is therefore well worth spending a great deal of time on the abstract so you get it right. Never concentrate all your efforts on the IMRAD (Introduction, Methods, Results and Discussion) part of the paper and simply knock off an abstract just before submission. This is not a good idea. A good abstract is difficult to write also due to the strict word count that is imposed by a journal’s Instructions for Authors.
There are two main types of abstract: the single-paragraph type and the structured type. They both do the same thing, as we have already seen, but the structured type is normally longer. While the single-paragraph abstract is generally around 250 words, the structured abstract can be up to 400 words long.
The titled sections of this second kind of abstract can include Objectives, Design, Setting, Patients or participants, Treatment, Results, Conclusions, Clinical relevance.
Anyone that has a vague intention of reading your paper will see the title and the abstract at the same time. It is therefore silly to repeat or paraphrase the title as the first sentence of your abstract. Quite simply, it is a waste of words. The Instructions for Authors of your target journal will be very strict regarding the total word count of the abstract, so never risk to exceed this limit with redundant repetition.
Use the past simple tense to describe your objectives, methods and results. Write numbers and percentages, not standalone p values.
Use the present simple tense when discussing or interpreting results and concluding.
Use an active form of the verb; avoid the passive voice.
Write concisely and clearly. Give it all you have. The abstract is a stand-alone document that is able to transmit the principal message of a biomedical manuscript while tempting peers to read the full paper. Be sure to get it right.

 

References

  1. Pitkin RM, Branagan MA, Burmeister LF. Accuracy of data in abstracts of published research articles JAMA 1999 1999; 281: 1110-1111.

'This is the ninth of a series of articles on this topic. Send any questions to michael.john@hsr.it who will answer them as part of this column'