Congress abstracts
A congress abstract is a publication of its own: it is written to the congress's structure and its word or character limits, supplied with tables and figures the abstract book has to print legibly, bound by the rules on authorship, prior publication and repeat presentation, and read before the poster or talk it wins a place for.
Meet the call's structure, language and submission rules
EMWA's call for posters at its May 2023 conference in Prague asked for abstracts based on original research, case reports or reviews in any field of medical writing, communication, editing or translation; structured as introduction, methodology, results and conclusion in fewer than 250 words, excluding title, authors and affiliations; in English with UK spelling; and submitted on the official form by email by the presenting author by 17 February 2023.
The same call put the obligations on the presenting author, who had to be registered for the conference, present the poster at the indicated times, and ensure, and say in the submission email, that every co-author was aware of, had contributed to and agreed with the abstract. An expert panel reviewed each abstract and selected it for poster presentation or rejected it, notifying authors with poster preparation guidelines; the data must not have been published as a full paper; and only presented abstracts were published, so an abstract whose presenter could not attend was withdrawn.
Write the abstract inside its word and character limits
Congress abstracts have constraints of their own, set out in a two-part grammar-and-style series in EMWA's journal that began in December 2015. Publishing data in a congress abstract is often the first opportunity physicians and pharmaceutical companies have to present clinical trial data, but congress guidelines usually set strict word or character limits, and the simple solution of publishing two or more abstracts is one that many congresses discourage and some actively forbid where the abstracts come from one study. Writers may therefore be pressed by authors and sponsors to fit as much data as possible into one abstract within the limits, and unless great care is taken the result is so data-heavy that the key data and messages are hard to find; the balance is between the congress guidelines and the authors' requests.
Abbreviations cut both ways, and the decision is not the same as for a manuscript abstract. They are a simple way of significantly reducing words or characters, and some that are not acceptable in a manuscript abstract are frequently used at congresses, as pts for patients and wks for weeks. Depending on the audience, defining commonly used abbreviations may not be strictly necessary, since constantly defining them detracts from the flow and may not help a reader who already knows them, and some congresses publish a list of abbreviations that may be used without definition. Against that, overusing abbreviations, especially uncommon ones, makes an abstract hard to follow, and under a word limit abbreviating may not shorten anything: abbreviating methotrexate to MTX saves no words and in fact costs one to introduce the abbreviation, though it does significantly reduce the character count. The series' rule is to use abbreviations with the reader in mind rather than only as a way of shortening the abstract.
Considered punctuation presents data concisely, and the series' worked example shows how. At Week 24, remission rates in treatment arms A and B were 55% and 45%, respectively becomes Week 24 remission rates: Arm A = 55%, Arm B = 45%, which conveys the same information more briefly and in a more digestible form and avoids the backtracking that respectively forces on the reader. Brackets, colons and semicolons used with consideration present data concisely, and under a word limit a slash without subsequent spacing may conserve words, so that responders/nonresponders and the corresponding values may count as one word. The approach is easy to overuse, not all data suit the format, and too much punctuation risks not appealing to the reader's eye. Brackets inside brackets should be avoided, so a comparison with a p value is punctuated as a single bracketed group rather than nested ones.
References cost words and characters, and congress practice differs from manuscript practice. Manuscript abstracts usually do not permit references while many congress abstracts include them, so only key references should be included. Where they are necessary, consider how much the reader needs in order to identify the publication: abbreviated journal titles and only the final page number of a page range, 1234–5 rather than 1234–1235, cut characters while leaving the reference identifiable, and removing et al. saves two words or five characters per reference.
Supply tables and figures the abstract book will print legibly
The series' second part, of March 2016, deals with what goes beside the text. It starts from the difference from a manuscript abstract: congress abstracts may often include tables or figures or both, many congresses allow these to be submitted alongside the abstract text, and this is often the most concise and clear way to present large amounts of data. The constraint that decides whether to use one is the word count, since including a table or figure may reduce the available character or word count, some congresses counting each as equivalent to 250 characters of body text, and the way round it is to upload a single multipanel image containing several tables or figures, which depending on the congress and its file size limits may count as one figure overall, with all titles and footnotes included inside the image file to save further characters.
The production risk is the argument for supplying a table as an image rather than as a table. A table supplied in table format will usually be transcribed and copyedited by the congress for the abstract book, and that introduces errors that can disturb the message and clarity of a complex table: row indenting may be translated incorrectly or not at all, important category headings may lose their bold, and a cell containing a lot of data will spill onto two rows with the line break falling in the wrong place. Additional columns can relieve the last of those, putting a value and its confidence limits in separate columns rather than one, and the decision to make is whether the risk of an erroneously translated table outweighs supplying it as a figure.
Supplying it as a figure brings its own rules. Take special care with superscript footnotes, which in the column's experience are sometimes not published clearly in the congress book, where square brackets print well, as [a] rather than a superscript a. Avoid over-compartmentalising data with grids and excessive row and column borders; keep tables simple and clean, with lines and shading used thoughtfully to aid comprehension of the data. Restrict horizontal borders, let explanatory rows span the whole table width where that helps, use shading carefully, put in-line footnotes rather than superscript ones because they read more clearly, and include the title inside the image to conserve characters.
Figures have a short list of their own. Many congresses have no specific rule against colour, and colour may be inherent to the figure, as in photographic material or imaging scans, while for graphs and charts its use should be encouraged wherever it improves understanding, with red-green and blue-yellow contrasts avoided for colour-blind readers. Graphs and charts should focus clearly on the data; figures may be resized with a loss of quality, so gridlines should generally be avoided and only major lines included, in a narrow width and a lighter shade. Consistency is the governing property: consistent shading for each data series and aligned text sizes for axis titles, data labels and other on-figure text, judged against the final size in the abstract book, because inconsistency is more noticeable than consistency, and a well-presented figure draws attention to the data partly by not drawing attention to anything else.
Two production facts apply to all of it. There is no opportunity to see proofs of a congress abstract, unlike a manuscript, so a table or figure has to be right when it is submitted; and to ensure it is published legibly, check the congress's minimum resolution and size requirements; many congresses give little guidance on them, and where the abstract book will be printed, aim for the standard print resolution of 300 dpi. There is no point overloading a table with data if the end product is illegible in the abstract book.
Follow the rules on authorship, prior publication and repeat presentation
Guidance for conference abstracts and presentations of company-sponsored research is not uniform, since each conference has its own recommendations and consistency is needed, and by June 2018 a group of editors and communicators had posted a draft, Good Practice for Conference Abstracts and Presentations (GP-CAP), as a preprint, with recommendations for researchers and for conference organisers and an invitation to comment before revision and publication. A news column in EMWA's journal that month took its four sections in turn.
On authorship, the draft distinguishes authors, contributors and study groups, and presenters and society sponsors, and recommends listing fewer than ten authors along with study group names; it also allows, in certain circumstances and if all authors agree, someone whose contribution does not meet the journal authorship criteria to present findings at a conference. On the abstract itself, it asks for a study identifier, a registration number for a clinical trial, a study name, a protocol number or a grant number, and states that most conferences will not consider findings already published in full, a requirement that must be respected, and that even where a conference permits it, presenting after full publication should be avoided.
Presenting the same findings twice has a name and a condition attached. It is permissible to present the same research findings at more than one conference where both the first and the subsequent conference allow it, the practice being called an encore abstract or encore presentation, but presentations of the same findings to the same audience should be avoided. On the presentation itself, author listing and sequence on posters and oral presentations should match the abstract, authors should not be added after the abstract is accepted, and where research findings change substantially between abstract submission and presentation in a way that affects the conclusions, the authors should alert the conference to the discrepancy. The draft closes on what a poster is: posters are not peer-reviewed by conferences and may not describe all aspects of the research, so they should not be viewed as a substitute for a full article in a peer-reviewed journal.
The recommendations were published in final form in 2019. Announcing them in EMWA's journal in September 2019, one of their authors described the gap they fill: industry-sponsored research is often presented at conferences, and although Good Publication Practice guides the development of full publications, it only touches on conference presentations, so no specific guidance had described best practice for them. The authors, all experienced publication professionals from medical communications agencies or pharmaceutical companies, drafted a framework, posted it as a preprint for public review and comment, and refined it before submitting it for publication. The final recommendations cover authorship and contributorship; transparency; the development, review and approval of abstracts and presentations; re-presentation of data as encore abstracts; copyright; and the appropriate citation of conference presentations; and although written with industry-sponsored research in mind, the authors consider them applicable to all research submitted to conferences and invite suggestions through the project's website.
Expect the abstract to reach readers before the poster does
The abstract is a separate publication event with its own audience. AMWA's journal printed the accepted poster abstracts for its annual conference in the issue before the meeting, expressly as a preview of what would be on display, in each year from 2007 to 2012, and described the poster session in the same terms throughout, as a forum for members to share new ideas and innovative thinking with their peers from 2007 to 2009, and from 2010 as a way for medical communicators to share recent innovations, advances and discoveries.
The same journal did not edit the abstracts it printed, and the reason it gave changed. In 2007 it printed them as submitted to preserve their integrity. In 2009, 2011, 2012 and 2013 an editor's note said the abstracts were published exactly as submitted and had not been changed during the journal's copyediting or proofreading processes, in order to maintain version control. Whatever the submitted abstract contains, a typographical error included, is therefore what readers meet in print.
EMWA publishes accepted and presented abstracts in its journal, open access under a CC BY licence, and from 2022 posts the abstracts of all its conferences on a public page of its website, with the full posters open to members only; the abstracts of its May 2022 poster session in Berlin were printed in the December 2022 issue, which introduced them as a snapshot of the latest thinking and research in the field.