The abstract: A ‘three-star’ opportunity

Authors devote huge amounts of time and thought to the writing and fine tuning of their papers. Unfortunately, this effort often does not extend to the paper’s title and abstract. This is a serious mistake.

In reality, the title and the abstract, whether available in print or online, are the only parts of a paper that most people will read [1]. As such, they are the only way that an author can entice a reader to either devote more time to their work or at least remember its most salient findings. This fact has serious consequences as research shows that abstracts alone are often used to inform clinical decision making, particularly in areas where access to information is limited [2–4]. Perhaps more immediately most journals pre-screen papers for review with as many as 25% being rejected by an editor following an assessment based on the title and abstract [5]. Interesting, well-written and sufficiently detailed abstracts are essential to enable the reader to make an informed judgement about the originality, significance and rigour of the research [3]. Ironically though, the abstract is often the part of the paper given least thought, the last part of the paper that is written, often rushed and with limited input from co-authors [5].

The disjuncture between the quality of a paper’s abstract and its body has been known by editors of both major and minor medical journals for years [6]. As an example, a paper by Berwanger et al. [6] reviewed 227 abstracts of randomised controlled trials (RCTs) in four leading medical journals (the New England Journal of Medicine; the Journal of the American MedicalAssociation; the British Medical Journal and the Lancet) and confirmed that abstract quality was poor, and particularly so, when it came to reporting the key mythological factors such as subject numbers, blinding, subject attrition, outcome identification, and effect sizes. These findings have been duplicated by others with respect to nursing journals and systematic reviews, at least as presented in conferences [7–9].

The CONSORT Statement provides recommendations for reporting RCTs, but contains limited advice on the preparation of abstracts –other than recommending a structured format [10]. In response to the poor quality of many journal and conference abstracts, and the limited guidance in this area, the CONSORT Group extended the CONSORT Statement in 2008 to include a checklist of essential items for use when reporting the main results of an RCT in a journal or conference abstract (the CONSORT for Abstracts: Table 1) [11]. The CONSORT for Abstracts contains 17 items, which the authors have stated can be sufficiently addressed in the standard 250–300 word abstract. There is evidence that the use of these guidelines has improved the quality, at least in terms of the number of items reported, in journals that have adopted and enforced the policy [12].

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) Group have also developed an abstract checklist for systematic reviews ‘PRISMA for Abstracts: reporting systematic reviews in journal and conference abstracts’ [13]. However, there is limited guidance on the best presentation of abstracts of qualitative research. A recent synthesis of recommendations on standards for reporting qualitative research gave little advice on the presentation of an abstract, other than a general overview of its structure [14]. Although the CONSORT for Abstracts was developed for RCTs, the general principles can be applied to abstracts of any research methodology, that is, the abstract should enable the reader to make an accurate estimation of the quality and importance of the research, its rigour, the key findings and their implications.

In recent years there has been a growth in national research assessment exercises. Governments use these assessment exercises to ensure public accountability for the allocation of research funding and as way to encourage research performance. Some of the more established research assessment exercises are: Excellence in Research for Australia; Evaluation of Research Quality, VQR, Italy; Evaluation of R&D Units, Portugal and Research Quality Evaluation in Sweden –FOKUS. In the UK the Research Excellence Framework (REF) [15] is a robust, internationally-recognised system for assessing the quality of UK university research, the results of which are used each year to allocate some £2 billion of government funding. University research submissions to REF panels and sub-panels are evaluated and star-rated (Table 2). Only research outputs rating three-star and above attract research funding. In the 2014 REF exercise there were more than 10,000 research outputs submitted to the sub-panel evaluating the allied health professions, biomedical sciences, dentistry, nursing, midwifery and pharmacy alone. Each was rigorously assessed and rated. The workload involved is immense and it would seem self-evident that an abstract that is well-written, and can display a project’s novelty, quality, and importance will assist the assessors in their efforts. While it may be difficult to upgrade the appearance of “two-star” research into three-star level with good writing, the opposite is true. A poorly written abstract with many omissions, could easily be interpreted as one or two-star research.

Common sections of a structured abstract and how they may addressed are outlined below, some journals may of course structure their abstracts differently.