PRISMA for reporting systematic reviews in journal and conference abstracts

Article type
Authors
Beller E1, Glasziou P1, Altman D2, Hopewell S2, Bastian H3, Chalmers I4, Gotzsche P5, Lasserson T6, Tovey D6
1Bond University, Australia
2University of Oxford, UK
3National Library of Medicine, USA
4James Lind Initiative, UK
5Nordic Cochrane Centre, Denmark
6Cochrane Editorial Unit, UK
Abstract
Background: Readers screen the title and abstract of articles to get informed and to determine whether it is worthwhile reading on. A well-written systematic review abstract should signal its systematic methodology and its relevance. As some people only have access to the abstract, the research should be conveyed as accurately and comprehensively as possible.

Objectives: To develop an extension to the PRISMA Statement for abstracts of systematic reviews; to elaborate on the rationale for each item.

Methods: We used the original PRISMA Statement and other sources of guidance to inform our selection of potential items to be included in systematic review abstracts. To refine and reduce the items in an essential checklist we used a modified Delphi consensus process by survey of 147 respondents. At a 2-day meeting, the literature on writing abstracts and the checklist items were discussed, and inclusion and wording were finalised.

Results: The final checklist contains reporting items related to the title, objectives, eligibility criteria, information sources, risk of bias assessment, included studies, synthesis of results, description of effect, strengths and limitations of the evidence, interpretation of the results, funding and registration. We explain the rationale for each item and provide illustrative examples of good and poor reporting.

Conclusions: While abstracts should not replace full articles in informing decision making, the abstract must stand alone in presenting a clear and truthful picture of the research. The PRISMA for Abstracts checklist may help authors provide abstracts that facilitate peer review for pre-publication and conference selection, enable efficient perusal of electronic search results, provide assessment of the validity of a systematic review and give an overview of the most important results.