All full reviews must include an abstract of not more than 400 words. It should be kept as brief as possible without sacrificing important content. Abstracts to Cochrane reviews are published in MEDLINE and the Science Citation Index, and are made freely accessible on the Internet, so will often be read as stand-alone documents. They should, therefore, summarize the key methods and results of the review and not contain any material that is not in the review. The content must be consistent with the text, data and conclusions of the review and not include references to any information outside the review. Links to other parts of the review (such as references, studies, additional tables and additional figures) may not be inserted in the abstract. An example is included in Box 4.3.a.
Abstracts should be made as readable as possible without compromising scientific integrity. They should primarily be targeted at healthcare decision makers (clinicians, consumers and policy makers) rather than just researchers. Terminology should be reasonably comprehensible to a general rather than a specialist healthcare audience. Abbreviations should be avoided, except where they are widely understood (for example, HIV). Where essential, other abbreviations should be spelt out (with the abbreviations in brackets) on first use.
It is important that Cochrane reviews of diagnostic test accuracy are identifiable in MEDLINE and other electronic bibliographic databases from their titles, abstracts and keywords. Indexing of diagnostic terms is currently poor, and the guidelines for the titles of Cochrane reviews rule out the use of phrases such as ‘systematic review’ or ‘diagnostic test accuracy’. Their titles may therefore be difficult to distinguish from reviews of interventions. “Screening for type 2 diabetes mellitus”, for example, may refer to an intervention review investigating the effectiveness of screening for type 2 diabetes mellitus in reducing morbidity and mortality associated with diabetes, but it could also describe a diagnostic test accuracy review investigating the accuracy of screening tests for type 2 diabetes mellitus. Diagnostic test accuracy reviews will be clearly identifiable by ‘flags’ within The Cochrane Library, but these will be lost when the abstracts are transferred to bibliographic databases.
To overcome this problem it is recommended that the phrase ‘diagnostic accuracy’ be included somewhere in the abstract – the ‘objectives’ section may provide a natural home for this phrase. This is a key phrase that is highly likely to be used in searching by researchers who want to identify reviews of diagnostic tests. Combining this phrase with the phrase for the Cochrane Database of Systematic Reviews as a journal title (for example, ‘Cochrane Database of Syst Rev’ in PubMed) will retrieve Cochrane reviews of diagnostic test accuracy. Ultimately, the introduction of specific database indexing terms into MEDLINE (such as Publication Types ‘diagnostic test accuracy review’ and ‘diagnostic test accuracy study’) should improve the efficiency of searching.
The content under each heading in the abstract should be as follows:
One or two sentences explaining the context, the purpose and rationale for the review.
A precise statement of the primary objective of the review, ideally in a single sentence. Where possible the style should be of the form ‘To determine the diagnostic accuracy of [Index test] for diagnosing [target condition] in [patient description]’. The exact phrase ‘diagnostic accuracy’ is important as this can then be used as a search term to retrieve Cochrane systematic reviews of diagnostic test accuracy from MEDLINE and the Science Citation Index, and other publications of these reviews in other electronic bibliographic databases.
List the sources and the dates of the search for each source, using the active form ‘We searched….’ For example ‘We searched MEDLINE from January 1950 to December 2006’. Search terms should not be listed here, and if a large number of databases have been searched only key databases and the number of databases should be stated. The date range of the search for each database should be given. For most databases such as MEDLINE, it should be in the form ‘MEDLINE (January 1950 to December 2006)’. Searching of bibliographies for relevant citations can be covered in a generic phrase ‘reference lists of articles’. If there were any constraints based on language or publication status, these should be listed. If individuals or organisations were contacted to locate studies this should be noted and it is preferable to use, ‘We contacted pharmaceutical companies’ rather than a listing of all the pharmaceutical companies contacted. If journals were specifically handsearched for the review, this should be noted.
Briefly list the main criteria used to select studies for inclusion in the review. Please include; type of study design (for example consecutive patient series and /or case-control studies); index and /or comparator tests; target condition; eligible reference standards, and key characteristics of the study population.
Describe how data extraction and quality assessment of studies were done, whether by more than one person, and whether multiple assessments were done independently. The description should be restricted to how data were extracted and assessed, and not include details of what data were extracted. It should be stated whether a meta-analysis was done and if so the statistical methods used should be named, and the summary statistics estimated stated. If comparisons are made between index and/or comparator tests it should be stated whether comparisons were direct (made within studies) or indirect (made between studies).
This section should begin with the total number of studies and participants included in the analysis, and brief details pertinent to the interpretation of the results, such as a summary of study quality, diversity in study characteristics and heterogeneity in study results. It should address the primary objective and be restricted to the main results. Wherever possible, accuracy should be expressed using summary statistics most likely to help someone make a decision about whether or not to use a particular test. Any summary statistics in the abstract should be the same as those presented in the review and in the summary of results table. Summary statistics should be stated together with confidence intervals.
The primary purpose of the review should be to present information, rather than to offer advice. The authors’ conclusions should be succinct, address the review question and draw directly from the findings of the review so that they directly and obviously reflect the main results. Assumptions should not be made about practice circumstances, values, preferences and tradeoffs, and the giving of advice or recommendations should generally be avoided. Any important limitations of data and analyses should be noted. Important conclusions about the implications for research should also be included.
Box 4.3.a: Example of an abstract for
a Cochrane Systematic Review of Diagnostic Test Accuracy: Magnetic
resonance imaging for provisional diagnosis of multiple sclerosis in patients with suspected
disease.