2.4.8 Reporting the Search Strategy: PRISMA-S
- Carolyn Phillips
The 2020 PRISMA Statement includes two requirements regarding the reporting of the search used for the conduct and completion of any systematic review:
‘Specify all databases, registers, websites, organisations, reference lists and other sources searched or consulted to identify studies. Specify the date when each source was last searched or consulted.’
‘Present the full search strategies for all databases, registers and websites, including any filters and limits used.’ (Page et al. 2021, p. 4)
These recommendations were supplemented in 2021 with the publication of PRISMA-S, which provided additional reporting details for systematic review searches (Rethlefsen et al. 2021). This 16-point checklist is designed to ensure that you have reported on different aspects of the search process, including the information sources, such as names and platforms of any databases, trial registries searched, whether databases were searched simultaneously, any citation searching techniques used and browsing methods (if any). You must also document the total number of records identified from each source (database or other source), how many records were deduplicated and what software was used. Moreover, you must provide the date when each search was last run as well as a reproducible search strategy for each source, including any limits or restrictions, and whether a search filter was used or modified. A helpful checklist is available at: https://osf.io/y765x
Thorough documentation of search strategies can help to demonstrate that a review is comprehensive and methodologically sound. Detailed search strategy reporting is essential for reproducibility and also facilitates future review updates. It is important that keywords and index terms, usually used in combination, such as MeSH for MEDLINE, Emtree for Embase and other controlled vocabulary, are incorporated into the search strategy. Both should be clearly identifiable and if key terms used, the field. This also allows both readers and future reviewers to evaluate the search strategy and potentially reproduce the search for themselves, as well as possibly incorporate new controlled vocabulary terms when they become available.
The set text available in JBI SUMARI can assist researchers in making sure that their search method reporting meets PRISMA standards. The set text asks researchers to enter important details regarding search terms, search strategies, information sources and any limiters that were used. Complete search strategies presented in the appendix should be labelled with the name of the database as well as the specific platform used to access the database (e.g. Embase via Ovid). Search strategies must include the appropriate database-specific syntax — it should be possible to rerun each search using the information included in the appendix. The number of search results retrieved for each search string can aid peer reviewers in their assessment of the search, which is why this is also a requirement for inclusion in the appendix.
Care should be taken when reporting database names and their corresponding platforms. One common error is to include an ‘h’ in the name of the PsycINFO database.
JBI reviewers may wish to report on additional aspects of the search methodology that are not addressed by the set text. For example, the review team may wish to report that a search strategy was peer-reviewed using the PRESS checklist. Logic maps, concept grids and other documentation related to the search may also be included. If text mining software, validated filters or search hedges, or previous search strategies have been used, these should be cited appropriately. Bethel et al. (2021) recommend the use of a Search Summary Table for reporting the conduct of a search strategy in terms of database effectiveness, sensitivity and precision.