Tailored PRISMA 2020 flow diagrams for living systematic reviews: a methodological survey and a proposal [version 3; peer review: 2 approved]

• Published in: F1000 research Vol. 10; p. 192
• Main Authors: Kahale, Lara A, Elkhoury, Rayane, El Mikati, Ibrahim, Pardo-Hernandez, Hector, Khamis, Assem M, Schünemann, Holger J, Haddaway, Neal R, Akl, Elie A
• Format: Journal Article
• Language: English
• Published: London Faculty of 1000 Ltd 2022; F1000 Research Limited
• Subjects: Coronaviruses; COVID-19; Libraries; Public health; Reviews; Systematic review; living systematic review; PRISMA statement; flow chart; research replication; research transparency; update; systematic review reporting standards; evidence synthesis; research methodology research reporting
• ISSN: 2046-1402; 2046-1402
• DOI: 10.12688/f1000research.51723.3

Background: While the PRISMA flow diagram is widely used for reporting standard systematic reviews (SRs), it was not designed for capturing the results of continual searches for studies in living systematic reviews (LSRs). The objectives of this study are (1) to assess how published LSRs report on the flow of studies through the different phases of the review for the different updates; (2) to propose an approach to reporting on that flow.
Methods: For objective 1, we identified all LSRs published up to April 2021. We abstracted information regarding their general characteristics and how they reported on search results. For objective 2, we based our proposal for tailored PRISMA approaches on the findings from objective 1, as well as on our experience with conducting Cochrane LSRs.
Results: We identified 279 living publications relating to 76 LSRs. Of the 279 publications, 11% were protocols, 23% were base versions (i.e., the first version), 50% were partial updates (i.e., does not include all typical sections of an SR), and 16% were full updates (i.e., includes all typical sections of an SR). We identified six ways to reporting the study flow: base separately, each update separately (38%); numbers not reported (32%); latest update separately, all previous versions combined (20%); base separately, all updates combined (7%); latest update version only (3%); all versions combined (0%). We propose recording in detail the results of the searches to keep track of all identified records. For structuring the flow diagram, we propose using one of four approaches.
Conclusion: We identified six ways for reporting the study flow through the different phases of the review for the different update versions. We propose to document in detail the study flow for the different search updates and select one of our four tailored PRISMA diagram approaches to present that study flow.