Cochrane Living Systematic Reviews: guidance, piloting and early evaluation

Article type
Authors
Synnot A1, Turner T2, Akl EA3, Schünemann HJ4, Kahale LA3, Quinn G5, McDonald S2, Green S6, Grimshaw J7, Sullivan K7, Buchbinder R8, Johnston R2, Wolfenden L9, Hodder RK10, Tovey D11, Soares-Weiser K12, MacLehose H11, Hilton J11, Elliott JH2
1Monash University and La Trobe University
2Monash University
3American University of Beirut
4McMaster University
5Royal United Hospital
6 Monash University
7Ottawa Hospital Research Institute
8Monash University and Cabrini Institute
9University of Newcastle
10Hunter New England Population Health
11Cochrane
12Cochrane
Abstract
Background: A number of Cochrane author teams are piloting Living Systematic Reviews (LSRs), with support from Project Transform and the Living Systematic Review Network. While LSRs promise to keep high-quality evidence syntheses continually up-to-date, they require some modifications to existing review authoring and editorial processes, and pose a number of technical and publishing challenges. As such, an evaluation of their feasibility, acceptability and ability to facilitate continual updating is warranted before wider implementation within Cochrane.

Objectives: To outline the Cochrane LSR pilot approach and report on pilot experiences to date in several LSRs, including implications for people and processes, as well as key barriers and facilitators.

Methods: The LSR Network has developed guidance on when to conduct an LSR, standard text for use in LSR protocols and guidance on how to identify, incorporate and present new data. An evaluation is ongoing, collecting quantitative data on workload implications (e.g. citations screened each month) and author and editorial team reflections via regular surveys, project documents and meeting minutes. We will also conduct semi-structured interviews with authors, editorial staff and other stakeholders at the conclusion of the pilots.

Results and conclusions: We will present the proposed LSR-specific review methods and early evaluative data from the current pilot Cochrane LSRs, including implications for Cochrane, authors and editorial teams, and related evidence products. Barriers and facilitators identified to date will be discussed, including the feasibility of monthly searching and study identification supported by machine learning and Cochrane Crowd.