Implementation of Summary of Findings (SoF) tables in Cochrane review groups: progress to date

Article type
Authors
Santesso N, D Oxman A, L Brozek J, E Vist G, J Schünemann H
Abstract
Background: The purpose of Cochrane systematic reviews is to facilitate health care decision-making by clinicians, patients, administrators, politicians, and the general public. Summary of findings (SoF) table provide the key pieces of information from a systematic review 'at-a-glance'. A number of factors play a role in the overall implementation of SoF tables including testing 'user friendliness', addressing methodological concerns, programming software, and building capacity in Cochrane Review Groups to support use of the SoF tables.
Objectives: To test methods for the implementation of SoF tables by providing training and support to authors to do SoF tables, and/or by dedicating staff to independently create SoFs for reviews.
Methods: Contact Review Groups to gauge interest in pilot testing and select at least 3 groups to participate in the pilot test. Develop teaching and training materials to use the GRADEpro software (a programme external to RevMan 5 to create SoF tables that can be imported to RevMan 5). Provide training and support to Review Groups, as well as direct support to review authors as needed. Obtain feedback and input from Review Groups and authors and record and measure time, and level of involvement and support required. Dedicate staff to create SoF tables for a variety of reviews and measure time, difficulty and support required.
Results: There is interest in the implementation of the SoF tables at the Review Group and author level. Determining the initial contact with Groups was debated; should authors be contacted and supported directly or should Review Groups take on this role? When developing the pilot test, it was also unclear whether training materials with instructions to use GRADEpro for use in general (e.g. for guideline developers, review authors, etc.) or more specific to Cochrane would be more useful. Feedback will be used to revise methods to implement the SoF tables. Measures of burden and time to Review Groups and authors will be considered and feedback into the training materials will be used to revise those materials. Measures from the time and difficulty for trained individuals to create SoF tables will also be used to inform the best methods for implementing SoF tables.
Conclusions: This pilot test provides useful information which will guide implementation of SoF in the future. Next steps include finalising results and a report back to the Cochrane Collaboration Steering Group to determine the best method for implementation of the SoF tables.