Lessons from four years of Cochrane’s Methods Support Unit: what are we asked and how do we answer?

Session Type
Workshop - training
Editorial processes and supporting review authors
Target audience
Authors of Cochrane reviews, from any background (including clinicians and consumers). Also people working in an editorial capacity
Level of difficulty

Background: Established in 2019, the Methods Support Unit (MSU) provides methods advice to editors and authors preparing Cochrane Reviews. We have helped people from 49 Cochrane Review Groups, dealing with over 1,100 requests. These requests range from advice on a single methodological/statistical issue through to a full review of a draft protocol or review. In addition, MSU has provided many hours of interactive training and authored guidance on key methods issues. As a result, MSU is in a unique position to understand the methodological and statistical concerns of Cochrane authors and editors, and to provide accessible training of immediate usefulness
Objectives: Our workshop will highlight the ‘Top Five’ issues encountered by authors and editors working on Cochrane reviews. These are the key issues that generate questions for the MSU team and that we encounter when reviewing manuscripts. We will explain how to handle these issues through a series of interactive exercises and discussions. As a result, this workshop will help to strengthen evidence production for the future by tackling ‘common errors’, and ensure Cochrane continues to produce robust and trusted evidence.
Description: We will use the most commonly asked questions as the basis for our workshop which will be delivered by experienced methodologists and statisticians. We will prepare materials and deliver a series of five 15-minute interactive sessions to teach attendees how to handle these issues. There will be plenty of time for questions and discussion. We will cover: • Issues arising from including cluster and cross-over randomised controlled trials. • Use and misuse of the Standardised Mean Difference effect measure. • Risk of Bias 2.0: Bias due to deviations from the intended interventions: how is this different from the ‘Blinding of participants and personnel’ domain in Risk of Bias 1? What counts as a deviation from the intended interventions? • How should the results of a Network Meta-Analysis be summarised? • Assessing imprecision in GRADE: what is an Optimum Information Size? We did not involve patients or healthcare consumers in preparing this workshop, but we will pitch the session to be accessible to the widest possible audience.