Replication is a corner stone of the scientific method yet replication of systematic reviews is too often overlooked, done unnecessarily or done poorly, as identified by Francis Collins, the Director of the National Institutes of Health (NIH). Excessive replication - doing the same study repeatedly - is unethical and a cause of ‘research wastage’. There is a lack of guidance for when or why to replicate systematic reviews.
Overarching Aim: The proposed research will use an evidence-driven, transparent process and implement consensus approaches to develop value-added guidance on: a) when to replicate and b) when not to replicate systematic reviews.
Objectives: We propose three sequential objectives:
1. To assess the extent of replication of systematic review topics and explore reasons for replication as well as issues such as discordance, agreement and bias.
2. To develop consensus-based recommendations on when to replicate systematic reviews of effectiveness (and when not), taking into account the needs and preferences of the various stakeholder groups.
3. To identify strategies for disseminating and implementing these recommendations to key stakeholder groups (e.g. patients, clinicians, journal editors, systematic review authors, systematic review users, systematic review organizations (such as Cochrane, Campbell, Joanna Briggs Institute), guideline developers, research funding agencies, KT researchers).
To achieve these objectives we will: assess evidence about when to replicate systematic reviews, document characteristics of overlapping systematic reviews, survey and interview authors of replicated systematic reviews, consult with global experts on replicability guidelines, prioritize and seek consensus on items for systematic review replication guidance, and develop collaborative knowledge translation plan including practical tools for specific knowledge users.
Proposed Papers and Leads
Peter, Sathya, Vivian
Draft in circulation
Framework and definitions for SR replication
Jeff Valentine, Mark Petticrew, Brigitte Vachon, Alison Riddle, Elie Akl, Brigitte Vachon, Lara Maxwell, Laura Weeks, Kednapa Thavorn, Zulfi Bhutta, Karla Soares, Bev Shea, Betsy Kristjansson, Sathya, Vivian, Peter T
Research Synthesis Methods,
J Clin Epi
Framework under development
SR replication for internal validity
Jeff, Jeremy, Zulfi, Terri, Bev, Sathya, Larissa, Peter
Elie, Janet, Lara, Howard, Gabriel, Betsy, Terri, Jeff, Peter T, Julian, Tamara, Larissa, Sathya, Vivian, Peter T
Stakeholder interviews & Survey
Peter T, Tamara, Vivian, Lara
Cochrane methods SR
Jennifer, Peter T, Zulfi, Vivian, Lara, Jordi
Cochrane & Campbell
Protocol reviewed; Responding to reviewer comments
Discussions underway for a possible Lancet commentary
Peter, Vivian & Jocalyn Clark
Janet, George, David, Sathya
To be discussed after main paper is submitted
Peter, Sathya, Vivian
David, Betsy, Howard, Kednapa, Brigitte, Sathya, Vivian, Peter T
Howard, Peter T, Vivian, Lara, Mark P, George, Janet, Peter W, Bev, Sathya, Vivian, Peter T
Series of other papers: editorials and think pieces?
Conferences and Meetings 2019-20
Meeting Dates & Venue
April 14-16, 2019
Using a conceptual value of information approach to decide when to, and when not to replicate systematic reviews
July 23-25, 2019
1. Vivian Welch
2. Jeff Valentine
1. When to, and when not to replicate systematic reviews: consensus meeting and checklist
2. Towards a framework to replicate systematic reviews
Sicily evidence based healthcare (EBHC)
November 6-9, 2019
To Replicate, or Not to Replicate: A Value of Information Approach to Decide Whether to Replicate Systematic Reviews
Campbell-what works global summit
October 16-18, 2019
Mexico City, Mexico
Vivian Welch, Jeremy Grimshaw, Howard White, Terri Pigott, Betsy Kristjansson, Peter Tugwell
Panel presentation (abstract submitted):
Developing guidance for when to, and when not to replicate systematic reviews
October 22-25, 2019.
Welch V, Grimshaw J, Rada G, Smith M, Pardo Pardo J, Soares-Weiser K, Tugwell P
Workshop (abstract accepted):
When should systematic reviews be replicated, and when is it wasteful?