Article type
Year
Abstract
Background:
Cochrane is recognised internationally as the benchmark for high-quality information about the effectiveness of health care and aims to produce high-quality, relevant, up-to-date systematic reviews. Methods used in Cochrane reviews are becoming more complex and diverse, as evidenced in the newly updated Cochrane Handbook. Methods once considered novel are quickly becoming mainstream (e.g., network meta-analysis (NMA) and inclusion of non-randomised studies of interventions (NRSI)), and new methods continue to emerge (Risk of bias 2 (RoB2) tool and methods for Systematic reviews WIthout Meta-analysis (SWiM). Evidence has shown that methods in systematic reviews are not always implemented appropriately. Cochrane must strive to maintain its reputation by supporting authors and review teams to use complex and novel methods accurately, while continuing to support those seeking guidance for more standard methods.
The Cochrane Methods Support Unit (MSU) was set up in October 2019 to improve the consistency and methodological quality of Cochrane Reviews through support, peer review and training for both new, more complex methods and standard methods.
The MSU team comprises three staff: a statistician, an epidemiologist and a methodologist. They work collaboratively with Cochrane Review Group (CRG) Networks by providing advice and support on requests from Associate Editors, Network Support Fellows or Network Senior Editors, and in response to queries escalated from the Community Support Team, Copy Edit Support or from the Editor in Chief directly.
Objectives:
To summarise the type, frequency and origin of requests for support from the Cochrane MSU and suggest.
Results:
In our first five months we have responded in total to 98 queries from all eight CRG Networks. This included comments on 28 reviews and 34 protocols. Requests originated from 23 review groups (44%). Methods commented on included: NMA (n=34); RoB2 (n=17); NRSI (n=8); standard intervention reviews (n=14); living systematic reviews (n=2); Risk-of-Bias 1 tool (n=1); individual patient data (n=1) and systematic reviews without meta-analysis (n=1). In addition, we have answered 31 statistical and methodology queries from the networks, Cochrane Fast Track and Cochrane Response. Methods tackled in these queries ranged from NMA (n=6), standard reviews (n=14), use of funnel plots (n=1), split body designs (n=1), inclusion of cluster and cross-over RCTs (n=3) and unit of analysis issues (n=2).
Conclusions:
The evolving nature of systematic review methods requires parallel evolution of knowledge within Cochrane, which is evident in the number and range of requests received. The Unit provides a broad range of support to the Cochrane community and is ideally placed to identify common errors and training needs to improve the exemplary use of both standard and more complex methods.
Patient or healthcare consumer involvement:
None.
Cochrane is recognised internationally as the benchmark for high-quality information about the effectiveness of health care and aims to produce high-quality, relevant, up-to-date systematic reviews. Methods used in Cochrane reviews are becoming more complex and diverse, as evidenced in the newly updated Cochrane Handbook. Methods once considered novel are quickly becoming mainstream (e.g., network meta-analysis (NMA) and inclusion of non-randomised studies of interventions (NRSI)), and new methods continue to emerge (Risk of bias 2 (RoB2) tool and methods for Systematic reviews WIthout Meta-analysis (SWiM). Evidence has shown that methods in systematic reviews are not always implemented appropriately. Cochrane must strive to maintain its reputation by supporting authors and review teams to use complex and novel methods accurately, while continuing to support those seeking guidance for more standard methods.
The Cochrane Methods Support Unit (MSU) was set up in October 2019 to improve the consistency and methodological quality of Cochrane Reviews through support, peer review and training for both new, more complex methods and standard methods.
The MSU team comprises three staff: a statistician, an epidemiologist and a methodologist. They work collaboratively with Cochrane Review Group (CRG) Networks by providing advice and support on requests from Associate Editors, Network Support Fellows or Network Senior Editors, and in response to queries escalated from the Community Support Team, Copy Edit Support or from the Editor in Chief directly.
Objectives:
To summarise the type, frequency and origin of requests for support from the Cochrane MSU and suggest.
Results:
In our first five months we have responded in total to 98 queries from all eight CRG Networks. This included comments on 28 reviews and 34 protocols. Requests originated from 23 review groups (44%). Methods commented on included: NMA (n=34); RoB2 (n=17); NRSI (n=8); standard intervention reviews (n=14); living systematic reviews (n=2); Risk-of-Bias 1 tool (n=1); individual patient data (n=1) and systematic reviews without meta-analysis (n=1). In addition, we have answered 31 statistical and methodology queries from the networks, Cochrane Fast Track and Cochrane Response. Methods tackled in these queries ranged from NMA (n=6), standard reviews (n=14), use of funnel plots (n=1), split body designs (n=1), inclusion of cluster and cross-over RCTs (n=3) and unit of analysis issues (n=2).
Conclusions:
The evolving nature of systematic review methods requires parallel evolution of knowledge within Cochrane, which is evident in the number and range of requests received. The Unit provides a broad range of support to the Cochrane community and is ideally placed to identify common errors and training needs to improve the exemplary use of both standard and more complex methods.
Patient or healthcare consumer involvement:
None.