Stakeholder involvement in the preparation of systematic and rapid reviews: a cross-sectional study

ID: 

18431

Session: 

Short oral session 12: Stakeholder involvement in evidence production, synthesis and use B

Date: 

Saturday 16 September 2017 - 14:00 to 15:30

Location: 

Meeting Room 1.61

All authors in correct order:

Mütsch M1, Feldmann J1, Puhan MA1
1 Epidemiology, Biostatistics and Prevention Institute, University of Zurich, Switzerland
Presenting author and contact person

Presenting author:

Margot Mütsch

Contact person:

Margot Mütsch
Abstract text
Background: To foster knowledge transfer and implementation it is considered essential to involve stakeholders in the preparation of systematic reviews (SR) and rapid reviews (RR). To date, little is known about the types of stakeholders involved and details of their engagement.

Objectives: In the field of health-services research, we aimed to assess main factors of the study focus and reporting characteristics of stakeholder involvement (SI) in published full SR (Cochrane and non-Cochrane) and in RR (RR and Dare RR published in the Cochrane database).

Methods: Based on the pre-designed protocol, searches were performed in Ovid Medline, Embase and the Cochrane databases. From the records retrieved between January 2011 and October 2015 a sample of 30 reviews for each of the 4 groups was randomly selected resulting in a total of 120 reviews. The standardised data-extraction forms assessed 9 epidemiological, study focus-based variables and 13 reporting characteristics with 10 targeting SI.

Results: From a total of 57 822 articles, 533 (0.9%) were identified as Cochrane SRs, 56 986 (98.5%) as non-Cochrane SRs, 208 (0.4%) as RRs and 95 (0.2%) as Dare RRs. Among non-Cochrane SRs 13% (4/30) were based on a study protocol and 17% (5/30) of RRs. Reporting of potential conflicts of interest was not included in 33% (10/30) and 27% (8/30) for non-Cochrane SRs and RRs, respectively.
We found SI in 13% (4/30) of Cochrane SRs, 17% (5/30) of non-Cochrane SRs, 40% (13/30) of RRs (40%) and 80% (24/30) of Dare RRs (80%). Overall, 33% (15/46) of these articles mentioned positive effects of SI.

Discussion: RRs and Dare RRs involved stakeholders more than twice as frequently than non-Cochrane and Cochrane SRs and they involved them at different stages and in a greater variety per review. SRs often target decision makers as audience but rarely involve them directly in their production.

Conclusion: Overall, stakeholder engagement was not general practice in either review type (with the exception of Dare RRs). Especially for SRs, there is still a great potential to improve reporting as well as to engage stakeholders in practice-academia partnerships.