Key points are not available for this paper at this time.
Background There is growing interest in theory-driven, qualitative and mixed-method approaches to systematic review, such as realist and meta-narrative review. These approaches offer the potential to expand the knowledge base in policy-relevant areas. However, the quality of such reviews can be difficult to assess. Objectives The aim of this project was to produce methodological guidance, publication standards and training resources for those seeking to undertake realist and/or meta-narrative reviews. Methods/design We (1) collated and summarised existing literature on the principles of good practice in realist and meta-narrative systematic reviews; (2) considered the extent to which these principles had been followed by published and in-progress reviews, thereby identifying how rigour may have been lost and how existing methods could be improved; (3) used an online Delphi method with an interdisciplinary panel of experts from academia and policy, to produce a draft set of methodological steps and publication standards; (4) produced training materials with learning objectives linked to these steps; (5) refined these standards and training materials prospectively on real reviews in progress, capturing methodological and other challenges as they arose; (6) synthesised expert input, evidence review and real-time problem analysis into more definitive guidance and standards; and (7) disseminated outputs to audiences in academia and policy. Results An important element of this study was the establishment of an e-mail mailing list to bring together researches in the field ( www.jiscmail.ac.uk/RAMESES ). Our literature review identified 35 and nine realist and meta-narrative reviews respectively. Analysis and discussion within the project team produced a summary of the published literature, and common questions and challenges into briefing materials for the Delphi panel, comprising 37 and 33 members (for realist and meta-narrative reviews respectively). Within three rounds this panel had reached a consensus on 19 (realist) and 20 (meta-narrative) key publication standards, with an overall response rate of 90% and 91% respectively. The Realist And Meta-narrative Evidence Syntheses – Evolving Standards (RAMESES) publication standards for realist syntheses and meta-narrative reviews were published in open-access journals and quickly became highly accessed. The RAMESES quality standards and training materials drew together the following sources of data: (1) personal expertise as researchers and trainers; (2) data from the Delphi panels; (3) feedback from participants at training sessions we ran; and (4) comments made on RAMESES mailing list. The quality standards and training materials are freely available online ( www.ramesesproject.org ). Discussion The production of these standards and guidance drew on multiple sources of knowledge and expertise, and a high degree of a consensus was achieved despite ongoing debate among researchers about the overall place of these methodologies in the secondary research toolkit. As with all secondary research methods, guidance on quality assurance and uniform reporting is an important step towards improving quality and consistency of studies. We anticipate that as more reviews are undertaken, further refinement will be needed to the publication and quality standards and training materials. Limitations The project’s outputs are not definitive and in the future updating and further development is likely to be needed. Conclusion An initial set of publication standards, quality standards and training materials have been produced for researchers, users and funders of realist or meta-narrative reviews. As realist and meta-narrative reviews are relatively new approaches to evidence synthesis, methodological development is needed for both review approaches. Funding The National Institute for Health Research Health Services and Delivery Research programme.
Wong et al. (Mon,) studied this question.