An introduction to narrative synthesis of quantitative data: a participatory workshop using a worked example from public health

Thomson H1, McKenzie J2, Doyle J1, Brennan S3
1Cochrane Public Health, 2Statistical Methods Group, 3Australasian Cochrane Centre, Monash University

Objectives: To facilitate learning:

- through experience about how to conduct a narrative synthesis.

- about the challenge of ensuring transparency and consistency in narrative synthesis.

Description: Narrative synthesis is a common synthesis method used to synthesize data when meta-analysis is not possible or appropriate. Around 10% of Cochrane Reviews use narrative synthesis for some or all of the data. In public health reviews, where there are often multiple sources of heterogeneity, narrative synthesis is frequently the main synthesis method. Despite this, narrative synthesis is often poorly reported, and the links between the data and the conclusions of the synthesis are often not transparent. The persistence of these problems may, in part, be due to the lack of support or clear guidance to promote rigour in narrative synthesis.

We will start with a presentation on the principles and challenges of narrative synthesis, and useful tools to improve transparency. Participants will have the opportunity to trial the knowledge presented using extracted and tabulated data from a published Cochrane Review of a public health intervention, from which they will prepare a brief narrative synthesis of one or more outcomes. Small group discussion will feed into a final discussion with the whole group about the issues in preparing a narrative synthesis, including promoting transparency and replicability.