Integrating quantitative and qualitative evidence in systematic reviews: a mixed-methods approach

Article type
Authors
C Berg R, Marie-Louise Denison E, Fretheim A, Lewin S
Abstract
Background: Few worked examples of mixed-methods approaches to systematic reviews exist, and little guidance is available. However, the commission for a systematic review on the facilitators of and the barriers to the continuation of Female Genital Mutilation (FGM) mandated inclusion of qualitative data alongside quantitative data to produce an appropriate analysis of a context-bound practice. Objectives: Using an FGM review that is currently underway as case, we present a mixed-methods approach for integrating quantitative and qualitative evidence in a systematic review. Methods: We examined guidelines about and published examples of mixed-methods reviews, eventually drawing on Cochrane’s suggestions of ways in which qualitative evidence can best contribute to a systematic review and a framework developed by the EPPI-Centre, which combines ‘views’ studies alongside trials. Results: The resulting mixed-methods approach allows qualitative and quantitative evidence to be considered within one multi-stage process (Figure 1), i.e. one review with three syntheses in which we operate towards the integrative end of the synthesis continuum. The review question impels the process and results in a comprehensive literature search. (This created challenges with developing a flexible database of search results.) Meta-analysis of quantitative data extracted from ‘predictor’ studies is conducted in parallel with narrative analysis to synthesise the textual data extracted from qualitative ‘perspectives’ studies. In the final step, the findings from the qualitative synthesis are integrated with those from the quantitative meta-analysis. This mixed-methods approach is especially appropriate for complex, context-bound practices like FGM. The approach: 1) grounds the intricate character and context of the practice within the lives of people affected by it; 2) expands the scope, relevance and utility of the review; and 3) provides comprehensive understanding. But the approach is time-consuming and methodologically challenging. Conclusions: Mixed-methods approaches to synthesis of quantitative and qualitative evidence are scarce and we present one promising approach.