Making sense of diverse research evidence: reflections on the use of the interweave synthesis approach to mixed-methods synthesis

Article type
Authors
Thompson-Coon J1, Gwernan-Jones R1, Garside R1, Nunns M1, Shaw L1, Moore D2, Lourida I1, Abbott R1, Melendez-Torres G1
1University of Exeter Medical School
2Graduate School of Education, University of Exeter
Abstract
Background: outputs resulting from mixed-methods syntheses are expected to have more relevance and utility for decision makers who are unlikely to be interested in understanding whether something works without considerations of context, accessibility and feasibility. Methodological guidance for mixed-methods synthesis continues to emerge and evolve but broadly involves a sequential, parallel or convergent approach according to the degree of independence between individual syntheses before they are combined.

Objectives: to describe and reflect on three case studies in which we used novel and innovative methods to draw together the findings from individual but related quantitative and qualitative syntheses to aid interpretation of the overall evidence base.

Methods: our approach moved beyond making a choice between parallel, sequential or convergent methods to interweave the findings of individual reviews and offers three key innovations:
1) the use of inter-subjective questions to understand the findings of the individual reviews through different lenses;
2) immersion of key review authors in the entirety of the evidence base; and
3) commencing the process during the synthesis stages (or earlier) of the individual reviews.

Underlying these methods is the process of ‘building bridges’ and exploring differences between and across review findings; an approach that is fundamental to all evidence syntheses but usually occurs at the individual study level, rather than between and across the syntheses themselves.

Results: allowing adequate time for debate, reflection and iteration was fundamental to progress. Enabling a full understanding of the breadth of evidence contributing to the overarching synthesis across as many members of the review team as possible greatly facilitated this. As review authors tend to be familiar and comfortable with either the quantitative or the qualitative evidence, we needed to develop mechanisms for the efficient sharing of understanding across the entirety of the evidence base. Necessary differences in the focus of research questions and the scope, range and nature of the target body of evidence available created many challenges, some of which were apparent at the outset of the individual reviews, others emerged as the reviews progressed. Using knowledge gained during the scoping processes or considering the potential for overarching synthesis during the design stage of multi-review projects, or both, may help to optimise opportunities for identifying links across individual reviews.

Conclusions: our experience suggests that adapting methods for exploring and identifying patterns and links between and across individual studies to interweave the findings between and across reviews is a valuable addition to mixed-methods synthesis methods.

Patient or healthcare consumer involvement: healthcare consumers and other relevant stakeholders were embedded within the research teams conducting all three case studies.