Background: Evidence based practice relies on robust methods of integrating research evidence. Traditional methods which rely on synthesis of trial data have limited ability to explain why or how an intervention exerts an effect. Recent years have witnessed sustained growth in methods to synthesise qualitative data, seeking to address this gap in understanding. However, exploring complete understanding of intervention effectiveness requires an approach that allows the synthesis of both quantitative and qualitative data. Methods of integration are in their infancy. Some methods, such as those pioneered by Harden and Thomas (2005), bring the results of a synthesis of qualitative and quantitative data together in the final stages of the review process. The qualitative data is used to explain the findings of the quantitative data. However, such approaches may not optimise the use of qualitative findings to inform the quantitative analysis. They may also inaccurately assume that findings from one setting can be transferred to another.

Methods: We carried out a participatory realist review which synthesized quantitative and qualitative published evidence with unpublished expertise from practitioners on culturally supported approaches to promoting health literacy. Findings from this review were used to co-produce a model of how community engagement and peer support promote health literacy.

Results and Conclusions: We shall present existing models of integrating qualitative and quantitative evidence, their strengths and weaknesses, and use our worked example to describe an innovative method used in the synthesis of quantitative and qualitative research to evaluate models of peer support to improve health literacy. The methods adopted in the worked example allowed us to integrate qualitative data in a way that informed subsequent data selection and analyses. Advantages and disadvantages of these methods will be presented.