There are two broad approaches that can be used to integrate qualitative and quantitative findings:
Multilevel syntheses: Qualitative evidence (synthesis 1) and quantitative evidence (synthesis 2) can be conducted as separate streams or separate, but linking, reviews and the product of each synthesis is then combined (synthesis 3) (see, for example, Thomas et al. (Thomas 2004)).
Parallel syntheses: Qualitative evidence (synthesis 1) and quantitative evidence (synthesis 2) can be conducted as separate streams or separate but linked reviews. The qualitative synthesis (1) can then be used in parallel and juxtaposed alongside to aid the interpretation of synthesized trials (synthesis 2) (see, for example, Noyes and Popay (Noyes 2007)).
Multilevel and parallel syntheses both require a separate systematic review of evidence, which at a later stage is synthesized with, or juxtaposed alongside, the synthesis of trials. Guidance on the conduct of narrative synthesis (Popay 2006b) contains a toolkit for bringing together findings from different study designs within different methods and approaches. Further methodological work is required on the processes by which evidence from studies using different qualitative methods and generating a range of types of evidence can be synthesized and combined with quantitative findings on effect without compromising the need to minimize bias (Lucas 2007).