An empirical study series to investigate the research synthesis of complex health care interventions and related methodological issues

Purpose: This thesis aimed to evaluate how theory-orientated approach to research synthesis of complex health care interventions may facilitate better understanding of intervention mechanisms. Thesis intended also to evaluate how qualitative research compliments a systematic review and meta-analysis of complex health care interventions, especially what participants’ perceive as effective intervention features and how this compares with systematic review and meta-analysis evidence. By combining these different approaches thesis aimed to improve reporting of reviews of complex health care interventions by providing more detailed information about intervention mechanisms that appear to be associated with a successful intervention. Methods: The thesis was built on a series of empirical studies. Multiple bibliographic databases and references of retrieved articles were searched for relevant review articles, randomised controlled and qualitative studies. Random-effects meta-analyses were conducted to estimate effectiveness of psycho-educational smoking cessation interventions, while behaviour change techniques used in the studies and their suitability to change behavioural determinants were evaluated using a framework by Michie et al. (2008). Thematic analysis was conducted to explore qualitative studies, while narrative analysis was used to bring the different case studies together. Results: Psycho-educational interventions significantly increased point prevalent and continuous smoking cessation, and despite superficial differences, interventions appear to deploy similar behaviour change techniques. Qualitative research suggested considerable variation in patients’ expectations and experiences of psycho-educational interventions, but combined results suggested that many of the techniques used in psycho-educational interventions appear to be well matched with patients’ experiences of successful interventions. Conclusions: Although questions remain about how to distinguish characteristics of an effective intervention, the theory-orientated approach to systematic review and meta-analysis was able to provide a detailed analysis of the intervention techniques to help in the design of future interventions. This approach, however, is labour intensive in its present form