Combining Normalisation Process Theory and logic modelling to enhance impact in a complex intervention: a critical reflection

This paper arises from observing that complex interventions in public health, even when devised from high-quality, trial-based evidence, often fail to achieve expected outcomes and impact in practice. Addressing this theory-practice gap will be discussed in the context of Medical Research Council (MRC) Guidance, Process evaluation of complex interventions (Moore et al., 2014). MRC guidance (2014) refers to using Normalisation Process Theory (NPT) (May & Finch, 2009) and also to using logic modelling (Kellogg, 2004) but does not give any examples of them being used together. Combining this theory (NPT) and method (logic modelling) in an innovative way are the focus of this paper. A researched example of the implementation of a complex intervention based on National Institute for Health and Care Excellence (NICE) Public Health guidance: 26 Quitting smoking in pregnancy and following childbirth (2010) will be used to illustrate the proposed method.AimTo consider the contribution of NPT together with logic modelling to improving the impact of a complex intervention.Methodological discussionNPT shows promise as an analytical tool, since it focuses on eliciting the implementation process of an intervention, and its feasibility and sustainability in context. A logic model can be used to hypothesise the intervention process; allowing for comparison with the findings from a thematic analysis, followed by analysis of the intervention and implementation using NPT. Elements of the theory-practice gap, largely overlooked in trials evidence, were clarified through eliciting other active ingredients and mechanisms of delivery using NPT and logic modelling.ConclusionNICE guidance (2010) is principally based on evidence from trials and would benefit from being combined with knowledge derived from logic models and research based on social theories, such as NPT. Routinely incorporating NPT, or similar, and logic models should be considered when implementing trial-based interventions into complex, public health systems.