Skip header navigation

Evaluating public health interventions in complex systems: challenges and future directions

22 Feb 2017, 12.15PM–1.15PM
Lecture Theatre A1, Cottrell Building
Dr Graham Moore

Graham Moore is Deputy Director of DECIPHer, Cardiff University. DECIPHer is a UKCRC centre of excellence for public health research, focused on health and wellbeing of young people.

Graham leads DECIPHer’s complex intervention methods programme, and was recently lead author on the MRC’s guidance for process evaluation of complex interventions.

Substantively, Graham’s work focuses on two main strands: 1) the effects of schools and school based health interventions on socioeconomic inequalities in young people’s health and education and 2) tobacco control policy and the denormalisation of smoking among young people.

He has recently joined the NIHR Public Health Research Funding Board, and is a member of the Cancer Research UK Prevention review panel.

There is increasing recognition that randomised trials and natural experiments of public health interventions do not tell us, in binary terms, whether or not an intervention “works”. Rather, effect sizes provide us with an estimate of the balance of benefit vs harm following the introduction of an intervention in a particular time and place. An understanding of mechanisms and context is then vital to make sense of the extent to which evaluation effects may translate to a new context.

Mainstream “does it work” thinking has increasingly been challenged from positions underpinned by critical realism and complex systems thinking. Realist evaluation (Pawson and Tilley 1997) has highlighted the contingency of intervention effects on their contexts, and the potential for the same course of action to activate different mechanisms in different spatial or temporal contexts.

Complex systems perspectives have challenged dominant definitions of complex interventions, moving away from viewing complexity as a property of intervention components and the interaction between them, and toward viewing complexity primarily as a property of the contexts in which interventions are delivered, conceiving interventions are events within complex systems (Hawe et al. 2009).

Much work has been conducted to operationalise the implications of realist perspectives for evaluation (Bonell et al. 2012; Fletcher et al. 2016). However, there is a need for further work to understanding the practical implications of a complex systems lens for how we do evaluation.

This talk reflects on what it means to adopt Penny Hawe and colleagues’ view an intervention as an event within a complex system, and the implications of this perspective for how we evaluate interventions in complex systems.

Scroll back to the top