Facilitating complex systems research for public health

How do we turn the tide on seemingly intractable conditions such as diabetes and heart disease when their causes are so entwined with the environments we live in?

Public health researchers and policy makers have long understood that the causes of ill health are complex. Yet the biomedical research model, with its focus on linear associations between cause and effect, remains at the forefront of research and practice. In recent years, the case has been made for a change of direction and the need for a ‘complex systems model of evidence for public health’. In April, I attended a thought-provoking workshop at the Health Foundation exploring the barriers to complex systems research and evidence generation. It was hosted by Harry Rutter, Natalie Savona and Martin White who have worked extensively in this space. 

The complex systems approach

Research on the social determinants of health has helped highlight the multi-layered processes that underpin health and disease, many of which sit outside the health sphere. Often these factors, from social networks to green spaces, are considered in isolation.

The dominant methodological approaches of epidemiology and clinical practice, which assume a direct association between x and y, are not sufficient to understand the messy real world in which we live. The hope is that a complex systems approach will improve our understanding of the determinants of disease and allow the design and evaluation of interventions in a more holistic way, ultimately allowing systems to be reorientated towards a beneficial set of outcomes.

There is a lot of interest in the complex systems approach, but questions persist around how best to employ these methodologies and how broadly useful they are. A strong theme from the workshop was the need for greater clarity on the definition of ‘complex systems research’, which can mean different things to different communities. There were also calls for consensus on the methodological approaches in this field. Currently funders, publishers and policy makers find it difficult to judge the quality of evidence under the umbrella of health-related systems research.

Barriers to the approach

The question of evidence is a particularly thorny one, as this approach does not fit neatly into the hierarchy of evidence within the medical research field.

By its nature, complex systems research needs to be adaptive and may not be defined by a measurable effect on a specific outcome. An iterative research process may, to funders and publishers at the biomedical end of the health spectrum, appear ‘softer’ than some traditional scientific methods. Greater clarity on what this approach means in practice, and the methods that could be applied to different research questions, would be extremely valuable. Rather than re-inventing the wheel, we should be learning lessons from disciplines that already study systems, like environmental sciences or political theory.

Another important barrier to complex systems research is the orientation of our funding and academic systems, which make it difficult to find support for inter-disciplinary research projects. Designing or evaluating a ‘healthy city’ would not be possible without engaging urban planners, designers, social scientists and residents, so there is a need to foster more collaborative partnerships. The new UK Prevention Research Partnership (UKPRP) is one attempt to adjust available funding structures to facilitate inter-disciplinary research on the prevention of non-communicable diseases that will take a complex adaptive systems approach. The first round of Consortium Development Grants has just been announced.

To date, the value of the complex systems lens seems to be in improving our understanding of ‘how’ an intervention works. The approach provides a framework for investigating the intended and unintended consequences of an intervention within a system, which can be extremely powerful. Systems maps can also be a useful discussion tool for practitioners as they are accessible to a wide range of stakeholders. The challenge now is to go beyond merely breaking the system down into a sum of its parts and building it back up again; this is still ultimately a reductionist approach.  

I am excited about the potential of a complex systems approach to public health research but also eager to see greater methodological development around the evidence that could be generated.

Sophie Hawkesworth is a Senior Portfolio Developer in Population Health at Wellcome, who are a funding partner of the UK Prevention Research Partnership

Sophie contributed to a roundtable at the Health Foundation on using complex systems thinking to inform practice.