‘Natural experiments’ could hold the key to better public health research
24 March 2016
Conducting studies of ‘natural experiments’ such as policy changes and infrastructure projects could hold the key to identifying key factors essential for improving public health, a seminar has heard.
Public health researchers are struggling to build the evidence base policy makers need because, in the complex real world, different interventions work in different ways for different people, said Cambridge University’s Dr David Ogilvie.
Dr Ogilvie was keynote speaker at the seminar ‘Physical activity for public health: in pursuit of rigorous evaluation in the real world’, jointly organised by the Prevention Centre, the Prevention Research Collaboration at the University of Sydney and the Sax Institute.
In studying physical activity, for example, researchers often focused on whether certain types of behavioural intervention worked rather than on strategies to address the more fundamental social and environmental reasons people were inactive in the first place, he said. Understanding how and why a new piece of infrastructure made people exercise more – even if that was not an intended consequence of the project – could provide generalisable findings to inform future policy direction.
“There are often multiple outcomes of interventions reached through multiple causal pathways. We need to understand why some things work in some places and not in others – what are the critical underlying functions of successful strategies that can be addressed in different ways in different places,” said Dr Ogilvie, leader of the MRC Epidemiology Unit’s physical activity and public health research program at the University of Cambridge.
“By drilling for these underlying functions and understanding more about ‘What’s going on here?’ we may ultimately be able to generalise – to go beyond the question ‘Did it work’ to ‘What should we do about this?’”
His team has conducted a number of large natural experimental studies, with key findings going on to inform local planning as well as England’s national physical activity framework, NICE guidance and parliamentary committees.
The Commuting and Health in Cambridge study measured people’s physical activity following the introduction of a busway with a walking and cycling route, using baseline and annual follow-up questionnaires with key outcomes validated using combined heart rate and movement sensor and GPS data. It found that the closer people lived to the busway, the more likely they were to increase the time they spent cycling to work over the duration of the study.
Another study, iConnect, followed the behaviour of people in towns where obstacles to walking or cycling had been removed, for example by building bridges. It found that people were more likely use the new routes for recreational purposes than for commuting to work, but the new routes nevertheless encouraged people to walk and cycle more and to increase their overall level of physical activity over time.
Panellist Elizabeth Flynn, Assistant Secretary of the Preventive Health Policy Branch of the Australian Government Department of Health, said interventions that impacted public health were sometimes implemented because of the real life experience of a politician or a presentation by a passionate researcher, rather than on scientific evidence alone.
“We aim for evidence-based policy, but there is also accidental discovery. Research using natural experiments is probably more valuable to policy makers than small scale studies that aren’t always scalable and can’t be replicated,” she said.
– Helen Signy, Senior Communications Officer