Understanding complex systems and context, while retaining generalisability of findings, are key issues, as the next stage of our research tries to genuinely involve the public.
By Matt Egan
SPHR at the London School of Hygiene & Tropical Medicine (SPHR@L) has dealt with a lot of different topics and subject areas: the use of evidence in local government; the use of internet platforms to encourage knowledge transfer; community empowerment; e-cigarettes; healthy ageing – for example. Here we focus on some of the alcohol work we have done, and discuss how local approaches to tacking problems related to alcohol illustrate many of the wider challenges facing researchers interested in informing public health decision-making.he NIHR School for Public Health Research has reached the end of its first five year programme. This is a good time to reflect on some of the work that has taken place, and consider some of the challenges and opportunities ahead, as a new programme is developed for the next five years.
Our study of local drinking – just one of SPHR@L’s many projects – helps to illustrate three major challenges that we’re addressing so that the School really makes a difference.
Three challenges
First, it’s shown to us how our work needs to think about systems that people fit into. Promoting healthy drinking isn’t simply about informing people about the dangers of alcohol. Camden’s drinkers are part of a wider system comprising, among others, retailers, licensing laws, producers, local and central governments, residents and visitors. Each part influences the others. As public health researchers, we must understand complexity, interconnectedness and how whole systems operate.
Second, good public health research should have meaning beyond its specific context: any findings we produce should resonate beyond Camden. It’s no good us getting to grips with the complexities of drinking in north London if we don’t also have something useful to say from it to help folk south of the river. In short, we must avoid sacrificing generalisability in pursuit of complexity and context.
Our third challenge is to work out how the “public” should be involved in public health research. That’s not just so the “public involvement” box is ticked in funding applications. It’s to hold a mirror up to our work and spot what’s missing. When we hosted a forum for Camden residents last summer, we recognised how very differently people see what’s going on. One person might be speaking about a great night out in Camden. But that could be the flipside of some others experiencing outsiders urinating in the street, drunken people making the streets unsafe and leaving a mess that council taxpayers pay to clean up.
- Complex systems
Let’s take each of our three challenges, one by one. First, there’s researching complex systems. One of our projects, “Reducing the Strength” examines the efficacy of voluntary restrictions on the alcohol content of beer sold in shops. The initiative in Camden and Islington impacts on street and homeless drinkers, on service providers and on retailers. But each of these groups also impacts on the others, as well as people beyond. That raises the question of how the whole system operates and the interconnectedness of different elements that are potentially affected by the intervention.
This is just one example of how we are stepping away from a tendency in public health research to fashion a grand, linear narrative in which a single magic bullet is responsible for a clear and quantifiable outcome. We find ourselves increasingly trying to map systems, to embrace complexity, even if that process is messy, troublesome and awkward.
Why does embracing complexity matter? Partly because this approach fits much better the landscape in which public health policy and practice now operates. Historically, the evidence culture among public health academics has been based on scientific methodology and epidemiology. We’ve focussed on quality and demonstrating singularity of cause and effects, rather than complexity. This approach remains important. However, it can mean setting aside a lot of findings and may lead to a selective understanding of a full range of findings. Some rich research, showing how an intervention can have a host of impacts, can easily be lost.
In contrast, public health policy makers and practitioners, sitting in local authorities, are often more concerned with feasibility, and with whether insights have relevance in their geographical area, whether they address problems that require solutions in that area, whether findings are relevant to what they, as practitioners – as people working in local government – can actually do. We’ve realised that the public health research culture has needed some updating so that it is useful to all the people who need it for their work.
How does this work in practice? Rather than assuming that practitioners and policy makers simply need enhanced methods/more magic bullets, we’ve tried to understand better the constraints under which they work. To that end, we’ve shadowed local government practitioners, observing them, running consultations with them, and holding focus groups. As a result, we recognise that the options open to public health practitioners are limited by many factors, including statute, finance, local and national politics and, sometimes, by the unexpected. They are required to perform certain actions but they also have choices about whether to do others, in intensive or non-intensive ways. One of our tasks is to offer them evidence highlighting possibilities that they can act upon.
For example, local authorities are empowered to define areas known as Cumulative Impact Zones (CIZ) within which licensing regulations can be more rigorous. CIZs tend to be created in areas where alcohol problems are particularly acute. Local authorities then have many options – how large the area should be; which premises to apply regulations to; whether to use the rules to reject as many premises as possible or use them to negotiate the types of premises wanted. It helps them if we can provide evidence of how other local authorities have used these powers, to highlight how someone in a similar job in a similar area has successfully taken a certain action.
- Generalisability
All of this leads to our second challenge, that good public health research should have meaning beyond its specific context. This is clearly more difficult when research is also tackling complexity. However, even if precise findings are not generalisable because of the specificity of the context, some of the key theories that underpin those findings may be still be generalisable. For example, in our “Reducing the Strength” project, we found evidence to support the theory that the closer area approached 100 per cent of retailers cutting the alcohol in super strength beers, the less likely it was that drinkers would circumvent the initiative by going to a different shop for alcohol. This might be a more useful, generalisable finding than one, for example, that precisely measured impacts on crime rates in one specific locality where ‘Reducing the Strength’ was implemented – given that there is no reason to assume those precise impacts will be replicated in other, different, places.
- Involving the public
Our third challenge is to work out how the “public” should be involved in public health research. It builds on work we have already done on partnered research with public health practitioners. This “co-production” has already led to co-authorship of an academic paper on “Reducing the Strength” on which Colin Sumpter, who was Public Health Strategist for Camden and Islington, was lead author.
Our work in Camden shows the value of our focus on involving the public. Academics have a voice in alcohol use research. So does industry and the media. But the public voice is often missing. That’s a problem, because, ultimately, it’s the public’s health that matters to us and they can sometimes highlight what’s missing in our research agenda.
The experience of bringing members of the public together into a free forum and asking them open-ended questions has been that people hold strong, diverse views that they will share respectfully. Getting these events right – reaching out to people, briefing them and encouraging them to think about the issues in advance – is rewarding but takes time and resources. Our consultation confirmed that the public are not homogeneous and showed how alcohol policy can have many different kinds of winners and losers. Also, it demonstrated that, if researchers are going to involve people, we should be ready for surprising messages.
We will be doing more public consultation in the future. It will be an important element in the second stage our NIHR SPHR work, as we continue to grapple with researching complexity, context, relevance and usefulness for the people we serve.
Dr Matt Egan is an Associate Professor at London School of Hygiene and Tropical Medicine, with an interest in evaluating the health impacts of complex social interventions.