A new approach to defining ‘place’ that’s supporting action on health inequalities Q&A with Luke Munford and Hardeep Aiden

Why is it important to understand health in terms of smaller areas of place?

Luke: The environment where people live, work and play is very important for their health. But the way we currently measure the health of a place is based on geographic definitions that aren’t always that meaningful when it comes to supporting decision making about how to improve health. 

Typically, health data is reported across quite a big area, like a local authority. The problem with this is that it hides variation. For example, Kensington and Chelsea is one of the most affluent local authorities in the country. But some pockets of Kensington have extremely poor health. This is masked if you only look at the bigger picture.

Typically, the smallest geographical area we’re able to break things down to is the Lower-layer Super Output Area (LSOA), which is an area of around 1,500 people. But these arbitrarily drawn boundaries can still hide a lot of variation. If 50% of the population in that area have really bad health and 50% have excellent health, the average will be middle health, which doesn’t represent anybody.

Tell us about your work to improve our understanding of the health of places

Luke: To better understand inequalities, we need better definitions of small areas. Our research has focused on working out the best way to break down the data reported at a big geographical level into smaller clusters that give a more meaningful picture of local health and inequalities. 

Starting in Greater Manchester, using census data and working with the Office of National Statistics, we looked at ways we could redraw the boundaries of those small areas, grouping neighbourhoods in ways that meant the health of local people was more similar. We’ve then started testing our model out in other areas. 

Hardeep: At the heart of this work is a really simple idea. Imagine these areas as children’s building blocks. Right now there are lots of random colours stuck together. Some of the red pieces are stuck to the yellow pieces. Other yellow pieces are stuck to some blue pieces. Luke is taking these building blocks and recombining them in more logical ways. 

When we talk about ‘place’ in statistical analysis, it's often divorced from people's understanding of what place means. Grouping local areas together that have more in common means this often aligns better with how people naturally think of their local areas. 

How could place-based leaders use your research in practice to improve health?

Luke: The way we’ve previously thought about place-based health inequalities might be wrong because we've based our understanding on averages of possibly disparate areas. Knowing where the hotspots of poor health really are means we can target funding in a more nuanced way.

If you look at the association between health and wider social and economic outcomes, using our new measures the relationships become much stronger. It could be that we have been underestimating the true effect of health on certain social and economic factors. 

Hardeep: Hopefully, we will end up with a more accurate picture, which chimes with people’s experiences of the places they know. This work should allow policymakers to make decisions that reflect what people are actually seeing in their communities, as opposed to an arbitrary grouping of postcodes. 

You’ve already spoken to some combined authorities about your work – how are they using it?

Luke: We know inequalities are getting worse in the UK and we need to develop genuine place-based solutions. By making health data more meaningful, we can help leaders target investment to areas of really high need and poor health. There seems to be a strong appetite for that from people we've spoken to. 

We've worked closely with Greater Manchester Combined Authority throughout this project. We’re further testing our model using administrative health records, giving us detailed information on people’s contacts with primary care. We also talked to the community, incorporating the opinions of people who have lived in these areas for years.

Using our new model is giving a slightly different picture of health inequalities in Greater Manchester, which is useful for decision makers. Some new areas are coming out as having worse health than we initially thought. And conversely, there are some with slightly better health. 

Hardeep: We’re excited about this work at the Health Foundation because we can see how it could be translated into practical action across the country. 

If you’re thinking of building a new school, extending a tram line, or building new facilities, having accurate data will help you understand where it could have the most impact on health. We probably underestimate the true effect of these things at the moment because we don't currently look at impact at community level. 

At a population level we understand that employment can improve health outcomes. But employment in an area where self-reported health is poor can improve outcomes really significantly. And we can only measure that if we’ve got good data. 

The Health Foundation is funding further work on this – what will you be doing next?

Hardeep: The original project, funded as part of our Social and Economic Value of Health programme, used data from the 2011 census. The next stage of Luke’s work will be updating the model with more recent data from the 2021 census, and testing it out in a few more areas. 

We've talked about making this work as relevant as possible to local authority stakeholders. In this next phase the project will explore how to engage those audiences to use this work to solve problems they face.

In parallel, we're funding a PhD studentship to look at how we visualise data coming from this and other projects. 

Luke: That will be really helpful. Maps have been a really powerful tool for explaining this work to people. If you can show someone a map of their borough and say, ‘This very dark red area is grouped with this green area,’ that's when people start to realise that the groupings have been a bit arbitrary.

Once we've finished updating data for Greater Manchester and Durham and the Tees Valley, we can apply our model to the whole country. 

Then we plan to publish all of our code on an open-source platform so all local authorities will be able to plug in their raw data. That will create maps and code to show the new areas, and they can see whether it makes a difference to their understanding of health inequalities in their local communities. 

This is not an academic exercise, we really want this work to be useful and to help local areas take action on health inequalities.  

This content originally featured in our email newsletter, which explores perspectives and expert opinion on a different health or health care topic each month.