Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

The authors of this blog reflect on ways to characterise research sites used as case studies for an investigation on the role of social prescribing link workers in primary care.

The study we are conducting has involved generating data in seven GP surgeries (our research sites) geographically spread across England. Each surgery has at least one link worker attached to it. This individual may be employed by a Primary Care Network or by a local voluntary sector organisation. A purported aim of social prescribing is to address health inequalities by targeting people most needing non-medical support. In this blog, we explore how we have started comparing socio-economic differences across research sites.  

What are the areas around each research site like?

As we bring our data together and compare our findings across sites, we are interested in factors that might make each site different to others. For example, if a surgery is in a rural area, transport networks for patients might be challenging, limiting the link worker’s options to connect people to local community activities. Another factor of interest is the relative levels of socio-economic deprivation of people living in areas surrounding each research site. This might affect the type of patients a link worker sees or the availability of activities or support in the area. There might be high levels of unemployment or low income, meaning people cannot afford to participate in clubs or groups. At some sites involved in our research, link workers are spending initial appointments ensuring that patients have access to food banks or supporting them to claim social welfare payments before they can consider linking or signposting them to community groups or activities.

How can we characterise levels of deprivation?

We want to characterise our sites in useful and descriptive ways, but also use measurements that allow for comparisons. We discussed ways of characterising deprivation with our co-applicants and have made some decisions based on their advice and experience of other studies. Primarily, we looked at compound measures such as the Indices of Multiple Deprivation (IMD) and individual metrics such as Health literacy scores.

The IMD is derived from government statistics. It defines ‘deprivation’ as unmet need in a broad range of areas covering income, employment, education, health, crime, housing, and living environment. Poverty is defined as meaning lack of financial resources, whereas deprivation is a wider concept embracing different types of unmet needs rather than just financial impoverishment. 

Health literacy and numeracy scores, developed by the University of Southampton, use data made available by Health Education England. In the UK, it is estimated that 16.4% of adults read and write at, or below, the level of a nine-year-old and, critically, 43% of adults do not understand written health information. Health literacy comprises both an individual’s ability to understand and use information to make decisions about their health, and healthcare providers’ ability to make information easy for people to understand. Low health literacy is thought to be connected to less positive health outcomes. If patients have poor health literacy, they might not know what services are on offer, have limited knowledge of how to access services, or not understand how to use health service or voluntary sector organisations’ information to support their health. Healthcare professionals might not be aware of these needs when explaining how patients can look after their health. Importantly for our study, this might also affect whether patients accept a referral to a link worker and the work that link workers can do with them.

Do deprivation measures divide up the country in a way that is useful for our study?

The IMD and health literacy scores are ways of considering characteristics of people who might be referred to link workers. The IMD measure is calculated at the Lower-layer Super Output Area (LSOA) level; a small geographical area used for the National Census and designed to make areas comparable across the country. These LSOAs consist of around 400-1200 households, encompassing between 1,000 and 3,000 persons. This allows IMD scores to be aggregated to council ward and local authority levels as well as other geographical areas. The tool comparing health literacy represents statistics at the local authority level only. None of these areas match the primary care populations served by link workers.

When we looked at our sites, patients using a doctor’s surgery (and therefore connected to a particular link worker) might live in different local authorities or LSOAs with different deprivation levels. At one of our research sites, we found people referred to the link worker came from two local authority areas with very different health literacy levels; hence, while the surgery itself is in a local authority area with health literacy at the national average, some patients came from a local authority with the lowest levels of health literacy in the country. This might mean the link worker sees people with very different understandings of health information and abilities to access support services.

The map below shows another of our research sites. The arrow points to the GP surgery (our research site). The red circle indicates the approximate patient catchment area for the surgery. The circle covers parts of five LSOAs (shown by different coloured areas and blurred for anonymity). Each of these LSOAs have different IMD scores. The darkest blue represents the most deprived 10% of areas in the country, and the lightest green depicts the least deprived 10% of areas. The stark difference in IMD scores shows how widely experiences of deprivation can vary across small populations – such as the population in one GP surgery.

  Map of a research site

What does this mean for how we characterise our sites?

As the above map indicates, there is a continuum of deprivation experienced across GP practice populations. Hence, as with health literacy levels, using one assessment of deprivation does not sufficiently cover the complexity and nuance of health inequalities affecting people who might be referred to link workers.

Is there a potential solution?

We have been working with Richard Blackwell at the Southwest Academic Health Science Network (SW AHSN), who has undertaken extensive evaluation work with statistics in primary care. Richard has calculated a Weighted Population IMD for each GP practice, or surgery, in England using the LSOA of the patient’s address (taken from the NHS Digital: Patients Registered at a GP Practice data) along with the IMD score for that LSOA. The index balances the numbers of patients experiencing different levels of deprivation so that, for example, the inclusion of just a few patients from an area of deprivation will not affect the overall deprivation score for the surgery. Using this information, we have found one of our sites has one of the least deprived population of patients in the country, while another site in the same local authority has average deprivation. This gives us some context when considering the role of link workers in these two practices that are so close to each other.

As with all measures, we need to be aware of its limitations. This measure relates to average health inequalities experienced across the GP surgery, and not the average for patients referred to a link worker. Link workers involved in our study state that patients referred to them are often complex cases; people experiencing severe socio-economic problems and severe or enduring mental health issues. Additionally, measures are not themselves without error. For example, in some predominantly wealthy areas there are pockets of deprivation, and people from these areas might be patients most likely to need support from social prescribing.

Consequently, we will still seek ways to supplement the Weighted Population IMD with other sources of both quantitative and qualitative data. Qualitative data will allow us to paint a picture of a research site in more detail than can be provided by measures such as IMD and health literacy scores. Qualitative data can also reveal explanations behind the numbers; for example, in describing the types of issues that patients bring to link workers and characterising features of the areas in which patients live.

Conclusion

No measure is perfect, but we have learned from this exercise of trying to characterise research sites and the populations they reach. Narrative description is still essential, but we now have more helpful means to quantify differences between sites. We have demonstrated important nuances in trying to understand individual differences using aggregate measures. This has suggested it might be helpful in future studies to look more closely at the patients within a practice who are referred to social prescribing; we might collect individual patient scores on health literacy and the IMD of patients’ home postcode, for example.

This process has shown the value of having a wide range of expertise on a study team. The topic of measurement stimulated a discussion about deprivation, what this means for social prescribing, and alerted us to various tools. It also led to a valuable connection with a colleague in the SW AHSN who was able to help us find new ways to characterise deprivation.

We would like to thank other members of our co-applicant team who contributed to this discussion and made helpful suggestions from their wide experience about how to characterise deprivation at out sites: Kamal Mahtani (joint lead applicant) and co-investigators: Geoff Wong, Caroline Mitchell, Sabi Redwood, Beccy Baird, Catherine Pope, Joanne Reeve.