Geographic information

Geodemographic profiling benefits stop-smoking service

From the time John Snow used a map to determine the source of a cholera outbreak in London’s Soho in 1854, there has been a tradition of using maps to analyse public-health priorities. GIS-based needs assessment is relatively well developed in the United Kingdom.(1)

Yet the full potential of the technology has remained underexploited by the public sector, some continuing to view it primarily as a mapping tool. In recent years however, there has been renewed interest in the use of GIS and specifically geodemographics, for spatial analysis.(2,3) The increased use of these tools has coincided with the development of public-health approaches, including health equity audits (HEAs), health impact assessment and social marketing.(4)

Geodemographic (GD) classification systems are useful tools for learning more about a population, enabling the identification of target groups according to a range of health-related factors. They provide a large amount of detail on socio-economic indicators and sociocultural behaviour with reference to location. There are a number of commercially available GD systems, including Mosaic, ACORN, and P2 People & Places.

HEA is a practical approach to help PCTs to meet their commitments to reduce health inequalities by targeting resources at those with the greatest need. HEAs require measurement and analysis of whether those with the greatest need are using the services provided. Findings need to be clearly presented to ensure effective decision making to remedy demonstrated gaps in delivery.

PCTs have identified the need to share specific expertise on methodologies for HEA.(5) However, very few methodological papers or reports of actual HEAs have been published. A search of published and unpublished literature on HEAs suggested that the use of GIS and GD systems in HEA is very limited. We set out to explore their use as tools for HEA.

Smoking cessation was selected for the HEA because reducing prevalence of smoking is a major priority for Nottingham, as that among adults in the city is estimated to be very high 35% compared to a national average of 27%.(6)

Nottinghams New Leaf Stop Smoking Service offers counselling one-to-one, in groups and at drop-in clinics with trained advisors at a variety of locations across the city, including GP practices, community centres and leisure centres. Clinics are held on different days and at various times to ensure the service is available to the maximum number of clients.

Using GIS and GD systems for health equity audit

We used Mosaic Public Sector to determine the groups with highest levels of smoking in Nottingham, and therefore with the highest need for the New Leaf service. Mosaic is a geodemographic tool that uses over 400 data items (54% from the census) to identify natural groupings that exist in the population.

Each postcode, and household, is assigned to one of 11 distinct lifestyle groups (and 61 types within the groups) by measuring how close it is, statistically, to an average value for each of the data items.(7) The levels of smoking within each group are indexed by using the British Market Research Bureau (BMRB) Target Group Index (TGI) survey, a continuous survey based on a sample size of around 25,000 interviews per annum.

We used two methods to characterise users of the New Leaf service. Firstly, the postcodes of all clients accessing New Leaf between January and December 2005 were linked to a Mosaic group. We then mapped these same postcodes using MapInfo, a GIS software programme.

We combined these sources of information to determine whether there was equity of service provision.

Findings of the geodemographic analysis

Table 1 (below) shows the ranking of Mosaic groups by the proportion of these groups in Nottingham (1), and a Nottingham index (2) a ratio of the proportion of the group in Nottingham to that in the England population (2).

Table 1.

This shows that the largest groups in Nottingham are groups G, E, D and F. Nottingham has three times more people in group G compared with the average UK population (index of 305). This group has the highest index for smoking (3), with an index value of 125; those in group G are 25% more likely to smoke than average. Groups D, F and H are similarly large in Nottingham with high levels of smoking. Group E, although the second largest population in Nottingham, has an average level of smoking.

Between January and December 2005, 8090 people accessed New Leaf. We were able to match 7892 of these users postcodes to a Mosaic group to ascertain the relative proportions of people from these groups accessing New Leaf (4).

The majority of those using the service (34.4%) were from group G. The access index a ratio of the proportion of the group in Nottingham to the proportion using New Leaf shows that group G are 73% more likely to use the service than average (6). There was only a small reduction in those in group G and F using the service who were still not smoking at four weeks (four-week quitters).

Figure 2, meanwhile, shows that there is a high correlation between service need and use, as measured by the Mosaic Smoking Index and New Leaf Access respectively.

Figure 2.

Combining GIS and geodemographic analysis

Figure 3 shows where the five groups with the highest smoking live in the city.

Figure 3

Mapping the 7892 New Leaf user postcodes showed cold and hot areas of service uptake (Figure 4). By comparing these two maps, we determined that there was a good overlap of areas of high need and service use.

However, gaps were also evident in certain areas for example, those where uptake was low, considering the groups living there. Although groups D and F live in central Nottingham, there is a low uptake of the service in this area.

Figure 4. Levels of uptake of the stop-smoking service

Conclusion of the HEA of Nottingham’s smoking cessation service

This HEA showed that the New Leaf service is doing very well in targeting the service according to high need, with a strong correlation between high smoking, access to the service, and the four-week quitting outcome.

By juxtaposing the GD information onto a GIS map of service uptake, we were able to identify specific geographic areas where there were gaps. We have since used GIS mapping to identify the GP practices in these low-uptake areas and are now working with them to increase the number of people with access to the service.

In addition, we are exploring the use of social marketing in targeting these specific groups of people, as their Mosaic profiles provide considerable detail, including smoking behaviours, which may prove useful in this.

Using GIS and geodemographics for HEA

A key identified barrier to PCTs undertaking HEA is the availability of data, lacking or limited for many issues, and the perception that it involves time-consuming data analysis.(5)

Our experience was that if postcode data is available, GIS and GD systems can rapidly provide information on health needs and service use in great geographic detail. We believe that the visual impact of the maps, with information being provided in a geographically identifiable way, enabled us to engage with the smoking cessation service manager to address identified gaps probably more effectively than using traditional analysis.

It has also helped us to sell the concept of HEA in wider, non-public health forums. And it has provided us with detailed information about the service users that we need to target to address these gaps.

GIS and GD systems and public health

There has been increasing interest in the use of GIS and GD systems as analytical tools for public health practice, including health-needs assessment, health equity audit and social marketing.

They show promise in providing multidimensional, readily available information that is accessible and meaningful to managers and partners, enabling them to make service changes to ensure equitable provision.

We would welcome further published case studies to improve our understanding of how best to utilise them in a rigorous, meaningful way. We also welcome further work analysing the methodologies of the different GD systems and the differences between them.(4)

References

1. McLafferty S. GIS and health care. Ann Rev Public Health. 2003; 24: 2542.
2. Longley P. Geographical information systems: a renaissance of geodemographics for public service delivery. Progress in Human Geography 2005; 29(1): 5763.
3. Mathieson S. Picture this. Health Serv Jnl. 2005; 115(5957): 46.
4. Dedman D, Jones A, Tocque K, Bellis M. Social marketing: geodemographics in action. Synthesis 2006; 6. Liverpool: North West Public Health Observatory, 2006. Available at: www.nwph.net/nwpho/Publications/
   Synthesis_6_Socialmarketing.pdf [Accessed 04.01.2007].
5. Research Works Ltd. Understanding the barriers to completing health equity audit in PCTs: findings from a qualitative study. London: Health Development Agency, 2005.
   Available at: www.nice.org.uk/page.aspx?o=525703  [Accessed 04.01.2007].
6. Twigg L, Moon G, Walker S. The smoking epidemic in England. London: NHS Health Development Agency, 2004. Available at: www.nice.org.uk/page.aspx?o=502811  [Accessed 04.01.2007]
7.  Experian Mosaic Public Sector: putting the citizen at the heart of government. Nottingham: Experian, 2006. Available at:
   www.business-strategies.co.uk/upload/downloads/
   mosaic%20public%20sector% 20factsheet%20for%20website.pdf [Accessed 11.01.2007]

Jeanelle de Gruchy, Consultant in Public Health Medicine, Nottingham City Primary Care Trust Health Equality Directorate

Jean Robinson, Head of Information, Nottingham Health Informatics Service