USING GEODEMOGRAPHICS
Linking patient data with geodemographics give public health researchers insight in the socio-economic profile of health conditions and areas. Diseases can be profiled using postcode types, and the same disease profile can be attributed to another area coded with the same postcode type. This powerful way of comparing diseases in areas and types derives from the statistical techniques used to crate the geodemographic classifications. If Geodemographics is the "analysis of people by where they live" (Sleight 1997, p.16), Health geodemographics could be defined as "the analysis of people's health by where they live".
The use of small area classification systems in private sector is commonly diffused and accepted as a standard marketing tool. Additional information is often linked to these systems to obtain better discriminatory power in selecting potential consumers for targeted commodities.
This additional information on consumer behaviours and attitudes is generally defined lifestyle data and comes from different survey types. Typical surveys are made by mail or by telephone. In the lattermost the interviewer asks specific questions on consumers' personal preferences and life habits. The information collected is used to model and predict lifestyle data.
In developed countries lifestyle and behavioural risk factors are the most important causes of ill health (Davey Smith 2003). Using lifestyle data in geodemographic profiling could be useful and can improve health promotion and screening campaigns. Healthy life choices are at the core of public health policies today (Department of Health 2004). Moving high-risk groups to change behaviour on a number of areas like diet, smoking and exercise have shown to have immense health benefits and savings for the public health care budgets. So-called lifestyle diseases, e.g. heart disease, diabetes (type 2), account for an increasing proportion of health care demand in most developed countries. So a better understanding of high risk population groups of not only their health care needs, but also their circumstances would be useful tools for public health decision making.
But there are some caveats in using lifestyle data, related to the statistical assumption underneath sampled surveys and to the nature of a geodemographic classification.
Linking national surveys with postcode types can lead to an erroneous generalisation of responses. Survey responses although well sampled represent only a small portion of the total population. Consequently, every conclusion drawn on lifestyle data must be considered in the light of response "representativeness" of the total population.
Attributing lifestyle data to a postcode type implies also a deterministic geographical relation that might not be correct. National surveys do not take in account local geographical variations in consumers behaviours. In a health geodemographic framework local patterns identify areas at risk for a particular condition.
Although the postcode is a very detailed level to work, there is still some bias from ecological fallacy. Some households living in a postcode might not belong to the type assigned by a geodemographic classification system. Therefore they do not have the lifestyle linked to that type assumed to be representative of the whole postcode.
With all these caveats in mind the maps contained in the session "Estimating risk behaviours" show "risk" surfaces for some lifestyle data coming from the Target Group Index (TGI) consumer survey appended to Mosaic UK types.
Target Group Index (TGI)[1] is a continuous general consumer survey, which has been carried out in Great Britain since 1969. The survey is carried out over approximately 25,000 respondents aged over 15 years and covers over 4,000 brands, 500 product/service fields and 200 publications.
Consumer surveys like TGI may be used in targeting of public health campaigns, because they can give insight into areas of diet, exercise, smoking and alcohol consumption, media use, which would not otherwise be accessible to a public health intelligence department.
In the Estimating hospital admissions section we present some exploratory spatial data analysis of estimated hospital admissions for some selected health conditions. The data used are Mosaic UK types appended to Hospital Episode Statistics (HES) using the method described in a paper by Richard Webber (Webber 2004). Hospital admission data are related to those conditions that originate a hospital episode. Using HES data to estimate disease risk might not take in account all those conditions that do not end in a hospitalisation or that are undiagnosed.
REFERENCES [TOP]
Davey Smith, G. 2003. Health Inequalities - lifecourse approaches. Bristol: The Policy Press.
Department of Health, 2004. Choosing Health: Making healthy choices easier: URL.
Sleight, P. 2004. Targeting customers. How to use geodemographics and lifestyle data in your business. Henley-on-Thames: WARC.
Webber, R. 2004. Neighbourhood inequalities in the pattern of hospital admissions and their application to the targeting of health promotion campaigns. CASA working papers 90: URL.
Notes [TOP]
[1] For more information visit also http://www.eurodirect.co.uk/Pages/MICROVISION_Using_Direct_Marketing.html
