Abstract

Social area classification focuses on segmenting geographical units into groups based on the socio-economic characteristics of their residents. The benefits of geodemographic classification systems for informing health
intelligence are increasingly being exploited. In this paper, we present a framework for evaluating the usefulness of geodemographic segmentations. We embark on a demonstration to compare six classification systems at our
disposal. Alongside Lorenz curves, we have also employed a statistic called the Ginicoefficient. We use this to measure discriminatory power for each of the different systems for four health conditions including admissions for Asthma,
Coronary Heart Disease (CHD), Chronic Obstructive Pulmonary Disease (COPD) and Diabetes.