Toward congruence between theory and practice in small area analysis and local public health data

Advances in computer hardware, software and database interfaces have provided opportunities for collation, manipulation, analysis and display of spatial data on an unprecedented scale. Demands for small area data in public health, fuelled in part by an increasing emphasis on benchmarking in relation to Year 2000 objectives but also in response to state and federal programmes to involve local communities in the assessment and planning process have simultaneously generated an unprecedented demand for these analyses and data presentations. This paper discusses four areas where geographic, cartographic and statistical theory and methodology need to be brought to bear on the development of applications involving small area health data. These areas are: (i) the theoretical conceptions of space; (ii) managing the inherent variability of rates and frequencies; (iii) attribution of events or cases to areas or to points; and (iv) the application of sound principles of cartographic design to the presentation of results.