B1062
Title: Spatial disaggregation of disease count data
Authors: Craig Anderson - University of Glasgow (United Kingdom) [presenting]
Abstract: Disease mapping focuses on estimating the spatial pattern of disease risk across a geographical region that has been subdivided into a set of administrative districts. The disease data consists of aggregated disease counts at this district level and traditionally this means that inference is also restricted to this geographical level. This standard inference can be susceptible to the modifiable areal unit problem (MAUP), whereby the estimated risk surface is affected by the arbitrary choice of district boundaries. The district-level count is really an aggregation of point level disease cases, and therefore if a different spatial partition of the region was selected we could observe different results. the aim is to address this problem by outlining a method for producing disaggregated disease risk estimates based on a regular 1km x 1km grid. The method is illustrated using an application to respiratory hospital admissions in Glasgow, Scotland.
