Using GIS to Model Incidence, Prevalence and Spread of Non-legal Drug Use

Current methods for estimating the incidence and prevalence of non‐legal drug use tend to be retrospective and are not capable of forecasting spatial characteristics. This paper details the development of a GIS model for forecasting and displaying spatio‐temporal trends in non‐legal drug use. It builds upon a current model that estimates levels of drug use, using GIS to develop and visualise the spatial dimension and to predict levels of use for unknown locations using radial basis functions with a thin spline interpolator. The model is calibrated against known data for the onset and spread of heroin use. Results of validation and cross validation of the interpolated surfaces give some confidence to the accuracy of the method. It provides a first stage in a process to develop more complex models that might be used to examine the introduction of new drug use practices, socio‐economic characteristics of different drug use habits, harm reduction measures or to further examine the role of space in the spread of drug use.