Statistical Modelling for Real-Time Spatial Surveillance and Forecasting
Peter J. Diggle, Barry S. Rowlingson
CHICAS, School of Health and Medicine, Lancaster University, Lancaster, Lancashire, United Kingdom

A generic problem in public health surveillance is to monitor, and if possible forecast, the spatial distribution of disease risk, and in particular to identify as quickly as possible any surprising departures from the expected distribution. Statistical modelling has a key role to play in distinguishing between random fluctuations in incidence and genuine changes in risk. To be useful to public health practitioners, it is also essential that the results of the statistical modelling are dissseminated in real-time, in user-friendly formats. We shall describe two examples of surveillance problems, one in a developed and one in a developing country setting, with a focus on the interface between statistical model development and software implementation.

Keywords: Public health surveillance; Spatial statistics; Environmental epidemiology; Tropical disease epidemiology

Biography: Barry Rowlingson is a Senior Research Associate in the School of Health and Medicine at Lancaster University. Working with Professor Peter Diggle, he has been developing solutions for spatial statistics applied to epidemiology for over 20 years. His current interest is in delivering complex analytical techniques in a user-friendly way with open-source tools.