Bayesian disease mapping for public health

Lawson, A. and Lee, D. (2017) Bayesian disease mapping for public health. In: Srinivasa Rao, A. S.R., Pyne, S. and Rao, C.R. (eds.) Handbook of Statistics. Elsevier, pp. 443-481. ISBN 9780444369684 (doi:10.1016/

Full text not currently available from Enlighten.


Disease risk varies in space and time due to variation in many factors, including environmental exposures such as air pollution, and the prevalence of lifestyle behaviors such as smoking. Quantifying and explaining this spatio-temporal variation in disease risk is vital to improving public health, as for example it allows risk factors to be highlighted to the general public, as well as allowing informed decisions to be made about the future allocation of health resources. Research in this area is collectively known as disease mapping and has a variety of aims including estimating the spatio-temporal pattern in disease risk, explaining the variation in the risk by covariate factors, identifying high-risk subregions (clusters) for focused action, and surveillance of disease outbreaks. The data used in this field are typically counts of the numbers of disease cases in a set of nonoverlapping areal units, and in the simplest case these data are available for a single time period. However, recent work has extended this simple case by analyzing disease count data for multiple time periods (space–time modeling) or for multiple diseases for a single time period (multivariate modeling). This chapter provides an in-depth review of the disease mapping field, focusing on the four key areas listed above in spatial, spatio-temporal, and multivariate disease domains.

Item Type:Book Sections
Glasgow Author(s) Enlighten ID:Lee, Professor Duncan
Authors: Lawson, A., and Lee, D.
College/School:College of Science and Engineering > School of Mathematics and Statistics > Statistics
Published Online:09 August 2017

University Staff: Request a correction | Enlighten Editors: Update this record