Bayesian Hierarchical Models for the Evaluation and Design of Surveillance Systems: an Application to Scrapie Surveillance in Great Britain

There is a need for regular evaluation of surveillance strategies. The emergence of new diagnostic tests and new sources of data, changes in the spatio-temporal distribution of diseases and other factors must be periodically assessed to guarantee that the objectives of the surveillance effort are met. Underlying this evaluation process is the need to increase the efficient use of resources.

OBJECTIVE

We have developed a flexible model which can evaluate surveillance strategies at different hierarchical levels. It identifies key elements in the performance of the surveillance and recommends optimal sampling designs.