Network Analysis for Surveillance Design and Evaluation
Abstract
The way in which an infectious agent or an information is disseminated within an animal production system is influenced by the structure of the network of contacts through which the infectious agent is transmitted between hosts, or information is shared among stakeholders. The analysis of these networks is therefore necessary to comprehensively design and evaluate surveillance programmes. In this chapter, we provide an introduction of key methods to characterise the structural features of a network influencing the diffusion of a disease or an information through it, and we illustrate the relevance of these methods in the context of animal health surveillance design and evaluation. First, we define a network; we present the specific features of network data, how networks are constructed, and the challenges posed by network data collection. Secondly, we introduce metrics used to characterise the position of a node, and the cohesion of the overall network, or a subset of this network. Finally, mathematical models simulating the transmission of infectious agents through a network of potentially infectious contacts are described. Examples of applications of these methods for the design and evaluation of surveillance programmes are provided as case studies.