Inferring transmission routes for foot-and-mouth disease virus within a cattle herd using approximate Bayesian computation

Abstract

To control an outbreak of an infectious disease it is essential to understand the different routes of transmission and how they contribute to the overall spread of the pathogen. With this information, policy makers can choose the most efficient methods of detection and control during an outbreak. Here we demonstrate a method for assessing the contribution of different routes of transmission using approximate Bayesian computation with sequential Monte Carlo sampling (ABC-SMC). We apply this to infer parameters of an individual based model of within-herd transmission of foot-and-mouth disease virus (FMDV), incorporating transmission through direct contact and via environmental contamination. Additionally, we use ABC-SMC for model selection to assess the plausibility of either transmission route alone being responsible for all infections. We show that direct transmission likely contributes the majority of infections during an outbreak of FMD but there is an accumulation of environmental contamination that can cause infections within a farm and also have the potential to spread between farms via fomites.