We analyse data from two foot-and-mouth disease experiments for which previous studies have indicated lower levels of virus in the blood of sheep infected in the later stages of the epidemic. By using a non-Markovian stochastic compartmental model in a Bayesian approach, coupled with Markov chain Monte Carlo techniques, we are able to relax earlier assumptions regarding possible pathways of infection, and to use the data to reconstruct the infectious network. Thus, the complex interactions among level of viraemia, individual infectiousness and temporal position in the epidemic process can be investigated.
展开▼
