Controlling a non-linear epidemic PDE model with delayed detection: a model simplification approach

This paper investigates the problem of controlling an epidemic spread in an animal flock. The disease is characterized by a long and variable incubation period, during which individuals are infectious but do not show clinical signs. The disease detection is therefore delayed. The epidemic process and population dynamic in the flock are modelled by a system of non-linear integro-differential transport-reaction PDE, structured according to health status, age and time remaining before detection. We show that for realistic parameter values this system can be approximated with guaranteed error bounds by a simpler model, and use this simpler model to design flock management strategies.