Know more

The objective is twofold :

• master modeling and model analysis tools to study / parameterize complex models
• understand the influence of model assumptions on model predictions in animal health epidemiology and economics.

The results of this theoretical work then allow us to simplify the finalized models developed elsewhere in the team without deteriorating the quality of model predictions according to the application objectives pursued.

Research Program

The work carried out over the period 2006-2010 involved:

• The use of the Bayesian approach to estimate the parameters of an epidemiological model (application to the spread of Coxiella burnetii in cattle herds)
• The impact of Markovian assumptions in epidemiological models
• The influence of the transmission function in epidemiological models on the simulated spread of a pathogen
• The use of the singular perturbation method to reduce the size of a model (application to the propagation of bovine viral diarrhoea virus (BVDV) in cattle herds)
• The use of global model sensitivity analysis methods (application to the spread of BVDV and contagious bovine peri-pneumonia in cattle herds and Salmonella in pig herds)
• The impact of host infection duration on the simulated spread of an epidemic in a host metapopulation 
• The relationship between host migration intensity and the simulated duration of infection of an animal metapopulation 
• The impact of individual farmer's decision to voluntarily enter a certification or vaccination scheme on the regional prevalence of infection (application to BVDV) 

Work in progress and prospects concern:

• The effect of inter-individual host heterogeneity (susceptibility, excretion, immune response) on the simulated spread of pathogens
• The integration of risks related to animal movements and proximity between herds into epidemiological models using the most appropriate spatial methods.
• Theoretical modelling of farmers' behaviour in relation to animal health disorders
• The coupling of epidemiological and economic models to evaluate decentralized decisions at the collective level.

Partnerships

Scientific partners
UR MIA-Jouy, UR Smart, ANSES Ploufragan, University of Rennes 1, University of Caen, UMR INRA-CIRAD ERRC, University of Warwick (UK), Scottish Agricultural College, University of Utrecht (NL)

Contact

Pauline Ezanno
Catherine Belloc
Stéphane Krebs