Aurélie Courcoul-Lochet

Abstract :

Q fever is a worldwide zoonosis caused by Coxiella burnetii which induces reproductive disorders in livestock. Ruminants are also recognized as the most important source of human infection. Therefore, the control of this infection in cattle is crucial to limit both the infection in livestock and the zoonotic risk. The objective of this thesis was to better understand the natural course of the infection within dairy cattle herds in order to propose effective control measures. A stochastic individual-based model in discrete time was conceptualised to represent the C. burnetii spread within a dairy herd. Its main epidemiological parameters were assessed from field data using a Bayesian approach. As a great heterogeneity between shedder cows, known to impact infection dynamics, has been described, the shedding routes and levels were explicitly represented in a variant of the first model. The most influential parameters of the infection dynamics, identified through a sensitivity analysis, were the levels of shedding, the characteristics of the bacterium in the environment and some physiological features of cows. Lastly, the long-term effectiveness of three different vaccination strategies in reducing the shedders prevalence, the number of abortions, the environmental bacterial load, and in leading to infection extinction was tested by simulation. A 10-year vaccination programme for both cows and heifers was found to be the most effective one. Besides providing a better understanding of C. burnetii infection dynamics, this work can help prioritizing needs of research and designing effective control programmes for Q fever in cattle.