Skip to Main content Skip to Navigation
Theses

Attribution des cas de salmonelloses humaines aux différentes filières de production animale en France. Adaptabilité et robustesse du modèle bayésien d'attribution par typage microbiologique

Abstract : Salmonella are the first bacterial cause of food-borne infections in France. These zoonotic pathogens occur in a variety of animal sources and are the aim of interventions from the farm level on. Identification of the relative contribution of the different sources to human cases is essential to implement an efficient policy to fight these infections. The bayesian microbial subtyping attribution approach allows determining the number of cases related to a particular source, and is thus particularly appropriate to this aim. The principle of this model consists in comparing the distribution of microbiological types (serotypes or subtypes) in human cases and in animal sources, weighted by the amount of sources consumed and taking into account the specificity of both the sources and the serotypes. The objectives of the thesis work were to identify and collect the data that are necessary to run such a model and to assess its robustness and the impact of the quality of the data on the results. The French surveillance system for Salmonella was thus described and analysed and a database sufficient to allow using the model was built. The application of the model in different European Member States showed the limits of the approach. As expected, the model proves to be very sensitive to the prior informative information that its overparameterized structure requires. We proposed a more appropriate parameterization based on the data. The use of the model on surveillance data of both passive (proportions) and active (prevalence) nature then demonstrated the robustness of this model. Thus, the source attribution approach could be implemented in France. The methodological limits are to be further explored but the perspectives of use, particularly to attribute the cases according to antibiotic resistance, justify mustering the necessary means. Keywords: source attribution, Salmonella, public health, Bayesian modeling, fo
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-00485441
Contributor : Emmanuelle Chiffoleau <>
Submitted on : Thursday, May 20, 2010 - 5:23:32 PM
Last modification on : Friday, October 23, 2020 - 4:34:49 PM
Long-term archiving on: : Thursday, September 16, 2010 - 2:38:18 PM

Identifiers

  • HAL Id : tel-00485441, version 1

Collections

Citation

Julie David. Attribution des cas de salmonelloses humaines aux différentes filières de production animale en France. Adaptabilité et robustesse du modèle bayésien d'attribution par typage microbiologique. Biologie cellulaire. Agrocampus - Ecole nationale supérieure d'agronomie de rennes, 2009. Français. ⟨tel-00485441⟩

Share

Metrics

Record views

713

Files downloads

7658