Skip to Main content Skip to Navigation

Modélisation de données de surveillance épidémiologique de la faune sauvage en vue de la détection de problèmes sanitaires inhabituels

Abstract : Recent studies have shown that amongst emerging infectious disease events in humans, about 40% were zoonoses linked to wildlife. Disease surveillance of wildlife should help to improve health protection of these animals and also of domestic animals and humans that are exposed to these pathogenic agents. Our aim was to develop tools capable of detecting unusual disease events in free ranging wildlife, by adopting a syndromic approach, as it is used for human health surveillance, with pathological profiles as early unspecific health indicators. We used the information registered by a national network monitoring causes of death in wildlife in France since 1986, called SAGIR. More than 50.000 cases of mortality in wildlife were recorded up to 2007, representing 244 species of terrestrial mammals and birds, and were attributed to 220 different causes of death. The network was first evaluated for its capacity to detect early unusual events. Syndromic classes were then defined by a statistical typology of the lesions observed on the carcasses. Syndrome time series were analyzed, using two complimentary methods of detection, one robust detection algorithm developed by Farrington and another generalized linear model with periodic terms. Historical trends of occurrence of these syndromes and greater-than-expected counts (signals) were identified. Reporting of unusual mortality events in the network bulletin was used to interpret these signals. The study analyses the relevance of the use of syndromic surveillance on this type of data and gives elements for future improvements.
Document type :
Complete list of metadatas

Cited literature [111 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Wednesday, June 29, 2011 - 9:28:24 AM
Last modification on : Friday, October 23, 2020 - 4:36:56 PM
Long-term archiving on: : Friday, September 30, 2011 - 2:21:02 AM


Version validated by the jury (STAR)


  • HAL Id : tel-00604435, version 1



Eva Petit. Modélisation de données de surveillance épidémiologique de la faune sauvage en vue de la détection de problèmes sanitaires inhabituels. Médecine humaine et pathologie. Université de Grenoble, 2011. Français. ⟨NNT : 2011GRENS006⟩. ⟨tel-00604435⟩



Record views


Files downloads