In an infectious animal disease outbreak, effective management of the event requires timely and accurate information collection, processing, storage and distribution. This thesis focuses on the tools to assist information collection and management. The first study describes the comparison of questionnaire methodology for the information collection in the initial epidemiologic investigation of a Canadian federally reportable disease. The second study defines attributes of an animal disease outbreak information management system (IMS). The studies were performed within a one-year period (July 2013-July 2014). The first study performed two comparisons to determine differences in the information quality (completeness and accuracy) between differing questionnaire methodology and modes of completion (hard copy and electronic). The study was conducted with 24 Canadian Food Inspection Agency (CFIA) inspectors and veterinarians using a fictitious Canadian reportable disease scenario. The first comparison used a hard copy of a Canadian Food Inspection Agency (CFIA) questionnaire designed to be applicable (or generic) for all highly infectious reportable disease investigations with a supplementary disease specific section compared to an electronic disease specific reportable disease questionnaire. There was no significant difference in the information quality (N = 22; P = 0.09). The mean difference in completeness and accuracy scores was 3.5% (95% CI -0.6, 7.6). The second comparison focused on the hard copy disease questionnaire and assessed differences in information quality between using only the generic sections of the questionnaire compared to the supplementation of a disease specific section. A difference in information quality was determined (N = 24; P < 0.0001). The mean completeness and accuracy score for the generic only sections was 50.2% (95% CI 43.6, 57.2) compared to 80.2% (95% CI 76.2, 84.5) with the inclusion of the disease-specific section. The greatest difference in information quality occurred in the tracing specific information categories (P < 0.0001) with a mean difference of completeness and accuracy scores of 67.7% (95% CI 52.0, 83.4) for the trace-in (exposure history) category and 38.3% (95% CI 28.3, 48.3) for the trace-out (potential spread of disease) category. The absence of disease-specific questions were determined to be the primary factor in the difference in information quality. The second study determined a comprehensive list of user-defined attributes of an animal disease outbreak IMS and further identified the most important (key) attributes. A list of 34 attributes and associated definitions were determined through a series of focus group sessions and two surveys of Canadian animal health stakeholders. The animal health stakeholders included federal and provincial governments, veterinary academia and animal production industry representatives. The key attributes of an animal disease outbreak IMS identified were: ‘user friendly’, ‘effectiveness’, ‘accessibility’, ‘data accuracy’, ‘reliability’ and ‘timeliness’. ‘User friendly’ received the highest frequency of ranking as the most important attribute, followed by ‘effectiveness’. Information management was identified as the main purpose of an animal disease outbreak IMS with a median rating of 10 (rating scale of 0-10 with 10 = strongly agree). The occurrence of a federally reportable disease or a large-scale animal disease outbreak can have a great impact on the animal agriculture sector, regulatory government agencies and the economy. Information collection and management are essential to assist with the epidemiologic investigation and disease control measures. The study provided a novel opportunity to study information management for an animal disease outbreak from a Canadian perspective. The knowledge obtained will add value to the future development of tools and systems designed for information collection and management involving an animal disease outbreak.

This project was designed to develop a method for the collection of environmental samples during prolonged Norovirus (NoV) outbreak investigations, and to develop real-time RT-PCR assays to analyze environmental samples for GI and GII noroviruses. The collection and processing of environmental samples could provide epidemiological data to facilitate investigations of prolonged NoV outbreaks and could guide public health NoV intervention strategies. Real-time RT-PCR assays for the detection of GI and GII NoVs were developed by adapting the State Hygienic Laboratory clinical GI and GII assays to the AB 7500 Fast platform. Analysis of the GI assay performance yielded a dilution curve slope = 3.28, R2 = 0.999 and a calculated amplification efficiency of 102%. The GII assay yielded a dilution curve slope = 3.39, R2 = 0.999 and a calculated amplification efficiency of 97%. Amplification efficiencies determine the sensitivity and the limit of detection of real-time RT-PCR assays. Optimum efficiencies range from 95%-105%, with a 100% efficiency indicating exponential amplification of targeted nucleic acid. To develop a method for the collection of environmental samples, multiple swab types were tested to determine their ability to recover NoV from laboratory spiked environmental surfaces. It was determined that foam swabs moistened with viral transport media were most effective in recovering NoV from spiked surfaces. A field test of the environmental sampling method was conducted by sampling environmental surfaces in four restaurants in one Iowa community. NoVs were not detected in the environmental samples. The collection and processing of environmental samples when conducting an investigation of a prolonged NoV outbreak could provide additional information on the epidemiology of NoV transmission and infection.

La surveillance épidémiologique est le recueil systématique et continu d’informations sur l’état de santé des populations, leur analyse, leur interprétation et leur diffusion à tous les décideurs ayant besoin d’être informés. Un de ses objectifs est la détection des événements inhabituels, i.e. des épidémies, nécessitant la mise en place rapide de contre-mesures. Les objectifs de ce travail de thèse sont : (i) d’évaluer les principales méthodes statistiques de détection publiées et communément employées dans différents systèmes de surveillance épidémiologique, (ii) de proposer une nouvelle approche reposant sur la combinaison optimale de méthodes de détection statistique des épidémies et (iii) de développer une nouvelle méthode statistique d’analyse étiologique d’une épidémie à partir des données de surveillance épidémiologique collectées en routine par le système.Pour atteindre ces objectifs, nous évaluons les principales méthodes statistiques de la littérature, à partir d’un jeu publié de données simulées. Puis nous proposons une approche originale pour la détection des épidémies sur le principe de la combinaison de méthodes sélectionnées lors de l’étape précédente. Les performances de cette approche sont comparées aux précédentes selon la méthodologie utiliséeà la première étape. Enfin, nous proposons une méthode d’analyse étiologique d’une épidémie à partir des données de surveillance en employant des modèles statistiques adaptés aux séries chronologiques. / Public health surveillance is the ongoing, systematic collection, analysis, interpretation, and dissemination of data for use in public health action to reduce morbidity and mortality of health-related events and to improve health. One of its objectives is the detection of unusualevents, i.e. outbreaks, requiring the rapid implementation of countermeasures.The objectives of this work are: (i) to evaluate the main published statistical methods for outbreak detection commonly implemented in different public health surveillance systems, (ii) to propose a new approach based on the optimal combination of statistical methods foroutbreak detection and benchmark it to other methods; and (iii) develop a new statistical method for the etiological analysis of an outbreak from public health surveillance data routinely collected by the system. To achieve these objectives, as a first step, we evaluate the main statistical methods, from a published set of simulated public health surveillance data. Statistical methods have been evaluated for an operational purpose: for all simulated time series, we used the tuning parameters recommended by their authors for each algorithm when available. We propose sensitivity and specificity metrics suitable for these tools. Then we propose an original approach for outbreak detection based on combination of methods selected in the previous step. The performance of this approach is compared to the previous ones according to the methodology implemented in the first step.Finally, we propose a method for the etiological analysis of an outbreak from surveillance data by using statistical models suitable for time series analysis

Background Indigenous peoples with traditional lifestyles worldwide are considered particularly vulnerable to climate change effects. Large climate change impacts on the spread of infectious vector-borne diseases are expected as a health outcome. The most rapid climate changes are occurring in the Arctic regions, and as a part of this region northernmost Sweden might experience early effects. In this thesis, climate change effects on the lives of Sami reindeer herders are described and 30 years of weather changes are quantified. Epidemiology of the climate sensitive human infection tularemia is assessed, baseline serologic prevalence of tularemia is investigated and the disease burden is quantified across inhabitants in the region. Methods Perceptions and experiences of climate change effects among the indigenous Sami reindeer herders of northern Sweden were investigated through qualitative analyses of fourteen interviews. The results were then combined with instrumental weather data from ten meteorological stations in a mixed-methods design to further illustrate climate change effects in this region. In two following studies, tularemia ecology and epidemiology were investigated. A total of 4,792 reported cases of tularemia between 1984 and 2012 were analysed and correlated to ecological regions and presence of inland water using geographical mapping. The status of tularemia in the Swedish Arctic region was further investigated through risk factor analyses of a 2012 regional outbreak and a cross-sectional serological survey to estimate the burden of disease including unreported cases. Results The reindeer herders described how the winters of northern Sweden have changed since the 1970s – warmer winters with shorter snow season and cold periods, and earlier spring. The adverse effects on the reindeer herders through the obstruction of their work, the stress induced and the threat to their lifestyle was demonstrated, forcing the reindeer herders towards the limit of resilience. Weather data supported the observations of winter changes; some stations displayed a more than two full months shorter snow cover season and winter temperatures increased significantly, most pronounced in the lowest temperatures. During the same time period a near tenfold increase in national incidence of tularemia was observed in Sweden (from 0.26 to 2.47/100,000 p&lt;0.001) with a clear overrepresentation of cases in the north versus the south (4.52 vs. 0.56/100,000 p&lt;0.001). The incidence was positively correlated with the presence of inland water (p&lt;0.001) and higher than expected in the alpine and boreal ecologic regions (p&lt;0.001). In the outbreak investigation a dose-response relationship to water was identified; distance from residence to water – less than 100 m, mOR 2.86 (95% CI 1.79–4.57) and 100 to 500 m, mOR 1.63 (95% CI 1.08–2.46). The prevalence of tularemia antibodies in the two northernmost counties was 2.9% corresponding to a 16 times higher number of cases than reported indicating that the reported numbers represent only a minute fraction of the true tularemia. Conclusions The extensive winter changes pose a threat to reindeer herding in this region. Tularemia is increasing in Sweden, it has a strong correlation to water and northern ecoregions, and unreported tularemia cases are quite common.

Le chikungunya, la dengue et la bilharziose urogénitale sont des maladies vectorielles émergentes qui ont récemment trouvé des conditions favorables à leur transmission en France métropolitaine.Santé publique France, l’Agence en charge de la surveillance de l’état de santé de la population française est en première ligne pour détecter et investiguer ces émergences afin d’orienter les mesures de leur prévention et de leur contrôle. Postulant que chaque épidémie constitue une situation d’« expérimentation naturelle », l’objectif de cette thèse était de montrer comment chaque investigation d’épidémie apporte l’opportunité d’acquérir des connaissances scientifiques sur la contribution des cas asymptomatiques à l’introduction, la dissémination et l’endémisation des maladies vectorielles de façon réactive.Notre méta-analyse d’études de séroprévalence per- et post-épidémiques suggère que, contrairement à nos attentes, la lignée de virus chikungunya qui a émergé en 2004 dans l’Océan Indien qui était associée moins d’infections asymptomatiques que les autres. Dans une étude de la séroprévalence de la dengue à Nîmes en 2015, nous avons montré que le potentiel de diffusion de la dengue en France restait actuellement limité. Les données du dépistage des personnes exposées au risque de bilharziose urogénitale en Corse montrant une fréquence élevée d'infections pré-symptomatiques, nous avons évoqué un risque d’endémisation de la maladie qui a justifié son inscription sur la liste des maladies à déclaration obligatoire.Ce travail de thèse démontre qu’une approche pragmatique basée sur une veille sanitaire sensible associée à des investigations épidémiologiques de terrain précoces peut contribuer à aussi bien à la lutte contre les émergences qu’à l’évolution des connaissances. / Conditions recently proved favourable to transmission of emerging vector-borne diseases, chikungunya, dengue and urogenital schistosomiasis in mainland France.Santé publique France, the Agency in charge of public health surveillance in France is at the forefront of detecting and investigating emerging infectious disease in order to guide prevention and control measures. Assuming that each outbreak constitutes a situation of "natural experimentation", the aim of this thesis was to show how outbreak investigations give the opportunity to acquire rapidly scientific knowledge on the contribution of asymptomatic cases to the introduction, dissemination and endemisation of vector-borne diseases.Through a meta-analysis of per and post-epidemic seroprevalence studies,we have shown that the chikungunya virus lineage that emerged in the Indian Ocean in 2004 is associated with a lower frequency of asymptomatic infections. In a dengue serosurvey in Nîmes in 2015, we showed that the diffusion potential of dengue in France is currently limited. Screening data of urogenital bilharziasis in persons exposed in Corsica showed a high frequency of pre-symptomatic infections suggestive of a risk of endemisation of the disease that justified its inclusion on the list of notifiable diseases.This thesis work shows that a pragmatic approach based on sensitive surveillance associated with early field outbreak investigations can significantly contribute to both emerging infections control and the advancement of knowledge.