Thesis (M.S.)--University of Missouri-Columbia, 2004. / Typescript. Includes bibliographical references (leaves 82-83). Also available on the Internet.

Introduction: An injury surveillance system was piloted in 2011 to monitor injuries in Canadian Forces. This evaluation of the key system attributes examined system performance. Methods: A retrospective chart review, a coding reliability study, a completeness of forms study and a key informant interview. Results: Sensitivity was 0.36 (95% CI: 0.28, 0.46). The system was missing patients over age 35. Kappa coefficients over 0.80 demonstrated good agreement. Completeness of forms study demonstrated high percentages of response for most questions and lower rates for questions related to using protective equipment, and consent for information sharing. Interviews proved acceptability to stakeholders, usefulness for identifying clusters and trends, simple and complete data collection, and flexibility. Conclusion: The injury surveillance system had good potential for several reasons: data collection did not require additional work in clinics; the system was well accepted and partially proved usefulness and timeliness in identifying unusual injury events.

Includes bibliographical references. / In context of increasing global migration and its correlation to heightened tensions around the meaning of a "foreign" body, this research questions the experiences of bodies crossing borders into the social and historical space of Cape Town, South Africa. Grounded in theories of surveillance, embodiment, and feminist geography of fear of crime, the study employed a feminist methodology using qualitative group interviews with international students from the Southern African Development Community at the University of Cape Town. The transcribed data was analysed through the participants' use of discourses and their descriptions of experiences. Questions arose around the meaning of surveillance and notions of respectability in transition. Furthermore, participants navigate amid new spaces of fear and insecurity in relation to their subjectivities, particularly as "foreigners". The research suggests that fear becomes a fundamental attribute of bodies in migration through which individuals mitigate through "passing" subverting expressions of embodied nationalities, knowledge gathering of the local terrain, among others. The challenges and techniques to overcome these fears become part of a process to re-establish the "self" in a foreign context.

The purpose of this thesis is to examine the nature of the relationships that exist between women and their use of cannabis. Currently, women’s use of cannabis is legally permitted within a federal, medical system of access; however, cannabis’ consumption outside of that framework is categorized as a criminal act. As a result, women are subjected to differing forms of cannabis stigmatization and surveillance across many socio-political contexts, often resultant in women’s losses of power and position. As Canada prepares to enact a legal cannabis framework, it is crucial that society and institutions understand the relationships which women have developed through cannabis use. Otherwise, the legalization of cannabis use will not -- in and of itself -- alleviate the systemic forms of stigmatization and oppression which continue to impact the lives of certain women because of their use of cannabis. A review of existing literature demonstrated that there is limited research which discusses the nature of women’s use of cannabis outside of a medical context. Using postmodern feminist and intersectional analysis, I conducted qualitative, semi-structured interviews with six women to gain insights into their experiences of using cannabis. Contrary to the medical and/or criminal cannabis discourse, the women’s stories reveal examples of unique and overlapping instances of cannabis use which differed from the traditional cannabis dichotomy. The findings of the women’s interviews create an alternative cannabis discourse, in which women’s use of cannabis is experienced as a fluid, multi-functional act with effects that satisfy experiences differently across diverse iv contexts, which extends our existing knowledge base. In relation to existing social work policies and practices, the finding implications are discussed. Ultimately, the thesis identifies opportunities for collaboration between social work and women, many of which could serve to disrupt the perpetuation of women’s stigmatization and surveillance in a legal cannabis framework. / Thesis / Master of Social Work (MSW)

The detection and identification of bio-threat agents and study of host-pathogen interactions require a high-resolution detection platform capable of discerning closely related species. This dissertation addresses the completion of the development of an array based platform and provides a robust pipeline for the discovery of unique bio-signatures for pathogens and their host. Our collection (library) of host and pathogen signatures has been greatly expanded to improve robustness and identification accuracy of an 'unknown' sample. The library containing measured bio-signatures for each species/isolate is complemented with computational methodologies to resolve the identity of the unknown sample as well as a mixture of organisms or a pathogen in a host background. Current approaches for pathogen detection rely on prior genomic sequence information. This research targets use of a broad based platform for identification of pathogens from field or laboratory samples on a high density Universal Bio-signature Detection Array (UBDA). This array is genome independent and contains all possible (49 combinations) 9-mer probes which are mathematically computed and genome independent. It works by comparing signal intensity readout to a library of readouts established by interrogating a wide spectrum of organisms. Each genome has a unique pattern of signal intensities corresponding to each of these probes. These signal intensities were used to generate un-biased cluster analysis patterns that can easily distinguish organisms into accepted and known phylogenomic relationships. Classification methods such as hierarchical clustering, Pearson's correlation matrix, principal component analysis and curve fitting regression methods were tested for pathogen specific use cases. Hierarchical clustering and Pearson's correlation matrix methods can establish phylogenomic relationships between highly diverse genomes. However, in order to assign a given sample to one or more groups, such as a pure isolate of a single species or composite mixture of multiple species, principal component analysis (PCA) was used. The test cases included identification of mixed samples, case study of field samples from state diagnostic labs and finally a surveillance method for viral and parasite carrying insect host vectors. Completion of these application challenges is meant to demonstrate the power and confirm confidence in the Universal Bio-signature Detection Array. / Ph. D.

West Nile virus (WNv) is an arbovirus and is transmitted by infected mosquitoes after feeding on the blood of birds carrying the virus. The Canadian WNv national surveillance system has just completed its tenth year of operation. The thesis is to evaluate the surveillance system and analyze multi-year human data. The evaluation includes the use of multiple lines of complementary methods such as the US CDC surveillance guidelines, Canadian Evaluation Framework, document review and a survey. Logistic and Poisson regressions were used for data analyses. WNv has become endemic in most parts of Canada since the virus occurred in 2001. The virus activity is peak around August. High numbers of human cases with WNv neurological syndrome identified pose a significant health concern due to the long term sequelae among affected patients. WNv national surveillance met its main objectives and there is a continual need for the surveillance.

La surveillance de la listériose en France s’est construite par étapes depuis les années 1980 sur deux piliers, la microbiologie et l’épidémiologie. Grâce à la création du Centre National de Référence des Listeria et à la mise au point de techniques de typage performantes, l’Institut Pasteur assure une surveillance microbiologique depuis 1987. Une surveillance épidémiologique initiée entre 1984 et 1992 par le Laboratoire National de la Santé, a été développée par le Réseau National de Santé Publique de 1993 à 1999, puis amplifiée par l’Institut de Veille Sanitaire à partir de 2000. Le premier objectif de cette thèse est de décrire les différentes phases de la construction de cette surveillance afin d’analyser leurs contributions respectives au cours de ces 30 dernières années. Cette construction s’est faite en 4 phases : 1. L’étape fondatrice de 1982 à 1992 a été la reconnaissance et la prise en compte du rôle des aliments dans la transmission de la maladie et dans la survenue d’épidémies. 2. La deuxième phase de 1993 à 2000 a été l’édification d’un système de surveillance opérationnel pour détecter et investiguer les épidémies en France. 3. La troisième phase de 2000 à 2005 a permis de consolider le système de surveillance et de le perfectionner en ajoutant un volet complémentaire avec des prélèvements alimentaires.4. Depuis 2005, nous sommes dans la quatrième phase avec comme objectif l’optimisation du système. Cette optimisation repose sur l’adaptation des outils de surveillance et d’alerte aux connaissances. Ainsi, après avoir montré que la durée d’incubation de la maladie varie selon la forme clinique de la maladie, nous avons proposé d’intégrer cette variation pour déterminer la période d’évaluation des expositions alimentaires à risque. L’analyse des performances du système a permis à deux reprises de proposer de nouveaux seuils de signalement plus spécifiques afin d’optimiser cette surveillance tout en réduisant son coût. Le deuxième objectif de cette thèse est de montrer la contribution des données de surveillance à une politique de santé publique. Un premier travail a consisté à mettre en perspective les variations temporelles d’incidence observées avec les différentes sources de données disponibles afin d’en analyser les déterminants. La phase de décroissance de 1987 à 1997 a été concomitante des mesures de contrôles prises par l’industrie agro-alimentaire et de la réduction de la contamination des aliments. La phase d’augmentation en 2006-2007 semble multifactorielle. L’augmentation de la prescription de traitements de réduction de l’acidité gastrique par des inhibiteurs de la pompe à protons pourrait être l’un des déterminants majeurs de cette augmentation.Dans une deuxième analyse, nous avons hiérarchisé les groupes à risque de listériose sur la base de l’estimation du taux d’incidence de listériose et de sa mortalité dans ces groupes. Cela a permis d’identifier les groupes les plus vulnérables : hémopathies, certains cancers (digestifs, cérébral et pulmonaire), maladie de Horton, cirrhose hépatique, les dialysés rénaux, les greffés, et les femmes enceintes. Une analyse épidémiologique des listérioses materno-néonatales (MN) a montré une association entre les régions avec une incidence plus faible de listérioses materno-néonatales et les régions où la séroprévalence toxoplasmique des femmes enceintes est la plus faible, ce qui suggère un effet positif des recommandations contre la toxoplasmose pour la prévention de la listériose MN. / Listeriosis surveillance was built up stage by stage in France since the 1980s on a twofold basis: microbiology and epidemiology. Thanks to the creation of the Listeria National Reference Centre (Centre National de Référence des Listeria), the Pasteur Institute has been doing microbiological surveillance since 1987. Epidemiological surveillance was initiated by the National Health Laboratory, then conducted by the National Health Network and further developed by the National Institute of Health Surveillance. This thesis aims first of all to describe the different stages in the setting up of this surveillance system in order to analyze their respective inputs during these last thirty years. The four stages are:1. From 1982 to 1992: awareness and recognition of the role of food in the transmission of listeriosis and as the source of outbreaks. 2. From 1993 to 2000: building a reliable surveillance system in order to detect and investigate outbreaks in France. 3. From 2000 to 2005: strengthening and perfecting the surveillance system by taking additional measures, such as food sampling.4. Since 2005, we have reached the fourth stage, designed to optimize the surveillance system. This optimization involves adapting surveillance and early warning tools to new knowledge and information. For instance, having established that listeriosis incubation periods vary according to the clinical form of the illness, we suggested the integration of the variation of exposure period factor when interviewing patients with the food questionnaire. On two separate occasions, analysis of the surveillance system performance results made it possible to modify the criteria for early warning so as to optimize surveillance by increasing its specificity whilst reducing costs.The second aim of this thesis is to illustrate how surveillance data can contribute to public health policies. A first study analyzed temporal trends, using all available data in order to give some explanation as to major trends. The first trend was a reduction of incidence from 1987 to 1997 that was concomitant with control measures by the food industry and a drop in food contamination. The increased trend observed in 2006-2007 appears to be due to several factors. The increased rate of sales of proton pump inhibitors medication could be the major factor in this increase. In a second study, we ranked groups at risk of acquiring listeriosis based on the incidence of listeriosis and its lethality in each group. This enabled us to identify the most vulnerable groups : hematological malignancy, some cancers (digestive, lung, and brain cancer), dialysis, cirrhosis, organ transplantation and pregnancy. Epidemiological analysis of listeriosis cases associated with pregnancy indicated an association between regions with low rate listeriosis associated with pregnancy and regions where toxoplasmosis prevalence of pregnant women is low. This suggests that recommendations for avoiding toxoplasmosis have a positive effect on preventing listeriosis during pregnancy.

Traditional disease surveillance systems are instrumental in guiding policy-makers' decisions and understanding disease dynamics. The first study in this dissertation looks at sentinel surveillance network design. We consider three location-allocation models: two based on the maximal coverage model (MCM) and one based on the K-median model. The MCM selects sites that maximize the total number of people within a specified distance to the site. The K-median model minimizes the sum of the distances from each individual to the individual's nearest site. Using a ground truth dataset consisting of two million de-identified Medicaid billing records representing eight complete influenza seasons and an evaluation function based on the Huff spatial interaction model, we empirically compare networks against the existing volunteer-based Iowa Department of Public Health influenza-like illness network by simulating the spread of influenza across the state of Iowa. We compare networks on two metrics: outbreak intensity (i.e., disease burden) and outbreak timing (i.e., the start, peak, and end of the epidemic). We show that it is possible to design a network that achieves outbreak intensity performance identical to the status quo network using two fewer sites. We also show that if outbreak timing detection is of primary interest, it is actually possible to create a network that matches the existing network's performance using 42% fewer sites. Finally, in an effort to demonstrate the generic usefulness of these location-allocation models, we examine primary stroke center selection. We describe the ineffectiveness of the current self-initiated approach and argue for a more organized primary stroke center system. While these traditional disease surveillance systems are important, they have several downsides. First, due to a complex reporting hierarchy, there is generally a reporting lag; for example, most diseases in the United States experience a reporting lag of approximately 1-2 weeks. Second, many regions of the world lack trustworthy or reliable data. As a result, there has been a surge of research looking at using publicly available data on the internet for disease surveillance purposes. The second and third studies in this dissertation analyze Wikipedia's viability in this sphere. The first of these two studies looks at Wikipedia access logs. Hourly access logs dating back to December 2007 are available for anyone to download completely free of charge. These logs contain, among other things, the total number of accesses for every article in Wikipedia. Using a linear model and a simple article selection procedure, we show that it is possible to nowcast and, in some cases, forecast up to the 28 days tested in 8 of the 14 disease-location contexts considered. We also demonstrate that it may be possible in some cases to train a model in one context and use the same model to nowcast or forecast in another context with poor surveillance data. The second of the Wikipedia studies looked at disease-relevant data found in the article content. A number of disease outbreaks are meticulously tracked on Wikipedia. Case counts, death counts, and hospitalization counts are often provided in the article narrative. Using a dataset created from 14 Wikipedia articles, we trained a named-entity recognizer (NER) to recognize and tag these phrases. The NER achieved an F1 score of 0.753. In addition to these counts in the narrative, we tested the accuracy of tabular data using the 2014 West African Ebola virus disease epidemic. This article, like a number of other disease articles on Wikipedia, contains granular case counts and deaths counts per country affected by the disease. By computing the root-mean-square error between the Wikipedia time series and a ground truth time series, we show that the Wikipedia time series are both timely and accurate.

L'objectif de cette thèse est d'optimiser la couverture visuelle d'une zone vaste et complexe de façon à ce que ses images puissent composer une mosaïque à partir d'un drone.Pour trouver les meilleurs points de passage, deux méthodes ont été étudiées : l'optimisation par essaims particulaires (PSO) et les algorithmes génétiques (GA).Notre étude a prouvé que le GA est la méthode offrant de meilleures performances en raison de ses performances et de sa capacité d'adaptation.Après avoir réalisé des expériences pour comparer les algorithmes, une hybridation de GA et PSO a été réalisée et étudiée.La méthode proposée peut être appliquée sur de grandes surfaces de formes irrégulières, comme les terrains agricoles, et fournit un nombre réduit de points de passage qui doivent être survolés par un véhicule aérien de type drone (UAV).Des essais ont été réalisés pour simuler le vol d'un drone dans un environnement intérieur, les images générées pendant la simulation sont utilisées pour représenter une image de la totalité de l'environnement sous la forme d'une mosaïque.La méthode proposée est également appliquée dans de vastes zones extérieures. Des images satellitaires sont utilisées pour visualiser la couverture du trajet qui a été planifié.Les expériences valident l'efficacité de la méthode proposée pour trouver le nombre et la position des points de passage. / The goal of this paper is to optimize the coverage of a vast and complexarea such that its mosaic image can be created. To find the best waypoints, twomethods have been investigated: Particle Swarm Optimization (PSO) and GeneticAlgorithms (GA). Our investigation proved that GA is a better method due toits performance and adaptability. After having performed experiments to compare the algorithms, a hybridization of GA and PSO is investigated.The proposed method can be applied on large areas with irregular shapes, such as agricultural fields, and it provides a minimized number of waypoints that must be flown over by the Unmanned Aerial Vehicle (UAV). The experiments were made to simulate the flight of the UAV in an indoor environment, and the images generated during the simulated flight have been used to show the final mosaic. The proposed method is also applied in the vast outdoor area using satellite images to visualize the final result of the coverage path planning. The experiments validate the efficiency of the proposed method for finding the number and the poses of the waypoints. The solution proposed to approach the problem of coverage path planning is rather different than the stateof the art by dividing the Coverage Path Planning on independent sub-problems to optimize and then using GA and later on GAPSO.

Thesis (Ph. D.)--University of Texas at Austin, 2006. / Vita. Includes bibliographical references.