Thesis (M.Sc. (Public Health))--Mahidol University, 1979.

Tese (livre-docencia) - Universidade Estadual de Campinas, Instituto de Geociencias / Made available in DSpace on 2018-08-04T14:43:27Z (GMT). No. of bitstreams: 1 Ferreira_MarcosCesar_LD.pdf: 11039567 bytes, checksum: 211ed0a607b655ff110573325fda72eb (MD5) Previous issue date: 2003 / Resumo: Este estudo apresenta um procedimento metodológico baseado em sistema de informação geográfica, para modelagem cartográfica e análise regional de dados epidemiológicos relacionados a doenças tropicais, utilizando como exemplo uma epidemia de dengue. A proposta apóia-se nos paradigmas da escola espacial da Geografia sintetizados no conceito de mapemática, que reúne em uma mesma abordagem espaço-tempo, a cartografia temática e a análise espacial aplicada em SIG. Tomou-se como universo de ensaio, a epidemia de dengue ocorrida em 2001 no noroeste do Estado de São Paulo, que infectou a população de municípios da mesoregião de São José do Rio Preto, cujos contextos espacial e epidemiológico, serviram de objeto para a experimentação do procedimento metodológico aqui proposto. Em síntese, o procedimento adotado baseia-se na fusão de dois paradigmas de investigação espacial: a modelagem do espaço da epidemia em objetos exatos e campos contínuos e a modelagem do [empo da epidemia nas dimensões escalares monotemporal e multitemporal. Estas categorias espaço-tempo são combinadas entre si, gerando-se quatro níveis de analise espacial e construção de mapas epidemiológicos. No nível monotemporal-objeto, os mapas elucidam a espacialidade da epidemia, evidenciando clusters, contágios espaciais entre municípios e anomalias locacionais de incidência. No nível multítemporal-objeto, utilizando-se sequencia mento cartotemporal, os mapas mostram a dinâmica espacial dos casos por municípios, segundo as categorias casos novos, casos mantidos e casos extintos durante a evolução da epidemia. Já na categoria monotemporai-campo, a epidemia é abordada em modelos digitais isopléticos, sem a segmentação do espaço em limites municipais, evidenciando a forma e a orientação preferencial de manchas na forma de nuvens de probabilidade de incidência da doença. Ainda sob esta categoria espaço-tempo, são construídos mapas de superfícies de mostrando a regionalização da epidemia, desprezando-se as variações locais e elucidando-se tendências predominantes em escalas menores. Na categoria multemporal-campos, é estudada a difusão espacial da epidemia em seqüências isopléticas espaço-tempo, e sintetizadas em mapas de vetores de mobilidade espacial do centro geográfico da epidemia. A fase final e sintética do procedimento apresentado trata-se da análise da difusão espacial da epidemia segundo o modelo de redes geográficas. Nesta etapa da investigação, são construídos mapas de nodalidade e de potencial de contágio entre núcleos urbanos por via rodoviária, adotando-se como referência modelos clássicos de acessibilidade e hierarquia urbana. O do procedimento inclui ainda, a análise estatística baseada na cartografia de probabilidades, seguindo-se os modelos de Poisson e Lambert-Gauss, e a análise comparativa entre mapas de indicadores da epidemia e mapas de indicadores socioeconômicos, buscando-se esclarecer, possíveis associações e correlações entre incidência de casos e variáveis demográficas e urbanas de municípios afetados pela enfermidade / Abstract: This study presents a methodology for cartographic modeling and regional analysis of dengue fever epidemics, based on spatial analysis techniques and geographical information system. Data from 109 counties organized in epidemiological weeks about a dengue fever epidemic occurred in 2001 in northwest of Sao Paulo state, were used to map incidence and spatial diffusion of cases. The methodology is based on a five levels approach: four levels, adding exact objects/continuous fields models and single/multiple times slices sequences, and a fifth level, based in network analysis of counties connection and disease probabilities mapping. At single time scale/exacts objects level, county clusters, spatial contagious of counties and local incidence rates were mapped. At multiple time scale/exacts objects level, spatial dynamics of the cases it was mapped in spatio-time sequencing model. Using the single time/continuous field level isoplethic and tendency surface maps it was produced. At the multiple times/continuous field level, spatial diffusion maps and spatial-time mobility of mean geographical center of dengue epidemics it were designed using a sequential maps model. At the last level of methodology, urban nodes connection are spatially analyzed using network road analysis techniques, to map potential of contagious between counties, spatial dispersion of epidemics between counties and the spread path of dengue over region as a whole / Tese (livre-docencia) - Univer / Livre-Docente em Geografia

Dengue mitigation is a major research area among scientist who are working towards an effective management of the dengue epidemic. An effective dengue mitigation requires several other important components. These components include an accurate epidemic modeling, an efficient epidemic prediction, and an efficient resource allocation for controlling of the spread of the dengue disease. Past studies assumed homogeneous response pattern of the dengue epidemic to climate conditions throughout the regions. The dengue epidemic is climate dependent and also it is geographically dependent. A global model is not sufficient to capture the local variations of the epidemic. We propose a novel method of epidemic modeling considering local variation and that uses micro ensemble of regressors for each region. There are three regressors that are used in the construction of the ensemble. These are support vector regression, ordinary least square regression, and a k-nearest neighbor regression. The best performing regressors get selected into the ensemble. The proposed ensemble determines the risk of dengue epidemic in each region in advance. The risk is then used in risk-based resource allocation. The proposing resource allocation is built based on the genetic algorithm. The algorithm exploits the genetic algorithm with major modifications to its main components, mutation and crossover. The proposed resource allocation converges faster than the standard genetic algorithm and also produces a better allocation compared to the standard algorithm.

Internet technology has affected humans' lives in many disciplines. The search engine is one of the most important Internet tools in that it allows people to search for what they want. Search queries entered in a web search engine can be used to predict dengue incidence. This vector borne disease causes severe illness and kills a large number of people every year. This dissertation utilizes the capabilities of search queries related to dengue and climate to forecast the number of dengue cases. Several machine learning techniques are applied for data analysis, including Multiple Linear Regression, Artificial Neural Networks, and the Seasonal Autoregressive Integrated Moving Average. Predictive models produced from these machine learning methods are measured for their performance to find which technique generates the best model for dengue prediction. The results of experiments presented in this dissertation indicate that search query data related to dengue and climate can be used to forecast the number of dengue cases. The performance measurement of predictive models shows that Artificial Neural Networks outperform the others. These results will help public health officials in planning to deal with the outbreaks.

The dengue virus has become widespread worldwide in recent decades. It has no specific treatment and affects more than 40% of the entire population in the world. In Thailand, dengue has been a health concern for more than half a century. The highest number of cases in one year was 174,285 in 1987, leading to 1,007 deaths. In the present day, dengue is distributed throughout the entire country. Therefore, dengue has become a major challenge for public health in terms of both prevention and control of outbreaks. Different methodologies and ways of dealing with dengue outbreaks have been put forward by researchers. Computational models and simulations play an important role, as they have the ability to help researchers and officers in public health gain a greater understanding of the virus's epidemic activities. In this context, this dissertation presents a new framework, Modified Agent-Based Modeling (mABM), a hybrid platform between a mathematical model and a computational model, to simulate a dengue outbreak in human and mosquito populations. This framework improves on the realism of former models by utilizing the reported data from several Thai government organizations, such as the Thai Ministry of Public Health (MoPH), the National Statistical Office, and others. Additionally, its implementation takes into account the geography of Thailand, as well as synthetic mosquito and synthetic human populations. mABM can be used to represent human behavior in a large population across variant distances by specifying demographic factors and assigning mobility patterns for weekdays, weekends, and holidays for the synthetic human population. The mosquito dynamic population model (MDP), which is a component of the mABM framework, is used for representing the synthetic mosquito population dynamic and their ecology by integrating the regional model to capture the effect of dengue outbreak. The two synthetic populations can be linked to each other for the purpose of presenting their interactions, and the Local Stochastic Contact Model for Dengue (LSCM-DEN) is utilized. For validation, the number of cases from the experiment is compared to reported cases from the Thailand Vector Borne Disease Bureau for the selected years. This framework facilitates model configuration for sensitivity analysis by changing parameters, such as travel routes and seasonal temperatures. The effects of these parameters were studied and analyzed for an improved understanding of dengue outbreak dynamics.

A dengue e a leptospirose estão entre as principais zoonoses de ocorrência no mundo. A primeira pelo elevado potencial epidêmico e a segunda pela alta letalidade. Na cidade de São Paulo, anualmente ocorrem dezenas de casos de leptospirose e centenas de casos de dengue, fazendo desses agravos alguns dos eventos de maior interesse da vigilância em saúde do município. Para melhor compreender o efeito de fenômenos climáticos e o impacto de medidas de prevenção e controle sobre a transmissão desses agravos ao longo do tempo, dois diferentes modelos estatísticos de estudo de séries temporais foram usados: 1) Função de Transferência, com erros dados por modelos ARIMA (ARIMAX), os quais foram modelados segundo a filosofia de Box-Jenkins e 2) Modelos Aditivos Generalizados (GAM) de regressão de Poisson, com estrutura de defasagem dada por funções polinômios PDL (Polynomial Distributed Lags). Os principais fatores climáticos associados ao aumento do número de casos de dengue na cidade de São Paulo foram a elevação da temperatura mínima do ar, dos níveis de precipitação pluviométrica, da densidade do vetor e a entrada de casos importados da doença, estimulada pelo feriado de carnaval. A chegada de frentes frias (temperatura < 16°C) e valores extremos de precipitação ( 70mm) reduzem o número de casos de dengue. Medidas de prevenção adotadas pelas equipes de vigilância em saúde do município também contribuem com a redução do número de casos. Os fatores associados ao aumento do número de casos de leptospirose foram o aumento da precipitação pluviométrica e da temperatura máxima do ar. O aumento das horas de brilho do sol reduz o número de casos. Os métodos adotados foram adequados aos objetivos do estudo e conseguiram captar as relações defasadas entre os fatores de interesse e a transmissão de dengue e de leptospirose na cidade de São Paulo. Tais técnicas também parecem adequadas como ferramentas a serem incorporadas à rotina da vigilância em saúde, permitindo fazer previsões do número de casos futuros e compreender as relações temporais entre as doenças e seus fatores determinantes e condicionantes / Dengue and leptospirosis are among the major zoonosis of occurrence in the world; the first because of the epidemic potential and the second due to high lethality. In São Paulo, dozens of leptospirosis cases and hundreds of dengue fever cases are registered annually, being some of the most important events to the municipal public health surveillance system. To understand the effect of climatic conditions and the impact of measures of prevention and control over the transmission of such diseases in the time context, two time series approaches were used: 1) Transfer Functions, with ARIMA error structure (ARIMAX), modeled by Box-Jenkins methods and 2) Additive Generalized Models (GAM) of Poisson regressions, with time structure given by Polynomial Distributed Lags (PDL). The most important climatic factors that increased the number of cases of dengue fever in the city of São Paulo were the elevation in air temperature, precipitation, vector density and the number of imported cases, which increased after carnival holiday (an important calendar event). The arrival of cold fronts from the south (air temperature < 16°C) and extreme precipitations (70mm) are factors that decrease the number of new dengue cases. The public health preventive interventions adopted by the municipality were effective in diminishing the dengue occurrence. The most important factors that increased the number of leptospirosis cases in São Paulo were elevation in maximum air temperature and precipitation. Largest amount of hours of sunshine decreased the number of new cases of leptospirosis. The methods used were adequate to the study objectives, the relations among the interest lagged factors and dengue and leptospirosis transmission in the city of São Paulo were satisfactorily modeled. Such techniques also seem appropriate as tools to be incorporated into the municipal health surveillance system, allowing the prediction of the number of future disease cases and understanding temporal relations between diseases and their determinants and conditioning factors

A dengue e a leptospirose estão entre as principais zoonoses de ocorrência no mundo. A primeira pelo elevado potencial epidêmico e a segunda pela alta letalidade. Na cidade de São Paulo, anualmente ocorrem dezenas de casos de leptospirose e centenas de casos de dengue, fazendo desses agravos alguns dos eventos de maior interesse da vigilância em saúde do município. Para melhor compreender o efeito de fenômenos climáticos e o impacto de medidas de prevenção e controle sobre a transmissão desses agravos ao longo do tempo, dois diferentes modelos estatísticos de estudo de séries temporais foram usados: 1) Função de Transferência, com erros dados por modelos ARIMA (ARIMAX), os quais foram modelados segundo a filosofia de Box-Jenkins e 2) Modelos Aditivos Generalizados (GAM) de regressão de Poisson, com estrutura de defasagem dada por funções polinômios PDL (Polynomial Distributed Lags). Os principais fatores climáticos associados ao aumento do número de casos de dengue na cidade de São Paulo foram a elevação da temperatura mínima do ar, dos níveis de precipitação pluviométrica, da densidade do vetor e a entrada de casos importados da doença, estimulada pelo feriado de carnaval. A chegada de frentes frias (temperatura < 16°C) e valores extremos de precipitação ( 70mm) reduzem o número de casos de dengue. Medidas de prevenção adotadas pelas equipes de vigilância em saúde do município também contribuem com a redução do número de casos. Os fatores associados ao aumento do número de casos de leptospirose foram o aumento da precipitação pluviométrica e da temperatura máxima do ar. O aumento das horas de brilho do sol reduz o número de casos. Os métodos adotados foram adequados aos objetivos do estudo e conseguiram captar as relações defasadas entre os fatores de interesse e a transmissão de dengue e de leptospirose na cidade de São Paulo. Tais técnicas também parecem adequadas como ferramentas a serem incorporadas à rotina da vigilância em saúde, permitindo fazer previsões do número de casos futuros e compreender as relações temporais entre as doenças e seus fatores determinantes e condicionantes / Dengue and leptospirosis are among the major zoonosis of occurrence in the world; the first because of the epidemic potential and the second due to high lethality. In São Paulo, dozens of leptospirosis cases and hundreds of dengue fever cases are registered annually, being some of the most important events to the municipal public health surveillance system. To understand the effect of climatic conditions and the impact of measures of prevention and control over the transmission of such diseases in the time context, two time series approaches were used: 1) Transfer Functions, with ARIMA error structure (ARIMAX), modeled by Box-Jenkins methods and 2) Additive Generalized Models (GAM) of Poisson regressions, with time structure given by Polynomial Distributed Lags (PDL). The most important climatic factors that increased the number of cases of dengue fever in the city of São Paulo were the elevation in air temperature, precipitation, vector density and the number of imported cases, which increased after carnival holiday (an important calendar event). The arrival of cold fronts from the south (air temperature < 16°C) and extreme precipitations (70mm) are factors that decrease the number of new dengue cases. The public health preventive interventions adopted by the municipality were effective in diminishing the dengue occurrence. The most important factors that increased the number of leptospirosis cases in São Paulo were elevation in maximum air temperature and precipitation. Largest amount of hours of sunshine decreased the number of new cases of leptospirosis. The methods used were adequate to the study objectives, the relations among the interest lagged factors and dengue and leptospirosis transmission in the city of São Paulo were satisfactorily modeled. Such techniques also seem appropriate as tools to be incorporated into the municipal health surveillance system, allowing the prediction of the number of future disease cases and understanding temporal relations between diseases and their determinants and conditioning factors