| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 8.41 MB | Adobe PDF |
Authors
Abstract(s)
Nascimentos prematuros são caracterizados como nascimentos ocorridos antes das 37 semanas completas de gestação. Em Portugal, tal como na maior parte dos países do mundo, a taxa
estimada de nascimentos prematuros tem vindo a aumentar ao longo dos últimos 20 anos e as
causas deste aumento ainda não foram devidamente estudadas. Uma das metodologias usadas
para análise espacial de doenças é o seu mapeamento. Esta metodologia aplicada aos nascimentos prematuros permite-nos identificar as regiões de Portugal onde estes acontecimentos são
mais prevalentes, e detetar possíveis fatores de risco associados a estes eventos que poderão
explicar a prevalência dos mesmos através da inclusão de covariáveis num modelo de regressão.
Recorreremos neste estudo a modelos hierárquicos bayesianos. Serão usados dados publicamente acessíveis no site do Instituto Nacional de Estatística (INE), registados num período de
10 anos, entre 2011 e 2020, divididos por 23 regiões de Portugal continental considerando a
Nomenclatura das Unidades Territoriais para Fins Estatísticos a nível III - NUTS III de 2013.
Esta dissertação pretende analisar a ocorrência de nascimentos prematuros em Portugal, usando
modelos hierárquicos bayesianos num contexto espacial, assim como identificar os fatores de
risco/protectores que possam contribuir para explicar o número de bebés prematuros.
Preterm birth is characterized as a birth that occurred before the 37th week of gestation. In Portugal, and in most countries, the estimated preterm birth rate has been growing in the last 20 years and the reasons for this growth have not been well studied yet. One of the methodologies used for spatial analysis of diseases is disease mapping. This methodology applied to preterm birth allows us to identify the regions of Portugal where these events are more prevalent, and detect possible risk factors associated with preterm births that might explain the prevalence in those regions using covariates in a regression model. We will use the bayesian hierarchical models. The data, which is publicly available on the website of the Instituto Nacional de Estatística (INE), will be collected for a period of 10 years, between 2011 and 2020, divided by 23 continental Portugal regions considering the nomenclature of territorial units for statistical purposes of 2013, NUTS III. This dissertation aims to analyze the preterm birth occurrences in Portugal, using bayesian hierarchical models in a spatial context, and in this way identify the risk/protective factors that might contribute to explaining the number of preterm births.
Preterm birth is characterized as a birth that occurred before the 37th week of gestation. In Portugal, and in most countries, the estimated preterm birth rate has been growing in the last 20 years and the reasons for this growth have not been well studied yet. One of the methodologies used for spatial analysis of diseases is disease mapping. This methodology applied to preterm birth allows us to identify the regions of Portugal where these events are more prevalent, and detect possible risk factors associated with preterm births that might explain the prevalence in those regions using covariates in a regression model. We will use the bayesian hierarchical models. The data, which is publicly available on the website of the Instituto Nacional de Estatística (INE), will be collected for a period of 10 years, between 2011 and 2020, divided by 23 continental Portugal regions considering the nomenclature of territorial units for statistical purposes of 2013, NUTS III. This dissertation aims to analyze the preterm birth occurrences in Portugal, using bayesian hierarchical models in a spatial context, and in this way identify the risk/protective factors that might contribute to explaining the number of preterm births.
Description
Tese de Mestrado, Bioestatística, 2023, Universidade de Lisboa, Faculdade de Ciências
Keywords
Nascimento prematuro Monte Carlo via cadeias de Markov Modelo hierárquico bayesiano Modelo espacial Teses de Mestrado - 2023
