A carregar...
Publicação
Modelação espácio-temporal da propagação da COVID-19 em Portugal Continental: evidências da importância de fatores geográficos
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 9.83 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
A COVID-19 é uma doença respiratória identificada na China no final de 2019 como
responsável por causar vários casos de pneumonia atípica. A sua dispersão, potenciada pelas redes
internacionais de transporte, levou a que esta se espalhasse rapidamente por todo o mundo, causando
uma pandemia com centenas de milhões de casos confirmados.
Abordagens geográficas a contextos epidémicos e pandémicos são comuns na literatura e a
utilização de Sistemas de Informação Geográfica (SIG) e técnicas de análise espacial mostram grande
aplicabilidade ao estudo da distribuição de doenças. Também estudos da COVID-19 baseados em SIG
são igualmente importantes para identificar e compreender atributos explicativos, ainda desconhecidos,
dos padrões de disseminação desta doença. Neste contexto, a presente dissertação de mestrado,
realizada no âmbito da colaboração em dois projetos de investigação acerca das dinâmicas espaciais
da COVID-19 em Portugal (https://www.comprime-compri-mov.com/), permitiu testar a aplicação de
novas métricas a este fenómeno.
Dos resultados obtidos foi possível confirmar as conclusões de outros autores, mas também
descobrir novas evidências acerca dos padrões espaciais de transmissão. Mais precisamente, a partir
de uma metodologia complementada por métodos estocásticos de análise bivariada, técnicas de
análise espacial e modelos de regressão linear múltipla e geográfica, inferiu-se que a propagação da
COVID-19 em Portugal Continental, associada a processos de difusão espacial de relocalização,
expansão hierárquica e por contágio, apresenta importantes relações com fatores decorrentes de
especificidades sócio territoriais dos municípios tais como: densidade populacional, dimensão média
das famílias, padrões de mobilidade, rendimento e exposição pela ocupação, que se mostraram
preditores significativos para a distribuição heterogénea dos casos de COVID-19. Ao mesmo tempo, a
proximidade e contiguidade entre municípios são propriedades geográficas com influência nos padrões
espaciais de incidência, assim como a hierarquia da rede urbana nacional. Além disso, constatou-se a
variabilidade espacial do efeito dos determinantes devido à existência de não estacionariedade
espacial. Não obstante a flutuação espácio-temporal da importância dos determinantes, a partir da
identificação dos seus limiares críticos o território foi classificado num índice de predisposição à infeção.
Esta classificação constitui-se como uma alternativa para a aplicação de medidas de Saúde Pública de
base territorial, traduzindo como diferentes dimensões societais influenciam a dispersão do
SARS-CoV-2 e refletem a existência de condições propícias à transmissão espacialmente
heterogéneas.
COVID-19 is a respiratory disease identified in China in late 2019 as responsible for causing several cases of atypical pneumonia. Its dispersion, enhanced by international transport networks, has led to a rapid worldwide spread, causing a pandemic with hundreds of millions of confirmed cases. Geographical approaches to epidemic and pandemic contexts are common in the literature and the use of Geographic Information Systems (GIS) and spatial analysis techniques show great applicability to the study of disease distribution. Also, COVID-19 GIS-based studies are important to identify and understand explanatory attributes, still unknown, of the dissemination patterns of this disease. This dissertation, developed as a follow-up on the conclusions of two research projects on the spatial dynamics of COVID-19 in Portugal (https://www.comprime-compri-mov.com/), allowed to test the application of new metrics to this phenomenon. From the results obtained it was possible to confirm the conclusions of other authors, but also to discover new evidence about the spatial patterns of transmission. More precisely, from a methodology complemented by stochastic methods of bivariate analysis, spatial analysis techniques and spatial and multiple linear regression models, it was inferred that the spread of COVID-19 in mainland Portugal, associated with relocation, hierarchical and contagious spatial diffusion processes, shows important relationships with factors arising from socio-territorial specificities of municipalities such as: population density, average household size, mobility patterns, income and exposure by occupation, which proved to be significant predictors for the heterogeneous distribution of COVID-19 cases. At the same time, proximity and contiguity between municipalities are geographical properties with influence on spatial patterns of incidence, as is the hierarchy of the national urban network. In addition, spatial variability of the effect of determinants was found due to the existence of spatial non-stationarity. Despite the spatio-temporal fluctuation of the importance of the determinants, the territory was classified in a predisposition index to infection based on the identification of common critical thresholds. This classification constitutes an alternative for the application of spatial-based Public Health measures, reflecting how different societal dimensions influence the dispersion of the new coronavirus and reflect the existence of spatially heterogeneous conditions conducive to SARS-CoV-2 transmission.
COVID-19 is a respiratory disease identified in China in late 2019 as responsible for causing several cases of atypical pneumonia. Its dispersion, enhanced by international transport networks, has led to a rapid worldwide spread, causing a pandemic with hundreds of millions of confirmed cases. Geographical approaches to epidemic and pandemic contexts are common in the literature and the use of Geographic Information Systems (GIS) and spatial analysis techniques show great applicability to the study of disease distribution. Also, COVID-19 GIS-based studies are important to identify and understand explanatory attributes, still unknown, of the dissemination patterns of this disease. This dissertation, developed as a follow-up on the conclusions of two research projects on the spatial dynamics of COVID-19 in Portugal (https://www.comprime-compri-mov.com/), allowed to test the application of new metrics to this phenomenon. From the results obtained it was possible to confirm the conclusions of other authors, but also to discover new evidence about the spatial patterns of transmission. More precisely, from a methodology complemented by stochastic methods of bivariate analysis, spatial analysis techniques and spatial and multiple linear regression models, it was inferred that the spread of COVID-19 in mainland Portugal, associated with relocation, hierarchical and contagious spatial diffusion processes, shows important relationships with factors arising from socio-territorial specificities of municipalities such as: population density, average household size, mobility patterns, income and exposure by occupation, which proved to be significant predictors for the heterogeneous distribution of COVID-19 cases. At the same time, proximity and contiguity between municipalities are geographical properties with influence on spatial patterns of incidence, as is the hierarchy of the national urban network. In addition, spatial variability of the effect of determinants was found due to the existence of spatial non-stationarity. Despite the spatio-temporal fluctuation of the importance of the determinants, the territory was classified in a predisposition index to infection based on the identification of common critical thresholds. This classification constitutes an alternative for the application of spatial-based Public Health measures, reflecting how different societal dimensions influence the dispersion of the new coronavirus and reflect the existence of spatially heterogeneous conditions conducive to SARS-CoV-2 transmission.
Descrição
Palavras-chave
COVID-19 Sistemas de Informação Geográfica Difusão espacial Regressão Geográfica Ponderada Saúde Pública