Mathematical modelling of epidemiological systems: from data and equations to public health implications
Carregando...
Arquivos
Data
2022-01-21
Autores
Franco, Caroline
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
A modelagem matemática de doenças infecciosas é um campo extremamente interdisciplinar no qual se faz necessário combinar conhecimentos obtidos a partir de diferentes especialidades, culturas e disciplinas científicas. Com base em dados biológicos e epidemiológicos disponíveis, modelos compartimentais podem ser construídos sob medida visando responder a diferentes questões científicas referentes à dinâmica de propagação de um doença dentro de uma dada população. Em nossos modelos descrevendo a dinâmica de COVID-19 e malária aqui apresentados, utilizamos diferentes estruturas para solucionar diferentes problemas. Através da abordagem introduzida pelo modelo CoMo, produzimos um modelo robusto para simular o efeito de diferentes cenários de intervenções não-farmacêuticas na redução de casos e mortes, durante a pandemia de COVID-19. Esse modelo foi utilizado para apoiar a implementação de políticas públicas a níveis nacional e internacional. Levando em conta heterogeneidades na rede de contatos entre domicílios, também contribuímos ao preeencher a lacuna metodológica existence entre modelos a nível populacional e domiciliar, usando idéias extraídas da teoria de percolação em redes. Ainda no contexto de modelos para COVID-19, desenvolvemos outras estruturas visando responder diferentes questões científicas. Simplicações na estruturação etária e na implementação de intervenções foram utilizadas para que se pudesse melhor avaliar e otimizar o efeito de intervenções únicas (viz. reabertura de escolas e vacinação) ou estimar parâmetros de relevância epidemiológica (viz. transmissibilidade de novas variantes). No contexto de malária endêmica, descrevemos modelos independentes para diferentes espécies, que são atestadamente úteis para informar tomadas de decisão no contexto de saúde pública. Visto que múltiplas espécies coexistem no meio ambiente, também levamos em conta a avaliação de sua carga combinada e o efeito de possíveis interações interespecíficas. Neste contexto, propusemos uma nova abordagem metodológica que viabiliza o desenvolvimento de modelos incluindo multiplas espécies e provamos analiticamente sua validade em regimes de interações fracas. Tal abordagem proporcionou o devido suporte teórico para estruturas de modelos já sendo adotadas para informar tomadas de decisão em saúde pública. Nossas contribuições englobam tanto o desenvolvimento de metodologias teóricas que fundamentam a construção de modelos mais robustos, quanto o desenvolvimento de tais modelos visando apoiar tomadas de decisão em saúde pública.
The mathematical modelling of infectious diseases is an extremely interdisciplinary field, where we need to amalgamate different areas of expertise, cultures and scientific backgrounds. Based on the available biological and epidemiological data, compartmental models can be tailored to investigate different research questions regarding a disease spread dynamics within a population. In our COVID-19 and malaria modelling exercises, we used different model structures to address different issues. Through the CoMo model framework, we have produced a robust model to compare non-pharmaceutical intervention scenarios and assess their relative effects on reducing cases and deaths, during the COVID-19 pandemic. This model was broadly applied to inform policy making at national and international levels. Taking into account inter-household connectivity heterogeneities, we have also contributed towards bridging the gap between population and household-level models for communicable diseases, using insights from network percolation theory. Within the COVID-19 modelling context, we have also developed other model structures to address different research questions. Simplified age-structuring and intervention implementations was used to explore and optimise the outcome of single interventions (viz. school reopening and vaccination) or estimate epidemiologically relevant parameters (viz. transmission rates for a new variant). In the context of endemic malaria, we presented single-species models that can and have been useful to inform health policymakers. Nevertheless, since multiple species can coexist in some parts of the world, it is valuable to evaluate their combined burden while taken into account their interactions. In this context, a novel multi-species modelling framework is proposed and its applicability to weak interaction regimes is proven analytically. Such framework provided theoretical support for model structures already being applied to support health policy making. Our contributions encompass both the development of theoretical methodologies to support more robust models and the development of such models, aiming to support health decision making.
The mathematical modelling of infectious diseases is an extremely interdisciplinary field, where we need to amalgamate different areas of expertise, cultures and scientific backgrounds. Based on the available biological and epidemiological data, compartmental models can be tailored to investigate different research questions regarding a disease spread dynamics within a population. In our COVID-19 and malaria modelling exercises, we used different model structures to address different issues. Through the CoMo model framework, we have produced a robust model to compare non-pharmaceutical intervention scenarios and assess their relative effects on reducing cases and deaths, during the COVID-19 pandemic. This model was broadly applied to inform policy making at national and international levels. Taking into account inter-household connectivity heterogeneities, we have also contributed towards bridging the gap between population and household-level models for communicable diseases, using insights from network percolation theory. Within the COVID-19 modelling context, we have also developed other model structures to address different research questions. Simplified age-structuring and intervention implementations was used to explore and optimise the outcome of single interventions (viz. school reopening and vaccination) or estimate epidemiologically relevant parameters (viz. transmission rates for a new variant). In the context of endemic malaria, we presented single-species models that can and have been useful to inform health policymakers. Nevertheless, since multiple species can coexist in some parts of the world, it is valuable to evaluate their combined burden while taken into account their interactions. In this context, a novel multi-species modelling framework is proposed and its applicability to weak interaction regimes is proven analytically. Such framework provided theoretical support for model structures already being applied to support health policy making. Our contributions encompass both the development of theoretical methodologies to support more robust models and the development of such models, aiming to support health decision making.
Descrição
Palavras-chave
Modelos matemáticos, Biologia Modelos matemáticos, Doenças transmissíveis, COVID-19, Malária, Mathematical models, Mathematical biology, Infectious diseases, Malaria, SARS-CoV-2