Evolutionary Dynamics Of Metapopulation-based Spatial Epidemiology And The Analysis Of Intervention Strategies

The evolutionary dynamics of spatial epidemiology is one of the hottest fields of interdisciplinary study combining several hard-core sciences such as the public health, and information science. Recent decades, the networking of human society prominently improves the public healthcare systems, thereby weakening the treat of infectious dis-eases. However, it is also quite ironic that the networking society also enhances the wide spread potential of emerging diseases due to the frequent mobility of individuals via transportation flows. For instance, in2002-2003, through the transportation of in-ternational airline network, the SARS virus was rapidly transmitted from Hongkong to more than30countries. Several years later, in2009, the A(H1N1) virus swept across the world through the public transportation networks again, and it only spent3-4-months covering more than200countries. These novel features of the large-scale spatial transmission of infectious diseases derive from the changing of human ecology in population structure as well as the mobility patterns.The networking metapopulation provides us a reasonable modeling scheme to characterize the large-scale spatial epidemic spreading. This model adopts the particle-network framework as well as the reaction-diffusion (or reaction-commuting) mecha-nism to capture the key elements driving the spatial transmission of epidemics. In this dissertation, we mainly utilize the metapopulation model to investigate the evolution-ary dynamics of spatial epidemiology. By using the theoretical and empirical analysis, computational simulation as well as handling real data, we study the intrinsic mecha-nisms and influencing factors of the spatial epidemiology, and systematically analyze the effectiveness of intervention strategies. The main contents of this dissertation are organized as follows:·With the empirical data, we analyze the joint scaling emergence of the large-scale spatial epidemic spreading from the theoretical and empirical aspects. We unveil that this feature derives from the broad heterogeneity of the infrastructure, which has not been clarified in previous literatures (the environmental factors are considered to be the driving element by previous works). By understanding the spreading pattern, we uncover the significant role of the heterogeneous structure of the system in characterizing the spatial epidemic spreading.This part of contents will be given in detail in Chapter2. · Based on the minimum metapopulation model, we develop a general theoretical framework to analyze the effect of intervention strategies on delaying the spatial invasion (previous method is only suitable to analyze the scenario of travel re-striction). We compare two categories of typical control measures, namely, the travel restriction and patient isolation, and unveil the important role of response time.This part of contents will be explained in detail in Chapter3.·We systematically study the effectiveness of the strategy of case isolation, which is based on the feedback relation between the risk assessment and the implement of strategy (this decision-making process has not been considered in previous works). By comparing the effects of the homogeneous and heterogeneous inter-ventions, we find that the latter one performs much better than the former one under some conditions, which, however, does not produce the optimal results in delaying spatial invasion. It deserves to highlight that this data-driven networking metapopulation model is the first one built in our country.This part of contents will be explained in detail in Chapter4.·In traditional metapopulation framework, the subpopulations are often assumed homogeneous mixing. Based on recent empirical results, we introduce two cate-gories of location-specific heterogeneous human contact patterns into the model-ing framework. We study their impact on the epidemic threshold, and extend this issue to study the scenario of networking metapopulation. By means of theoret-ical analysis and computational simulation, we find that the considered hetero-geneous contact patterns significantly decrease the epidemic threshold and thus favor the outbreak of diseases.This part of contents will be given in detail in Chapter5.