Understanding Human Contact Patterns And Epidemic Spreading Across Populations

The study of spatial spread of infectious diseases is a cutting-edge issue concern-ing the field of statistics, biology, medicine and system science. For decades, along with changes in human’travel efficiency, the spreading ability of space-borne diseases has been significantly improved, one typical example involves the H1N1 influenza 2009, found in Mexico in March, just after three months, the disease made its presence across the globe. In recent years, empirical studies based on real data, not only deepens our understanding of human mobility patterns, but also provides an access to gain a deep understanding of individuals’contact patterns. The contact refers to happenings where infection transmission could occur. Understanding of the contact patterns can help us better depict the spatial spreading of diseases. By empirical data analysis, theoretical analysis and Monte Carlo simulations, we study the contact patterns and spatial spread-ing patterns, and how the contact pattern may possibility affect the spreading. The main content of the paper is organized as follows:● Utilizing the log data of WiFi users in a Chinese university campus, we ana-lyze the contact patterns from the aggregated network and temporal network per-spectives. In the aggregated approach, we find that in the campus environment, individuals’contact frequencies follow a exponential distribution, which means that the population’s contact may be depicted by a unique characteristic contact rate. Also, from the temporal network perspective, we infuse the temporal dimen-sion into the construction of contact network. By quantitative comparison with the traditional aggregation approach, we find that the temporal contact network differs in many features, e.g., the teachability, the path length distribution. We conclude that the correlation between temporal path length and duration is influ-enced by the micro-dynamical features of human activities under certain social circumstance. In addition, our study provides a promising measure to identify the potential superspreaders by distinguishing the nodes functioning as the relay hub.● The location-specific factor has been reported as a potential source of the remark-able statistical variation in the number of disease incidence among populations. To study this issue, we introduce the heterogeneity characterizing the contact patterns of individuals, assuming that the individual contact feature is an inher-ent attribute of him, only relevant to his origin location. With the metapopulation model where subpopulations are coupled by the commuting flows. We demon-strate that the considered contact scenario dramatically impacts the dynamic be-havior of disease spreading in the whole system, by lowering the epidemic thresh-old conditions for an outbreak. Moreover, take a simple link-deletion measure as an example, we illustrate how the heterogeneous contact scenario may influence the efficacy of containment strategies, which indicates the necessity to reevaluate the efficacy of travel restriction measures that was considered to be ineffective.● The literatures studying the spatial spreading patterns mainly focus on the global or countrywide scale, while the intra-city community level has been less em-braced. With the laboratory-confirmed cases of H1N12009 in Hong Kong, for the first time we analyze the data from a systematic perspective with the network approach. We first infer the correlation between time series of the infectious at the DCCA level, and construct the correlation network and invasion tree from on the causality relationship. With analysis of the above two structures, we uncover some basic attributes of the spreading process at the community level.● Currently, several international research teams constructed the simulation plat-form in the framework of metapopulation model. These open platforms usually concern too much detail at the individual level, and have a high computational complexity. And simulations can only be performed on a cluster. In our work, we develop a simulation platform for the epidemic spreading on the metapopulation model. As the first open source simulation software, the platform finds a better balance between computational efficiency and accuracy, and is easily to be ex-tended. In addition, the computing kernel and user interface are separated, which facilitates further updating and maintenance of the platform.