Graph-based risk modeling for infectious diseases in social systems

Kurzfassung

The COVID-19 pandemic has changed people’s social lives since the outbreak in January 2020. The transmission of the disease through aerosols from infected individuals has been counteracted by governments worldwide. Companies have to adapt internal processes, which can lead to reduced capacity. This thesis presents a framework for determining the risk of infection in a social subsystem, such as a company. The basis for this is interaction data of the persons of the subsystem. The interaction protocol of this work was recorded over six weeks in a company in Italy using Bluetooth sensors. System-specific infection dynamics metrics for SARS-CoV-2, SARS-CoV-2-B.1.1.7, and Influenza are calculated and compared using specific workplace reproduction numbers and infection periods reported in the literature. The results show that for the social system considered, SARSCoV-2-B.1.1.7 has a 2.6 times higher risk of infection per social interaction than the original variant. Moreover, SARS-CoV-2-B.1.1.7 leads to 3.4 times as many secondary cases (920) as SARS-CoV-2 (270) if no countermeasures are taken. Social distancing turns out to be an effective countermeasure for the social subsystem under consideration, allowing a reduction of secondary cases of the British mutation to 360 and for SARS-CoV-2 to 69. It was also shown that targeted countermeasures based on topological network properties for a small fraction of individuals within the system can reduce the number of events in which an individual infects many more individuals at once by 25% by introducing social distancing measures for 15% of the individuals in the social subsystem under consideration (related to SARS-CoV-2-B.1.1.7). The framework can be used for any infectious disease transmitted through social interactions. It allows decision makers to evaluate different interventions, better understand social structures, and identify individuals within the social subsystem who are particularly at risk or transmitting infection.