Identifying Potential Superspreaders of Airborne Infectious Diseases in Construction Projects

Abstract

The spread of airborne infectious diseases has largely been driven by superspreading events, in which a single individual directly infects several contacts. Superspreading events that occurred at several construction sites around the world afflicted construction practitioners and forced the suspension of construction activities. To reduce the probability of superspreading events, this study developed a network-based computational framework based on a K-shell decomposition approach with the input of the topological interaction network of project participants to identify potential superspreaders in construction projects. The feasibility of the developed framework was evaluated with three numerical case studies: one sample case with a hierarchical structure with an average accuracy of 98.45%, one sample case with a matrix structure with an average accuracy of 92.25%, and an empirical case related to a COVID-19 outbreak in a construction project in Hong Kong with an accuracy of over 80.13%. This study recommends that all potential superspreaders, especially if they are employed by the main contractor, take rapid antigen tests (RATs) regularly. If all potential superspreaders are detected through regular RATs and all potential secondary cases are detected by contract tracing, up to 82.35% of infected cases can be prevented.