Assessment of the COVID-19 infection risk at a workplace through stochastic microexposure modeling

Background: The COVID-19 pandemic has a significant impact on economy. Decisions regarding the reopening of businesses should account for infection risks. Objective: This paper describes a novel model for COVID-19 infection risks and policy evaluations. Methods: The model combines the best principles of the agent-based, microexposure, and probabilistic modeling approaches. It takes into account specifics of a workplace, mask efficiency, and daily routines of employees, but does not require specific inter-agent rules for simulations. Likewise, it does not require knowledge of microscopic disease related parameters. Instead, the risk of infection is aggregated into the probability of infection, which depends on the duration and distance of every contact. The probability of infection at the end of a workday is found using rigorous probabilistic rules. Unlike previous models, this approach requires only a few reference data points for calibration, which are more easily collected via empirical studies. Results: The application of the model is demonstrated for a typical office environment and for a real-world case. Conclusion: The proposed model allows for effective risk assessment and policy evaluation when there are large uncertainties about the disease, making it particularly suitable for COVID-19 risk assessments.