Universities are grappling with how to resume on-campus education and research activities while mitigating the risk of COVID-19 transmission. In collaboration with the University of California, San Diego, Mathematica helped the university keep its campus community safe from the spread of the coronavirus. The agent-based model, which simulates the interactions of individual groups to assess their effects on the system as a whole, calculated what would happen under three strategies to reduce transmission of the virus: (1) mitigating risk (including wearing a mask and social distancing), (2) monitoring viral activity through testing, and (3) using public health interventions (such as isolation, contact tracing, and quarantine activities). The model simulated the spread on campus during an 80-day term using different combinations and levels of these strategies.
The model includes four components: (1) the breakdown of the university population, which includes students, staff, and faculty; (2) possible interactions members of the population; (3) transmission of the coronavirus through these contacts; and (4) disease progression of COVID-19 for those infected.
To visualize the potential impact of the mitigation strategies, Mathematica and UC San Diego developed an interactive COVID-19 university tool. Users can modify aspects of the two risk mitigation strategies that significantly affect the virus’s spread: (1) social distancing and mask wearing, and (2) testing frequency.
The findings from the model, as displayed in the visualization, show the following:
- Universities can create thoughtful policies, like less crowded living spaces, or hybrid instruction, but the social interactions among students and the community play a significant role in mitigating spread.
- Testing at higher frequencies with lower accuracy is better than testing at a lower frequency with higher accuracy.
Learn more about Mathematica’s agent-based models for mitigating COVID-19 here.
If you are interested in using simulation studies powered by agent-based models to help inform your institution’s safety protocols and policy decisions, please contact Andrew Hurwitz.