Modeling the drivers of oscillations in COVID-19 data on college campuses

PURPOSE: Incorporating human behavior in a disease model can explain the oscillations in COVID-19 data which occur more rapidly than can be explained by variants alone on college campuses. METHODS: Dampened oscillations emerge by supplementing a simple disease model with a risk assessment function, which depends on the current number of infected individuals in the student population and the institutional public health policies. After accounting for a rapid disease impulse due to social gatherings, we achieve sustained oscillations that follow the trend of 2020/2021 COVID-19 data as reported on the COVID-19 dashboards of US post-secondary institutions. RESULTS: This adjustment to the epidemiological model can provide an intuitive way of understanding rapid oscillations based on human risk perception and institutional policies. More risk-averse communities experience lower disease-level equilibria and less oscillations within the system, while communities that are less responsive to changes in the number of infected individuals exhibit larger amplitude and frequency of the oscillations. CONCLUSIONS: Community risk assessment plays an important role in COVID-19 management in college settings. Improving the ability of individuals to rapidly and conservatively respond to changes in community disease levels may help assist in self-regulating these oscillations to levels well below thresholds for emergency management.