Using Agent-Based Modeling to Examine Risk for COVID-19 Infection in Custodial Settings

Research on pandemics in institutional settings often assumes that all human interactions within a jail pose similar viral transmission risks. We developed an agent-based model (ABM) called Simulation Applications for Forecasting Effective Responses in Corrections (SAFER-C (TM)) to simulate nine scenarios of possible interactions and virus transmission among incarcerated individuals and jail staff and tested this assumption. We found that resumption of high-contact activities has a greater impact on the number of infections, while out-of-cell group sizes and initial vaccination rates had lower impact. This work emphasizes the importance of understanding and modeling human interactions in confinement facilities, as well as understanding, responding to, and limiting the mechanism of viral transmission in jails. Insights from ABMs provide correctional administrators with realistic options for managing responses.