morPOP: a fast and granular agent-based model of COVID-19 to examine school mitigation strategies in newfoundland & labrador

Dionne M. Aleman, Benjamin Tham, Sean J. Wagner, Justin Semelhago, Asghar Mohammadi, Paul Price, Jordan Bradfield, Randy Giffen, Proton Rahman, Jenna Lawrence

Abstract: The Medical Operations Research Lab's Pandemic Outbreak Planner (morPOP) is an agent-based simulation of pandemic disease spread, wherein each individual in the population is an agent with unique demographic and comorbidity characteristics; for Newfoundland & Labrador, there were ≈520,000 agents. Individuals interacted in various environments, and each infectious contact increased probability of infection. Individuals potentially changed behaviors to self-isolate or seek medical care when symptomatic. By comparing the resulting disease spread across proposed school mitigation measures, including physical distancing and masks, effective policies were identified, allowing public health officials to make evidence-based decisions about appropriate measures to enact. The model was unique in its ability to capture such a large population while requiring ≈25 s computation time for a 100-day simulation on a single processor, thanks to implementation in C++ with various cost-saving techniques. Parallelization support was provided by the Center for Health Informatics and Analytics (CHIA) at Memorial University.