Using an agent-based model of COVID-19 dynamics to support public health decision making

Guido España

University of Notre Dame

"Using an agent-based model of COVID-19 dynamics to support public health decision making"

In Bogotá, Colombia, more than 1.8 million cases of COVID-19 and 30,000 deaths had been reported by April 2023. During the critical phase of the pandemic, decision makers required estimates of the impact of different scenarios to design public-health interventions, such as school closures, face-masks, or the distribution of available vaccines. For instance, public schools were closed for in-person instruction in Bogotá during most of 2020. We used an agent-based model of COVID-19 and calibrated it to represent the epidemiological dynamics of COVID-19 in Bogotá, including SARS-CoV-2 variants, and capable of reproducing time-varying public health interventions, such as reduction in mobility, school closures, and vaccination programs. To inform school reopening during the first semester of 2021, we simulated school reopening at different capacities, and found that school reopening could have had a small impact on the number of deaths reported in the city during the third wave at 35% capacity of in-person instruction during the simulation period. Deaths were lowest when only reopening pre-kinder grades, and largest when secondary school was opened. The impact of opening pre-kinder at 100% capacity was below 10% of additional deaths. Finally, we also estimated the impact of vaccination in the city during the third and largest wave of COVID-19 in 2021. Our simulation results suggest that vaccination may have prevented more than 17 thousand deaths in the city.
Additional authors: Zulma Cucunubá, Universidad Javeriana, Bogotá, Colombia; Hernando Diaz, Universidad Nacional de Colombia, Bogotá, Colombia; Sean Cavany, University of Oxford, Oxford, England; Nelson Castañeda, Escuela Tecnológica Instituto Técnico Central, Bogotá, Colombia; Laura Rodriguez, GCFEP-Universidad del Tolima.

---