Evaluating trade-offs between COVID-19 prevention and learning loss: an agent-based simulation analysis

TL;DR

This study uses a simulation to compare how different strategies for preventing COVID-19 in schools affect both infection rates and learning loss.

Contribution

The paper introduces an agent-based model to evaluate trade-offs between infection prevention and learning loss in a post-Omicron context.

Findings

Quarantine of exposed students reduced infections but caused significant learning loss.

Test-to-stay achieved similar infection reduction with less learning loss.

Universal masking and vaccination reduced infections without harming learning.

Abstract

The COVID-19 pandemic presented significant challenges in educational settings. Schools implemented a variety of COVID-19 mitigation strategies, some of which were controversial due to potential disruptions to in-person learning. We developed an agent-based model of COVID-19 in a US high school setting to evaluate potential trade-offs between preventing COVID-19 infections versus avoiding in-person learning loss under different mitigation policies in a post-Omicron context. Mitigation policies included isolation alone and in combination with quarantine of exposed students, weekly testing of all students or testing of exposed students (‘test-to-stay’) under different scenarios of mask use and booster dose uptake. Outcomes were simulated over an 11 week trimester. We found that requiring a full 5 or 10 day quarantine of exposed students reduced COVID-19 infections by five to sevenfold relative to isolation alone, but at a cost of nearly 40% learning days lost. Test-to-stay achieved nearly the same level of infection reduction with lower levels of learning loss. Weekly testing also reduced COVID-19 infections but was less effective and incurred higher learning loss than test-to-stay. Universal masking and increased vaccination not only reduced infections at no cost to learning but also synergized with other strategies to reduce trade-offs.

Figures