The Center for Risk Analysis Informed Decision Engineering (RAIDE) has developed a COVID risk model to support mitigation decision making by the College of Engineering. This model was commissioned by the College of Engineering in March 2021 to provide input to the College’s decision making for how to prepare for the fall semester. Results of this model have been briefed regularly to College leadership. This is part of a broader RAIDE effort to provide risk analysis inputs to the decision making process.
It is often said (often attributed to George Box) that “All models are wrong, some models are useful.” This is especially true with modeling the spread of a novel virus like SARS-CoV-2. There are very large uncertainties about key aspects of the speed of SARS-CoV-2, and new variants only increase this uncertainty. Our approach to modeling this problem is grounded in a few key principals:
- Transmission and infection should be modeled at the level of individual class period. This allows class-level and room-level mitigation measures to evaluated.
- The entire semester should be simulated. This allows the propagation of the virus and associated infections across the semester to examined.
- The end point of interest cannot be just infection. Health outcomes such as hospitalization and death must be included. The likelihoods of these given infection are highly age-dependent.
- Transparency of assumptions. All models have assumptions, and this is especially true of any model of COVID risk. These assumptions should be made clear.
This page, which is a work in progress, will summarize the model and present model results and associated useful information.