OpenUp Model

The Columbia Covid-19 model allows users to examine the cost-effectiveness of infection control measures for Covid-19 in a university or school setting. The model computes the average daily cost associated with common infection control measures, such as testing for active infections, over a 90-day period. The model treats students and school employees as two separate but interacting populations. Each population is exposed to a daily risk of infection.

Please be careful to fully read the technical notes, including the inputs and assumptions, prior to using the model. The model and the paper have been peer reviewed once, and are now undergoing peer review in a second journal.

The model interface is available here. In this interface, the user should first define the number of infectious cases/100,000 population. In New York, this can be computed as the corrected number of confirmed cases/100,000 population multiplied by 7 (for the number of days that people remain infectious with Covid-19). The user should next enter the university population, input the sensitivity and specificity of the test that is under consideration, and modify the number of close contacts that are expected between university affiliates, both on and off campus. It is very important to obtain an accurate estimate of close contacts--particularly housemates--for the model to work properly. University administrators may wish to first survey students.

The underlying model, written in R, is available here.

This model is open source under a GNU 3.0. If you modify it to fit your context, we ask that you outline how the model was modified. If you find a bug, we ask that you report it to us.