When is it appropriate to model transmission of tuberculosis using a dose response model?

Abstract

The risk of transmission of tuberculosis (TB) in confined spaces is analyzed using the Gammaitoni and Nucci model and its variant, a dose response model. The dose response model with its additional parameters has the benefit of taking the immune status and susceptibility of an individual into account, as a separate term from the generation parameter. It is shown via a sensitivity analysis and a model algebraic identifiability analysis that there is little benefit in using the more complex dose response model unless the number of infectious TB particles in the air can be measured. This is because the uncertainties associated with the additional dose response model parameters are otherwise lumped into the parameter representing the generation of infectious TB particles.