Fitting dynamic models to epidemic outbreaks with quantified uncertainty: A primer for parameter uncertainty, identifiability, and forecasts

Chowell, Gerardo

In this article we have described and illustrated a relatively simple computational approach to quantify parameter uncertainty, evaluate parameter identifiability, assess model performance, and generate forecasts with quantified uncertainty. In the process we have employed simple phenomenological and mechanistic models to characterize epidemic patterns as well as estimate key transmission parameters such as the basic reproduction number R 0 and the effective reproduction number R t . This uncertainty quantification approach is computationally intensive and relies solely on case incidence series from an unfolding outbreak and allows considerations of different error structures in the case series data (e.g., Poisson vs. negative binomial).