Estimating the effective reproduction number of COVID-19 in North Carolina

Abstract

In conclusion, R_t can be a useful metric, but it is sensitive to differences between models and underlying data. In North Carolina, this was particularly evident during September to October, as well as after mid-November when the trend in cases changed dramatically. In light of these findings, R_t
estimates should be presented in a way that emphasizes the uncertainty inherent in these estimates. For example, by showing a range of potential values (an interval estimate), by showing comparisons with estimates from a range of models, or by showing an ensemble estimate that combines estimates from multiple models.