A1690
Title: Probabilistic reconstructions of epidemic curves using wastewater data
Authors: Justin Slater - University of Guelph (Canada) [presenting]
Abstract: Determining the current and historical prevalence of an infectious disease is crucial for understanding disease burden. Wastewater data is a relatively low-resource way to measure changes in prevalence and, unlike other forms of surveillance data, is less affected by preferential testing. However, how to best utilize wastewater data to estimate prevalence is an area of active research. A novel framework is proposed for reconstructing epidemic curves using wastewater data within a binomially thinned Poisson autoregressive framework. The effective reproduction number is modeled as the logarithmic derivative of the wastewater signal to make probabilistic inference about the true historical and current case counts. Several applications of this framework are demonstrated.
