Wastewater-based epidemiology (WBE) is a method to monitoring the prevalence of pathogens in a community or regions at-large using wastewater concentrations. During disease outbreaks, wastewater provides a rapid and effective early indicator of high-risk localities. WBE is a cost-effective surveillance tool, particularly as large-scale community testing efforts are unsustainable long-term. However, using wastewater data to guide public health decision-making requires understanding of factors contributing to uncertainty in disease prevalence estimation and the development of actionable metrics. We present two models to estimate COVID-19 cases from levels of SARS-CoV-2 in wastewater. The first model relates observed wastewater measurements through a convolution of the daily incidence of infections with the profile of SARS-CoV-2 RNA shedding in the wastewater by an infected individual days after infection or symptom onset. The number of daily cases is modeled with a Negative Binomial (NB) distribution in a Bayesian framework. The second approach uses a simple linear regression to estimate cases (dependent variable) from WW data (independent variable). To account for bias introduced by clinical testing, we propose a classification scheme to assess the appropriateness of model training periods as a function of clinical testing rates and evaluate the sensitivity of model predictions to training periods. The proposed modeling framework was applied to three Northern California communities served by separate wastewater treatment plants. The results showed that training periods with adequate testing are essential to provide accurate projections of COVID-19 cases.

Place: Math, 402 and Zoom   https://arizona.zoom.us/j/83758253931Password:  applied