Predicting respiratory-related ED visits using wastewater signals and reported cases: a hierarchical Bayesian model

During the COVID-19 pandemic, we saw a widespread adoption of wastewater-based surveillance as a passive, non-invasive tool for monitoring community-level transmission. Early efforts focused on correlating wastewater viral loads with clinical indicators such as reported cases and hospitalizations, often identifying a time lag between wastewater signals and clinical outcomes. As the field evolved, researchers began applying statistical and machine learning models to leverage these signals for short-term forecasting of disease trends. In this study, we contribute to the growing body of work by developing predictive models for SARS-CoV-2, Influenza A, and Respiratory Syncytial Virus (RSV) in British Columbia (BC). Using wastewater viral load as a predictor, we first constructed models to forecast reported case counts. Building on this foundation, we are developing a spatiotemporal hierarchical Bayesian model that integrates reported case data and wastewater signals to predict respiratory-related emergency department (ED) visits across BC. Our work highlights the value of wastewater-based data for early detection and response planning in public health systems.

Predicting respiratory-related ED visits using wastewater signals and reported cases: a hierarchical Bayesian model

Monday, July 14 at 6:00pm

Rebeca Cardim Falcao

"Predicting respiratory-related ED visits using wastewater signals and reported cases: a hierarchical Bayesian model"

#SMB2025 Follow