Real-time estimation of pathogen transmission dynamics from wastewater

Area of Science:

• Environmental microbiology
• Epidemiology
• Computational biology

Background:

• Wastewater surveillance effectively tracked SARS-CoV-2, but estimating transmission for other pathogens is challenging due to low concentrations and limited data.
• Existing methods require extensive data preprocessing, hindering real-time analysis.

Purpose of the Study:

• To develop EpiSewer, a Bayesian semi-mechanistic model for robust wastewater-based pathogen transmission inference.
• To enable direct estimation of effective reproduction number (R_t) and epidemic growth rate (r_t) from raw wastewater data.

Main Methods:

• Developed EpiSewer, a Bayesian model integrating infection dynamics, pathogen shedding, and measurement noise (outliers, non-detects).
• Applied the model to multi-pathogen wastewater surveillance data (SARS-CoV-2, IAV, RSV) from Swiss treatment plants over three seasons.
• Assessed model performance for real-time R_t estimation, forecasting, and robustness to low pathogen concentrations and sampling frequencies.

Main Results:

• EpiSewer directly inferred R_t and r_t from raw wastewater data, eliminating the need for preprocessing.
• Achieved consistent and robust R_t estimates even for pathogens with 10-50 times lower concentrations than SARS-CoV-2.
• Provided well-calibrated 14-day concentration forecasts with minimal bias across epidemic phases and maintained forecast accuracy under reduced sampling.

Conclusions:

• EpiSewer offers a robust framework for inferring transmission dynamics of various pathogens, including those with low wastewater concentrations.
• The model enhances public health surveillance by enabling real-time monitoring and forecasting with limited clinical and wastewater data.
• Facilitates effective wastewater-based epidemiology for a broader range of infectious diseases.