Model-based index reveals highly varied correlation between norovirus wastewater surveillance and gastroenteritis hospital admissions

Abstract Wastewater virus surveillance is a powerful complementary tool for epidemic monitoring and countermeasure, yet its quantitative implications, especially in a long-term perspective, have not been well studied. In this study, by monitoring the occurrence of human norovirus (NoV) GII in municipal wastewater for multiple years, we analyzed and modeled the long-term correlation between the wastewater surveillance results and gastroenteritis hospital admissions. We found the wastewater viral load correlated well with the patient viral load calculated from hospital admissions during an outbreak event, but the correlation became loosened in the subsequent non-outbreak period, implying the uncertainty introduced by mild and asymptomatic patients. Bayesian analysis revealed that the correlation varies greatly over time and is under the impact of factors with various timescales, each epidemic year and month has its own impact on the correlation. In the study period, a difference of up to 45 times was observed in the relative gap between wastewater viral load and estimated patient viral load. Our findings highlight that: 1) the uncertainty of wastewater surveillance can be attributed to factors under different temporal scales, and 2) considering the pathological and environmental impacts is crucial to the dependability when interpreting wastewater surveillance results.