Identification of sentinel upstream community sites for wastewater surveillance of SARS-CoV-2 in a large urban area

Abstract

Wastewater-based surveillance (WBS) captures the presence of disease in a community of people regardless of symptom status and supports public health interventions to mitigate the spread of disease. Wastewater-based surveillance can be applied to a variety of spatial scales and population sizes, particularly where households are served by municipal wastewater collection systems (e.g., large areas served by a single wastewater treatment plant (WWTP), smaller areas contained within a single neighbourhood, individual facilities). Since the onset of the COVID-19 pandemic in 2020, governments have had to make critical decisions on where, and at what scale, to implement WBS. Population size, health equity, and sampling access are some of the factors that are typically considered in these decisions; however, other population and sewer system characteristics may be important to consider when optimizing WBS resources. In this study, we undertook WBS for SARS-CoV-2 (the virus that causes the COVID-19 disease) at six community sites located upstream of a large WWTP in the City of Toronto, Ontario, Canada. We then used mixed effects modelling to explore the dominant drivers of spatio-temporal variability in the relationship between the wastewater signal and clinical cases for SARS-CoV-2 across these sites. The data collected over a 17-month period suggested that population density, pipe length, and 'dependency' - a community marginalization index that quantifies the number of seniors, children, and adults whose work is not compensated - played a significant role in judging whether a specific site could be used as a sentinel site. Though the number of upstream community sites was relatively small - and there were correlations between predictors - the length of data record allowed us to demonstrate which variables had the strongest explanatory power in a multi-model context. Community marginalization indices can be used - in addition to physical variables like population density and sewer pipe length, to inform sentinel site selection for WBS in urban community 'sewersheds'.

Keywords: Community marginalization; Mixed effects modelling; Public health; SARS-CoV-2; Sentinel site selection; Wastewater-based surveillance.