AUTHOR=Dhiyebi Hadi A. , Abu Farah Joud , Ikert Heather , Srikanthan Nivetha , Hayat Samina , Bragg Leslie M. , Qasim Asim , Payne Mark , Kaleis Linda , Paget Caitlyn , Celmer-Repin Dominika , Folkema Arianne , Drew Stephen , Delatolla Robert , Giesy John P. , Servos Mark R. 

TITLE=Assessment of seasonality and normalization techniques for wastewater-based surveillance in Ontario, Canada

JOURNAL=Frontiers in Public Health

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2023.1186525

DOI=10.3389/fpubh.2023.1186525

ISSN=2296-2565

ABSTRACT=Wastewater-based surveillance is at the forefront of monitoring for community prevalence of COVID-19, however continued uncertainty exists regarding the use of fecal indicators for normalization of the SARS-CoV-2 virus in wastewater. Using three communities in Ontario, sampled across 2021-2023, the seasonality of a viral fecal indicator (pepper mild mottle virus, PMMoV) and utility of normalization of data to improve correlations with clinical cases was examined. Wastewater samples from Warden, Humber AMF and Kitchener were analyzed for SARS-CoV-2, PMMoV and crAssphage. Seasonality of PMMoV and flow rates were examined and compared by Season-Trend-Loess decomposition analysis. The effects of normalization using PMMoV, crAssphage and flow rates was done by comparing the correlations to clinical cases by episode date (CBED) during 2021. Seasonal analysis demonstrated that PMMoV had similar trends at Humber AMF and Kitchener with peaks in January and April 2022 and low concentrations (troughs) in the summer months. Warden had similar trends but was more sporadic between the peaks and troughs for PMMoV concentrations. Flow demonstrated similar trends but was not correlated to PMMoV concentrations at Humber AMF and was very weak at Kitchener (r=0.12).  Despite the differences among the sewersheds, unnormalized SARS-CoV-2 (raw N1-N2) concentrations in wastewater (n=99-191) was strongly correlated to the CBED in the communities (r= 0.620-0.854) during 2021. Additionally, normalization with PMMoV did not improve the correlations at Warden and significantly reduced the correlations at Humber AMF and Kitchener. Flow normalization (n=99-191) at Humber AMF and Kitchener and crAssphage normalization (n=29-57) correlations at all three sites were not significantly different from raw N1-N2 correlations with CBED. Differences in seasonal trends in viral biomarkers caused by differences in sewershed characteristics (flow, input etc.) may play a role in determining how effective normalization may be for improving correlations (or not).  This study highlights the importance of assessing the influence of viral fecal indicators on normalized SARS-CoV-2 or other viruses of concern. Fecal indicators used to normalize the target of interest may help or hinder establishing trends with clinical outcomes of interest in wastewater-based surveillance and needs to be considered carefully across seasons and sites.