Wastewater-Based Epidemiology for SARS-CoV-2 Biomarkers: Evaluation of Normalization Methods in Small and Large Communities in Southern Germany

Abstract

In the context of the COVID-19 pandemic, wastewater-based epidemiology (WBE) emerged as a useful tool to account for the prevalence of SARS-CoV-2 infections on a population scale. In this study, we analyzed wastewater samples from three large (>300,000 people served) and four small (<25,000 people served) communities throughout southern Germany from August to December 2021, capturing the fourth infection wave in Germany dominated by the Delta variant (B.1.617.2). As dilution can skew the SARS-CoV-2 biomarker concentrations in wastewater, normalization to wastewater parameters can improve the relationship between SARS-CoV-2 biomarker data and clinical prevalence data. In this study, we investigated the suitability and performance of various normalization parameters. Influent flow data showed strong relationships to precipitation data; accordingly, flow-normalization reacted distinctly to precipitation events. Normalization by surrogate viruses CrAssphage and pepper mild mottle virus showed varying performance for different sampling sites. The best normalization performance was achieved with a mixed fecal indicator calculated from both surrogate viruses. Analyzing the temporal and spatial variation of normalization parameters proved to be useful to explain normalization performance. Overall, our findings indicate that the performance of surrogate viruses, flow, and hydro-chemical data is site-specific. We recommend testing the suitability of normalization parameters individually for specific sewage systems.

The dynamic relationship between COVID-19 cases and SARS-CoV-2 wastewater concentrations across time and space: Considerations for model training data sets

“…Concentration of SARS-CoV-2 in wastewater is affected by the number of infected individuals and their respective viral shedding dynamics, but also by precipitation events, infiltration and inflow ( McCall et al, 2022 ), industrial flow contributions, and many other factors ( Soller et al, 2022 ; Wade et al, 2022 ; de Araújo et al, 2023 ; Mitranescu et al, 2022 ). Flow rates at wastewater sampling sites can be used to adjust for dilution, but flow data is not always available, especially for samples collected from manholes or small wastewater treatment facilities where no flow meter is present.…”

“…Prior research has used correlation between wastewater and case data as a readout for the effectiveness of normalization methods, for determination of lead/lag times between datasets, and as a means to state the value of wastewater monitoring in general ( Greenwald et al, 2021 ; Ho et al, 2021 ; Mitranescu et al, 2022 ; Feng et al, 2021 ; Maal-Bared et al, 2023 ). However, statistical caveats of this analysis are often ignored – for example the fact that the correlation of two variables that measure the same phenomenon in a time series is inflated due to autocorrelation ( Hamed, 2009 ; Wolfe et al, 2021 ).…”

Ratio of HF183 to PMMoV in Sewage and Its Change When Exposed to Different Temperatures and Sunlight Conditions

“…Results from our study demonstrate that HF183 decays faster than PMMoV, meaning that PMMoV might be a better-normalizing marker than HF183 if the travel times are long. This is consistent with the findings of another recent publication, where it was reported that normalizing by the mean of two indicator biomarkers (PMMoV and crAssphage), produced better normalization results for COVID-19 cases than using either of the two virus biomarkers on their own …”

COVID-19 trends at the University of Tennessee: predictive insights from raw sewage SARS-CoV-2 detection and evaluation and PMMoV as an indicator for human waste

“…Other than PMMoV, Hsu et al (2022) found that Paraxanthine was a more reliable population biomarker than PMMoV. Mitranescu et al (2022) reported that the best normalization performance was achieved with a mixed fecal indicator calculated from both CrAssphage virus and PMMoV. It is crucial to note that these divergent conclusions were primarily attributed to the choice of matrix (solids versus liquid) ( D'Aoust et al, 2021 ; Wolfe et al, 2021 ) and variances in analytical methods ( Scott et al, 2021 ; Zhan et al, 2022 ).…”