Evaluation of sampling frequency and normalization of SARS-CoV-2 wastewater concentrations for capturing COVID-19 burdens in the community

Abstract

Wastewater surveillance for SARS-CoV-2 provides an approach for assessing the infection burden across a city. For these data to be useful for public health, measurement variability and the relationship to case data need to be established. We measured SARS-CoV-2 RNA concentrations in the influent of twelve wastewater treatment plants from August 2020 to January 2021. Replicate samples demonstrated that N1 gene target concentrations varied by 21% RSD between technical replicate filters and by 14% RSD between duplicate assays. COVID-19 cases were correlated significantly (rho≥0.70) to wastewater SARS-CoV-2 RNA concentrations for seven plants, including large and small cities. SARS-CoV-2 data normalized to flow improved correlations to reported COVID-19 cases for some plants but normalizing to a spiked recovery control (BCoV) or a fecal marker (PMMoV or HF183) generally reduced correlations. High frequency sampling demonstrated that a minimum of two samples collected per week was needed to maintain accuracy in trend analysis. We found a significantly different ratio of COVID-19 cases to SARS-CoV-2 loads in one of three large communities, suggesting a higher rate of undiagnosed cases. These data demonstrate that SARS-CoV-2 wastewater surveillance can provide a useful community-wide metric to assess the course of the COVID-19 pandemic.

Tools for interpretation of wastewater SARS-CoV-2 temporal and spatial trends demonstrated with data collected in the San Francisco Bay Area

“…For COVID-19 WBE to be useful for public health decision-making, a better understanding is needed of the variability of SARS-CoV-2 in wastewater and how it relates to the true incidence or prevalence of COVID-19 in the contributing population ( McClary-Gutierrez et al, 2021 ). Sources of target signal variability in wastewater include inconsistencies in sample collection and laboratory processing ( Ahmed et al, 2020d ; Feng et al, 2021 ), nucleic acid degradation based on travel time and conditions in the sewer ( Hart and Halden, 2020a ), and signal dilution due to rainfall and diurnal flow changes ( Zahedi et al, 2021 ). Researchers have addressed some of these sources of variability through normalization to biomarkers, increased sampling frequency, processing sample replicates, and smoothing/forecasting ( D'Aoust et al, 2021b ; Feng et al, 2021 ; Graham et al, 2020 ; McLellan et al, 2021 ; Nemudryi et al, 2020 ; Stadler et al, 2020 ).…”

Operationalizing a routine wastewater monitoring laboratory for SARS-CoV-2

“…Additionally, the pipeline presented here allowed meta-analysis of the data, revealing the variability present in each step of sample processing. While some amount of variation is due to lower quality or technical error, some appears to be inherent methodological variation or variation in the wastewater itself 27,28 . In the future, modeling the contributions of these variables to overall noise in the data could provide better confidence in the results.…”

Interpretation of temporal and spatial trends of SARS-CoV-2 RNA in San Francisco Bay Area wastewater

“…Bacteroides rRNA loads varied more spatially and temporally than crAssphage or PMMoV in this study (Figure 1), but Bacteroides-normalized SARS-CoV-2 N1 had a moderate correlation with clinical testing data ( =0.35). While measurement of Bacteroides rRNA gene in wastewater has been commonly applied for fecal source tracking and Bacteroides rRNA has been targeted to increase assay sensitivity (D'Aoust et al, 2021b;Feng et al, 2021), to our knowledge, no prior raw wastewater values have been reported in the literature for Bacteroides rRNA (Table S10). Similarly, no values were found in the literature for 18S rRNA concentrations in raw wastewater (Table S10).…”

“…For COVID-19 WBE to be useful for public health decision-making, a better understanding is needed of the variability of SARS-CoV-2 in wastewater and how it relates to the true incidence or prevalence of COVID-19 in the contributing population (McClary-Gutierrez et al, 2021). Sources of target signal variability in wastewater include inconsistencies in sample collection and laboratory processing (Ahmed et al, 2020d;Feng et al, 2021), nucleic acid degradation based on travel time and conditions in the sewer (Hart and Halden, 2020a), and signal dilution due to rainfall and diurnal flow changes (Zahedi et al, 2021). Researchers have addressed some of these sources of variability through normalization to biomarkers, increased sampling frequency, processing biological replicates, and smoothing/forecasting (D'Aoust et al, 2021b;Feng et al, 2021;Graham et al, 2020;McLellan et al, 2021;Nemudryi et al, 2020;Stadler et al, 2020).…”

“…More recently, four biological markers have emerged as promising candidates to normalize SARS-CoV-2 RNA signal for fecal content. Pepper mild mottle virus (PMMoV), a nonenveloped RNA plant virus, is commonly used for COVID-19 WBE (D'Aoust et al, 2021b;Feng et al, 2021;Whitney et al, 2021;Wu et al, 2020) but concentrations in sewage vary with season and local diet (Symonds et al, 2019). Another normalization biomarker is the cross-assembly phage (crAssphage), a non-enveloped, DNA virus that ubiquitously infects the human gut commensal bacteria Bacteroides (Edwards et al, 2019;Green et al, 2020;Stachler et al, 2017;Wilder et al, 2021).…”