Nationwide Trends in COVID-19 Cases and SARS-CoV-2 RNA Wastewater Concentrations in the United States

Abstract

Wastewater-based epidemiology has emerged as a promising technology for population-level surveillance of COVID-19. In this study, we present results of a large nationwide SARS-CoV-2 wastewater monitoring system in the United States. We profile 55 locations with at least six months of sampling from April 2020 to May 2021. These locations represent more than 12 million individuals across 19 states. Samples were collected approximately weekly by wastewater treatment utilities as part of a regular wastewater surveillance service and analyzed for SARS-CoV-2 RNA concentrations. SARS-CoV-2 RNA concentrations were normalized to pepper mild mottle virus, an indicator of fecal matter in wastewater. We show that wastewater data reflect temporal and geographic trends in clinical COVID-19 cases and investigate the impact of normalization on correlations with case data within and across locations. We also provide key lessons learned from our broad-scale implementation of wastewater-based epidemiology, which can be used to inform wastewater-based epidemiology approaches for future emerging diseases. This work demonstrates that wastewater surveillance is a feasible approach for nationwide population-level monitoring of COVID-19 disease. With an evolving epidemic and effective vaccines against SARS-CoV-2, wastewater-based epidemiology can serve as a passive surveillance approach for detecting changing dynamics or resurgences of the virus.

Separating Signal from Noise in Wastewater Data: An Algorithm to Identify Community-Level COVID-19 Surges

“…Our finding that normalizing wastewater levels did not substantially alter correlations between wastewater and case data aligned with some previous studies, but not others. In two previous analyses that included data from several sites, findings were similar to our study (16, 21). But in past analyses that largely focused on data from a single site, correlations between wastewater and case data were improved after normalizing wastewater concentrations (22, 23).…”

“…Whereas previous reports from statewide wastewater programs have focused more on urban communities with large WWTPs (16, 21), North Carolina’s statewide program includes rural communities with small WWTPs (the smallest of which serves 3,500 people). The underrepresentation of small WWTPs may explain why Duvallet et al found no relationship between the strength of correlation between PMMoV normalized wastewater and clinical case data and WWTP service population size (21). Our finding that normalizing wastewater levels did not substantially alter correlations between wastewater and case data aligned with some previous studies, but not others.…”

Population Normalization in SARS-CoV-2 Wastewater-Based Epidemiology: Implications from Statewide Wastewater Monitoring in Missouri

“…The reduced strength of the relationship between the PMMoV-population normalized SARS-CoV-2 load and COVID-19 incidence rate was consistent with previous studies that showed PMMoV had mixed or adverse effects on the correlation between wastewater measurements of SARS-CoV-2 and clinical cases. [35][36][37] To the best of our knowledge, there is only one nationwide study to date (conducted by Biobot Analytics) that is comparable with our study in terms of the temporal and spatial magnitudes of sampling. That study collected 2,433 samples from 55 WWTPs across the U.S. and found that PMMoV normalization did not always improve the correlation between wastewater measurement and clinical cases.…”

“…That study collected 2,433 samples from 55 WWTPs across the U.S. and found that PMMoV normalization did not always improve the correlation between wastewater measurement and clinical cases. 35 Feng et al showed that PMMoV normalizations reduced the correlations between SARS-CoV-2 concentration and COVID-19 incidence for 8 of 12 WWTPs and suggested that variability’s influence across measurement for human viral is stronger than that of differences in the fecal loads in the samples 36 . For some sewersheds where the normalizations by metadata population and without population normalization were better than the normalizations by caffeine and PMMoV estimated population, probably because most WWTPs serve rural areas in Missouri where the population does not fluctuate with time.…”

“…Many wastewater SARS-CoV-2 surveillance studies have been conducted across the U.S. during the last two years 35 , which provided an excellent network for the effective and long-term monitoring of SARS-CoV-2 and possibly other diseases in the future. Thus far, however, the normalization applied to the SARS-CoV-2 wastewater surveillance does not have a standardized approach.…”