Rapid identification of COVID wastewater surges in the absence of case data

Article Link

Abstract: Genetic testing of community wastewater (wastewater surveillance) is a valuable tool for following trends in the abundance of SARS-CoV-2 and other infectious disease pathogens over time. Wastewater surveillance is increasingly important in the absence of corresponding epidemiological data, particularly for infectious diseases with limited timely data on clinical case incidences. Due to the inherent noise in wastewater data, a single sample is not sufficient to identify a sustained trend in the abundance of a target. This challenge is magnified in resource-limited settings where samples may be collected only once or twice per week. In this work, we collected 24-h composite samples of wastewater daily from a single facility for nearly 4 years. We use this high-frequency data set to describe a method for identifying trends in SARS-CoV-2 abundance in wastewater based on a variety of collection frequencies. Our results indicate that collecting two 24-h composites per week for 2 weeks is sufficient to accurately identify a SARS-CoV-2 surge. We conclude that low-frequency wastewater sampling performs reasonably well in identifying trends in a timely fashion.

Methods: The study utilized wastewater surveillance to track SARS-CoV-2 trends over nearly four years at a single treatment facility serving 50,000 people. ​ Untreated wastewater was collected as 24-hour flow-based composite samples 5–7 days per week and transported to the lab on ice, stored at 4°C, and processed within 48 hours. ​ Biological particles were concentrated using Nanotrap Microbiome A particles, and total nucleic acids (TNA) were extracted using the MagMax Viral/Pathogen Nucleic Acid Isolation kit. ​ SARS-CoV-2 N1 and N2 targets were quantified using reverse transcriptase droplet digital PCR (RT-ddPCR) on a Bio-Rad QX600 system, with results reported as target copies per mL of wastewater. ​ A computational model was developed to identify surges in viral abundance using rolling windows of sample data, with parameters for window size, consecutive steps for surge detection, and collection frequency optimized for accuracy. ​ Results showed that collecting two 24-hour composite samples per week is sufficient to identify surges within 10 days, while three samples per week provided slightly better flexibility and faster detection. ​ All data and code are publicly available on GitHub.

Results: The study found that collecting two 24-hour composite wastewater samples per week is sufficient to identify surges in SARS-CoV-2 abundance within approximately 10 days, enabling timely public health responses. ​ Models using three samples per week provided slightly better flexibility and faster detection, identifying surges within an average of 8.7 days, but with only marginal improvements in accuracy compared to two-sample models. ​ Higher sampling frequencies did not significantly improve surge detection, likely due to increased noise in the wastewater matrix. ​ The study suggests that sampling frequencies of 2–4 days per week strike a balance between sensitivity and noise reduction, with evenly spaced collections across the week further improving model performance. ​ These findings support the use of limited sampling in resource-constrained settings to effectively monitor infectious disease trends.

Overview of wastewater testing workflow used in the current study. Samples are collected as 24-h flow-based composites at the treatment facility, brought to the testing lab on ice, and processed within 24 h of receipt. Biologicals are concentrated from 10 mL of the liquid fraction and total nucleic acid extracted directly thereafter. SARS-CoV-2 abundance is quantified using ddPCR on a Bio-Rad QX600 with a triplex assay that simultaneously quantifiesquantifiesquantifiesSARS-CoV-2 N1, SARS-CoV-2 N2, and human RNase P.