A comparison of SARS-CoV-2 wastewater concentration methods for environmental surveillance

Wastewater surveillance of SARS-CoV-2 may be a useful supplement to clinical surveillance as it is shed in feces, there are many asymptomatic cases, and diagnostic testing can have capacity limitations and extended time to results. Although numerous studies have utilized wastewater surveillance for SARS-CoV-2, the methods used were developed and/or standardized for other pathogens. This study evaluates multiple methods for concentration and recovery of SARS-CoV-2 and seeded human coronavirus OC43 from municipal primary wastewater and/or sludge from the Greater Seattle Area (March–July 2020). Methods evaluated include the bag-mediated filtration system (BMFS), with and without Vertrel™ extraction, skimmed milk flocculation, with and without Vertrel™ extraction, polyethylene glycol (PEG) precipitation, ultrafiltration, and sludge extraction. Total RNA was extracted from wastewater concentrates and analyzed for SARS-CoV-2 and OC43 with RT-qPCR. Skimmed milk flocculation without Vertrel™ extraction performed consistently over time and between treatment plants in Seattle-area wastewater with the lowest average OC43 C(q) value and smallest variability (24.3; 95% CI: 23.8–24.9), most frequent SARS-CoV-2 detection (48.8% of sampling events), and highest average OC43 percent recovery (9.1%; 95% CI: 6.2–11.9%). Skimmed milk flocculation is also beneficial because it is feasible in low-resource settings. While the BMFS had the highest average volume assayed of 11.9 mL (95% CI: 10.7–13.1 mL), the average OC43 percent recovery was low (0.7%; 95% CI: 0.4–1.0%). Ultrafiltration and PEG precipitation had low average OC43 percent recoveries of 1.0% (95% CI: 0.5–1.6%) and 3.2% (95% CI: 1.3–5.1%), respectively. The slopes and efficiency for the SARS-CoV-2 standard curves were not consistent over time, confirming the need to include a standard curve each run rather than using a single curve for multiple plates. Results suggest that the concentration and detection methods used must be validated for the specific water matrix using a recovery control to assess performance over time.