Improved passive sampling methods for wastewater to enable more sensitive detection of SARS-CoV-2 and its variants.

Wastewater-based epidemiology (WBE) can be used as a part of a long-term strategy for detecting and responding rapidly to new outbreaks of infectious disease in the community. However, wastewater collected by grab samples may miss marker presence, and composite auto-sampling throughout a day is technically challenging and costly. Tampon swabs can be used as passive collectors of wastewater markers over hours, but recovery of the captured markers is a challenge. Our goal was to improve tampon elution methods for virus detection and variant analysis to increase the likelihood of detection near the Limit of Detection (LOD) and to potentially detect new or rare variants in a new outbreak. Counts of SARS-CoV-2 N1 and N2 markers in grab samples were compared to markers eluted from tampons that had been immersed in 3 sewersheds for 4-6 h during June to December 2023. We compared tampon elution methods that used different elution volumes, pressure, and amounts of Tween 20, evaluated after automated magnetic bead purification and RT-ddPCR of SARS-CoV-2 markers. Overall, method "SwabM2" in which tampons were eluted by high pressure squeeze in a 50 mL syringe after adding 2 mL of 0.5 X TE + 0.075 % Tween-20 yielded a median four-fold higher concentration of final purified SARS-CoV-2 markers than paired grab samples and significantly more than other tested tampon elution methods (p < 0.0001). Method SwabM2 was more likely to yield enough extracted nucleic acids for sequencing and also gave higher quality variant sequences than two other tampon elution methods. Variant analysis captured the Fall 2023 transition of variants from XBB to JN and "H" lineages. In summary, we demonstrated a tampon-based wastewater collecting and elution method that yielded higher counts, more detections near the LOD, and higher quality variant sequences compared to both grab samples and other tampon-based passive-collecting wastewater methods.