Building-level wastewater surveillance of SARS-CoV-2 is associated with transmission and variant trends in a university setting

Abstract

The University of South Carolina (UofSC) was among the first universities to include building-level wastewater surveillance of SARS-CoV-2 to complement clinical testing during its reopening in the Fall 2020 semester. In the Spring 2021 semester, 24h composite wastewater samples were collected twice per week from 10 residence halls and the on-campus student isolation and quarantine building. The isolation and quarantine building served as a positive control site. The wastewater was analyzed using RT-ddPCR for the quantification of nucleocapsid genes (N1 and N2) to identify viral transmission trends within residence halls. Log₁₀ SARS-CoV-2 RNA concentrations were compared to both new clinical cases identified in the days following wastewater collection and recovered cases returning to sites during the days preceding sample collection to test temporal and spatial associations. There was a statistically significant positive relationship between the number of cases reported from the sites during the seven-day period following wastewater sampling and the log₁₀ viral RNA copies/L (overall IRR 1.08 (1.02, 1.16) p-value 0.0126). Additionally, a statistically significant positive relationship was identified between the number of cases returning to the residence halls after completing isolation during the seven-day period preceding wastewater sampling and the log₁₀ viral RNA copies/L (overall 1.09 (1.01, 1.17) p-value 0.0222). The statistical significance of both identified cases and recovered return cases on log₁₀ viral RNA copies/L in wastewater indicates the importance of including both types of clinical data in wastewater-based epidemiology (WBE) research. Genetic mutations associated with variants of concern (VOCs) were also monitored. The emergence of the Alpha variant on campus was identified, which contributed to the second wave of COVID-19 cases at UofSC. The study was able to identify sub-community transmission hotspots for targeted intervention in real-time, making WBE cost-effective and creating less of a burden on the general public compared to repeated individual testing methods.

Keywords: Building-level surveillance; COVID-19; SARS-CoV-2; University; Variants of concern; Wastewater-based epidemiology.

Graphical abstract

Fig. 1

Log₁₀ SARS-CoV-2 RNA copies per liter of wastewater for individual residence halls. Individual residence halls are represented by different colors in the stacked area plot to observe spatial and temporal trends of wastewater viral RNA concentration from January 6, 2021, to April 26, 2021. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2

(A) The log₁₀ SARS-CoV-2 RNA copies per liter of wastewater (black line, left axis) and the number of identified COVID-19 cases (grey area, right axis) identified in the Site 1 residence hall from January 6 to April 26, 2021. (B) To assess site-specific associations, a negative binomial mixed regression model was fit with the number of cases detected in the next 7 day at Site 1 as the response and the wastewater RNA copies/L as the predictor. Site-specific random intercepts and random slopes were included, along with an offset term for the residence hall occupancy at the beginning of the semester (Overall IRR 1.08 (1.03, 1.16), p-value 0.0126). Site 1 is shown as an example; see Fig. S6 for trends observed in Sites 3–11.

Fig. 3

(A) The log₁₀ SARS-CoV-2 RNA copies per liter of wastewater (black line, left axis) and the number of recovered COVID-19 cases (grey area, right axis) returning to the Site 1 residence hall from December 25, 2020, to April 26, 2021. (B) A negative binomial mixed regression model was fit with the number of priorly infected individuals who returned to Site 1 in the seven days before wastewater sampling as the response and the wastewater RNA copies/L as the predictor. Site-specific random intercepts and random slopes were included, along with an offset term for the residence hall occupancy at the beginning of the semester (overall IRR 1.09 (1.01, 1.17) p-value 0.0222). Site 1 is shown as an example; see Fig. S7 for trends observed in Sites 3–11.

Fig. 4

SARS-CoV-2 RNA concentration in isolation building wastewater and COVID-19 Cases. (A) Site 13 log₁₀ SARS-CoV-2 RNA copies per liter of wastewater is represented by the black line, and the number of confirmed positive occupants in the building is represented by the grey area plot. (B) A binomial regression model with the logistic link function was used to assess the association between the number of confirmed positive individuals in the building and the log₁₀ SARS-CoV-2 RNA copies per liter of wastewater. The number of occupants of the building (including both confirmed cases and close contacts without a positive test) was used as the number of trials for the binomial model (odds ratio 1.26 (1.13, 1.40), p-value <0.0001).

Fig. 5

Heatmaps showing the percentage of Variants of Concern (VOCs) found in positive wastewater samples. Sites are shown for each tested mutation. Dark orange shows that the wastewater sample at that time point consisted of 100% mutation and 0% wild-type while the green color represents 100% wild-type and 0% mutation. The 69–70 del mutation is associated with the Alpha strain. The E484K mutation is associated with Beta and Gamma strains. The K417N mutation is associated with the Beta strain. No VOCs were detected in Sites 7 and 8, so these sites are not included in the figure. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)