Figure 1

Figure 1. Spread of SARS-CoV-2 Omicron variant in community wastewater samples, Alberta, Canada, November 2021–January 2022. A, B) Percentage of Omicron RNA detected in community wastewater samples (data lines) compared with the...

Wastewater was collected from municipal treatment plants across the province as 24-hour composite samples up to 3 times per week. We isolated RNA from wastewater by using either ultrafiltration followed by RNA extraction (5), which was used to process 233 samples, or affinity binding columns that purify nucleic acids directly (21), used to process 209 samples (Figure 1). We applied the same method consistently at a given sampling site throughout the entire study period. We processed wastewater samples from 3 geographically disparate treatment plants in Calgary, Fort McMurray, and Lethbridge, comprising 11% of all samples in the study, by using both methods for comparison and revealed no significant difference (Mann-Whitney test: p = 0.46 [Calgary], p = 0.39 [Fort McMurray], and p = 0.59 [Lethbridge]) (Appendix Figure 1).

RNA quantification by qRT-PCR incorporated a newly designed set of assays that selectively amplify the BA.1 Omicron variant or the B.1.617.2 Delta variant by targeting mutations R203K/G204R and R203M in the N200 region of the nucleocapsid gene (M. Fuzzen et al., unpub. data, https://www.medrxiv.org/content/10.1101/2022.04.12.22273761v1). The R203K/G204R mutation in the BA.1 Omicron variant is also present in the B.1.1.7 Alpha variant. Clinical screening of cases indicated that the Alpha variant was no longer detected in Alberta as of July 2021 (22). We confirmed this finding by screening the wastewater samples from this study by using a separate assay that targets a D3L mutation in the nucleocapsid gene of the Alpha variant (23). We quantified total SARS-CoV-2 levels separately with widely used universal assays targeting the N1 and N2 regions of the nucleocapsid gene in the wild-type virus (3,4) and all other VOCs identified to date. We triplexed N200 assays for Omicron, Delta, and total SARS-CoV-2 together so we could estimate an Omicron-to-Delta ratio in each wastewater sample using the Omicron signal (R203K-G204R assay) and the Delta signal (R203M assay) (M. Fuzzen et al., unpub. data). This technique enabled us to track the emergence and prevalence of Omicron at the population level throughout the province.

Wastewater sampling and sample processing followed by identifying and quantifying SARS-CoV-2 is intrinsically more complicated than conducting the same PCR strategy on clinical samples (i.e., nasopharyngeal swabs). Directly comparing results between different treatment plants is not normally recommended because of intrinsic heterogeneities (e.g., physiochemical differences manifesting different PCR inhibition potential; different proportions of urban, industrial, and agricultural inputs to urban wastewater; different flow rates and distances affecting signal degradation) (24). Other factors, such as population movement between sewershed catchments, can also influence results. These limitations apply to total SARS-CoV-2 quantification and have led to evaluating different population normalization markers in wastewater sample analysis (25,26). However, the approach presented in this study for determining Omicron-to-Delta ratios within the same multiplex qRT-PCR reaction overcomes these issues, because RNA genomes derived from either variant are expected to react similarly to these factors. In this regard, we confirmed that subsets of samples with 100% Omicron showed good correlation between the R203K-G204R (Omicron) and N1 (total SARS-CoV-2) qRT-PCRs (Appendix Figure 1). We collected daily numbers of new cases of COVID-19 clinically diagnosed across the province by positive PCR test from the Data Analytics branch of Alberta Health Services and reported them using a 7-day rolling average (Figure 1).

Figure 2

Figure 2. Number of days required for the SARS-CoV-2 Omicron-to-Delta variant ratio to pass thresholds of 10% (A), 50% (B), and 90% (C) of community COVID-19 burden, Alberta, Canada, November 2021–January 2022....

Omicron variant SARS-CoV-2 was first detected in Alberta community wastewater during late November and early December (corresponding to the displacement of the Delta variant) (Figure 1; Appendix Table). In Calgary, 4 consecutive samples collected during December 5–9 revealed the sustained presence of 3%–9% Omicron (compared with >90% Delta) among infected persons contributing to the sewershed in this cosmopolitan city of 1.3 million. Omicron was first detected in wastewater in the capital city of Edmonton (population 1.1 million) on December 10 (15% Omicron, 85% Delta). The rate of increase of Omicron in the international resort town of Banff was higher than in larger cities such as Calgary and Edmonton (Figure 1, panel A), and Omicron surpassed 80% in samples taken 3 times a week during December 20–23 (Appendix Table). By that time, Calgary and Edmonton had just passed 50% Omicron, and the proportion of Omicron infections was growing in smaller bedroom communities adjacent to these 2 large urban centers (e.g., Okotoks, High River, Strathmore, and especially Airdrie, which are all <70 km from Calgary) (Figure 1; Appendix Figure 2). Communities that experienced the most delayed emergence of Omicron were smaller and more remote; Brooks (population 14,451, 190 km from Calgary) and Taber (population 19,070, 263 km from Calgary) did not reach high proportions of Omicron until late December (Appendix Table).

The general trend demonstrated by this analysis of objective wastewater evidence is that large cities experienced the emergence of a newly introduced virus before smaller centers farther away from cities, but with notable exceptions. Banff, located 127 kilometers west of Calgary, experienced a more rapid onset of Omicron infection than anywhere else in the province despite its resident population (13,427 persons) being <1% of the Calgary population and the smallest among monitored communities (Appendix Figure 2, panel B). Banff is an international resort community in Banff National Park, Canada’s busiest national park, which attracts >4 million visitors annually from around the world (27). Early detection of Omicron in Banff might correspond to attracting tourists at the onset of the ski season in November and December. Of note, the nearby and slightly larger mountain town of Canmore (population 27,664), located 105 kilometers west of Calgary (22 km east of Banff and outside the national park), experienced a much later emergence of Omicron infections. This delay is likely related to Canmore hosting fewer international tourists than Banff and featuring less high-density dormitory-style housing, where much of the worker population supporting Banff's tourism industry resides.

More remote communities located a greater distance away from Alberta’s large international airports exhibited later emergence of the Omicron variant (Figure 1; Appendix Figure 2). The Calgary International Airport serves 16 million travelers per year with direct flights arriving from 15 countries (28), compared with 8 million travelers and 6 countries for Edmonton International Airport (29). Plotting Omicron dynamics in Alberta municipalities as a function of distance from Calgary (Appendix Figure 2, panel C) suggests a link to international travel whereby incoming travelers introduce a new virus into a large densely populated urban center, enabling its spread. International travel in and out of Alberta increased sharply in November and December; 22,700 passengers came through Alberta in November and 28,800 in December, compared with only 8,400 travelers during the first 10 months of 2021 combined (30).

Fort McMurray offers an interesting example in relation to domestic air travel. Despite being a remote, relatively small (population 79,205) northern community farther from Calgary than any other municipality sampled in this study, Fort McMurray exhibited an Omicron emergence comparable to the rapid onset in Calgary. Fort McMurray has one of the busiest airports in Canada to accommodate shift workers commuting from across the country to work in the oil sands industry (31). This high level of contact with other parts of Canada is likely to result in rapid introduction of an emerging virus such as the Omicron variant. Workers traveling to Fort McMurray from other provinces or from major urban centers in Alberta likely contributed to accelerated Omicron emergence relative to other smaller or remotely situated Alberta municipalities.

The rapid emergence of Omicron in Calgary, Edmonton, Banff, and Fort McMurray is especially evident when this variant comprised lower proportions of the SARS-CoV-2 community burden. The time required for these 4 communities to surpass 10% Omicron was on average (+ SEM) 10.9 + 2.0 days faster than the other communities, highlighting the significantly earlier emergence of Omicron in these locations (p<0.0001 by unpaired t-test) (Figure 2, panel A). These significant differences are maintained by using Omicron cut-offs of 50% (6.3 + 2.7 days faster; p = 0.0331), and the trend is similar at 90% (5.0 + 2.6 days faster; p = 0.0686) (Figure 2, panels B and C). Later emergence of Omicron infections in the less populated outlying communities of Taber, Cold Lake, Lacombe, and Brooks is clearly evident using the 50% cutoff (Figure 2, panel B). In these 4 locations, Omicron surpassed Delta on average 9.8 + 1.8 days later than the other communities, highlighting the significantly slower emergence of infection in these smaller, more remote settings (p<0.0001). Some evidence suggests lower adherence to COVID-19 public health interventions in rural settings than in urban settings, including in Alberta (32,33), but less densely populated remote areas can experience slower spread of SARS-CoV-2 (34,35) because of less frequent interaction events, which could potentially contribute to the patterns we report in this study.

Wastewater results demonstrate that the emergence of Omicron was the driver of clinical cases increasing in December and January during Alberta’s fifth wave (Figure 1, panels A, B). During this time, COVID-19 public health surveillance shifted to much more focused PCR testing that prioritized essential workers, patients at risk for severe illness and eligible for early treatment, and patients in emergency departments with more serious illness (36). This shift resulted in PCR testing dramatically underestimating total disease burden in the population relative to earlier waves. Reported clinical cases still show a steep increase after the emergence and propagation of Omicron revealed by wastewater testing (Figure 1, panels A, B). These dynamics mirror the shift from Delta to Omicron in Alberta, confirmed clinically by screening subsets of samples using PCR for VOCs and genome sequencing, which revealed levels of Omicron to be >50% by December 16 and >95% by December 28 (22). This finding demonstrates that wastewater surveillance reliably provides information vital to public health officials.

VOC information derived from viral genome sequencing of clinical samples in Alberta is nonrandom (which is also the case in many other jurisdictions), placing emphasis on clinical cases of interest (e.g., outbreaks, hospitalizations) or incoming international travelers (37). Similarly, clinical PCR testing is susceptible to changes in testing policies, capacity limitations, or persons not getting tested (e.g., by personal choice or when infections are asymptomatic) (38). Wastewater testing offers an unbiased representation of disease prevalence, capturing all persons and groups contributing to the sewershed. This comprehensive coverage can be achieved for a tiny fraction of the cost of clinical testing on a per capita basis (39). In large cities such as Calgary and Edmonton, which have >1 million residents (Figure 2; Appendix Figure 1), monitoring wastewater for COVID-19 community burden costs only a few cents per person per year (based on testing 3 times per week in Alberta) and can provide objective information about community infection to public health authorities, policy makers, and the public in near-real time. COVID-19 clinical testing strategies and resources are becoming more targeted with jurisdictions such as Alberta turning to self-testing and less frequent public reporting. Wastewater monitoring offers an objective population-based surveillance metric of disease burden that continues to deliver real-time information on COVID-19 and could potentially be adapted for other emerging pathogens.