Wastewater Surveillance for SARS-CoV-2 RNA in Canada

A particularly popular application of wastewater-based COVID-19 surveillance has been on university and college campuses, primarily in U.S States, where academic laboratories engaged in this research had the opportunity to test their monitoring methods in well-controlled environments where clinical testing was often regular and comprehensive. These circumstances facilitated proficient comparisons between the two types of surveillance, demonstrated the potential value of wastewater’s early warning capability, and provided estimates of analytical sensitivity and biomarker load on a per case (capita) basis. Harris-Lovett et al. (2021) describe a consortium of 25 U.S. educational institutions that performed on-campus wastewater-based COVID-19 surveillance. An early publication by Betancourt et al. (2021) reported successful detection of COVID-19 cases in student residences at the University of Arizona (Tucson, 47 000 students) by means of monitoring sewers serving residences during late August (start of the fall semester) of 2020. Wastewater detection of SARS-CoV-2 RNA led to three students being confirmed positive by clinical testing who were relocated and quarantined. For this application the testing provided an 79.8% positive predictive value and an 88.6% negative predictive value. The latter value is important, often going unreported, and indicates that false positives (wastewater detection without accompanying recorded case detection), at least in this particular context, were found to be fairly rare.

Gibas et al. (2021) reported details about experience with practical realities and logistics for operating an on-campus student residence sampling network at the University of North Carolina (Charlotte, 24 000 undergraduate students). They used 19 on-campus sewer sampling sites to monitor 17 student residences, with sampling three times per week from late September to late November 2020, with rapid laboratory turn around within 26 to 30 hours of sample receipt allowing for residence lockdown within 36 hours of sample collection. Gibas et al. (2021) noted that this program tested 332 sewer samples (out of 475 samples attempted) over eight weeks vs. testing 3 000 students, three times per week for a total of 72 000 clinical samples. They concluded this program could detect one asymptomatic case in a resident student population of 150 to 200.

Karthikeyan et al. (2021) undertook a major campus monitoring program at the University of California (San Diego, 9 700 undergraduate and graduate students living on campus) during the fall 2020 term. They used a large-scale GIS (geographic information systems)-enabled building-level wastewater monitoring system associated with the on-campus residences of 7 614 individuals using 68 automated samplers to monitor 239 campus buildings. They developed an extremely rapid turn-around on wastewater analyses using an automated, high-throughput wastewater processing pipeline with capacity of processing 96 wastewater samples in 4.5 h. This program benefited from a requirement for all students to undergo clinical testing every other week. Over the period from 23 November to 31 December 2020, 59 cases were diagnosed among on-campus students residing in buildings monitored by the wastewater program and 84.5% (n=50) of these individual case diagnoses were preceded by positive wastewater samples (either in the days prior or the day of diagnostic testing). The monitoring program was judged to be able to detect a single asymptomatic case in a building of 415 residents.

Brooks et al. (2021) performed sewer sampling at a residential college in Maine based on weekly grab samples of sewers serving residences housing 605 students and twice weekly 24 h composite sampling from sewage lift stations spanning a 13-week period from late August to late November 2020. A sewage lift station is a facility that pumps wastewater from a lower elevation to a higher elevation. They identified two COVID-19 outbreaks among resident students with 76% of cases identified in the weekly grab samples (< 7 d) before clinical confirmation of cases.

Scott et al. (2021) performed sewer sampling at Tulane University (New Orleans, with over 14 000 students) from the middle of August to the end of November 2020 with weekly grab samples from nine sewer locations yielding 117 samples over the duration of surveillance. Wastewater surveillance provided complementary data to document an outbreak in student residences in early November, but weekly sampling was likely too infrequent to provide a clear early warning.

Reeves et al. (2021) performed sewer sampling at 20 locations on the University of Colorado (Boulder, 30 000 undergraduate students) campus from late August to late November 2020 collecting a total of 1 512 samples. They considered six possible scenarios for their sampling regime in which sewer samples found SARS-CoV-2 RNA to be: (1) absent, (2) low and stable, (3) low and increasing, (4) high and increasing/stable, (5) high and decreasing, and (6) decreasing to absent. Reeves et al. (2021) concluded that sewer sampling could provide an early warning in scenarios (1), (2) and (3). Specifically, detection of increasing concentrations during the first two weeks of sampling led to public health officials contacting residents and employees of identified buildings within 12 h of result reporting directing them to submit saliva samples for testing. This action revealed individual cases not identified by a routine clinical testing program.

Wang et al. (2022) performed wastewater COVID-19 surveillance on the campus of Emory University (total student population over 15 000) from the middle of July 2020 to the middle of March 2021 using weekly Moore swab (Liu et al. 2022) samples from 25 sewer sites serving student residences. They found that weekly sampling using Moore swab sampling was not sensitive enough (only 6 of 63 times) to reliably detect one or two sporadic cases in a residence building, but during the spring 2021 semester SARS-CoV-2 RNA was detected in wastewater from most student residences from one to two weeks before COVID-19 cases grew rapidly on campus. Liu et al. (2022) provided a detailed assessment of the advantages, limitations and costs of the Moore swab sampling approach that was used.

In Canada, Corchis-Scott et al. (2022) performed a sewer surveillance program for a student residence at the University of Windsor (normal student population of over 16 000) initially for seven weeks from early February to late March 2021 based on three samples per week using a grab sample that yielded negative results. This was modified to using passive sampling with a modified Moore swab that provided a positive detection of SARS-CoV-2 RNA only two days after starting this sampling approach. A feminine hygiene tampon was used as an absorbent swab to collect an integrated sample. The swab was suspended in the sewer for about 20 h at a time. The detection was confirmed to be the Alpha (B1.1.7; an emerging and not widespread VOC at the time) variant using a variant-specific assay (Section 4.5) leading to a case finding program the following day that confirmed two cases among 200 students tested. The confirmed cases were quarantined to a separate residence and an on-campus outbreak was likely averted. Wastewater surveillance has been performed on other Canadian universities (e.g., University of Guelph) with results posted on public-facing websites and in other cases with reporting to administration and public health agencies (e.g., University of Waterloo, University of Toronto, University of British Columbia)

Wastewater surveillance of healthcare facilities has been widely practiced during the pandemic, but generally with a different purpose for active treatment hospital facilities than for long-term care facilities (LTCF)/nursing homes, because COVID-19 cases being treated are expected in the former while COVID-19 cases must be avoided in the latter with its highly vulnerable aged population. Canada’s COVID-19 death toll is strongly influenced by early deaths that occurred in LTCF. As of December 2021, LTCF residents accounted for 3% of all COVID-19 cases and 43% of COVID-19 deaths in Canada (CIHI 2021).

Colosi et al. (2021) undertook a proof-of-concept study for wastewater SARS-CoV-2 surveillance at a LTCF with residential complexes housing 105 and 66 occupants. Wastewater surveillance results were validated first using hospital wastewater known to contain SARS-CoV-2 RNA then against clinical testing of LTCF residents for an 8-week period. Across all hospital and LTCF samples collected after methods validation, they obtained 25 true positives, 0 false positives, 9 true negatives, and 1 apparent false negative, yielding an apparent sensitivity of 96.2% and an apparent specificity of 100%. Given an intent of detecting possible COVID-19 cases entering a LTCF, Colosi et al. (2021) noted a concern that their surveillance could not distinguish new cases from convalescent patients previously infected who were still shedding the virus. From this perspective, if convalescent virus shedding was considered to be a false positive, sensitivity was 100%, but specificity was only 45%. This highlights the importance and the need of establishing baselines and adequate sampling frequency to be able to distinguish new cases (high shedding) from convalescent shedding. Work towards this goal is continually evolving and a clearer path has been recently elucidated by Welling et al. (2022) who found that two consecutive day detections in wastewater is most predictive of case detection in the context of building-level surveillance.

Xu et al. (2021) validated their methodology for detecting SARS-CoV-2 RNA in wastewater from a Hong Kong hospital treating COVID-19 patients. Gonçalves et al. (2021) studied wastewater from a small hospital in Ljubljana, Slovenia, sampling one composite wastewater sample per day from June 1 to 15, 2020, starting before any patients with COVID-19 were being treated (it went from 1 up to 4 COVID-19 patients during the 2-week sampling period). They found that they could detect SARS-CoV-2 RNA when only one COVID-19 patient was being treated. The Jørgensen et al. (2020) study at a Danish hospital estimated being able to detect a COVID-19 prevalence rate as low as a 0.02%–0.1% (i.e., between two virus-shedders per 10 000 persons and one virus shedder per 1000).

In Canada, Acosta et al. (2021) studied wastewater directly from three tertiary-care Calgary hospitals with a combined total of more than 2 100 in-patient beds between August and December 2020. Tertiary-care refers to a hospital providing specialized medical treatment referred from primary and secondary health care providers. They noted that many hospitalized COVID-19 patients and certainly those with severe enough symptoms to require intensive care would not be using toilet facilities feeding the sewer. Rather, they would be diapered and faecal wastes would be handled as medical biohazard waste. Accordingly, this approach was sensitive to the detection of new hospital-acquired infections, revealed by wastewater data. This is despite studies suggesting that 52%-56% of hospital employees avoid defecating while at work in a hospital, undermining the ability of sewer sampling to monitor staff for COVID-19 infection. Toileting patterns are thus a fundamental factor in determining the ability of sewer sampling to monitor staff at any facility for COVID-19 infection making behavioural factors an important consideration in study design.

Other outbreak risk locations that might offer promise for future wastewater surveillance included industrial plants, most notably meat processing facilities, and prisons. Despite that potential, published reports of implementing wastewater surveillance at such sites are scarce, owing to facility operators choosing to keep such information confidential. Dyal et al. (2020) reported that by the end of April 2020, the U.S. had 115 meat or poultry processing plants in 19 states report COVID-19 outbreaks accounting for 4 913 cases and 20 confirmed deaths. Alberta Health Services (AHS 2021) reported that Alberta had experienced COVID-19 outbreaks at five meat and poultry facilities between April and November 2020. Pokora et al. (2021) performed a cross-sectional epidemiology study to determine risk factors for COVID-19 infection at 22 German plants totalling over 19 000 employees of which seven plants with more than 10 COVID-19 cases had a disease prevalence of 12.1%.

Piché et al. (2022) summarized clinical cases in Canadian correctional facilities between 1 December 2021 and 28 February 2022 when there were over 12 000 newly reported COVID-19 cases (8375 prisoners and 3961 staff). These cases during this time frame correspond to 55% of the total cases in these provincial/territorial and federal correctional institutions since the onset of the pandemic. Arora et al. (2020) reported on wastewater surveillance in Jaipur, India that included a WWTP serving the city centre that showed a hotspot in May 2020 that was attributed to a jail being located in the sewershed serviced, but no details were provided to focus on wastewater from the jail.

Recognition that international transmission of COVID-19 was enhanced by travel led to a number of international travel bans early in the pandemic. This reality prompted a few investigations into the potential for wastewater surveillance as a complementary data source to evaluate risk for COVID-19 transmission by disembarking passengers from international travel modes. Ahmed et al. (2020b) evaluated two wastewater samples (one influent, one treated) from a cruise ship, with only crew onboard about a month after passengers had disembarked (estimated that 24 passengers may have been infected with COVID-19) and three wastewater samples from separate international flights: Los Angeles to Brisbane (117 passengers, 26 April), Hong Kong to Brisbane (19 passengers, 7 May) and New Delhi to Sydney (185 passengers, 10 May). The results of this pilot study that evaluated a number of sample preparation and analytical approaches documented many of the challenges (e.g., not every passenger on a flight defecates during the flight, a high proportion of paper in airplane wastewater) facing wastewater surveillance of these sources. Only the cruise ship influent wastewater sample and wastewater from the first flight provided consistently positive results. At the time of this study, it was estimated that over 60% of COVID-19 cases in Australia were infected overseas and it had restricted international air travel for non-Australians since late March 2020. Albastaki et al. (2021) studied wastewater from 198 incoming aircraft from 59 airports on all six continents at Dubai International Airport, United Arab Emirates before September 2020, finding that 13.6% had positive signals for SARS-CoV-2.

Ahmed et al. (2020b) performed a follow-up study on wastewater from 37 long distance charter flights to Darwin, Australia, arranged for re-patriating Australians from overseas between the middle of December 2020 and the end of March 2021. All passengers were quarantined for 14-day post arrival during which clinical testing identified 112 cases of COVID-19. Wastewater from 24 (64.9%) of the flights tested positive for SARS-CoV-2 RNA and results demonstrated a positive predictive value (PPV) of 87.5% and a negative predictive value (NPV) of 76.9%, suggesting flight-wide surveillance results complementary to clinical testing.

Ahmed et al. (2022a) collected an aircraft wastewater sample from a November 25 flight from Johannesburg, South Africa to Darwin, Australia and retrospectively, detected RNA fragments of the Omicron variant. Omicron was declared a VOC by the WHO the day after, on 26 November 2021 and caused massive waves of COVID-19 infection around the world over the following months. Although all passengers on this flight were tested prior to boarding and after disembarking, a single passenger was confirmed on 29 November to be infected by the Omicron VOC by genetic sequencing of a clinical swab sample collected after arrival. Agarwal et al. (2022) reported detection of the Omicron VOC in wastewater samples collected from wastewater on 2 and 23 of November from the Frankfurt International Airport as well as at the Frankfurt city WWTP.

There is a wide range of different applications of wastewater surveillance that can be applied to institutions. We have provided only a relevant sampling of what has been described in accessible publications that should provide some perspective on what is possible and what challenges need to be overcome.

Community-scale monitoring programs at municipal wastewater treatment facilities are often designed to capture trends, so these applications necessarily need to be able to estimate SARS-CoV-2 RNA concentrations in the wastewater over time. To achieve a representative sample, community wastewater is typically collected through deployment of industrial autosamplers that collect a time- or flow-weighted composite sample over 24 h (Fig. 6). Depending on available resources and the intended purpose of the surveillance program, these composite samples can be collected at a wide range of frequencies. Initial programs aimed at proof of concept sampled from random daily to bi-weekly, but it has become clear at least three times per week (CDC 2022a) is required to track COVID-19 trends and offer any realistic chance of providing early warning for public health decision-makers and the public.

For the most common surveillance purpose of tracking trends of SARS-CoV-2 in wastewater or detecting its emergence in communities with low COVID-19 prevalence, sampling at WWTPs or centrally located pumping stations is the most widely deployed approach. Because SARS-CoV-2 RNA and (or) viral particles can be excreted in faeces by infected individuals, and has been shown by several groups to preferentially partition to the solids/colloidal phase of wastewater (D’Aoust et al. 2021a; Graham et al. 2021; Peccia et al. 2020), sample preparation often focuses on concentrating solids from pre-treated “raw” wastewater. Some researchers have advocated sampling primary sludge (collected after the primary treatment process of sedimentation) to leverage the concentration of solids and maximize detection of SARS-CoV-2 (D’Aoust et al. 2021a, 2021b; Graham et al. 2021; Peccia et al. 2020). However, not all municipal wastewater treatment facilities necessarily have accessible sampling locations for primary sludge (e.g., lagoons). Moreover, an understanding of system-specific WWTP hydraulics and operational conditions is especially important when primary sludge is collected, because the age of the sludge (residence time distributions) and return flows in the system are important considerations for interpretation. The wastewater matrix of choice will have further implications for sample analysis strategies (e.g., PCR inhibition – see QA/QC section).

Sampling from a sewer network monitoring upstream of a WWTP has been pursued where there is an objective source to monitor wastewater for SARS-CoV-2 provided wastewater flows of sufficient depth can be sustained. Autosamplers have been deployed, with some consideration for winterization needs to avoid frozen sampling ports. Sewer sampling also encounters problems with the sampling device becoming fouled with extraneous items that are flushed down toilets despite warnings against such actions (paper, tampons, condoms). In some circumstances, where deploying a large autosampler is not feasible (due to depth of manhole), correlations from grab samples with composite samples can be established to provide estimates. Using a consistent time of day and time of week for grab sampling is an important consideration.

Further upstream at the facility level, some sewer access locations may require discretion where industrial autosamplers cannot be used (e.g., access in an office setting). In such cases, portable units have been developed and deployed (lightweight briefcase-sized autosampler from CEC Analytics or collecting in-building time-weighted composite samples). In many upstream sampling locations (and most facility level surveillance) where wastewater flows are more intermittent and cannot be sustained, passive samplers (Bivins et al. 2022; Habtewold et al. 2022; Hayes et al. 2021, 2022; Liu et al. 2022; Rafiee et al. 2021, Schang et al. 2021) have been deployed to provide qualitative (presence/absence) results. Consequently, the goal of passive surveillance in this context is typically for qualitative “first detection” rather than tracking trends quantitatively. Habtewold et al. (2022) and Hayes et al. (2021) evaluated different sorbent materials for use in passive samplers.

Access to upstream sewer locations generally requires additional determination of responsibilities and coordination with the responsible utility to ensure strategic sampling locations with respect to the corresponding population contributing to the sewershed, as well as the essential, rigorous implementation of safety precautions (e.g., confined space hazards, access through manhole in a roadway), and training of facility personnel to collect wastewater samples.

Regardless of the choice of PCR platform (discussed below), the sample processing step (i.e., extracting the RNA to be later probed by PCR, Fig. 7) contributes to most of the variability and uncertainty in SARS-CoV-2 concentration estimates from wastewater (Chik et al. 2021; Griffiths et al. 2021; Pecson et al. 2021).

To achieve optimal precision and sensitivity to track trends in the wastewater, laboratories undertaking wastewater surveillance activities across Canada have deployed a wide range of processing methods to extract sufficient quantities of RNA. Various physical-chemical approaches to obtain the RNA extract include chemical precipitation, affinity binding columns, filtration, sedimentation, and centrifugation. Accordingly, the degree of concentration achieved, the targeted fraction(s) of wastewater captured by the approach, and interaction effects depending on the specific wastewater matrix examined, contribute to substantial variability between different methods compared to the precision that may be achievable using a single method.

In the Ontario inter-laboratory program, split sample testing was conducted amongst participating laboratories every two months. Despite obvious method-specific biases, the level of intra-laboratory variability (i.e., precision of concentration estimates) achieved allowed participating laboratories to discern similar trends reflecting the quantitative differences between samples over multiple rounds (Chik personal communication 2022).

Following RNA extraction, the principles applied for detection and quantification of SARS-CoV-2 RNA in wastewater is consistent with PCR-based approaches used for clinical samples. PCR was invented almost 40 years ago, and the inventor shared the Nobel Prize for chemistry in 1993. This method has revolutionized clinical diagnostics and biotechnology. Application of PCR to water and wastewater and various environmental matrices has seen some commercial applications and has been central to many lines of research across multiple disciplines. As described above, RT-qPCR is in common use amongst laboratories engaged in wastewater surveillance. However, a major advance in the last ten years has been the advent of RT-digital droplet PCR (RT-ddPCR) technology and its more recent iteration RT-digital PCR (RT-dPCR) which have the potential to provide greater analytical sensitivity and precision. While all of these platforms typically rely on measurement of fluorescence being generated as the polymerase chain reaction proceeds, RT-qPCR converts its signal by comparing outputs to fluorescence signals from a series of PCR reactions with reference standards of known quantities of target nucleic acids, i.e., a calibration curve or “standard curve”. Inherently RT-qPCR facilitates relative quantification of the target of interest. On the other hand, d/dd-PCR relies on partitioning a sample into thousands of individual reactions. The occurrence or absence of a fluorescence signal in each partition after PCR cycling is used to facilitate absolute quantification of the analyte and therefore precludes the need for a calibration curve. While d-/dd-PCR may afford better precision of concentration estimates at low-levels of SARS-CoV-2 in wastewater by eliminating systematic biases attributable to RT-qPCR and the need for standard calibration, RT-qPCR is the more established technology, and these platforms are widely deployed in laboratories across Canada (specifically probe-based methods which can provide superior analytical sensitivity and specificity). RT-qPCR platforms generally also allow for greater sample throughput.

Multiple processing and analysis pipelines exist across different laboratories and even within laboratories. Until the variability between methods can be reduced or compensated appreciably, site-specific surveillance activity for monitoring trends at a given location should be conducted by a single laboratory using a single method (Chik et al. 2021; Griffiths et al. 2021; Pecson et al. 2021). For this reason, it is imperative that sample analyses not be randomly transferred between laboratories, nor should quantitative results between different laboratories (or different sample processing methods applied within the same laboratory) be assumed to be equivalent or aligned. Sample storage conditions can also affect quantitative results, such that storage should be consistent and any storage effects ascertained. Site-specific methodological consistency still enables longitudinal data sets for a given location to be established, which is arguably one of the most important uses of wastewater surveillance because it enables trend analysis for a given community to provide evidence that can inform public health actions (see Section 5).

Imposing a standardized, common sample processing approach would certainly help reduce variability between different laboratories processing wastewater samples for SARS-CoV-2 surveillance. In retrospect this would have been difficult to achieve in Canada over the past two years. As investigators faced multiple challenges responding to rapidly changing conditions there was a disincentive for adopting new methods. In fact, given the bottom-up, grassroots approach that has made this venture successful across the country, the lack of standardization may have been a factor that contributed to the rapid emergence of testing programs in Canada. Progress towards standardization is desirable although the path for this is not yet clear. Reliance on existing laboratory infrastructure and limitations of supply chain issues encountered earlier on during the pandemic present obstacles. In Québec, a locally standardized protocol was shared by key laboratories, facilitating troubleshooting efforts when an unusual result was encountered. In Alberta and British Columbia, provincial public health laboratories and the National Microbiology Laboratory (Alberta Precision Laboratories albertaprecisionlabs.ca, B.C. Centre for Disease Control bccdc.ca, National Microbiology Laboratory of the Public Health Agency of Canada, canada.ca/en/public-health/programs/national-microbiology-laboratory.html) provide potential platforms for development and dissemination of standardized procedures as they have done historically for clinical microbiology procedures. However, in the broader Canadian context, practical geographical and jurisdictional constraints effectively preclude timely coordination of wastewater surveillance activities through a central, federal laboratory in Canada, such as is currently done in the Netherlands. RIVM, the National Institute for Public Health and the Environment laboratory monitors wastewater from 300 WWTPs, four times per week, covering 17 million Dutch residents (rivm.nl/en). These constraints have largely dictated the patchwork of methods that are currently deployed for wastewater surveillance applications in Canada. Analogous issues are faced in the United States and reportedly presented a challenge with the Centers for Disease Control and Prevention (CDC) delaying the roll out of clinical testing protocols for COVID-19 in the early stages of the pandemic (Evans and Clayton 2020).

Despite there being no single standard method for quantification of SARS-CoV-2 biomarkers in wastewater in the short-term, longitudinal application of a consistent method by a qualified laboratory will yield useful quantitative estimates to facilitate tracking of temporal trends, provided that stringent QA/QC procedures are followed (e.g., Ahmed et al. 2022c, Chik et al 2021, MECP 2022). Given the predominance of RT-qPCR platforms deployed for wastewater surveillance across Canada and other parts of the world, a key focus has been on ensuring the quality of the standard calibration curve and streamlining best practices for the choice of standard materials.

Another area of focus is ascertaining whether PCR (an enzyme-dependent method) is inhibited by physical and chemical characteristics of the wastewater matrix, leading to underestimated levels or false negatives. However, QA/QC procedures for these methods were not widely standardized before the COVID-19 pandemic and few operational definitions have been adopted among laboratories. These realities have complicated efforts to bring the datasets together in a comparable manner. In Ontario, to support the range of process pipelines used (which generally consist of sampling, concentration, extraction, detection, and normalization) that are deployed amongst 13 academic institutions conducting surveillance across the province (see case studies in Part 1 of the Supplementary Information), Ontario Clean Water Agency led the development of technical guidance (MECP 2022) in collaboration with international experts and stakeholders to establish minimum performance expectations based on streamlined operational definitions and the current state-of-the-knowledge. This process included recommendations for streamlining the choice of a certified quantified RNA standard material among laboratories in Ontario, and verification of assay sensitivity to identify SARS-CoV-2 variants of concern.

Given the importance of QA/QC, the implementation of quality management frameworks in the laboratories responsible for generating these data is underscored. Quality management frameworks provide checks and balances at various levels of laboratory operations to ensure the generation of data that are reliable and fit-for-purpose. However, a key challenge to implementing quality management frameworks broadly is the need to tailor them for the sectors relied upon for generating these data.

In industry, accreditation frameworks are used to verify that commercial laboratories have appropriate quality management systems and can perform the tests according to their scope of accreditation. This includes stringent requirements for data quality, documentation, personnel, infrastructure, and participation in inter-laboratory comparisons to demonstrate proficiency. However, early conversations with accreditation bodies during the pandemic suggested that they lack tailored accreditation checklists for PCR methods in environmental matrices, ostensibly due to lack of client demand (Alex Chik, personal communication, 2021). Coupled with the fact that results from these types of tests are not used in clinical diagnosis and do not report substances that might be directly harmful to public health (in contrast to other substances, e.g., heavy metals), there is no strong incentive from either government or industry to develop accreditation programs tailored to wastewater surveillance. That said, recent advances have started to address this gap: the Ontario technical guidance (MECP 2022) as developed and first published in August 2021, and the American Council of Independent Laboratories (ACIL) released a draft accreditation checklist (acil.org/news/597076/Wastewater-Surveillance-for-SARS-CoV-2.htm) in February 2022. The American Public Health Association (APHA) has also released additional guidance in March 2022, and an International Standards Organization (ISO) ongoing working group (ISO/TC 147/SC 4/WG 26, SARS-CoV-2 in wastewater - aphl.org/aboutAPHL/publications/Documents/EH-2022-SARSCoV2-Wastewater-Surveillance-Testing-Guide.pdf) has been formed.

The desire and opportunity for standardization must be balanced with a framework that allows flexibility, and alignment with client goals. The innovation cycle in the wastewater surveillance field is very short, thanks to significant interest (and accompanying investment) from both government and industry. Arguably, this has been aided by the lack of regulatory bodies and accreditation that can disincentivize rapid knowledge dissemination and translation. Academic laboratories are typically engaged in the ongoing development and optimization of new methods and technologies, rather than strict adherence to prescribed QA umbrellas regulating existing methods and technologies. University research groups generally employ personnel who lack professional designations (in contrast to certified medical laboratories). For these reasons, academic laboratories typically do not participate in existing accreditation frameworks for quality management. Consequently, surveillance programs dependent on predominantly academic laboratories may require greater oversight and emphasis on internal QA/QC policies and processes (e.g., coordination of inter-laboratory studies, split sampling, establishing minimum requirements for documentation, and reporting) by the entity administering the surveillance program. There are cases in Canada where laboratory staff and leads are cross appointed between academic institutions and provincial public health laboratories (e.g., Alberta, British Columbia). The current pandemic has emphasized the need for coordinated networks to ensure methodological advancements (often undertaken by academic laboratories) are consolidated in a manner that facilitates rapid adoption of best available practices for these tools in industry and public health laboratories that can scale-up to meet the needs of large-scale surveillance networks. Protocol changes should be undertaken cautiously to mitigate disruption to longitudinal data generation so that the ability to compare trends within a given surveillance program over time is maintained. In any case, this reality presents one of the challenges to achieving standardized methods across jurisdictions.

The use of wastewater surveillance data to establish trends was deemed “very feasible” by experts early on (WRF 2020). There is general confidence that significant changes in the SARS-CoV-2 RNA signal in a wastewater source can be tracked over time. These trends can provide a useful indication of the impacts of interventions implemented in the community served by the sewershed that is sampled. However, like other types of environmental data, wastewater surveillance data are inherently “noisy”. This means that establishing trends requires careful interpretation that considers variability in the data attributable to extraneous and (or) confounding factors.

To overcome the noise in the data and help elucidate trends, a range of approaches have been taken. Simple data smoothing techniques such as calculating moving averages to more complex smoothing algorithms have been applied on SARS-CoV-2 wastewater surveillance data (Arabzadeh et al. 2021; Ai et al. 2021), with and without normalization to adjust for variability attributable to the faecal content captured within a sample (e.g., D’Aoust et al. 2021a ; see section 4.4.3 below). Pileggi et al. (2022) presented a quantitative statistical linear trend analysis approach, based on recommendations by US CDC, to systematically use points of inflection to segment wastewater surveillance time series data with associated linear regression to establish whether or not an observed trend in a given segment was statistically significant. While there is not a universally accepted approach for trend analysis, it requires an important trade-off to be considered: approaches that are more intuitive (“manual” expert interpretation) are more time-consuming, whereas more systematic algorithms that are automated might not integrate other factors that can influence trend interpretation.

Since the Research Summit in April 2020, various studies globally have shown that SARS-CoV-2 wastewater signal fluctuations often trended with clinical case fluctuations in many systems. Although work is ongoing to examine whether wastewater surveillance data can be used to yield credible estimates of COVID-19 cases, there is a growing body of evidence that suggests wastewater surveillance trends can provide an unbiased estimate of changes to disease prevalence and spread in a community. This is important, as changes and trends at the community level have great value for informing public health officials and the public. Wastewater surveillance data is often interpreted alongside other conventional epidemiological metrics corresponding to population served by the sampled sewershed. A coordinated effort is required from municipalities (provision of sewershed boundaries) and public health units (e.g., clinical case testing, vaccination statistics) to facilitate exploration of these trends. Many public dashboards across Canada have been established during the COVID-19 pandemic (see case studies in Part 1 of the Supplementary Information), featuring wastewater signal data along with corresponding clinical testing data.

Just as with clinical samples, a PCR signal specific to SARS-CoV-2 RNA in wastewater does not report how much, if any, infectious virus is present; rather it only indicates the presence of small fragments of its genetic footprint. The detected “raw” RT-qPCR signal, regardless of type of specimen tested is a continuous variable generally expressed as a cycle threshold value (Ct; ranging from 1–40, where higher Ct indicates lower amounts of the target present in the sample). It is generally accepted that SARS-CoV-2 viral load in nasopharyngeal specimens is linearly correlated with infectious viral load (Puhac et al. 2022), and this is likely true of faecal shedding as well. Thus, individuals with high viral loads, on average will be expected to shed more SARS-CoV-2 RNA during the course of their infection. This analogue information (amount of RNA) is used by clinical laboratories in a standardized analysis process to render a binary outcome, namely, “positive” or “negative” calls. Throughout the COVID-19 pandemic, a RT-qPCR specimen that is “positive” has been designated an active SARS-CoV-2 case. In the clinical context, these cases are treated as discrete variables (i.e., counts) in public health reporting.

In contrast, a wastewater sample can be considered a pool of multiple clinical specimens with contributions of viral RNA shed from multiple infected individuals who are contributing to the sewershed and the composite wastewater samples being collected. At present, this pooled sample is viewed as extremely convoluted and cannot provide any insight into which individual (i.e., age, vaccination status, etc.) or how many individuals might be infected. The wastewater-based RT-qPCR signal is a continuous variable and at its core represents an estimate of the number of specific RNA fragments per volume of wastewater (i.e., a quantifiable concentration). This eponymous feature of RT-qPCR is achieved via detection of increasing fluorescence that varies directly with each PCR amplification cycle, resulting in exponentially increasing fluorescence that is expressed as Ct. Knowledge of the Ct value allows the original number of RNA fragments present in the sample to be estimated based on standard curves of reference. Wastewater surveillance via RT-qPCR thus provides quantitative information with a reasonable expectation that higher concentrations are indicative of a greater number of infected individuals excreting the virus into wastewater in that sewershed catchment area. However, in the absence of reasonable and relevant estimates of the amount of viral RNA excreted per person, wastewater-derived units expressed as equivalent COVID-19 cases could be grossly over- or under-estimating this parameter. There may be other tractable strategies to minimize this error (e.g., by adjusting the model using wastewater signal at a time period when clinical testing was high and case estimates were more accurate). Presently, however, the authors are of the opinion that there is no rationale or strong empirical evidence available that serves as a basis for translating the concentration of specific genetic fragments in wastewater into a number of infected individuals in a given sewershed. However, even if equivalent case numbers cannot be estimated with reasonable certainty, it is reasonable and valuable to interpret upward/downward trends in the wastewater signal as increase/decreases in the number of active cases in that community.

Any source of wastewater dilution will reduce the concentration of RNA being measured. Efforts to account for this dilution include characterization of different physical, chemical and (or) biological parameters to estimate the amount of excreta captured in a sample which varies with the size of the contributing population in a given sewershed. Municipal wastewater can be diluted by a variety of sources free of SARS-CoV-2, including: stormwater in municipalities where a sanitary sewer is not isolated from any stormwater inputs; groundwater infiltration into sewers (i.e., from rain and snow events), other sources of non-toilet wastewater (household greywater - as a component of municipal sewage greywater refers to all household water discharges through the sanitary sewer including that from laundry, bath, shower and kitchen. Li et al. (2022) have argued that laundry, bath, and shower wastewater that will contain sputum is as important as wastewater containing faecal sources because of the high concentration of SARS-CoV-2 in sputum) and other non-sanitary industrial and institutional discharges to a sewer network (all of which have both liquid and solid components that contribute to dilution or changes in organic load). These sources of dilution can vary over time from sample to sample, affecting the magnitude of the wastewater SARS-CoV-2 concentration estimates. Earlier studies on impacts of stormwater on pathogens in sewers have demonstrated how complex relationships can be (Tolouei et al 2019a, 2019b). Where that quantitative magnitude is being relied upon to demonstrate trends that are correlated with the number of COVID-19 cases in the population being sampled, variation in dilution over the period analyzed will weaken that correlation. For example, different waves of COVID-19 in Canada have coincided with the 2021 and 2022 spring seasons (when snowmelt enters the wastewater system in some municipalities). Concern around diluted signal (whether in liquid- or solids-based processing pipelines) has led to the evaluation and adoption of monitoring for a variety of substances in wastewater that can be interpreted as indicators of faecal contribution to wastewater composition. The rationale being that expressing SARS-CoV-2 signal as a proportion of the amount of a human biomarker (or multiple biomarkers) in a sample will correct for any non-human contributions to the amount of material used in the assay. This has the effect of increased precision of the signal estimates over time.

While there is no doubt that random dilution of wastewater with water that is free of SARS-CoV-2 will interfere with being able to track quantitative trends in excretion of SARS-CoV-2 by means of concentration measurements in wastewater, the solution to this issue is not as clear as is sometimes assumed. Normalizing RT-qPCR results from wastewater with a parameter that is known to represent the faecal content of wastewater has been investigated extensively by different teams. Candidate analytes have been drawn from a list of parameters that have been useful for tracking sanitary sewage discharges within receiving waters (Bivins et al. 2020; Jmaiff Blackstock et al. 2019) such as: Escherichia coli (faecal bacteria in all humans), BacteroidesHF183 (faecal bacteria in all humans), crAssphage (viral phage infecting Bacteroides and found in 50% of a large set of human faecal samples), pepper mild mottle virus (PMMoV, one of the most abundant RNA viruses present in human faeces), faecal sterols, bile acids (present in all human faeces), caffeine, common over-the-counter pharmaceuticals, common artificial sweeteners, common personal care products, and optical brighteners/fluorescent whiteners (indicative of laundry wastewater).

CrAssphage and PMMoV have been the most widely used faecal content markers by wastewater surveillance groups around the world. In Canada, some laboratories are using PMMoV normalization while others are not, with no consensus approach for how to account for dilution in either the liquid or solid fractions. When normalized to PMMoV concentrations, D’Aoust et al. (2021a) reported improved correlation between wastewater-based SARS-CoV-2 RNA signal and reported clinical cases using a processing method based on solids. Graham et al. (2021) reported no change which is unsurprising given that PMMoV levels did not change significantly within the study time period, while Feng et al. (2021) using a processing method based on influent liquid where PMMoV partitions, rather than solids, found a decreased correlation. Although PMMoV has been widely used and is recognized as a viral indicator that is readily recovered from wastewater, the fact that its presence is determined by diet (PMMoV is endemic to many pepper cultivars around the world and is present in many processed foods that are part of the human diet, but levels of PMMoV differ according to source) means it may not be the best indicator of faecal content in wastewater. This is a conspicuous concern when attempting to use PMMoV to normalize for faecal content in near-source applications where fewer individuals contribute to the signal. Small changes in diet over time could lead to wild variation in PMMoV levels and subsequent inaccuracies in reported SARS-CoV-2 signal. A faecal or human excreta biomarker that is more closely linked to metabolic activity (e.g., faecal sterols, bile acids) may prove superior. This topic needs to remain an active area of research.

Xie et al. (2022) reported success in normalizing the SARS-CoV-2 signal using acesulfame, a widely used dietary sweetener. This parameter has been widely used as a tracer for wastewater impacts on receiving waters. While possibly more consistent than PMMoV, acesulfame is still governed by dietary consumption. Cluzel et al. (2022) have reported success in normalizing for sanitary sewage content with a mathematical approach incorporating the following markers: ammonia, conductivity and chemical oxygen demand.

There is clearly value in being able to compensate for random dilution of human excreta by water known or expected not to contain SARS-CoV-2 RNA (e.g., stormwater, groundwater infiltration, industrial and commercial wastewaters, and their associated solid components) to arrive at more accurate and precise estimates of SARS-CoV-2 RNA as a fraction of assay input. The ideal indicator would be equally persistent as SARS-CoV-2 RNA in wastewater, would not be confounded with other non-human faecal water sources from communities, and its recovery should be comparable to that of SARS-CoV-2 RNA. Different sewersheds will have different requirements that could influence the choice of the “ideal” indicator and may require a period of baseline monitoring to establish a case for choosing one vs. another. There’s a distinct possibility that multiple normalizers could be employed to arrive at better estimates of signal. There is a real risk that as clinical testing for COVID-19 is further reduced, fewer opportunities will be available to benchmark different normalization methods and researchers will have to depend on retrospective analyses of potentially compromised archived samples. Another important question to resolve from this research is whether similar normalization schemes can be applied to targets other than SARS-CoV-2, which would be desirable for a wastewater-based platform with expanded applicability.

Wrangling disparate data into a common format facilitates collaborative research, large-scale analytics, and dissemination of results to stakeholders. Early in the pandemic, researchers quickly understood that a data model (i.e., defining data elements and their relationships) would be needed so that, for example, wastewater surveillance data generated in Edmonton can be accessed, understood, analyzed, and interpreted by a research group in Montreal, and beyond. In the first year of the pandemic, Dr. Doug Manuel created the “Ottawa Data Model”, an important component of the open science approach used by Ottawa’s wastewater surveillance project since summer 2020. This work defined minimum data elements and metadata associated with wastewater-based SARS-CoV-2 surveillance. Today, this renamed Public Health Environmental Surveillance Open Data Model (PHES-ODM) has broadened its data dictionary to encompass several environmental sample types beyond wastewater. It is entirely open source and its development is supported by the research community and CoVaRRNet (covarrnet.ca/wastewater-surveillance-group/). Several entities managing wastewater surveillance across 23 countries have so far adopted this data model or are planning to implement it and include: Ontario’s WSI, the Canadian National Microbiology Laboratory, and the European Union. The PHES-ODM will also incorporate data dictionaries of other international repositories including the US CDC’s National Wastewater Surveillance System. A data repository will be built on top of PHES-ODM and will be accessible to the public. In Ontario, the WSI manages a closed database that includes combined wastewater and clinical analytics such as trend reports. The province’s 34 public health units can access this information through a centralized web portal. In the near future, this information will be accessible to the public. Currently in Ontario, the Ontario Science Table serves aggregated wastewater trends over time for over 100 sampling sites on their public online dashboard (covid19-sciencetable.ca/ontario-dashboard/). Many public health units across the province are also providing local wastewater data and analytics through their own dashboards and (or) through those set up by the academic laboratories doing the servicing (e.g., 613covid.ca/wastewater/). Similar dashboards are being used in other jurisdictions across Canada (see Supplementary Information Part 2: a curated list of Canadian sites is also accessible and around the world, Naughton et al. 2021). Many researchers around the world are also directly reporting and interpreting wastewater results via social media.

The importance of direct and two-way communication between the data collection teams and the public cannot be underestimated. Although anecdotal, researchers have reported that some members of the public might put more trust into wastewater surveillance data and may even alter their own behaviour based on local trends in wastewater surveillance metrics. This communication can benefit from good relationships and mutual understanding between scientists and media outlets that play a large role in disseminating information and updates to the public. Not only must the relationship between researchers and the public be cultivated and maintained, so must relationships between researchers and local public health units. This is due to the fact that wastewater surveillance data reported in isolation can be difficult to interpret, but can be made easier through regular, two-way communication and mutual accountability between these parties.

Infection with SARS-CoV-2 consists of viral replication within host cells, a process that is prone to the introduction of mutations in the viral genome. Mutations allow for natural selection resulting in the emergence of novel genetic variants of the virus being shed by people and (or) animal hosts. Currently, several organizations (i.e., WHO, USCDC, ECDC, UKHSA, PHAC) classify variants based on their potential or known risk to public health. VOC can exhibit any combination of enhanced transmission (spread) in communities, increased pathogenicity, or ability to escape immunity (failure of protection from infection or disease) relative to the ancestral (wild-type) virus and thus pose an increased risk to both regional and global public health. Early discovery and tracking of emerging variants have the potential to enable public health risk reduction in both pandemic and pre/post-pandemic contexts by:

A summary of VOCs and variants of interest (predicted to behave as a VOC but for which epidemiological evidence is very preliminary or unclear) that have been classified by WHO is provided in Table 4. There are currently two methodologies in use in Canada to track VOCs (and genetic diversity of SARS-CoV-2 in general) through wastewater and estimate their prevalence in communities. The basic technology platforms employed are the same as those used for identifying VOCs in clinical specimens, namely RT-qPCR and genomic sequencing. Specialized allele-specific (AS) RT-qPCR assays are needed to quantify the specific mutations diagnostic for VOC, whereas sequencing strategies fish-out and then assemble the recoverable portions of the viral gRNA (consisting of a string of approx. 30,000 nucleotides) and sgRNA (Lightbody et al. 2019). Sequencing can therefore survey the frequency of multiple mutations in an unbiased manner, i.e., these mutations do not need to be known in advance. Furthermore, retrospective interrogation of these genomic data is possible, allowing newly identified mutations to be investigated in past samples.

Multiple waves of COVID-19 have been driven by the emergence of VOCs. A pan-Canadian strategy to survey and track the genetic evolution of SARS-CoV-2 through clinical genomic surveillance started early in the pandemic through the creation of the CanCOGeN consortium in April 2020 (Gooderham 2021). While clinical genomic surveillance was an established method to follow genetic mutations in populations via testing of nasopharyngeal samples, recovery of SARS-CoV-2 genomes from wastewater samples connected to multiple different COVID-19 infections is much more complex. Successful recovery of a consensus genome from wastewater was reported in the summer of 2020 (Nemudryi et al. 2020) and served as a proof-of-concept study, illustrating the potential of using this ‘metagenomic’ surveillance approach method to follow viral variation at the municipal level. This study used long-read sequencing of amplified genomic fragments, while others have used short-read sequencing strategies to pursue the same objectives. An early example of the latter approach demonstrated significant genetic diversity in wastewater consensus SARS-CoV-2 genomes in California that were correlated to the genetic diversity found in corresponding clinical samples (Crits-Cristoph et al. 2021).

In Canada, Lin et al. (2021) undertook metagenomic sequencing to monitor VOCs by analyzing RNA fragments present in wastewater at the British Columbia Centres for Disease Control in Vancouver. In Québec, N’Guessan et al. (2022) reported prevalent SARS-CoV-2 variant lineages in wastewater and clinical sequences from three cities. At the National Microbiology Laboratories (NML) in Winnipeg Landgraff et al. (2021) operationalized metagenomic sequencing wastewaters across Canada in Winter 2021, and then continued routinely sequencing wastewater from a number of municipalities and facilities across Canada to track of SARS-CoV-2 variants. This genomic information is relayed to public health units in an ad hoc manner as there is currently no formal mechanism for this kind of information sharing and processing.

In contrast with metagenomic sequencing, AS RT-qPCR, requires that a diagnostic mutation (allele) be known for the targeted VOC. So far, these have been identified thanks to rapid depositing of genomic information via GISAID and subsequent analysis and crowd-sourced interpretation via the Github community. This highlights the relative utility of both clinical and wastewater-based genomic sequencing vs. RT-qPCR; sequencing requires more effort but is a pre-requisite for PCR assay development. Once a diagnostic mutation is identified, development of the AS RT-qPCR assay may need a lead time of up to two weeks prior to implementation for reagents to be received and minimal validation to be carried out. Currently there are a growing number of verified wastewater-based AS RT-qPCR assays available (Graber et al. 2021; Peterson et al. 2022; Fuzzen et al. 2022), with new ones becoming available typically in response to a new variant or a new wave, as described below. Variant-specific assays can be applied in different combinations to probe for increases in emerging VOCs and (or) decreases in endemic (existing) variants. Although a single mutation may not be 100% specific for a given viral lineage in a clinical sample, owing to the defining characteristics of a given VOC (e.g., ability to spread more rapidly than existing variants in a given population; Hubert et al. 2022), any increase in frequency of this mutation in a wastewater sample is an excellent proxy marker for the emerging VOC. As with clinical samples, confirmation of the presence of a putative VOC and estimation of its proportion relative to all SARS-CoV-2 in wastewater can also be performed by sequencing, by adopting a metagenomic strategy.

The main strengths of AS RT-qPCR are: (1) An ability to probe a sampling site at high frequency to generate real-time information; (2) ease of implementation by any laboratory running standard SARS-CoV-2 RT-qPCR assays on RNA samples from wastewater; (3) short turn-around time (equal to that of standard SARS-CoV-2 RT-qPCR) and, (4) affordability (on average twice the cost of the standard SARS-CoV-2 RT-qPCR on a per reaction basis). Different technical approaches are used in AS RT-qPCR. These have been used for many years in basic biological research laboratories and are also employed in clinical diagnostics. As with any PCR assay development, methods and results must be carefully scrutinized to minimize the chance of false positives or over-interpretation. The same QC measures used for standard RT-qPCR assays are also employed with AS RT-qPCR experiments, which can similarly be performed on raw influent, raw solids, or primary sludge and can be expected to achieve similar sensitivities as standard wastewater-based RT-qPCR.

In mid-December 2020, Public Health England (now the UK Health Security Agency; UKHSA) published a technical report describing a SARS-CoV-2 variant that emerged from Southeast England (Kent) in fall 2020 and that was rapidly spreading (PHE 2020). A more detailed analysis was also posted (Rambaut et al. 2020b). The WHO was also notified of the identified VOC, now known as Alpha (WHO 2020). By the end of January 2021, the first Canadian cases of Alpha infections were identified by targeted clinical genomic surveillance of travelers. Detection and tracking of this variant was made possible owing to a substantial clinical genomic surveillance program in England coupled with the ability to survey Alpha incidence by using a faster PCR-based proxy test made possible by a mutation that leads to “S-gene dropout” (i.e., one of the PCR tests no longer could detect this variant, because its mutation profile changed the genomic region that the S-gene PCR assay was diagnostic for). Public Health Ontario and other Canadian jurisdictions also employed AS RT-qPCR VOC screening to determine variant prevalence, with complementary genomic sequencing of a subset of clinical specimens to confirm the PCR results. When used in conjunction with other PCR assays unaffected by the mutation, this and other AS RT-qPCR assays (for various VOCs) allow PCR testing to reveal information about the presence and proportions of different variants in wastewater (e.g., Peterson et al. 2022; Hubert et al. 2022; Fuzzen et al. 2022), in lieu of more comprehensive and conclusive genome sequencing.

Canadian researchers were among the first in the world to establish robust and accurate wastewater-based methods to not only detect but also quantify the proportions of SARS-CoV-2 RNA signals attributable to different VOCs beginning with the Alpha variant in January 2021. The fragmented nature of RNA in this matrix, (i.e., as gRNA and smaller sgRNA) from faeces mixing and transiting sewer pipes downstream to the sampling points necessitates a different strategy. This need arises because PCR assays being used on nasopharyngeal samples for clinical diagnoses may not demonstrate the same sensitivity and allele specificity (ability to distinguish the mutation from background) to reliably measure the VOC signal as a proportion of total SARS-CoV-2 signal in wastewater. The sensitive and specific tests developed for clinical samples should not be expected, a priori, to perform similarly in the wastewater context. Moreover, Canadian researchers have observed poor allele specificity (cross-talk) and analytical sensitivity of commercially available assays that have been made available more recently, leading to overestimation of VOC prevalence in wastewater (personal communication, Dr. Shelley Peterson, PHAC). An Alpha variant-specific qRT-PCR assay which targets a mutation (N:D3L) close to the N1 region of the genome was developed by Graber et al. (2021) and applied to wastewater in early January 2021 to detect one of the first Alpha outbreaks in Canada at a long-term care facility in Barrie, Ontario. As Alpha spread in communities such as Ottawa, researchers were able to follow its incidence (how quickly it was supplanting the prevailing variant) in near real-time with results being provided to Ottawa Public Health within a day or two of sampling. Retrospective analysis of that period found that the estimates of Alpha incidence and prevalence derived from wastewater by RT-qPCR closely correlated with the estimates provided by clinical testing (via RT-qPCR screening for S:N501Y+/E484- allele positivity, which was a proxy marker for Alpha at the time) but that the wastewater results did not suffer from the same data reporting lags as the clinical testing did (Graber et al. 2021). Throughout the COVID-19 pandemic a variety of RT-qPCR assays have been applied across Canada to detect and monitor emerging variants of concern, including Delta and Omicron (e.g., Fuzzen et al. 2022; Hubert et al. 2022).

The wastewater sample matrix poses unique challenges that have made methods development key to now-established wastewater-based AS RT-qPCR and metagenomic sequencing capacity throughout Canada. Unlike clinical samples where there is (with the exception of co-infections) a single viral variant represented at relatively high concentration and with a putatively intact genome, SARS-CoV-2 RNA in wastewater is present at relatively low concentration and is fragmented. These fragments represent contributions from multiple infections, each of which could in theory represent a different variant, making detection and re-assembly of viral genomes from wastewater technically challenging and an ongoing area of research in the context of SARS-CoV-2. The mixture of variants in wastewater and corresponding constellation of mutations that can be identified by sequencing do not necessarily derive from the same SARS-CoV-2 viral genomes present in the viruses circulating in the community. Sophisticated bioinformatics tools are needed to decipher and disentangle this information. This current state of progress highlights the various technical challenges and knowledge gaps that wastewater-based metagenomic sequencing and AS RT-qPCR testing and research continue to address. Both approaches require significant expertise and knowledge to deliver reliable results.

Accordingly, wastewater-based sequencing and PCR analysis methods are rapidly evolving such that analysis and interpretation are not standardized. There is potential to over-interpret findings or arrive at erroneous conclusions regarding the absence/presence of a given viral lineage. Metagenomic SARS-CoV-2 sequencing in wastewater today can provide higher specificity than AS RT-qPCR but can suffer from lower analytical sensitivity. Because of the relatively high cost of sequencing (compared to PCR), multiple samples are generally run as a batch, prior to subsequent computational analysis. These factors generally preclude high frequency reporting of sequencing results from wastewater (currently every week or fortnightly in Canadian jurisdictions). There is currently a lack of properly benchmarked studies comparing sequencing to AS RT-qPCR. The lower sensitivity for sequencing may be lineage dependent and might also be affected by the viral diversity in the samples vis-a-vis the number of different variant lineages contributing to a wastewater sample. Notwithstanding these limitations, tremendous advances in wastewater-based sequencing of SARS-CoV-2 RNA have been achieved since 2020 and continued innovations are expected. Overall, the complementary nature of metagenomic sequencing and AS RT-qPCR in wastewater surveillance means the two strategies can be deployed effectively in a context- or question-specific manner.

Complementary AS RT-qPCR assays and metagenomic sequencing are being strategically used today in Canada to identify the emergence of known variants, and to act as a proxy estimate of their incidence and prevalence in a sampled population (how quickly it spreads in a population) but also have the potential to identify unknown variants. Metagenomic sequencing can be used to confirm AS RT-qPCR results as viral sub-lineages may not be easily distinguished using AS RT-qPCR. Because of its relative ease of implementation and fast time-to-reporting, AS RT-qPCR can be used to identify the likely presence of a particular VOC at the facility-, neighbourhood-, or city-level in near real-time (within eight hours of sampling). Metagenomic sequencing can identify diagnostic mutations, knowledge of which can be used to design new AS RT-qPCR assays should clinical genomic surveillance be insufficient for this objective or miss new mutations. Metagenomic sequencing can also be strategically located at facility-, neighbourhood-, or city -level to monitor for emerging, unknown variants (i.e., variants that have not yet been identified clinically) and known variants at country or provincial entry points. Ad hoc monitoring of airplane toilet pump-outs in Australia (Ahmed et al. 2022a, 2022b) and airport wastewaters in Germany (Agrawal et al. 2022) have successfully detected VOC in travellers prior to the detection of community transmission using both metagenomic sequencing and AS qRT-PCR. It could also be possible to infer the emergence of an unknown variant through monitoring signal drop-out for a given allele using AS RT-qPCR.

Since late 2021, Ontario research groups, through the province’s Wastewater Surveillance Initiative (WSI), have been performing metagenomic sequencing in many regions of Ontario including transportation hubs. Low frequency sequencing at sentinel sites is used strategically together with high frequency AS RT-qPCR at multiple locations. The Ontario VOC data and trends are routinely reported to the affected public health units. VOC signatures in wastewater collected from all major urban centres in Canada and other strategic locations are monitored on a regular basis through NML. Furthermore, Canada’s Coronavirus Variants Rapid Response Network (CoVaRRNet) – a network of interdisciplinary researchers from institutions across the country – includes a wastewater surveillance priority area and is performing metagenomic sequencing of wastewater samples provided from across the country. AS RT-qPCR assays are being used in several locations across Canada (AB, SK, ON) to monitor VOC prevalence. Some jurisdictions have reported “cryptic” SARS-CoV-2 variants in wastewater (Smyth et al. 2022), although there has not yet been an instance of a VOC first identified through wastewater. It is likely this will happen given the reductions in clinical genomic surveillance around the world. Indeed, researchers in Québec have shown that it is more likely that a variant will be detected through wastewater than through an equivalent clinical sampling effort (N’Guessan et al. 2022). The strategic use of wastewater-based VOC tracking capacity in both Alberta (Hubert et al. 2022) and Ontario (Arts et al. 2022) as part of the relatively advanced surveillance programs in these provinces, enabled tracking of the emergence of Omicron from the time of the first cases identified in November 2021, through to the peak in infections of sub-lineages BA.1 as clinical testing in both provinces became severely restricted. This showcased the scalability of the wastewater testing platform, giving both a high-level overview of Omicron attack rates at the provincial level, as well as more regional- and municipal-level situational awareness. In both instances, wastewater closely reflected available clinical estimates of VOC over time. Because new variants sometimes contain similar diagnostic mutations to prior variants (e.g., Omicron BA.1 and Alpha B.1.1.7 share a common mutation used in diagnostic PCR assays), these nuances must be accounted for (Hubert et al. 2022). Designing and implementing multi-plex assays that truly allow different variants to be disentangled is essential for wastewater monitoring using AS RT-qPCR.

The British Columbia Centres for Disease Control (BCCDC) Public Health Laboratory (PHL) leveraged an existing collaboration with Metro Vancouver focusing on enteric viruses in wastewater since 2018 so that methods for the quantification of SARS-CoV-2 in wastewater were developed in May 2020. Following participation in the CWN inter-laboratory study (Chik et al. 2021), the BCCDC PHL team further optimized its methods for detecting and quantifying SARS-CoV-2 in wastewater. Starting in October 2020, wastewater samples were collected weekly from five WWTPs in the metro-Vancouver area capturing close to 50% of BC’s population and spanned its two largest health authorities. The wastewater results have been integrated with clinical case counts by a medical geographer at the sewershed level. Public health epidemiologists have compared the concentration of SARS-CoV-2 in wastewater at each WWTP to the incidence of COVID-19 cases in the corresponding wastewater catchment area.

On a weekly basis, the wastewater data, epidemiological graphs, and key messages are compiled and reported weekly for Medical Officers of Health and epidemiologists at the regional health authorities since March 30, 2021. Wastewater data is also incorporated into the bi-weekly BC COVID-19 Data Summaries since August 14, 2021 and in the BC Situation Report (BCCDC 2022) since November 28, 2021. To make the data and information available to the general public and to help facilitate the dissemination of the SARS-CoV-2 in wastewater data, Metro Vancouver launched an online page providing an interactive map that allows the public to view SARS-CoV-2 concentrations at each WWTP over time.

In collaboration with Dr. Ziels and Xuan Lin at UBC, methods were quickly developed to test wastewater samples (Lin et al. 2022) for variants of concern (VOCs). These methods have been deployed for both Metro Vancouver WWTPs (since January 2021) and the UBC project (since September 2021) and successfully detected Alpha, Gamma, Delta, and Omicron VOCs.

In the first half of 2020, two experienced Alberta research teams began implementing SARS-CoV-2 wastewater monitoring programs. Dr. Xiaoli Pang at the Provincial Laboratory of Public Health (cross-appointed to the University of Alberta) began testing for SARS-CoV-2 RNA in WWTP samples from across Alberta, later expanding to include long-term care facilities in Edmonton. In parallel, preliminary studies at the University of Calgary were initiated by an inter-disciplinary team with expertise in environmental microbiology and virology, wastewater engineering and clinical microbiology including Calgary’s head of infectious diseases Dr. Michael Parkins. The Edmonton and Calgary teams agreed to submit separate proposals to a Canadian Institutes of Health Research (CIHR June 2020) competition and both secured approximately $500k each for one year pilot studies. Wastewater testing became established in different Alberta municipalities, urban neighbourhoods, and hospitals by mid 2020, followed by long-term care facilities in Edmonton funded by the COVID-19 Immunity Task Force (covid19immunitytaskforce.ca).

In Calgary, wastewater monitoring was performed in hospitals, a setting with a high degree of testing of patients and healthcare workers and thus a very good understanding of transmission dynamics. Accordingly, SARS-CoV-2 levels in hospital wastewater enabled differentiation of new nosocomial outbreaks of COVID-19 against a high background of patients admitted to hospital with COVID-19 infection. This indicated that the vast majority of RNA shedding into wastewater was associated with early onset of disease (Acosta et al. 2021).

Testing in several Edmonton long term care facilities studied the cost-effectiveness and early warning potential of wastewater surveillance (Lee et al. 2021) finding that undetected clinical cases could be revealed from wastewater surveillance. Nodal sampling in neighbourhood sub-catchments throughout Calgary demonstrated links between COVID-19 infection and social determinants of health during Alberta’s second and third waves in late 2020 and early 2021 (Acosta et al. 2021). Li et al. (2023) reported a Probit analysis of over 1 800 wastewater samples collected from 12 Alberta WWTPs (reported in Section 5.2) for over a year that provided estimates of case detectability in relation to population size.

By 2021, with two successful wastewater monitoring programs up and running, the Edmonton and Calgary teams began collaborating more closely. The groups formally coalesced and secured funding from the Alberta government creating a single PanAlberta monitoring program to cover large and medium sized municipalities throughout the province as well as selected institutions. WWTP samples taken three times per week were sent by courier to either the Edmonton or Calgary groups for RT-qPCR testing with rapid turnaround times of 24 to 48 hours. By late 2021, this program covered more than 80% of the province’s population and ∼95% of its urban population. Also in 2021, in partnership with data sharing experts from the University of Calgary’s Centre for Health Informatics (CHI), wastewater results began being published on CHI’s COVID-19 tracker website (covid-tracker.chi-csm.ca/). By 2022, this website’s wastewater page was getting up to 8 000 visits per day.

A pilot study in Saskatoon was initially funded by the University of Saskatchewan-led Global Water Futures (GWF) program and supported through in-kind contributions of personnel and sampling equipment by the City of Saskatoon, postdoctoral fellows, and students. Based on three weekly samples, viral loads in Saskatoon’s wastewater remained low throughout July, August, and September 2020, but began to rise exponentially in October and November 2020 providing a leading indicator of impending surges in case numbers. The team informed Saskatoon’s population of upcoming potential increases (and decreases) in positive cases primarily through press releases and media interviews. Wastewater data were shared with the Saskatchewan Health Authority and the Saskatchewan Ministry of Health. Provincial modelling teams used the information from wastewater surveillance to refine their models that helped forecast future health risks associated with COVID-19. A first-of-its-kind study with Indigenous communities was initiated in partnership with the Indigenous Technical Services Co-operative (ITSC), which included five First Nations with one each from Agency Chiefs Tribal Council, File Hills Qu’Appelle Tribal Council, Saskatoon Tribal Council, Touchwood Agency Tribal Council, and Yorkton Tribal Council.

The University of Ottawa, in collaboration with the Children’s Hospital of Eastern Ontario’s Research Institute (CHEO-RI) and the City of Ottawa, performed the first measurement of SARS-CoV-2 viral signal in Canadian wastewaters on April 8, 2020 (D’Aoust et al. 2021a). With sufficient data, SARS-CoV-2 wastewater surveillance was found to provide useful information such as early detection of disease incidence in the community, shown during the beginning of the second resurgence of COVID-19 in Ontario (July 2020). Specifically in the summer of 2020, SARS-CoV-2 wastewater surveillance was shown to forecast increases in clinical cases of COVID-19 by 48 hours, and increases in COVID-19-related hospitalizations by 96 hours (D’Aoust et al. 2021b). Ottawa Public Health rapidly became further involved in the novel surveillance system, ultimately requesting testing seven days a week and an analysis turn-around-time of 24 hours, which was attained in September 2020. In addition, the first public-facing dashboard (613covid.ca) of SARS-CoV-2 surveillance in Canada was put online in Ottawa in September 2020 in collaboration with the University of Ottawa, CHEO-RI, the Ottawa Hospital Research Institute, and Ottawa Public Health.

The daily testing frequency and rapid turn-around-time demonstrated improved understanding of COVID-19 surveillance data in the City of Ottawa with the wastewater data being triangulated with clinical data. In response to the use of wastewater surveillance data, the City of Ottawa, University of Ottawa, and CHEO-RI contributed to an Ontario Science Advisory Table Science Brief (Jüni et al. 2020) which resulted in planning of an Ontario-wide SARS-CoV-2 Wastewater Surveillance Initiative (WSI).

Led by Ontario’s Ministry of the Environment, Conservation and Parks (MECP), the Ontario WSI was established as a provincial program that is comprised of a network of 13 academic and research institutions along with involvement of PHAC-NML and now extends to 170 locations capturing over 75% of Ontario’s population. This wastewater surveillance network is clearly the largest in Canada. Wastewater surveillance efforts have emerged as a critical measure of community spread of COVID-19 that is independent of clinical testing, attracting considerable media attention and increasing public awareness (December 2021-February 2022). In Ontario, Public Health Units and their respective Medical Officers of Health have since used wastewater surveillance data to assist in planning, public messaging, and directing resources.

At the University of Waterloo, Dr. Mark Servos, a Biology Professor recognized that his laboratory’s experience in wastewater and environmental research could be adapted to detect SARS-CoV-2 in wastewater influent. Early in the pandemic (April 2020), he and his team returned to the laboratory to focus on developing methods that could be applied to conduct wastewater surveillance for Ontario communities. Once the methods were developed in the summer of 2020, with the support of municipal and Public Health Unit partners, pilot programs were initiated at several sites to test and validate the approach. Within a few weeks this grew into a formal surveillance program for these regions with multiple sites covering a population of more than 2.6 million people. Eventually, this program has grown under the Ontario WSI to cover the regions of York, Peel and Waterloo. All three Public Health Units (PHUs) /Regions early in the pandemic established mechanisms to disseminate the results to senior management as well as the public by means of dashboards (see Part 2 of the Supplementary Information). This group, in collaboration with the Ottawa group and NML developed refinements that could be used with PCR (Fuzzen et al. 2022) to allow tracking of VOCs, a critical advance that has allowed the tracking of the take-over by Omicron from Delta in late 2021 when provinces had largely abandoned wide-spread clinical testing in the population, leaving wastewater surveillance as a more unbiased means of tracking VOC waves.

Researchers in Québec, using their own research funds, began surveillance of SARS-CoV-2 in wastewater in March 2020 (with the earliest samples collected in February 2020). They did not initially have access to their laboratories because of large outbreaks in Montréal early in the pandemic. A frozen archive (stored at −80 °C) of samples from the early days of the pandemic in the Montréal was created. Once permission was granted to return to the laboratory, researchers joined nation-wide initiatives such as the CWN-led interlaboratory study (Chik et al. 2021) prior to selecting a final protocol for analyzing archived and fresh wastewater samples.

In December 2020, a grant from the Fonds de recherche du Québec (FRQ), the Trottier family Foundation, the Molson Foundation, and the National Centre for Electrochemistry and Environmental Technologies (CNETE) for a total of $1.7 million enabled the launch of the CentrEau-COVID 6-month pilot project with collaborators from municipalities, seven universities, local public health, and the Institut national de santé publique du Québec (INSPQ). Internal funding through McGill’s Mi4 program enabled the sequencing of samples collected from Québec’s CentrEau-COVID pilot project (N’Guessan et al. 2022) and comparison with clinical samples to find that wastewater sampling was highly efficient for the detection of VOCs.

Wastewater monitoring for SARS-CoV-2 in Nova Scotia has been led by Dr. Graham Gagnon and Dr. Amina Stoddart at Dalhousie University’s Centre for Water Resources Studies. Their work has been conducted through support and partnership with Research Nova Scotia, Halifax Water, LuminUltra Technologies, Genome Atlantic, and many other municipal and industrial partners. Sustained weekly sampling was undertaken at four WWTPs in Halifax Regional Municipality processing 92% of the wastewater in the region as well as WWTPs in Sydney, Antigonish, and Wolfville contributing weekly samples which were quantified at partner Universities. Passive samplers were also collected up to three times per week from targeted sewershed locations in Halifax Regional Municipality and other communities across the province totalling over 30 sites across Nova Scotia. The research team developed a passive sampling device consisting of a small spherical cage about the size of a softball — named the COVID-19 Sewage Cage (COSCa) — which can be 3D-printed for about $1. The device was ultimately outfitted with an electronegative filter that adsorbs viruses and viral remnants including the biochemical species that is/are associated with SARS-CoV-2 RNA signal derived from wastewater. (Hayes et al. 2021, 2022) and it has been used at sites in France, Australia and across Canada (B.C., Ontario, Northwest Territories).

In November 2020, the Water Resources Management Division of the Department of Environment and Climate Change, and the City of St. John’s made a proposal to the Department of Health and Community Services to begin sampling wastewater from the Riverhead WWTP. HCS gave approval for this surveillance in February 2021, when there were known cases in the St. John’s area. The province now monitors wastewater for the presence of COVID-19 virus in 17 separate sewershed catchment areas, including residents from 14 communities, representing about 46% of the provincial population. In 2021, wastewater surveillance was useful as an “early warning” system for detecting COVID-19. For example, Public Health issued a public advisory for the Town of Deer Lake in November 2021 when the wastewater suddenly showed a strong presence of SARS-CoV-2 RNA. The notification prompted symptomatic residents to seek testing which led to the identification and isolation of previously unknown cases. In September of 2021, the province released its Wastewater Surveillance for COVID-19 dashboard to share wastewater data publicly.

One of the most important lessons learned was to wait until there was buy-in for wastewater surveillance for COVID-19 from public health officials. As the end user of the data, it was vital that the public health decision-makers be part of the conversation. The establishment of a provincial working group that meets every two weeks to discuss results, issues, and new advances was also instrumental in helping guide the development of the wastewater surveillance program in Newfoundland and Labrador.

The Federal Fall Economic Statement 2020 allocated $37.4M to support the advancement of innovative approaches to COVID-19 detection from which approximately $12.8 million was allocated over a period of 2.5 years to establish a wastewater monitoring program in Canada. NML had identified Statistics Canada as a key partner early in the pandemic because the existing Canadian Wastewater Survey - CWS covering Metro Vancouver, Edmonton, Toronto, Montreal and Halifax already existed. Through a pilot program between the two organizations, wastewater surveillance covering ∼23% of the Canadian population was established by fall 2020. Building on this success, a long-term agreement between the organizations was signed in the spring of 2021 and has formed the core of the national program since then.

NML was vital to the first Canadian interlaboratory study done in partnership with CWN because it provided the logistics for the common sample collection, spiking with known quantities of inactivated virus and internal standards and shipping to seven other Canadian laboratories. NML conducted a second study with the Ontario Ministry of Environment, Conservation and Parks, and the Ontario Clean Water Agency in February 2021 where 29 laboratories across Canada participated in assessing detection of SARS-CoV-2 RNA in wastewater. Since those early days, PHAC has worked with partners to directly support wastewater surveillance at 65 sites (as of April 28, 2022). PHAC estimates that between provincial and other programs, as well their own approximately 60% coverage of the Canadian population was achieved in March 2022. PHAC has set a target of 80% coverage of the Canadian population by the end of 2022.

PHAC arranged with the Government of the Northwest Territories to deploy molecular testing to Yellowknife (Daigle et al. 2022) that was able, with sample preconcentration to detect and rapidly report a consistent SARS-CoV-2 signal in community wastewater that was subsequently confirmed in samples shipped to NML. After a detection on April 16, 2021, daily sampling was implemented leading to increased clinical testing focused on recent travelers to Yellowknife that identified a cluster of six COVID-19 cases over the period April 20 to 26.

Signals of SARS-CoV-2 RNA measured in wastewater, as long as they are obtained in a rigorous manner of sample collection, preparation and analysis, can be accurate and provide an independent and complementary source of relevant information that can be generally free from inevitable biases arising with clinical test results which are caused mostly by clinical testing policies.

Wastewater should capture excretion of SARS-CoV-2 RNA from all residents who are served by the sewer system being sampled, except those who are incontinent or whose faecal waste does not enter the wastewater system. This coverage of wastewater surveillance would include otherwise marginalized populations including those who cannot, or decline being tested for personal reasons. Clinical testing for COVID-19 has not and generally cannot service all groups in society equally. There is a need for surveillance systems to provide equity to marginalized populations, for which wastewater surveillance can contribute to providing more equitable access to community public health data. Widespread and voluntary cooperation of wastewater treatment plants was typically demonstrated, though some communities may be under-resourced and unable to participate or are otherwise resistant.

During the COVID-19 pandemic in Canada, asymptomatic and pre-symptomatic cases in overall communities have not been commonly tested by clinical surveillance, except for potential high risk exposure circumstances and for travel requirements. Asymptomatic and pre-symptomatic cases contributed substantially to community transmission. Evidence that asymptomatic cases can excrete SARS-CoV-2 RNA, and pre-symptomatic cases can excrete SARS-CoV-2 RNA for days before onset of any symptoms, while clinical testing is unlikely to detect such cases raises the expectation that wastewater monitoring done frequently and reported rapidly can provide an early signal of such cases in the population being sampled.

Particularly when sampling at WWTPs, a large population can be monitored with a single daily composite sample. In most Canadian jurisdictions, such samples need to be regularly taken for environmental regulatory monitoring and (or) effective WWTP operations. In such cases, the additional personnel burden for sampling is limited to splitting the sample and preparing it for shipping to the analytical laboratory. The cost per unit of population sampled varies according to the size of the population served but is negligible compared with the costs per individual for clinical testing even after taking account of the slightly higher cost for wastewater sample preparation.

Wastewater surveillance for SARS-CoV-2 RNA is inherently scalable for WWTPs, involving the same sampling and analytical cost for a large community as for a small one. Sample shipping costs will be higher for communities distant from the analytical laboratory. Resource demands are higher per sample within sewer networks. The cost-effectiveness is clearly greater in terms of cost per individual for large populations sampled. Ethical issues will inevitably become a greater concern when smaller populations are sampled because of the possibility of identifying infection in small groups of individuals. Wastewater surveillance for SARS-CoV-2 RNA is capable of informing targeted public health interventions at different scales whether implemented at building, neighborhood, or city-scales. Wastewater surveillance also offers future scalability in terms of its potential to monitor other public health risk targets from the same sample. Institutional investments in wastewater sampling infrastructure now will enable adaptation of methods for new and complementary public health surveillance goals. The ability to scale up quickly will always be limited by existing laboratory capacity.

As long as the procedural quality control and quality assurance measures are satisfied and sufficient wastewater testing frequency is achieved, evidence from many Canadian and international locations has shown the ability of wastewater surveillance for SARS-CoV-2 RNA to show trends for the virus RNA signal in community wastewater. In addition to detecting increases in infections, reporting of trends may provide insight into the success of public health interventions implemented to curb new infections.

In some locations, wastewater signals are being incorporated into public health warning systems (e.g., through trend classification, CDC 2022b). While these metrics are not standardized, they demonstrate potential utility of wastewater data for high-level classification systems. As models that relate wastewater measurements to other metrics of COVID-19 improve, wastewater data can be more systematically incorporated into informational and decision-making frameworks.

Numerous studies in Canada and around the world have demonstrated that wastewater surveillance performed on samples from sewershed nodes and building outflows can provide site-specific information—i.e., by flagging emerging “hotspots” when COVID-19 prevalence is otherwise low, and (or) by monitoring specific priority buildings, like residential living complexes. This site-specific information can then in turn justify site-specific interventions, such as targeted communications to a population or mass testing of all individuals residing in a given location and document trends if representative composite sampling is possible. Effectiveness of this approach requires that the frequency of sampling is high enough and sample turn-around-time and reporting is short enough, to inform public health action provided ethical considerations discussed in Sections 5.7 and noted in limitation 6.2.4 are respected.

Several applications reported internationally and in Canada have shown that wastewater surveillance for SARS-CoV-2 RNA can be effectively adapted to detect the proportion of variants in a wastewater sample. These examples have allowed an accurate assessment of how a given VOC is becoming dominant within the system being sampled, information that should be useful to public health officials in knowing what VOCs are going to be showing up in the healthcare system. Although more demanding in terms of analytical facilities, metagenomic sequencing may allow for detection of new variants provided the RNA signal is strong enough to provide a clear signal.

The nature of wastewater surveillance for SARS-CoV-2 RNA provides it with a demonstrably more efficient ability to track the community dynamics of virus shedding spatially and over time such that reasonable inferences about geographic patterns of disease distribution can be developed.

The emergence of the Omicron variant overwhelmed the ability of clinical testing to track the dynamics of that wave of COVID-19 infection in most jurisdictions. The availability of wastewater surveillance for SARS-CoV-2 RNA provided the only near-real time tracking of the first and second Omicron waves in those locations that were performing such wastewater surveillance. Routine wastewater surveillance can also fill gaps in public health data caused by increasing use of at-home test kits for which results are not reported.

Provided that results are made available to the public in a comprehensible manner, wastewater surveillance for SARS-CoV-2 RNA can enhance public understanding about the occurrence of COVID-19 in a community. Effective engagement of the public about the status of the pandemic is important. Particularly as policies such as mask mandates and physical distancing are relaxed, wastewater data can help inform individuals what preventive actions to take given their personal risk profile and tolerance.

Unlike clinical testing of individuals who are asymptomatic and are not seeking medical care, wastewater surveillance for SARS-CoV-2 RNA, itself, is normally, entirely non-invasive to individuals. Only when wastewater provides evidence of undetected infection and individual clinical testing becomes necessary the latter is invasive. At that point the individuals identified are likely at higher risk.

The ability of wastewater surveillance for SARS-CoV-2 RNA to generate large quantities of data in a cost-effective manner has created a number of substantial databases in Canada. These can be subjected to retrospective analyses for model development and multivariable assessment to seek better understanding of the dynamics of the COVID-19 pandemic at a given monitored location in relation to public health interventions there. This source of evidence opens the prospects for developing much better evidence-informed insight about public health interventions than have been available for previous pandemics. Likewise, the availability of archived samples in many cases may allow pursuit of new research answers to questions that are not yet evident.

Attempts to relate wastewater surveillance signals to clinical test results have not always been based on assured knowledge about the physical boundaries for each category. The catchment of a given sewer system or WWTP is not likely to be the same physical boundary as may be used for reporting clinical results. Even when locational data are investigated, there are inevitably problems with individuals contributing to wastewater in a different location than where they may have their clinical test result reported. This is particularly a challenge for tourist destinations, educational institutions with students from afar, commuter populations who live and work in different sewersheds, and industrial locations with a fly-in worker population. Similar, but possibly less recognized challenges exist for clinical testing where individual test location or home residence need not represent where the individual has been exposed.

Ideally, wastewater surveillance for SARS-CoV-2 RNA would provide data that could reliably be translated into estimates for COVID-19 prevalence in the community. The signal obtained for SARS-CoV-2 RNA in wastewater is dependent on a number of unknown factors such as the rate of SARS-CoV-2 RNA shedding (either as intact virus, viable or not, or as RNA fragments) that can be attributed to a given case. Such numbers will also depend on the stage of infection (initially asymptomatic, active, or recovering), possibly different VOCs and characteristics of the patient (age, health status, vaccination and booster status, etc.). There will also be expected degradation of the RNA signal with sewer travel time to the sampling point and a variety of factors in wastewater that may inhibit the PCR signal. Models that estimate community prevalence may account for many of these factors, but such models would need to be developed and validated across jurisdictions and relevant local factors to facilitate their application.

The reasonable expectation that wastewater surveillance for SARS-CoV-2 RNA provides an early warning of COVID-19 emergence in a community, based in part on viral shedding by asymptomatic patients who will not likely be subject to clinical testing, is absolutely dependent on the frequency of wastewater sampling, rapidity of analysis, and reporting efficiently. Where the necessary resource commitments to achieve these requirements, early detection has been reported. Many initial reports of early warning were based on retrospective analysis of archived wastewater samples and comparison with case data for the time of collection of the archived sample. In simplest terms, a weekly wastewater sampling program, with additional days for sample handling, analysis and reporting cannot be expected to deliver an effective early warning unless clinical testing of such a system is also infrequent and delayed in reporting. Because resources were committed early on to achieve sufficient frequency of sampling, rapid analysis and reporting in Ottawa, early warning from wastewater surveillance for SARS-CoV-2 has been demonstrated there and in other Canadian locations that have also committed to satisfying these requirements.

Consideration of ethical issues was limited in the beginning of wastewater monitoring for SARS-CoV-2 RNA programs in Canada. This is still evident in most international publications. For example, Hoar et al. (2022) present an overview of how academic research can align with a transition of wastewater monitoring to routine public health surveillance in the U.S. with no mention of ethics. Canada has published a set of ethical guidelines for wastewater surveillance for SARS-CoV-2 RNA (CWN 2020a, Hrudey et al. 2021). The smaller the population that is under surveillance, the greater the likelihood that ethical considerations will arise.

In addition to institutions (hospitals, prisons, and universities) that treat their own waste, households that have their own private holding tanks or septic system will not necessarily be covered unless their septic tank is pumped out and delivered to a community wastewater system that is under surveillance. Pump-outs are only done intermittently and the RNA signal may be degraded, so there may not be timely nor accurate representation of such households in the community system sample. Although this can include many rural communities, these could be high risk (based primarily on risks of serious outcomes, but also applies to risks of high exposure) or marginalized populations that would benefit from this type of surveillance, such as remote indigenous communities, or migrant farm workers. The same populations that are not being served by clinical testing, due to access issues, may also be missing out on wastewater surveillance. The current ability to cover small and remote communities, or individual homes not on sewerage systems, is limited by logistics, capacity, and cost, leaving a gap in total population coverage.

Although tracking of VOCs has proven to be a major positive feature of wastewater surveillance for SARS-CoV-2 RNA in Canada, not every laboratory that has been involved has necessarily been able to adopt the modified procedures necessary to track VOCs even though Canadian researchers have been very creative and collaborative in sharing the necessary knowledge to be able to track recent VOCs.

As documented by Canadian inter-laboratory studies, quantitative differences will be found across different laboratories analyzing the same wastewater sample because of differences in procedures used in sample processing and analysis. Likewise, there is not yet a national consensus on how to present quantitative wastewater surveillance data including a lack of consensus about means for normalizing faecal strength of wastewater. These factors mean that results for different communities produced by different laboratories are not readily comparable. Policymakers need to be provided an opportunity to understand the inherent variability and uncertainty in the analytical results while also appreciating the timeliness of such surveillance.

The existing “silos” among academic researchers, many in environmental science and engineering, public health professionals and wastewater utility personnel have interfered with achieving common understanding about the meaning of results from wastewater surveillance for SARS-CoV-2 RNA. Important progress has been made at some locations and efforts by the CWN and NML have helped in some cases, but further work is needed.

Sampling wastewater from manholes or other sewer access points is much more challenging and resource-intensive than sampling at WWTPs. Manholes are potentially dangerous, confined space locations that require professional occupational health and safety supervision for sampling. These sites are often in public spaces (roadways) that make security of samplers a problem. Difficulties also arise in winter with factors such as: freezing, low or intermittent flow as spatial granularity increases (e.g., wastewater outflow from a single building), presence of materials flushed into sewers that can clog samplers etc. Passive samplers can block smaller sewer pipes or be lost to the monitoring system if not anchored properly

In a number of cases, municipal or institutional personnel have been reticent about allowing sampling for the purposes of wastewater surveillance for SARS-CoV-2 RNA to be performed on their premises. Reasons have differed but having a clear level of support for this activity by municipal, provincial/territorial and federal governments might assist in recruiting such sites.

Wastewater SARS-CoV-2 data seen to date, within Canada and internationally, show a high level of variability from one day to the next and between sites, even when analyzed by the same laboratory. Trends are generally discerned by using three to seven day rolling averages to smooth out the day-to-day fluctuations. To date there is not a clear understanding of what are the most important factors driving the observed variability.

Sampling at a WWTP is generally and successfully done at the entrance to primary treatment or using primary sludge. Likewise, choosing municipalities according to population coverage is reasonably logical. Beyond that, choosing sampling sites for sampling within a sewer network or to characterize institutions has been more challenging to rationalize. Sampling within building sewers will raise even more challenges.