Saint John dropped from national COVID-19 wastewater monitoring site over data accuracy concerns

Saint John has been temporarily removed from the national COVID-19 wastewater surveillance dashboard over concerns about the accuracy of its data.

Public Health Agency of Canada officials believe "the sampling point and treatment of the sample prior to testing was the origin of the test variance," said spokesperson Anna Maddison.

This suggests "the reported values are not reflective of the community burden," said spokesperson Nicholas Janveau.

Asked what, if any, guidance the federal agency provided to Saint John or the 26 other jurisdictions across the country conducting wastewater monitoring about where to install sampling equipment or how to conduct sampling, Maddison did not answer directly.

"PHAC scientists are continually improving scientific approaches to wastewater surveillance," she said in an emailed statement. Once an investigation into the situation is complete, the agency will "decide how to approach wastewater testing in Saint John and offer advice accordingly to other jurisdictions based on this experience."

The agency continually monitors all sites for any data discrepancies and none have a similar issue, Janveau said.

'Possible issue' noted on April 19

Public Health of Canada became aware of a "possible issue" with the Saint John data on April 19, about two weeks after results for the city, dating back to March 2, were added to the national dashboard.

Scientists at the national microbiology lab in Winnipeg launched an investigation that same day, said Janveau. It's expected to take "several weeks."

A notice advising the public was posted on the national dashboard on May 2. 

"We have temporarily removed Saint John from the dashboard, due to possible issues affecting data accuracy," it says. "Once these issues have been investigated and resolved, Saint John will once again be included in the dashboard."

The Department of Health did not respond to questions about whether the Dr. Georges-L.-Dumont University Hospital Centre lab in Moncton, which also tests New Brunswick's wastewater samples, picked up on any issues with the Saint John data.

Spokesperson Sean Hatchard said only that the project is a collaboration between many partners, and the department is "pleased this issue was identified quickly."

"When new projects launch, there can be an adaptation period where some issues may arise," he noted in an emailed statement.

"The department is taking steps to rectify this situation to better inform the public of the COVID-19 circulating in their community." Hatchard did not elaborate, but said the department has "full confidence" in the team's ability to continue to identify and rectify issues, should they occur.

1 of only 6 N.B. sites monitoring wastewater

The Saint John wastewater data issues come as the province has cut back its updates on COVID deaths, hospitalizations and case numbers to monthly instead of weekly, and as fewer PCR (polymerase chain reaction) lab tests are being conducted because people with symptoms now require a referral from a health-care provider.

Some experts have called for an increase in wastewater monitoring to track virus spread and any new variants.

Wastewater results can serve as an early warning of increased community infection levels and risks.

People who are infected shed the virus in their feces in the form of a genetic material called ribonucleic acid, or RNA, before they show symptoms of COVID-19. This can be found in raw sewage and can detect the virus in a community up to 10 days prior to clinical testing, according to the Public Health Agency of Canada.

Saint John is one of only six New Brunswick cities conducting COVID wastewater monitoring.

Alternative testing performed

The federal agency became aware of the Saint John issues by reviewing the test data and visually inspecting samples, according to Janveau.

As part of the testing process, the wastewater is tested for both SARS-CoV-2 as well as "another prevalent virus," said Maddison. "This second virus helps to confirm that the test process worked.

"In this case, the control virus was low or absent for a prolonged period of time, helping to flag an issue with the samples."

To identify the issue, the federal agency performed an alternative testing procedure on a subset of samples and was able to recover both SARS-CoV-2 signal and the other virus, used as internal quality control, said Janveau.

"This suggests our original testing approach was not able to accurately recover SARS-CoV-2."

"Stakeholders" were informed via email on May 1, said Maddison. These included provincial Public Health authorities, who receive the testing reports, and Saint John's wastewater treatment plant operators, who submitted the samples.

Asked why they and the public weren't advised sooner, Maddison said they were informed "once confirmed," and she reiterated that a "public disclaimer" was added to the dashboard the following day.

Now collecting deeper samples

A visual inspection of the samples suggests they might be collected from a point in the wastewater treatment process that is affecting test quality, according to Janveau. The samples are "uniquely clear to the eye," which was the first indication of a possible issue, said Maddison.

The sampling in Saint John occurs within a treatment process called primary clarification, during which solid material is removed from wastewater, she said. The federal agency's testing method relies on the presence of SARS-CoV-2 attached to suspended solid material. 

"While adequate test performance can be gained from samples collected from primary clarifiers, it appears to depend on where samples are precisely drawn from," Maddison said.

The agency is now working with Saint John Water to revise its sample collection procedures.

Asked about the issues, Kendall Mason, director of Saint John Water, told CBC, "We are simply trying to get the sampling equipment installed and commissioned to meet manufacturing specs."

"After initial testing (commissioning) it was determined that the sample collection point depth has to be modified, submersed deeper … to obtain an accurate sample."

The sample collection point was adjusted, and new samples were collected and submitted for analysis in late April, early May, said Mason.

Retesting archived samples

Public Health Agency of Canada is also testing archived samples using the "alternative method," said Janveau.

The original testing involved extracting SARS-CoV-2 from solid material collected from wastewater, while the revised method recovers virus that is suspended within the liquid portion of wastewater, according to Maddison. "We believe it works better because the viral recovery method is better suited to the condition of the sample," she said.

Officials hope to review the test results for the retrospective data this week, and then decide how to proceed, said Janveau.

"We are considering 'backfilling' the dashboard with updated test data depending on data consistency."

The federal agency declined to release the Saint John COVID levels at issue.

As of April 6, the national dashboard indicated the seven-day rolling average of COVID-19 viral load — expressed as the number of viral gene copies found in a millilitre of raw sewage — had remained below one since the city started testing on March 2.

On March 27, the viral load was 0.16, unchanged from four days earlier, when it had jumped from 0.005.

Some jurisdictions, such as Vancouver's Lulu Island, have seen readings over 1,000 copies/ml during the peak of the pandemic.

The other New Brunswick sites monitoring wastewater for COVID-19 include: Moncton, Fredericton, Bathurst, Campbellton and Miramichi.