Skip to main content

Why variant statistics don't tell the full story of Omicron's spread in Canada


Since the first case of Omicron was reported in Canada, labs across the country have been trying to keep up with its spread by tracking the newest COVID-19 variant of concern.

But a surge in demand for COVID-19 PCR tests over the last few weeks has not only created a backlog in certain provinces – it’s also making it tougher for labs to keep up with processing positive cases in order to determine their strain.

“The problem is the sheer number of tests that have to be done,” Dr. Gerald Evans, chair of Queen's University's infectious diseases division in Kingston, Ont., told on Jan. 5. “It does take some time to be able to push out that result to tell you, is this Omicron or is this Delta.”

Omicron cases can be distinguished from Delta cases through PCR tests that identify an S-gene target failure, said Evans. The procedure involves detecting parts of the S-gene, or spike protein, to determine whether or not it is present. With the Omicron variant, he said, the S-gene can’t be detected. While the Alpha variant carries a similar trait, chances are low that anyone would currently test positive for Alpha, Evans said.

Still, the best way to confirm the variant of a viral sample is through whole genome sequencing, said Evans, a process that is much more expensive and takes longer to complete, as it involves analyzing the full genetic makeup of a viral sample. Evans is the medical director of infection prevention and control at the Kingston Health Sciences Centre, one of only a handful of centres in the province that does whole genome sequencing.

“When you test an isolate…of SARS-CoV-2 and it’s got an S-gene target failure, I would say you can be about 90 to 95 per cent sure that’s Omicron,” he said. “However, to be 100 per cent sure, then you have to do whole genome sequencing.”

Ontario uses a combination of PCR testing and whole genome sequencing to determine the strain of COVID-19 viral samples, said Dr. Marek Smieja, an infectious diseases physician and professor at McMaster University in Hamilton. The province began implementing this combined strategy in December.

Due to the high demand for testing, only select viral samples are currently being sent to Ontario’s labs for PCR testing and whole genome sequencing in order to determine their strain, Smieja said. Because of this, total COVID-19 case numbers being reported each day may not necessarily match up with the variant breakdowns provided.

“Most of our labs [in Ontario] no longer try to figure out which variants it is for every single positive,” Smieja told in a phone interview on Jan. 5. “It's extra work at a time that all labs are doing everything they can to simply test lots of people and to get the PCR results back in as little time as possible.”

Only a proportion of positive cases are sent from Ontario labs to either Public Health Ontario or one of four hospital-based labs, and all five labs will then sequence those cases, said Smieja. The province is currently performing whole genome sequencing on about five per cent of all positive samples, which translates to about 3,000 to 3,500 sequences per week, he said. The province is also prioritizing whole genome sequencing for specific cases, Smieja said, including those hospitalized with COVID-19, international travellers who are infected, and those involved in COVID-19 outbreak situations.

Earlier in the year, when there were fewer positive cases being logged, it was possible for his lab to do whole genome sequencing on all samples, Evans said.

“We would be looking at maybe 30, 40, 50 samples a week,” he said. “Right now, if we were trying to do it, it would be 1,000 to 2,000 isolates a week, so it's a matter of volume.

“You can subject whole genome sequencing to all samples when you have low numbers, but once you get into these big numbers, it would overwhelm the capacity of the system to do it.”

The increased demand for testing also poses a challenge for labs when it comes to resources, said Evans.

“We just don't have enough slots and testers around to do that many people here in the region every day,” he said. “So there's capacity limits in obtaining the test, and then capacity limits in running the test in the lab.”


According to Dr. Nathan Zelyas, a medical microbiologist with Alberta Precision Laboratories, the western province also uses a combined approach of variant screening and whole genome sequencing when testing for COVID-19 variants. Both processes are performed by Alberta’s public health laboratory, which has two sites – one based on Edmonton, where Zelyas works, and another in Calgary.

“The labs have become very busy, basically working 24/7,” Zelyas told in a phone interview on Wednesday. “We've had people redeployed to the [public health] labs from other labs to help out with just COVID-19 testing in general.”

In the past, labs in Alberta were able to perform variant screening tests on all positive COVID-19 samples simply because they had the capacity, said Zelyas. Since then, demand for testing in the province has skyrocketed, and about 1,000 to 1,500 samples are currently being screened per day, he said.

“When we're in times like current, in the middle of this really large wave [with] really unprecedented numbers of positives that we've never seen or experienced before, we're still doing a lot of variant testing, but the capacity just isn't there to do that variant screening of all positives,” he said.

Even fewer samples are being put through whole genome sequencing, with about 800 to 1,000 sequences currently being performed per week, Zelyas said. Variant screening can take an average of about two to three days to complete, while whole genome sequencing can take up to two weeks. These timelines also depend on overall caseload and the time it takes to transport samples.

Zelyas said that Alberta is likely also conducting whole genome sequencing for about five per cent of positive cases within the province, similar to Ontario. As of Dec. 24, the province also began prioritizing variant screening for subsets of the population instead of all samples. Subsets include those involved in outbreaks as well as hospital cases, international travellers and some smaller community samples to identify what’s circulating in the province.


While tens of thousands of COVID-19 tests are being performed and logged by labs in Ontario each day, Smieja said current figures are almost certainly an underestimation of the true number of positive cases. And of that portion, only a small percentage are being tested to determine their variant.

“We know we're missing many, many cases of COVID today because many people no longer have timely access to testing,” he said. “That’s unfortunately what happens at peak, there's just too much COVID going around to be able to diagnose it all.”

But being able to pinpoint the actual figures being reported for each COVID-19 variant becomes less important as more time goes on, said Evans, particularly when one variant begins to dominate. Therefore, he said, it isn’t worth getting too caught up on the numbers.

“[This is] simply because we know that as each of these different variants like Alpha, Delta, and Omicron arose, they simply out-competed the other variant that was around,” he said “Paying attention to what variants are what really isn't that important at this point.”

The goal of testing for variants is less about finding the exact number of people infected with each strain, and more about monitoring for new variants that may come along, said Zelyas.

“Once we get to a point where everything is coming up Omicron, the stuff that we're going to be most interested in is those ones that aren't screening as Omicron,” he said. “Those are the ones that we prioritize for whole genome sequencing, because those are the ones where we're really looking out to see if there's something novel popping up that we haven't tested before.”

According to Smieja, the rate at which tests are being conducted in Ontario now would still likely identify any new variants that might emerge.

“[The province] estimates that…any variant that's more than about one per cent of the circulation of COVID should be detectable,” Smieja said. “In Ontario, we're doing over 3,000 [sequences] a week, so we estimate that it’s very unlikely that we would miss a variant that [accounts for] more than one per cent [of cases].”

It’s hard to find that magic number in order to know whether or not enough variant testing is being done, said Zelyas. But when Alberta had shifted from screening all cases for variants to screening only a proportion of them in the past, not much of a difference was noted when looking at the ratios, he said.

“We looked at the trajectory from when we stopped screening all to when we started screening all again, which was back in the spring of last year,” he said. “We found that the ratios of the variants that we were detecting from the screening, they were very much in line with what we would have expected if we had screened all.

“From the variant or the lineage level perspective, we're probably not missing a whole lot by having a proportion of samples tested overall for variants.” Top Stories

Stay Connected