New Brunswick

New Brunswick's mid-December COVID modelling did not account for Omicron variant

Soaring COVID-19 case numbers have already blown past modelling numbers released by the New Brunswick government just two weeks ago — numbers, it turns out, that did not take into account the fast-spreading Omicron variant.

Data released Dec. 19 predicted 400 daily COVID cases by end of January

Dr. Jennifer Russell told a Dec. 19 news conference that the province's modelling included the effects of the more contagious Omicron variant. (Government of New Brunswick)

Soaring COVID-19 case numbers have already blown past modelling numbers released by the New Brunswick government just two weeks ago — numbers, it turns out, that did not take into account the fast-spreading Omicron variant.

University of New Brunswick physicist Dr. Sanjeev Seahra says he provided the modelling on Dec. 19 and at the time, there was still not enough data to incorporate Omicron into the projection.

Even so, Chief Medical Officer of Health Dr. Jennifer Russell said two days later that the numbers did reflect Omicron's growth.

"As you can see from this slide, when Omicron replaces Delta as the dominant strain in this province, as is happening elsewhere, we expect the number of COVID-19 infections will double every three days," Russell said Dec. 21.

She displayed Seahra's graph, which projected a sharp spike in cases heading into 2022.

"At that pace, our modelling suggests that we could see up to 250 new cases each day by early January and as many as 400 new cases a day by the end of next month," Russell said.

In fact, New Brunswick surpassed the 250-case mark within two days, on Dec. 23 and it surpassed 400 cases on Dec. 29, more than a month earlier than Russell's projection. As of Monday, the seven-day average for daily cases was 656.

Seahra, the chair of UNB's department of mathematics and statistics, says that at the time he did the modelling, Omicron's properties weren't well-known enough to be included.

"On that particular date, Omicron was a very new variant," he told CBC News. "We didn't know the properties back then and we still only have very rough estimates of the properties right now. It is new and emerging science.

"It's possible to construct models with the new variants but to have full confidence in the projections you need more data. You need longer data." 

He said he watched the Dec. 21 briefing in which Russell said the graph depicted Omicron replacing Delta and driving "exponential growth" in cases.

Sanjeev Seahra, professor and chair of UNB's department of mathematics and statistics in Fredericton, did the modelling. He said there was too little Omicron data to include in the modelling released in December. (University of New Brunswick)

"I guess I was watching it knowing where that graph was coming from, so I didn't find it confusing myself, but maybe I'm not the right one to judge." 

Russell told CBC's Information Morning Fredericton on Tuesday that her description of the graph was "a bit of a misstep on my part."

She confirmed the graph did not take into account the Omicron variant, but "I had already seen projections at the federal level and knew that when you put Omicron in the mix it was going to be much higher than that."

"The point ... was to communicate to the public that things are going to get worse and we need to prepare for that."

As of Tuesday morning, the Omicron variant represented between 86 and 100 per cent of COVID-19 cases, depending on the health zone, she added.

Russell said on Dec. 21 that "the jury is still out" on whether Omicron produced milder cases and said rapid, exponential growth would still lead to more severely sick patients straining the health-care system.

At that briefing, Russell used the graph to explain how dire a threat Omicron posed to the province.

Referring to early indications that the variant produces milder illness in general, Russell said "the jury is still out on that" and said rapid, exponential growth would still lead to more severely sick patients straining the health-care system.

At the same briefing, Premier Blaine Higgs announced the government was moving to Level 2 of its winter COVID-19 plan, with more restrictions. 

Russell also announced that Public Health was scaling back contact tracing and asking New Brunswickers who tested positive to notify their contacts themselves.

Seahra, a mathematician, said he uses standard, long-accepted modelling to make COVID-19 projections for the province, incorporating various data points into equations to create forecasts.

The province 'revised' winter plan chart released before Christmas, when the decision was made to move from Level 1 to Level 2 at 11:59 p.m. on Dec. 27. (Government of New Brunswick)

He then compares those to real data and refines the assumptions in the model until it lines up with reality, then projects the trends into the future. 

Seahra said the Dec. 19 model used data going back to Christmas 2020 and accounts for the original, so-called "wild" COVID-19 variant as well as the Alpha and Delta variants, each of which has different properties.

He said he is now "tinkering" with his model to incorporate information on Omicron, but it's still too early to be sure of how it will behave. 

With Omicron's dominance in the province, the Dec. 19 model is "certainly out of date," Seahra said. "Things change all the time." 

Department of Health spokesperson Bruce Macfarlane said the model has been updated since Dec. 21. 

He pointed out that last Friday Health Minister Dorothy Shephard said the province could see 1,000 cases a day within a week. There were 922 new cases Monday.

The absence of Omicron data from the Dec. 19 modelling was first reported by Sackville radio station CMHA-FM. 

With files from Information Morning Fredericton