As N.S. lifts COVID-19 restrictions, here's an inside look at its predictions for hospitalizations
Shaina Luck | CBC News | Posted: July 7, 2022 9:00 AM | Last Updated: July 7, 2022
Throughout the pandemic, analysts have attempted to model scenarios for hospitalizations
As Nova Scotia lifts all remaining COVID-19 restrictions for the general public this week, newly released documents show how the province's hospitalization forecasts evolved as restrictions were lifted.
Using access to information legislation, CBC obtained COVID-19 modelling updates prepared by the Department of Health and Wellness' analytics division.
The updates show the province's forecasts for how many people would need acute care hospitalizations and ICU admissions between January and late May. The models were updated at least 13 times over that period as new information emerged.
Tara Moriarty, an infectious disease researcher and associate professor at the University of Toronto in the faculty of dentistry and medicine, said all modelling during the pandemic must be updated often. "The situation changes all the time, and we have to keep changing," she said.
Moriarty is part of a group of scientists called COVID-19 Resources Canada. The group runs a weekly online discussion supported by the Public Health Agency of Canada, where scientists volunteer to answer questions about COVID-19 for the public.
"Models are usually pretty good for the coming month or so, and then after that if there are uncertainties about new things that can happen, that can change what those models are," she said.
An important change in the Nova Scotia models happened between the version dated February 23 and the next version dated March 28.
The February model showed the number of patients in acute care steadily dropping through March, April, and May.
However on March 28 the analytics group presented a new version that showed the number of patients in acute care rising again in April to more than 50 patients, and levelling out in May and June.
This happened after the reopening measures which started with phase one on Feb. 14 and progressed to full re-opening in phase three by March 21.
The modelling group analyzed the reproduction number of the virus, a measure of how quickly infections are spreading. A reproduction number higher than one indicates cases are increasing, and lower than one shows they are decreasing.
The presentation stated the reproduction number began to climb "soon after the reopening phase 1 (Feb. 14), reached about 1.8-2.0 level by March 15 (Phase 2), and stayed at that level for 10 days (Phase 3)."
The March 28 model accurately predicted the number of people in acute care during what the modelling group called the "post-reopening wave" would not top 90 patients at one time, which was approximately how many people were admitted during the Omicron wave peak.
The model also predicted an "extended plateau of hospitalization."
At a briefing for the media on Monday, Dr. Robert Strang, Nova Scotia's chief medical officer of health, defended the province's choices in reopening.
"Over two years we learned lots and lots about the virus," he said. "We learned much more about how it was transmitted and many of the measures we put in place initially weren't actually necessary to limit transmission."
Strang also noted vaccines are strongly protective against severe disease.
"Our response from Public Health has always been to minimize overall harms, not just to minimize harms from COVID. All of those tight restrictions had significant impacts, whether it's financial, mental health. We're now seeing significant health impacts from delayed access to care because of some of the restrictions that have been in place in our health-care facilities. We continue to work to find a balance that limits overall harms."
The modelling group noted that during the Omicron wave the province didn't have an accurate count of how many people were getting sick, and therefore it wasn't possible for the models to predict daily case numbers. Because of this, the modelling group focused on forecasting hospitalizations and ICU admissions.
None of the models attempted to predict the number of deaths.
What happens now?
As of Wednesday, Nova Scotia lifted all COVID restrictions in the community, though a few restrictions in high-risk places like long-term care homes remain.
"I don't anticipate that lifting the restrictions will have a major impact in and of itself," Strang said in response to a question about how the move would affect the health-care system. "We have to look at the bigger picture, not just the health system."
The models from late May give a glimpse into possible future scenarios where acute care and ICU patients decrease in number through July and early August.
The late May model stated that the reproduction number dropped below one in April, and continued to drop through May. It predicted by August there will be on average five patients in acute care and three in ICU.
"We will still see some cases in the community (travel-related, random/sporadic, or localized outbreaks)," the presentation cautioned.
In an email this week, the Department of Health confirmed the hospitalization modelling was last updated on June 28, and that model was in line with earlier predictions.
Isolate when sick
Tara Moriarty, the infectious disease expert and researcher at the University of Toronto, said she's concerned about the spread of the BA.5 Omicron sub-variant, and she believes it was "not great timing" this week to remove the mandatory self-isolation requirement in Nova Scotia.
However, she said what happens next will depend on how the public behaves. "Making it mandatory may not have been having a large effect on people's behaviour," Moriarty said.
"It may be that lots of people are isolating, and whether it's mandatory or not, [Public Health] don't believe it will have an effect. But in general you absolutely want to message to people that you should be isolating when you're sick."
The projections released to CBC were created before the announcement of this week's changes. The province says those models did not take into account the announcement made Monday, however they assumed "minimal public health restrictions and good public awareness and compliance."
They do not show how waning immunity from vaccinations will affect hospitalizations, which the modelling group said would have to be taken into account in the future.
Strang said that's an area the National Advisory Committee on Immunization (NACI) is studying and his department is incorporating its guidance. "Last week they released some recommendations on fall boosters that we're starting to look at now," he said.
One scenario generated in early May shows the modelling group considered the possibility of an early summer outbreak between June 15 and July 30 that would see some community spread for about four weeks.
"This outbreak shows minimal impact on acute care admissions," the modelling presentation stated.
Strang said his department knows there is a possibility of a resurgence of newer Omicron sub-variants later in the summer, but there isn't a specific plan developed for that possibility.
"We know all the measures we've used in the last two and a half years," he said, noting the department would evaluate the severity of any outbreak. "We will go back to the tools we've used before."
MORE TOP STORIES