How Alberta's COVID-19 numbers compare to the rest of Canada
CALGARY -- We hear the numbers every day telling us how many people are infected with COVID-19 and how Alberta compares to the rest of Canada.
But how important are those numbers really? Is there even a point comparing ourselves to other provinces? What are the most important indicators we should look at when assessing how COVID-19 is affecting Albertans? We put those questions to three experts in biostatistics and immunology.
Since the pandemic was declared in mid-March, the Cumming School of Medicine’s Centre for Health Informatics has been tracking a wealth of statistical data about COVID-19, not only in Alberta but across Canada.
“I think that if you focus on a single number, you don't get the full picture of what's really happening. You need to think of each of those things as a pixel in the picture. That is the COVID-19 situation in Alberta,” said Tyler Williamson, Associate Dean of the Centre for Informatics. “It's just one piece of that information. It's not the whole picture.”
Because COVID-19 has a 10 day to two week incubation period, daily statistics are lagging indicators, meaning they give researchers information about what was happening in COVID spread two weeks ago, not what’s happening right now.
“The most important things to look at, are unfortunately, the things that are in the rearview mirror most of the time. And that’s a problem, “ said Wiliamson “The most important things that we want to look at, of course, are ... hospitalizations, mortality, for the same reasons that we've been told over and over again, right, the last thing we want is to overwhelm the (healthcare) system.”
Williamson's comments were echoed by Dr. Craig Jenne, a microbiologist and immununologist with the Cumming School of Mediciine in Calgary.
“I think some of the key statistics are obviously healthcare capacity, and often this is a bit of a mystery, because decisions have to be made long before we reach capacity. Knowing that, again, activities we do today, or changes we make today will not have an effect for two or three weeks,” said Jenne. “We have to know where the hospitalizations right now are going to be, for example, in the middle of November, for decisions made today,”
But Jenne warned readers of statistics have to look beyond the simply the numbers, pointing to the current spike in COVID-19 hospitalization numbers in Alberta, which are higher now than at any previous time.
“There's a couple of reasons for that. One is we did have a hospital outbreak, so these are some of those cases are not people in the community that are requiring hospitalization but infection of people already in hospital, but that doesn't explain all of it.” said Jenne
“Those rising hospital numbers are a strong indicator of how difficult it is becoming to keep the virus out of susceptible populations," he said. "And that's simply because more virus in the community means the more likelihood somebody walks through the front door or is admitted and has virus positive and then the virus enters the facility.”
Bio Statistician Ryan Imgrund posts COVID-19 stats daily on his Twitter account @Imgrund. One of the key stats he follows is the reproductive number, or R(t). That number indicates how many other people an infected COVID-19 patient will reinfect before being cured of the disease. If it is greater than one there is the potential for exponential growth.
On October 20 Alberta's R(t) number was 1.56.
“1.56 is not a value that you want to be at when you're looking at R(t) because one case leads to 1.56 cases, in other words, 100 cases leads to 156 cases in just four days,” said Imgrund “Then those 156 cases will lead to about 220 cases in another four days and it gets out of control really, really quickly.”
While all biostatisticians agree the R(t) number is important, many say the way it is calculated can be subjective, and lead to unreasonable fear in members of the public with only a passing knowledge of statistics.
“People sometimes see things as really bad, causing fear and panic, and then we don't get to their perceived threat level. Then they switch the other way, sort of like crying wolf," saidJenne “Tthey will see an R(t) value that says that this is growing exponentially. And the very next week it starts to come down. And the view is ‘Well that wasn't so bad now was it? I think we have to be careful with overinterpreting the data in any direction up or down.”
Jenne says while the data is important, users of it need to ensure they are seeing the whole picture the numbers paint.
“It's important to look at the data as a whole big picture, rather than trying to sort of cherry pick individual numbers, that that either, you know, bring a sense of concern or shock, or worse, I consent to a sense of complacency.”
Williamson also warns against looking at caseloads without examining testing data at the same time.
“We see bigger numbers right now, but we're doing way more testing now than we were before. And, and so, you know, there's a little bit of ‘seek and you shall find,’” said Williamson. “If we just keep testing, we're gonna find more infections.”
All three biostatisticians CTV spoke with reccomended examining the “positivity numbers”. That is the percentage of people who have been tested that receive a positive diagnosis. Using that number allows readers of the data to assess if more cases are being discovered because of increased testing, or if the cause is actual virus spread.
By the numbers
Since the pandemic began Quebec has consistently seen the largest numbers comprising 47 per cent of Canada’s cases , followed by Ontario with 32 per cent and Alberta pulling up third with 11 per cent of the cases in Canada.
Right now, Alberta is the province with the third-most active cases of COVID-19. Tuesday, Quebec reported 877 cases, Ontario slightly fewer at 821, and Alberta third with 323.
When you weight the numbers for population, Quebec still tops out at 87.1 cases per 100,000 but Ontario drops to fourth, behind both Manitoba and Alberta where here we push the bar past 60 cases per 100,000.
Alberta has done the most per capita testing in Canada - almost 375,000 tests per 1M people, but we're not far ahead of Ontario and Quebec. The numbers fall off quickly after that.
Most experts agree the positivity proportion is a better indicator. Alberta's positivity level spiked early and levelled off around June. Right now, it sits at 1.78 per cent, but like all the COVID-19 statistics, it is a lagging indicator, usually reflecting what happened about two weeks ago.
If we are always looking back, how do we know what's ahead? Epidemiologists often look to the reproductive number. Here, Alberta isn't doing so well, sitting at second-highest in the country with an R(t) value of 1.56.
Alberta has seen it's highest-level of new cases in the past week, and it's also seeing higher hospitalization this month - the most since the pandemic was declared and it's risen sharply through October. This, according to most health officials, will be the key metric used to decide if further lockdowns are necessary.