New data on ‘excess’ deaths hint at scope of virus’s toll
But Ontario’s exclusion from the Statistics Canada report obscures the national picture
The first glimpse of all deaths across Canada during the time of the country’s COVID-19 battle shows how the pandemic footprint in Canada’s two westernmost provinces may be even bigger than previously acknowledged.
And data for the provinces hardest hit by COVID-19 is yet to come.
The scope of the coronavirus’s true potential toll remains largely obscured, because of the preliminary nature of the newly released Statistics Canada data, the early time period it covers — and the gaping black hole of information regarding Ontario, the province that was second hardest hit by COVID-19.
Following scrutiny over the lack of information, StatCan on Wednesday released preliminary data on the total death numbers in nine provinces between January and March.
Though the data is from a time period before the height of the pandemic in any province and, importantly, includes deaths not related to COVID-19, it offers the first suggestions of how many people may have died — directly or indirectly — as a result of the pandemic.
The data showed 318 “excess” deaths in Alberta during that time period, above what would be expected had no pandemic taken place. The data also showed some weekly death increases in B.C., where more extensive data previously released by public health officials concluded the province saw 170 “excess deaths” in March and April.
DEATHS continued on A10
In every other province for which data was provided, deaths either went down or remained close to the number for the same week in 2019.
B.C. provincial health officer Dr. Bonnie Henry, who last week reported on total death numbers compiled by the Centre for Disease Control in the province, said that some of the increase in deaths at the beginning of 2020 are explained by the number of people who have succumbed to the COVID-19 disease.
What scientists are still trying to determine is how many more deaths could be indirectly related to the pandemic — such as people who did not visit the emergency room out of fear of COVID-19, or people who had their elective surgeries cancelled.
“What we are looking at now is getting the cause of death information for every single death during that period of time,” she said on Wednesday.
Henry said she didn’t know why the data released by Statistics Canada appeared to show no excess deaths for most of the country, unlike B.C.’s data, but noted the time frame measured likely played a role.
Other provinces are further behind on answering this question. Due to a combination of delays processing data and epidemics that started later than the western provinces, it’s still unknown how many “excess” deaths may eventually be counted in Canada’s hardesthit provinces: Ontario and Quebec. Ontario did not release any death data to Statistics Canada in time for the release, and Quebec’s data for the months of January-March reflect only a time before the COVID-19 crisis began to spread in earnest.
Ontario’s Ministry of Government and Consumer Services did not respond to a request from the Star to clarify whether it provided this data to Statistics Canada.
Statistics Canada says information about deaths in Ontario and New Brunswick is often received more than 60 days following the death. Since this data is incomplete, it has been suppressed from the data set. Information is also missing on Nunavut, which the agency also deemed incomplete, and Yukon, which has reportedly not provided this information to Statistics Canada since 2017. In Quebec — which, as of Wednesday evening, has reported 3,200 deaths due to COVID-19 and accounts for the majority of Canadian deaths — the data shows the province is seeing an overall drop in deaths, with 1,257 fewer deaths compared to 2019.
Speaking on Quebec’s very low mortality numbers, Colin Furness, an infection control epidemiologist at the University of Toronto, speculated that a mild flu season may explain the dip.
Once Canada “shut everything down, sent people home, put masks on everyone — the flu goes away, too,” he explained.
“It’s also possible that the reporting is incomplete, that they sent the data and that it’s going to have to get corrected.”
With the data missing causeof-death information, it’s difficult to know the true reason for Quebec’s dip in deaths. “If we had cause of death, then we would have a sense of where these negative deaths are showing up,” Furness said.
What’s certain is that once death data is available for April and May, the numbers will increase. But that’s not the only reason the April data may prove more revealing.
In March, “cases were just getting going,” Furness said of the coronavirus pandemic. But in April, coronavirus-related deaths started building, “because people have to be sick before they die.”
Better data would separate out the cause of death and age group of those who had died. “Without cause of death and age group ... we’re not really going to be able to get the excess death answer that we wanted,” Furness said.
Post-mortem testing would provide better information about who has actually died from the virus, adding context to a novel virus we still don’t fully understand. “We really have an obligation not just to test people and write it down, but to actually … create data and use data to learn what this virus is and does, and we’re not,” he said.
Wednesday’s report comes just days after Ontario refused to publicly release the province’s mortality rates for all causes of deaths — including those from COVID-19 — for the first quarters of 2019 and 2020.
Last week, the Star began tracking down death statistics from every province and territory in an effort to capture the national death tolls over the early months of both years. All regions agreed to provide the information, except Ontario.
It’s unusual for Statistics Canada to provide interim mortality data. Historically, there is a reporting lag after a calendar year when all provinces and territories submit death counts to the federal agency tasked with collecting, then publishing the data. For instance, it may take up to a year in Ontario to register a death, meaning a death that occurred in May 2019 can still be registered this month.
Statistics Canada’s most recent official death tallies are from 2018.
The agency confirmed that it had received “some” data from Ontario for 2019, and limited death data for 2020. The data was deemed “insufficient to be able to release or to conduct a comparative analysis for the period from January to March. Receipt of data from Ontario is ongoing,” spokesperson Peter Frayne said.
The agency has also said that the need for up-to-date realtime data on Canada’s population, society and economy prompted them to move to release partial data earlier than usual.
“Statistics Canada stepped up its data collection and its dissemination of insights on the impacts of COVID-19 on businesses and individuals,” Frayne said.
Frayne also confirmed that additional data can be expected, including in jurisdictions where incomplete data was provided. “When additional data are received from other jurisdictions, we’re committed to making this information available in a timely fashion,” he said.
On Monday, the Ottawa-based Canadian Institute of Actuaries called on the federal government and public institutions “to expand the quality and depth of the data collected and made available in relation to COVID-19.”
Actuaries use mathematical and statistical methods to assess risk for insurance, financial and other industries; it’s a sciencebased profession that relies on data. By studying historical and current mortality data, for instance, actuaries can forecast the amounts of money needed to fund financial security systems such as pension plans.
Marc Tardif, president of the actuarial organization, reviewed the Statistics Canada death report with his team. He said while the agency’s template to chart excess deaths was “perfect,” the submitted data was lacking and required “lots of caveats” from Statistics Canada.
“In risk management and decision making, our health authorities and our governments are highly dependent, in our opinion, with ( having) as (much) precise data as possible,” Tardif said. “So if they can’t get precise data, the risk they’re running is making decisions that are not as data-based as we think they should be.”
Keith Walters, chair of the institute’s research council, said Ontario’s exclusion from the Statistics Canada table is a glaring data gap.
“We need to get Ontario in there,” said Walters, adding that sharing provisional information during a health crisis is more useful than a “blank.”
“For those of us used to doing analysis, we would probably say it’s better to report with appropriate disclosure (i.e. caveats about future data adjustments) around what the information’s basis is,” Walters said.
“We’d rather know what you have, with the basis that you have to produce it, than be blank. So even if there are known (data) weaknesses, it’s better to work with what you know and be able to draw lessons from that.”
Ottawa epidemiologist Ann Jolly said it’s unfortunate not all provinces and territories provided useful data for the Statistics Canada report, considering the practice of counting a community’s dead goes back centuries.
“The original purpose of death registration was to monitor deaths from disease, where it was used first in estimating cholera cases in England in the 1800s, and later became the standard globally,” said Jolly, a University of Ottawa associate professor.
“It is truly unfortunate that we are unable to access prompt accurate information on deaths of Canadians.”
However, Jolly said Statistics Canada’s decision to publish interim death data was a positive move. “It is possible to do these kinds of things, so we should be doing it,” she said. “We have the technology, we have the data, we just have to do it.”
Canada does not have a central death registry that is regularly updated for the public. That puts Canada out of step with many countries like the United States, Belgium, Germany, Norway and Ecuador, which during this pandemic are publicly releasing all-cause (including COVID-19) mortality rates in a timely manner — data that can help determine if excess deaths are occurring over a certain period compared to previous years.
Faster reporting of all deaths during the pandemic can reveal a variety of trends, including whether unexplained deaths rose quickly before COVID-19 was declared a pandemic, or if delayed surgeries or hospital visits for other health issues potentially cost lives, experts say.
“We really have an obligation not just to test people and write it down, but to actually … create data and use data to learn what this virus is and does, and we’re not.” COLIN FURNESS INFECTION CONTROL EPIDEMIOLOGIST, UNIVERSITY OF TORONTO