Protecting Canadians from COVID-19: How are we doing?

A serosurvey measures antibodies among a group of people. In a similar manner, a thermometer measures the outdoor temperature. The measured temperature is useful, but to really know how we will feel outdoors, a formula or model is used to combine temperature readings with other environmental data, such as wind in the winter and humidity in the summer. In the case of serosurveys, the CITF has developed a statistical model to combine antibody results with other epidemiological data, such as confirmed cases and vaccine coverage, so that we can know how many Canadians are protected from COVID-19.

CITF Modelling uses results from CITF-supported serosurveys and published serosurvey results. This work is conducted in close collaboration with SeroTracker, which documents published serosurveys and plays a central role in modelling. Many of the serosurvey results used in modelling are provided by Canadian Blood Services (see details of methods and data below).

January 2022 | Natural antibody wane reduces immunity from infection, reinforcing importance of vaccination

In this month’s analysis, we modelled the effect of waning antibodies, known as seroreversion, to account for the natural loss of antibodies that occurs through time. Our calculations, based on the most recent research on time to antibody wane, found that as of November 30, 2021, an estimated 1 in 21 Canadians (4.7%) still had detectable antibodies due to a previous infection. This proportion ranged from a low of 1 in 69 (1.5%) in the Atlantic provinces to a high of 1 in 13 (7.7%) in Alberta. The data were prior to widespread exposure to the Omicron variant.

Notably, this estimate is lower than our previous month’s estimate of seroprevalence (7.8%), which did not incorporate seroreversion rates, suggesting that antibody wane has a profound effect on the accumulation of people with detectable antibodies in the Canadian population. This being the case, it is evident that vaccination is crucial to increase the level of immunity throughout the population.

Double the number of reported COVID-19 infections across Canada

Accounting for seroreversion also made it possible to estimate the number of true infections in the population prior to Omicron. We assumed that seroprevalence due to infection results from infections 7 to 10 days before the serosurvey or earlier. Although 1.8 million cases of COVID-19 had been reported by November 21, 2021, our modelling estimates that there had been 3.1 million infections in Canada.

Our analysis combined seroprevalence results from 35 studies with data on confirmed cases of SARS-CoV-2 infection to determine the proportion of Canadians infected with COVID-19.

How many Canadians have antibodies to SARS-CoV-2?


Estimated proportion of Canadians who have antibodies, not taking into account antibody waning


Estimated proportion of Canadians who have antibodies, taking into account antibody waning

1.8 million

Number of cases reported

3.1 million

Estimated real number
of infections

Canadian seroprevalence from infection lower due to antibody waning

When accounting for antibody waning, about 4.7% of Canadians have SARS-CoV-2 antibodies as the result of a previous COVID-19 infection. This is in contrast to the previously reported unadjusted estimate that 7.8% of Canadians had SARS-CoV2-antibodies by the end of November 2021.

Earlier in the pandemic, antibody waning, or seroreversion , did not have a significant effect on estimates of seroprevalence in Canada; antibody levels in individuals who were infected with SARS-CoV-2 persisted for several months. However, as the pandemic progressed into April 2021, the prevalence of antibodies due to infection began to plateau, and then decrease in July of 2021. These results suggest that seroreversion is important to account for when measuring levels of antibodies due to infection.

Confirmed cases Estimated seroprevalence

Rise over time in Canadian Infections

When conducted in the months immediately following infection, antibody testing detects nearly all people with prior SARS-CoV-2 infection. In contrast, COVID-19 cases are usually confirmed among people with symptoms, who represent a fraction of all infections. Consequently, there can be a difference between confirmed cases and true infections. Estimated true infections have steadily risen through time, as have confirmed cases. However, estimated infections have increased at a higher rate because of the accumulation of mild or asymptomatic cases, which are generally not tested. This situation has led to a constant increase in the number of undetected infections.

Confirmed cases Estimated seroprevalence

Seroprevalence due to infection, confirmed cases,
and deaths by province

The epidemic has swept across Canada in waves, striking different provinces (or regions) at different times. The charts below animate this phenomenon, showing how provinces have ranked by seroprevalence due to infection (top) and total true infections (bottom) over time.

By the end of November 2021, Alberta and Manitoba had the highest number of estimated true infections per 100,000 citizens as well as highest proportion of the population with antibodies to SARS-CoV-2 due to infection. In general, seroprevalence and estimated infections were comparable across provinces.

Driving the regional differences in current estimated infections is the total number of cases through time. The highest number of confirmed cases per capita has been in Alberta and Saskatchewan, and the lowest in the Atlantic Provinces and Territories (anti-N seroprevalence data not available). The magnitude of undocumented cases can be inferred from the color bar on the right. Alberta and Saskatchewan were the provinces where true infections were underestimated the most.

Confirmed cases per 100k Estimated number of people with antibodies due to infections per 100k

Background information

Why are serosurveys important?

SARS-CoV-2 serosurveys are studies that aim to estimate the proportion of the population with antibodies to SARS-CoV-2, the virus that causes COVID-19. The presence of SARS-CoV-2 antibodies is strongly correlated with having protection against a serious COVID-19 infection. When a very high proportion of the population has antibodies, the virus spreads much more slowly, However, judging from recent seroprevalence findings, and seroreversion or reduced immunity against new variants, it may not be possible to vaccinate enough people to effectively block SARS-CoV-2 transmission in the Canadian population, as we have done for other vaccine-preventable infectious diseases.

How can you tell if people had antibodies due to an infection or due to vaccination?

Serosurveys can measure the presence of different types of SARS-CoV-2 antibodies in large groups of people in a cost-efficient manner. SARS-CoV-2 infection and the COVID-19 vaccines used most commonly in Canada cause the body to produce antibodies to the virus’ spike protein (abbreviated as anti-S). Infection also induces antibodies to the virus’ nucleocapsid (abbreviated as anti-N).

While having both anti-S and anti-N does not seem to provide better protection than anti-S alone, it does allow researchers to measure whether a study participant had COVID-19 in the past – generally, at least 3 weeks or more prior to testing. See our FAQ page for more information.

Antibody Abbreviation Interpretation
anti-S alone Vaccination induced antibodies
anti-S and anti-N Infection induced antibodies

How does CITF estimate seroprevalence?

Because not every Canadian can, or wants to, donate frequent blood samples (such as from a finger prick), estimating updated seroprevalence requires information from many sources.

  • The CITF uses serosurveys from CITF-supported research groups and published serosurveys documented by SeroTracker. This includes Statistics Canada, and Canadian Blood Services.
  • Cases and testing data is coming from the Covid19 Canada Open Data Working Group
  • The correlation between seroprevalence and total reported cases up to one week prior to the midpoint of the serosurvey sampling date range is used in each region to smooth and extend the information provided by serosurveys.

For more information about our data sources and statistical methods, see our technical reports.