Disease burden and threat in Canada
Our reactions to COVID-19 should recognize threat isn’t spread equally across population
In the past few months, most of us have experienced acute anxiety about the threat of contracting the virus SARS-CoV-2, developing the disease COVID-19, becoming seriously ill, requiring hospitalization, ventilation and dying.
A technical report published by the Public Health Agency of Canada (PHAC) on April 9 estimated that with stronger mitigation controls, up to 10 per cent of Canadians would be infected, with 44,000 deaths; and with weaker mitigation controls, up to 50 per cent would be infected with 220,000 deaths. Expressed as deaths per thousand, this latter estimate (5.8/1,000) is just under the estimate for 1918-20 Spanish flu pandemic (6.7/1,000). Are these projections of disease burden meaningful?
Disease burden in pandemics is measured in three ways: 1. by the number of infections: a count of individuals (cases) testing positive for SARS-CoV-2; 2. by the number of deaths: a count of individuals assigned COVID-19 as the cause of death; and 3. the case fatality rate (CFR): the percentage of individuals dying among those with infection.
The number of infections is not a useful measure of disease burden for three reasons.
One, the number of infections depends on the number of individuals tested each day — the more individuals tested, the more cases of infection identified. Two, concern about transmission of the virus to health care practitioners and the urgency to help those who are ill mean that individuals seeking health care with suspected infection are priorities for testing. These individuals, more likely to test positive for the virus and at greater risk for death due to pre-existing medical conditions, exaggerate the number of infections linked to deaths among those tested. Three, we know little about the symptom profile of the 90 to 95 per cent of individuals testing positive for the virus but not dying.
Until general population surveys of individuals with the virus are implemented and variations in symptom response are clarified, the burden of disease associated with infection will remain unknown.
Loss of life is the ultimate and most meaningful indicator of disease burden in pandemics.
On May 13, there were 5,304 deaths in Canada attributable COVID-19. However, it is not the count of deaths but their distribution that is important. Residents of longterm care homes account for more than eight in 10 of these deaths. Among Canadians hospitalized with COVID-19, less than one per cent are under 20 years old with no deaths; two-thirds are aged 60 or older, accounting for 95 per cent of all deaths; and about 74 per cent have one or more pre-existing medical conditions.
The case fatality rate (CFR) is a misleading measure of disease burden because it carries over the distortions associated with counts of infection, inflating estimated risk for death. Our CFR is currently 7.2 per 100. The CFR used in the PHAC report was 1.2 per 100. The actual CFR could be less than half of that.
A threat is a perception that something bad is going to happen. Watching the counts of infections and deaths multiplying each day has created a mesmerizing sense of threat for everyone. This threat has been brought to life by local death counts and individual stories of personal tragedy — the loss of a relative, a friend, a notable figure — as well as stories of extraordinary heroism and sacrifice — health care providers doing their job in the face of great risk and desperate need.
A burden is something we must bear — grief, pain, loss of function and, ultimately, death. Although the threat of COVID-19 is being perceived by everyone, the burden of the disease is not experienced equally. Unlike previous pandemics, the young have been spared, but not the elderly or those with predisposing health conditions. We have known for months about the vulnerability of these groups yet attention to them has only come after the recent intense media coverage of long-term care homes.
The estimates of infections and deaths included in the report by PHAC magnify threat for everyone and are wholly unsuitable baselines for determining the number of lives saved by government policies.
This focus on threat has made for public acceptance of extreme, restrictive and disruptive policies, even though the estimates are based largely on unproven assumptions and poor quality information.
Furthermore, this focus on threat has deflected attention from an important task in a pandemic: determining if there are subgroups in the population at excessive risk for dying because of COVID-19; and if there are, putting in place specific, targeted interventions to lower their probability of exposure to SARS-CoV-2.
The characteristics of those dying because of COVID-19 is the best information we have on this disease and we need to use it in planning for the next steps in dealing with this pandemic.
Michael H. Boyle is Professor Emeritus of the Michael G. DeGroote School of Medicine of McMaster University, Hamilton. Laura Duncan is an assistant professor (part time) at McMaster.