TORONTO -- As epidemiologists scramble to develop models to project the future course of the COVID-19 pandemic, health officials are trying to keep track of the increasing number of deaths from the coronavirus disease.

While it may sound simple enough, the reality is far from it.

That’s due, in part, to the difficulty in trying to accurately identify every victim of COVID-19. In a hospital setting, it’s fairly easy in the sense that patients with symptoms are swabbed and tested for the respiratory illness.

If that patient dies, they’re included in the official daily death counts for the disease. Even if that individual had another underlying health problem, such as a heart condition, which was then exacerbated by the virus and resulted in a heart attack, their death would still be recorded as a death from COVID-19.

That’s according to the World Health Organization’s definition of what constitutes a COVID-19 death. The world health body has recently released a new ICD (International Classification of Diseases) code for COVID-19 to be included on death certificates in order to capture mortality data for the disease.  

Dr. Prabhat Jha, the director for the Centre for Global Health Research at St. Michael’s Hospital in Toronto, explained the new code is supposed to be applied to cases in which the patient tested positive for COVID-19, the patient had a probable case of COVID-19, but lab results are still pending, or if the patient tested positive for COVID-19, but there is another cause of death.

“That’s the default because in an epidemic situation, you don’t want to miss any deaths if you’re tracking where this is going,” the epidemiologist, who specializes in death statistics research, told during a telephone interview from Toronto on Friday.

Tracking mortality rates for the disease becomes substantially trickier, however, when people suspected of having COVID-19 die before they are tested.

Dr. Christopher Milroy, a forensic pathologist and the director of the Eastern Ontario Forensic Pathology Unit at the Ottawa Hospital, said it’s possible to conduct detailed death investigations and autopsies for these cases, but that won’t happen because they don’t have the resources. What’s more, a shortage of testing kits means the living get priority during a pandemic.

“Unless you’re going to test every one, autopsy every one, you’re never going to be 100 per cent [certain],” he told during a telephone interview from Ottawa on Friday.

The death rate may appear higher in places where they’re only testing people with symptoms in the hospital as opposed to areas where testing is more widespread, such as in Germany and Iceland.

“The problem with mortality stats at the moment is they’re partly based on the number of people who tested positive and then died, and if we haven’t tested everyone… we don’t know the exact mortality, which is why people are varying in their figures.”


While both Jha and Milroy said it’s difficult to know the exact number of deaths from the COVID-19 pandemic as it’s happening, they said a clearer picture will likely emerge in its aftermath.

In order for that to happen, Jha said it’s important for government officials to double down on their efforts to keep track of all of the deaths in a province or city, not just those related to COVID-19, and publish them in a timely manner so researchers have a common denominator to work with when they’re monitoring the pandemic.

For example, Jha said if a senior dies from pneumonia in a nursing home, that death should be recorded and published on the province’s online registry ideally the same day, instead of a week later, which is typically the case. He said those daily reports should include the number of people dying, the ages of those people, and hopefully, a cause of death.

“Then you can look at the curves and say ‘Wait a minute, there was a big increase in 80 year olds dying and a big increase in pneumonia deaths dying at the same time that we had the COVID outbreak so those are probably COVID deaths,” he explained.

Instead of relying on modelling projections, Jha said they can use the patterns identified in the overall mortality data to forecast the trajectory of the pandemic.

“If you track daily deaths, then you've got some sense of where the epidemic is going and whether it's turning the corner,” he said. “It requires the data systems to not be on the kind of the usual path, they have to be a little bit more on the war footing.”

Milroy said because it’s always challenging to get accurate statistics during an epidemic, he said Canada and other countries like it that are keeping track of mortality data won’t know the full extent of it until months or even years later.

“I think the bottom line is: we’ll know more, unfortunately, after this event is over,” he said.