Inconsistent COVID death data useless says ethicist

A prominent Canadian ethicist says Hamilton public health was right to remove six deaths from its pandemic count.

Although, an infectious disease specialist says the deaths shouldn't be hidden from view altogether.

Both raise concerns about the reliability and consistency of COVID death data.

If you've got different criteria being used in different places, then you haven't got data, you've got junk," said Arthur Schafer, founding director of the Centre for Professional and Applied Ethics at the University of Manitoba. It's really important that it be accurate ... and comparable across time and across jurisdiction otherwise we can't hold governments accountable and we can't make decisions as to which policies we support as individuals or as communities."

Public health completely reversed its definition of a COVID death on Nov. 4. Up until that time, the death of an active case of COVID would be counted no matter what. Now, the virus has to be a contributing or underlying cause of death." In addition, there must be no clear alternative cause of death."

The change resulted in six deaths being withdrawn, leaving Hamilton listing a total of 416 pandemic deaths as of Tuesday.

It can't be ethically right to list deaths as COVID deaths when you know that COVID didn't cause them or contribute in any way," said Schafer. That can't be morally right. I think it robs people of the clarity and information they need when they're trying to figure out risks and benefits and how they're going to live their lives."

A member of the Ontario COVID-19 Science Advisory Table agreed that certain deaths shouldn't be included. Those who die of another cause who were infected with COVID but never became ill from the virus shouldn't be a pandemic death.

It will make it look like more people died of COVID but they really died with COVID," said Dr. Andrew Morris, a professor of infectious diseases at the University of Toronto and director of the Antimicrobial Stewardship Program at Sinai Health and the University Health Network.

At the same time, he wasn't convinced excluding the deaths from the reporting was right either.

Ideally we see both and can see how much this makes a difference." said Morris.

Haldimand-Norfolk has found a work around by reporting two categories of deaths - COVID related and unrelated.

Hamilton appears to have rejected this solution and removed the deaths from its reporting altogether.

Four were in outbreaks in palliative care at St. Peter's Hospital in March and November, one was in an outbreak on 5 West at Hamilton General Hospital in February and one was in an outbreak at Emmanuel House hospice in April 2020.

Keeping them in the count would have been misleading," said Schafer.

The most precious resource that public health possesses is the trust of the public," he said. If the public believes that the ... death rate caused by COVID has been exaggerated knowingly and deliberately, the trust public health enjoys will suffer. Public health officials have a duty to be scrupulously open and honest and transparent."

However, Morris says the answers aren't black and white.

This is a really complicated issue especially because cause of death is so poorly documented in Canada," said Morris. The problem with all of this - we are really not good at ascribing cause of death very well."

He gives the example of someone with COVID dying of a stroke.

Is this secondary to COVID or coincidental," he said.

Adding to the confusion is that Hamilton's definition of a COVID death no longer matches the province - it's completely opposite.

The province states on its website: Deaths are included whether or not COVID-19 was determined to be a contributing or underlying cause of death."

Public Health Ontario (PHO) defines it on its website as, People with laboratory confirmation of COVID-19 who have died."

However, PHO provided all of Ontario's public health units with a different definition: A COVID-19 death is defined as a death resulting from a clinically compatible illness in a COVID-19 case, unless there is a clear alternate cause of death that is unrelated to COVID-19."

The above definition also matches the World Health Organization.

By making the change, Hamilton was getting in line with the rest of the province's public health units, said Dr. Bart Harvey, an associate medical officer of heath. It's not clear how Hamilton came to be doing things differently in the first place.

However, PHO acknowledged in a statement Tuesday that its instructions get interpreted and applied" by Hamilton public health.

It should be standardized," said Schafer. The criteria should be well known."

It's especially important considering, Death rates are used for comparative purposes so we can judge or evaluate which policies are working and which are failing," he said.

If we're using a totally different method of calculating the death rates so now we're comparing apples to oranges, the data becomes useless."

Joanna Frketich is a Hamilton-based reporter covering health for The Spectator. Reach her via email: jfrketich@thespec.com