Canada’s public data on COVID-19 is (mostly) a mess. Here’s how to find useful information

0
64


The COVID-19 pandemic in Canada ushered in the golden age of amateur graphics on epidemiological curves. Wherever you turn, someone has a different graphic.

The problem is, much of it is based on limited, incomplete, and often out-of-date public data, which means it can lead to wrong conclusions about what’s really going on.

This is not all bad news. Public health agencies are slowly improving by providing detailed and up-to-date information. Delays in testing are starting to disappear and the capacity of laboratories is increasing. But it remains very difficult to find detailed data sources and accurately compare jurisdictions, and most graphs should always be viewed with skepticism.

Here is a guide to known issues with our data right now, and how to sort through the mess to find what’s really useful.

Pay less attention to the number of cases, more to hospitalizations

Most of the graphs you see are based on the number of “confirmed cases”, which are tests for COVID-19 performed by a qualified laboratory. But these counts present a very limited picture of a country’s situation – and sometimes very distorted.

The number of cases is still about two weeks late due to the delay in people developing symptoms and being screened. They also significantly underestimate the actual number of cases due to the limited testing capacity.

However, the number of cases in Canada is particularly difficult to analyze due to the differences in provincial testing. BEFORE CHRIST. and Alberta, for example, began testing on a large scale and then tightened its criteria for high priority cases. Ontario and Quebec are the opposite: they started slowly and are now testing faster and more widely (although Quebec has progressed much faster).

Also, beware of misleading peaks in the number of daily cases. They are usually due to a reduction in arrears or a change in the way tests are handled, not a real daily increase in cases.

A patient was brought to the emergency department of the Verdun Hospital on April 2, 2020 in Montreal.


Ryan Remiorz / The Canadian Press

For all these reasons, the number of cases should be treated as a single indicator of the situation in a country. Hospitalizations and deaths are more reliable parameters, although there are also caveats for both.

People hospitalized for breathing problems have high priority for testing, so these numbers are less affected by different testing standards and delays. Hospitalizations – especially intensive care statistics – are also crucial to track, as COVID-19 becomes particularly deadly when hospitals are short of capacity.

Provincial health agencies have been slow to provide hospitalization information on their websites, but most are doing so now. For the best example, the Quebec health data agency now displays a tracking chart of hospitalizations and intensive care cases over time. Ontario released a hospitalization count for the first time on Thursday.

However, hospitalizations are a slower indicator than confirmed cases, so be aware that you are seeing the situation as it was more than two weeks ago.

Deaths from COVID-19 are also a more accurate picture of the spread of the disease, but this indicator is even more lagged behind hospital admissions. Deaths can also be disproportionately affected by outbreaks in long-term care homes.

Find the most local data

The Public Health Agency of Canada recently started publishing a daily epidemiological summary of COVID-19, including breakdowns by age group and common symptoms.

But the report comes with a warning: they only have detailed information on barely half of the cases, as they depend on the provinces that submit case reports. Of the 9,017 cases confirmed in their April 1 report, the federal authorities only had information on age and sex for 5,590 of the cases, and only had hospital status for 3,177 of the cases .

A general rule of thumb is that the higher a government body, the less up-to-date it is.

This was clearly illustrated in Ontario, where provincial medical officials have been baffled by reporters wondering why provincial figures are so far from local public health units.

For example, on Wednesday, journalists asked the Ontario Deputy Chief Medical Officer of Health why local health units collectively reported 65 COVID-19 deaths, when the official number in Ontario was still 37. “I am a little surprised by this gap, I must say, “Barbara Yaffe replied. She said the province’s numbers depend on the health units that enter the data.

If you are collecting data, first check the most local levels of the health authority. Many are now reporting their own COVID-19 data, and it may take days before that data reaches higher levels.

Follow the experts

Many people have the basic knowledge of coding or software to create their own graphics. This does not mean that they know only one thing in epidemiology.

But there are a lot of people who are subject matter experts and comment publicly on the data, often on their social media accounts, although some have created websites. If you want to know what’s going on, find infectious disease experts, intensive care doctors, statisticians and public health officials, and pay close attention to what they say.

As a recent example, Ontario is now (finally) releasing an open data set for its individual COVID-19 cases. At first glance, putting it in a graph might show that Ontario is bending the curve.

But David Fisman, epidemiologist at the University of Toronto’s Dalla Lana School of Public Health, ran this data through its own modeling with adjustments for known data breaches. His conclusion? COVID-19 is still running in Ontario.

“Unfortunately, we now have to retract our five years,” Fisman tweeted. “But it is important to know where we are … We are not out of the soup yet. “

• Email: [email protected] | Twitter:



LEAVE A REPLY

Please enter your comment!
Please enter your name here