Canada Succeeds on Key Covid Metric

The map shows that in Canada 8839 deaths have been attributed to Covid19, meaning people who died having tested positive for SARS CV2 virus.  This number accumulated over a period of 148 days starting January 31. The daily death rate reached a peak of 177 on May 6, 2020, and is down to 11 as of yesterday.  More details on this below, but first the summary picture. (Note: 2019 is the latest demographic report)

Canada Pop Ann Deaths Daily Deaths Risk per
2019 37589262 330786 906 0.8800%
Covid 2020 37589262 8839 60 0.0235%

Over the epidemic months, the average Covid daily death rate amounted to 7% of the All Causes death rate. During this time a Canadian had an average risk of 1 in 5000 of dying with SARS CV2 versus a 1 in 114 chance of dying regardless of that infection. As shown later below the risk varied greatly with age, much lower for younger, healthier people.

The Key Covid Metric

With easing of lockdowns and increased testing in many places, epidemiologists are focusing on a key metric to inform public policies: Positivity. The positivity metric is the rate (%) of people who test positive out all people sampled. The significance is that (by definition) a presumed case is a person who tests positive once. If a second test comes back positive it is a confirmed case. The metric is not perfect for two reasons.

The first problem is false positives from the testing procedure itself or from errors in the data processing and reporting. For this we have to hope that quality assurance protocols are being followed and mistakes corrected along the way.

The larger issue appeared in Florida recently when officials discovered that numerous batches of samples were reported 100% positive and other batches 100% negative. While the latter result is expected sometimes, all people testing positive seems unlikely. Behind this is the reality that in many situations (eg hospital ICU) a single patient will be tested many times with many positive results in the course of monitoring that individual’s clearing of the virus. Obviously a batch of samples from that ICU might legitimately be 100% positive.

But it is also true that 10 or 20 positive tests from one patient should not be reported as 10 or 20 new cases. In some jurisdictions, officials say they go to the effort to link test results to the individuals tested, and can distinguish between number of cases and number of positives. In other places, cases and positives may be the same number. Thus confirmed cases could be only 1/2 of the total positives, or less.

How is Canada Doing?

Recoveries are calculated as cases minus deaths with a lag of 24 days. Daily cases and deaths are averages of the seven days ending on the stated date. Recoveries are # of cases from 24 days earlier minus # of daily deaths on the stated date. Since both testing and reports of Covid deaths were sketchy in the beginning, this graph begins with daily deaths as of April 24, 2020 compared to cases reported on March 31, 2020.

The line shows the Positivity metric for Canada starting at nearly 8% for new cases April 24, 2020.  That is, for the 7 day period ending April 24, there were a daily average of 21,772 tests and 1715 new cases reported. Since then the rate of new cases has dropped down, now holding steady at ~1% for the last month. Yesterday, the daily average number of tests was 42,191 with 363 new cases. So despite double the testing, the positivity rate is not climbing.

Another view of the data is shown below.

The scale of testing has increased and is now exceeding 40,000 a day, while positive tests (cases) dwindled to 1%.  The shape of the recovery curve resembles the case curve lagged by 24 days, since death rates are a small portion of cases.  The recovery rate has grown from 83% to 97% steady over the last 3 weeks. This approximation surely understates the number of those infected with SAR CV2 who are healthy afterwards, since antibody studies show infection rates multiples higher than confirmed positive tests. In absolute terms, cases are now down to 363 a day and deaths 11 a day, while estimates of recoveries are 345 a day.

Note: We are expecting an initial report from the National Immunity Task Force any day now regarding a major program of testing random blood samples for SARS CV2 anti-bodies.

Aside:  While preparing this post I was watching CBC channel, and the scroll bar had the text:  Canada passes 100,000 Covid cases.  I thought, what was the point of that? Then I realized:

Canada Mask2

My bad. That is the Health Minister of PEI solemnly announcing in July their first Covid case in two months.

Background Updated from Previous Post

In reporting on Covid19 pandemic, governments have provided information intended to frighten the public into compliance with orders constraining freedom of movement and activity. For example, the above map of the Canadian experience is all cumulative, and the curve will continue upward as long as cases can be found and deaths attributed.  As shown above, we can work around this myopia by calculating the daily differentials, and then averaging newly reported cases and deaths by seven days to smooth out lumps in the data processing by institutions.

A second major deficiency is lack of reporting of recoveries, including people infected and not requiring hospitalization or, in many cases, without professional diagnosis or treatment. The only recoveries presently to be found are limited statistics on patients released from hospital. The only way to get at the scale of recoveries is to subtract deaths from cases, considering survivors to be in recovery or cured. Comparing such numbers involves the delay between infection, symptoms and death. Herein lies another issue of terminology: a positive test for the SARS CV2 virus is reported as a case of the disease COVID19. In fact, an unknown number of people have been infected without symptoms, and many with very mild discomfort.

This discussion takes the assumption that anyone reported as dying from COVD19 tested positive for the virus at some point prior. A recent article by Nic Lewis at Climate Etc. referred to evidence that the average time from infection to symptoms is 5.1 days, and from symptoms to death 18.8 days. A separate issue, of course, is that 95+% of those dying had one or more co-morbidities contributing to the patient’s demise. Setting aside the issue of dying with/from Covid19, it is reasonable to assume that 24 days after testing positive for the virus, survivors can be considered recoveries.

Previous Post May 20, 2020

It is really quite difficult to find cases and deaths broken down by age groups.  For Canadian national statistics, I resorted to a report from Ontario to get the age distributions, since that province provides 69% of the cases outside of Quebec and 87% of the deaths.  Applying those proportions across Canada results in this table. For Canada as a whole nation:

Age  Risk of Test +  Risk of Death Population
per 1 CV death
<20 0.05% None NA
20-39 0.20% 0.000% 431817
40-59 0.25% 0.002% 42273
60-79 0.20% 0.020% 4984
80+ 0.76% 0.251% 398

In the worst case, if you are a Canadian aged more than 80 years, you have a 1 in 400 chance of dying from Covid19.  If you are 60 to 80 years old, your odds are 1 in 5000.  Younger than that, it’s only slightly higher than winning (or in this case, losing the lottery).

As noted above Quebec provides the bulk of cases and deaths in Canada, and also reports age distribution more precisely,  The numbers in the table below show risks for Quebecers.

Age  Risk of Test +  Risk of Death Population
per 1 CV death
0-9 yrs 0.13% 0 NA
10-19 yrs 0.21% 0 NA
20-29 yrs 0.50% 0.000% 289,647
30-39 0.51% 0.001% 152,009
40-49 years 0.63% 0.001% 73,342
50-59 years 0.53% 0.005% 21,087
60-69 years 0.37% 0.021% 4,778
70-79 years 0.52% 0.094% 1,069
80-89 1.78% 0.469% 213
90  + 5.19% 1.608% 62

While some of the risk factors are higher in the viral hotspot of Quebec, it is still the case that under 80 years of age, your chances of dying from Covid 19 are better than 1 in 1000, and much better the younger you are.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s