May 13: Ontario is finally lowering new daily deaths

Daily new COVID-19 deaths in Ontario might finally be going down.

5-day rolling average to smooth out fluctuations

“Might” because we’ve seen declines before that reverse. We need to give it another week to be sure, but it seems like even the number of deaths in Ontario (blue line) is decreasing. That, sadly, isn’t the case for Quebec, which has been relatively flat the entire month of May.

For Ontario, however, I am fairly confident that the number of deaths is declining as well, looking at the number of patients in ICU that are also declining (that chart is shown later).

Same data as the above chart, but daily data instead of a 5-day average

For reasons beyond my understanding, Ontario (and Quebec) are both struggling to sustain a healthy rate of testing. Ontario has been fluctuating:

This makes it hard to compare new infections, since that’s affected by the number of tests:

One way I found to make sense of it is to compare what percentage of tests came positive:

The 4-day moving average evens out fluctuations, this rate has been coming down (although it has flattened in recent days).

One clear way of determining the peak is to look at the load in ICU:

Ontario peaked on April 26th and has been declining since then. There were 252 ICU patients on April 26th. Today’s update reported 189 ICU patients.

This data gives me confidence that we are in a decline.

In another set of good news for Canada, the number of deaths increased by < 3% today, that’s the lowest so far.

Here’s the zoomed-in version:

Obviously, as the total number of deaths goes up, the percentage of increase will come down new deaths are constant. There-in lies our success: with time, you expect new deaths to rise until you’ve controlled the infection. The fact that the number of new deaths seems to be declining as well is a great sign.

This chart actually shows that while Canada did a good job controlling cases, it saw higher deaths for the given number of cases. Germany had 1.8 times as many deaths in Canada in the early stages. That number has fallen below 1.3x now. That’s mostly because of the outbreaks in for-profit longterm care and nursing homes.

Looking at the data for Ontario, I am confident that we’re in a decline. Now the question rests on Quebec:

It’s not clear yet whether Quebec is declining or not. IF you look at the history, the decline might be a real decline or it might be a part of the regular fluctuation.

Quebec only ran 9,000 tests (well below its goal of 14,000), which makes it hard to understand the issue there.

Same chart as above, but with a 5-day rolling average

This 5-day average chart smooths out the fluctuations. You can see that Quebec had a decline in early April, but started going up again after April 13th.

The optimist in me wants to believe that Quebec peaked on May 3rd. We will know for sure in another week.