Monday, March 16

I am sick again with a cough, headache, and runny nose – but no fever. The BC CDC has a very useful self-assessment tool that told me I do not need COVID-19 testing. However, it also informed me that “the BC Ministry of Health strongly encourages any individuals experiencing symptoms (fever, cough, sneezing, sore throat, or difficulty breathing) to stay home from work and/or school, and avoid going out in public where possible”, so I am stuck at home.

On the outside, the COVID-19 pandemic is advancing quickly. Italy is the hardest hit country outside of China, with 2,158 deaths and 27,980 confirmed infections. Due to the steepness of the infection curve, the health care system is overwhelmed, and doctors are being forced to make triage decisions.

The Washington Post has a very good animation demonstrating how diseases spread through the population. Keep in mind that this is a model, and does not perfectly reflect how COVID-19 is spreading. As one example of the limitations of the model, one researcher noted that “some of the dots should disappear”.

Since I am at home, I decided to look at the number of COVID-19 cases in BC and Canada more closely. I used information from Wikipedia and CTV News, which agree with each other very closely.

Using March 1 as Day 0 (Day 29 is Monday, March 30, when we should return to school), the exponential curve looks like this:

Note how well the data fits the curve – an R² value of 1.0 would means that the data matches the exponential growth trendline perfectly. The fit did not change much depending on what date I chose for Day 0.

In BC, I chose March 9 as my Day 0 (Day 21 is Monday, March 30, when we should return to school). I chose this date because it is where the data starts to become less noisy. The trendline is very similar to the trendline for all of Canada.

According to this simple exponential model, on March 30, BC will have about 1,025 confirmed cases of COVID-19, and Canada will have about 4,214 confirmed cases.

It is important not to extrapolate too far out into the future – at some point the growth will slow, and the number of cases will no longer follow the exponential growth model I have used here.