The COVID-19 pandemic

[tomh009]

In the past week there have been 300 cases (-18) of B.1.1.7 (UK), four cases (-2) of B.1.351 (SA) and 10 cases (+4) of P.1 (BR) variant.

So, it turns out that this really is quite misleading. Not all samples can be tested for the N501Y mutation (which all of the above variants have); the number screened for N501Y seems to fluctuate around 1,000 per day, or only about 2% of the total number of COVID-19 tests.

The percentage of these that test positive for N501Y has gone up from about 15% in mid-February to over 30% recently.

These samples then go through genome sequencing to determine whether they match one of the above variants; this takes multiple days. It's not clear whether all N501Y-positive are sequenced. If they do sequence all of them, then it appears that the variant positivity rate is somewhere below 20%.

But applying the 30% N501Y positivity rate and the 20% variant positivity rate means that maybe somewhere between 5% and 15% of the COVID samples (that are tested for N501Y) are positive for one of the variants. The case numbers posted make it look like the variants make up only about 0.1% of the cases in Ontario, but I think, based on the data above, that the actual percentage of variants may be between 10x and 100x that.

I'm going to stop posting the variant numbers until I can determine some more meaningful numbers.