Back to school to work out test rates

Every time the first minister gives her daily coronavirus address and says she knows there are people out there who watch these numbers very closely, I feel she is speaking directly to me.

This time she was explaining a change in how the government calculate the testing positivity rate. This metric became more prominent in May when the World Health Organisation stated that countries should aim to have a positivity rate of less than 5%.

Unfortunately there was no general consensus as to how this percentage should be calculated.

Sadly, to understand these changes I have to use words that if you’ve made wise life choices you may not have heard since school: Numerator and denominator.

The new methodology makes changes to both in the calculation. The previous numerator (the number at the top of the fraction) was the number of new cases, whereas the new method makes a subtle change to using the number of positive tests. These numbers can be different if someone is tested (and is positive) more than once on the same day.

The denominator (the one at the bottom) here was the trickier number to define. The Scottish Government made the initial decision that it should be the number of newly-tested individuals.

The rationale behind this was an attempt to adjust for the number of people who are tested regularly as part of their jobs. Only their first test would be counted in the

“newly tested individual” category. Unfortunately, this causes issues with the usefulness of the metric at this stage.

As more and more people are tested, the pool of people able to appear as a “newly tested individual” shrinks and has the potential to inflate the positivity rate.

The new methodology would see the total daily tests reported used as the denominator. Using this new metric sees the positivity rate drop by more than 10% on the old method – but at 6.4% it’s still above the WHO red line of 5%.