Thread by @Tim_Bousquet

A short explanation of why I don't track testing positivity rates in Nova Scotia /thread

The change, almost day to day but certainly week to week, over who gets tested makes any positivity comparison over time meaningless.

A year ago, NS was using PCR testing for only about symptomatic people a day. There was no asymptomatic testing, and no antigen testing.

Over the next year, that changed. First close contacts and then travellers were added to the PCR testing. Eventually, the labs (to their great credit, btw) got up to testing tens of thousands of people a day, the vast majority of whom were asymptomatic.

Throw into that mix Lisa Barrett's pop-up testing sites, doing antigen testing, on some days doing as many 5,000 tests.

But then the backlog hit, and for a week or so the primary assessment centres stopped testing asymptomatic people completely. But those people could get tested at the pop-up sites, and if they tested positive there, they'd be sent for PCR testing.

Now the backlogged is clear, and the primary assessment centres are back to doing PCR tests for asymptomatic people, but many people are still regularly going to the pop-up sites for testing (this is a good thing).

All of which is to say that over time, there's been no standardized pool of people being given PCR tests. The people going for PCR testing has changed so dramatically that no comparison over time can draw a meaningful conclusion.

That's not to say that positivity rates *couldn't* be used for that purpose — I argued early on that the province should be doing randomized testing at worksites and schools, so that some idea of background covid rates could be reached. But that never happened.

A positivity rate today says something about today, and maybe about this week, but I'm not sure exactly what it says, even then. It certainly doesn't say anything at all in terms of comparison with, say, January.

I think "positivity rates" were used in the US as a sort of bullshit stat to justify reopenings (for example, of schools in NYC), but it never made a lot of sense to me. But because it was in the media a lot, people thought it must be really important.

Nova Scotia's strategy changed over time. I'll write about this in detail at some later point, but it was basically a political decision (and the right one) that happened almost by chance, in recognition of the public's demand for zero or near-zero cases.

In short, early on it was about "flattening the curve" — having a low but "acceptable" level of covid in the wild so that the ICUs weren't overwhelmed. That evolved into a strategy of eradication. And it served us quite well — until April, anyway, but it's still the strategy.

This mostly successful strategy has nothing at all to do with positivity rates. Nada.

So, for me, the positivity rate is a stat looking for a purpose. It's not very important for me. Your milage may vary, and that's fine. I'm just explaining my view, and why I don't track or make a big deal out of positivity rates.