Interesting. Even after all this time I can't say I fully understand how testing and % positivity works (aside from low being better!). So if we want schools to open all the healthy people should run out and get tests to drive % positive down?
Aside from pre-surgical cases and athletes, who is being tested to drive the 95% negative goal? Contacts of people that are positive? (Proving whether measures work or not.) Routine tests for front-line workers? (I haven't heard of this happening in my state.)
It's a measure of community spread and if we have a handle on where the virus is or if it's just everywhere in the community.
Yes, we should be testing everyone who had a contact with someone who tested positive. Then, each of those that test positive repeat until there are not more positives. Isolating all of the positives. This cycle needs to be fast enough to be practical. Still some will slip through.
Yes, we should be doing some level of random surveillance testing in the community to catch things that slipped through. Also fast enough to execute the tracing loop on a positive result.
A higher % positive, the more general community spread is occurring. A lower number means there's a better handle on isolating and reducing the spread.
If we could have everyone go out and get a test, isolate until they got result, then continue to isolate if they're negative, that would be wonderful. We would know where the infection was and could contain it. The problem is, "everyone" is a really big number, and there's no way to do it fast enough. When we talk about increasing capacity, we're talking about solving this and solving it at speed. While we report on the raw number of tests, we're not reporting on the speed. Without the speed, we're missing half the picture, and it certainly feels like our testing capacity isn't performing at the level the raw numbers would suggest it should be.
It's a basic OODA loop.
https://en.wikipedia.org/wiki/OODA_loop
We're trying to see an infection, isolate, trace, test, repeat. The virus is trying to infect, jump to next host, repeat. Whichever loop is faster determines the speed of the spread.