Normally, it is the percent positive out of all who tested.
It bears no relationship to the positivity of the general population. If the "scare factor" was high, a higher number of people will test than normal, tbus a higher percen tof those who are not actually positive. (e.g., testing with any light sniffle, or even less) If "low", the opposite. We can't measure "scare factor" nor the percent of people who came in due to excess fear.
Other factors effect "who tests". Government may start requiring testing for participation in certain things, e.g., travel, and percent participating and thus needing a test would vary.
There is actually, usually, some relationship between the percent positive in testing and the spread of Covid in a population. But the variation is so huge that percent positive cannot give you a good estimate of Covid spread. The exception: if percent positive is very low or very high, these are indications that concern should be lower or higher, respectively, (since again, there is some relationship between testing positives and positives in a population, though it is far from exact.)
An example: Country (or county, etc) A could have 10 percent positive tests. County (or county) B could have 20 percent positive tests. But their Covid rates could be the same. It was just that fewer people tested in Country B, and those coming in tended to be sicker and have Covid more often. But if Country (or county) C has 90 percent positive tests, it's safe to assume Covid is more rampant there than in Counties A or B, though you can't say exactly how much more rampant.
|