Quote:
Originally Posted by lewdog
Oh wow, that's great. It seems in some areas that cases or positive percentage of tests maybe up, but hospitalizations are down. The virus may be weakening in the USA.
|
The %positive rate is up because testing is down. It's a bullshit statistic when used alone. If I test 100 people and have a 10% positivity rate it doubles to 20% if I test only 50 people and gets cut in half if I test 200 people.
Then you factor in the fact that most people who get tested are people that think there is a need to be tested such as they are sick or have been exposed to someone who tested positive. Therefore the pool of test cases is already slightly skewed to one side of the probability scale.
Unless you are testing at random and at a consistent rate I think the %positive metric is fairly misleading.
Now if I remember correctly I believe it was KU who randomly tested Teachers and Students alike and had a 1.4% positivity rate.
I make a stink about this because the JoCo Dept of Health's lead doctor used the %positivity rate to make the statement that "we have to assume 10% of students are positive" and that is completely inaccurate. At best he could only say that 10% of students that feel sick or have been exposed to someone who tested positive are positive. That's a far cry from 10% of all students are positive.
I mean if you have 30,000 students he is making an assumption that 3,000 are positive. When in reality if you tested kids who felt sick and\or were exposed to someone who tested positive you might come out to maybe a total of 3,000 kids if that and then only 300 of them would actually be positive. And that is being extremely generous as I would think given what I think I have seen the numbers in JoCo would probably more like 30 kids saying they feel sick and we will say another 30 that have been exposed to a person who was positive and you would end up with a whopping 6 positive cases.
That may a bit of an underestimate but I think it's close to the actual if I had to bet $'s.