I have analyzed this 10 different ways, trying to sort fact from fiction, politics, and agenda. I have some stuff I trust more than others obviously, but I keep coming back to the same conclusion. I was actually pretty excited about some of what was being suggested by these antibody tests, until i started digging into the tests being used and the methodologies of those conducting them. The majority of these antibody tests have such significant sensitivity and specificity errors that is is impossible to use the results the way they are using them. It's the equivalent of measuring the thickness of a sheet paper with a ruler.
For example: The Abbott Labs test I linked a while back is very good, but there are no results back from it. It is claimed to be 100% sensitive (catches every case) and 99.5% specific (has 5 false positives in 1000 tests). Even as good as it is, if it had been used in the Santa Clara county test its error rate would have been 33% of the total positives they found. Out of 3300 people they tested, they found 50 positive....*50*. With the best-in-class Abbot test, the false positives would be expected to be 17 of the 50. But they didn't use the Abbott test, they used one out of China that was far less accurate than that. To make it worse, the sample was highly salted by the way they got test subjects. They then took that small number of positives, found by a questionable test, from a salted sample, and extrapolated that to the entire county and made the claim they did. Sorry, that isn't science.
When results start coming back from the Abbot test and fall outside the margin of error of the test, I'll change my tune. I hope I get to.
