I take it you lack either the desire or the intellect to understand Bayes Law as it applies to clinical testing. positive results are NOT “almost always correct” and false positives are not “rare”. Moreover, “rare’ is an undefined arithmetic expression, and represents little more than someone’s subject impression.
A year ago we were told that the mutation rate for COVID was “low” so the likelihood of unfavorable mutant strains was “low.” Except there is and was nothing inherently “low” in the mutation rate EXCEPT in comparison to other viruses. But mutation rate is not the determining factor, it’s that rate TIMES THE OPPORTUNITIES TO MUTATE and every single case of COVID involves millions, billions, whole friggin Avogadro’s numbers of opportunity to mutate, and “low” multiplied by a $hittonfull was enough to produce a number of interesting mutant strains that are worrying us now.
Since you are either unwilling to read the ample literature (or view the video) on Baye’s theorem, or comprehend the same, we’ll do it the Baye’s Theorem for Idiots way, which if not totally correct will at least convey the flavor. Let us consider two groups of a million people. One group has a 5% rate of infectious COVID, the other a 0.5% rate of infectious COVID.
let us use a PCR test that has a sensitivity of 99.9% (which is higher than anybody really claims, especially early in the course of the disease when people are only recently infected). Let us say the PCR test has a specificity of 99%, which is FAR higher than PCR tests have.
OK, now let’s test those two million-person groups.
The 5% group will have 50,000 people who are true positives. The 99.9% sensitivity test will detect 49,500 of them, missing a mere 500. I guess 1 in 200 doesn’t sound bad, but look at the flip side. The 99% specificity means that of the 950,000 people that DON’T have COVID, the test will erroneously identify 950 of them as having COVID. So at a 5% prevalence rate the predictive value of a positive is 49,500/(49,500 + 950) or approximately 98%. Two out of a hundred are going to be false positives.
But now look at group two.
Only 0.5% of that group have COVID. That means that only 5000 of them have the disease. The test again detects 99.9% which is 4995 true positives and misses 5 true positives. It erroneously categorizes 1% of the 995,000 people who DON’t HAVE disease as positive meaning the predictive value of a positive has now declined to 4995/ (4995+ 9,950) or roughly .33.
Two out of three of your tests are going to be false positives.
now this was for illustration. The sensitivity and specificity of these clinical tests do not approach 99.9 or .99 (see below)
and in the population you are proposing to screen, the prevalence of what you are looking for - EARLY ASYMPTOMATIC CASES, is nowhere near 0.5%, far less 5%. One cannot make claims of “rare” in the absence of the population prevalence of the population being tested.
so in short, you are WRONG. I can explain it to you but I can’t understand it for you and if you are not going to take the effort to really research it you’ll just have to remain ignorant and no amount of posting info YOU clearly don’t understand - by either you or me - is going to change that.