[BGrahamMD]

“But that’s precisely what they’re being used for. People are being quarantined for two weeks after testing positive. Everyone who has come in close contact with that person is also being quarantined. Example: my friend’s five year old granddaughter was quarantined at home for two weeks after ONE child in her classroom tested positive. If that child’s test wasn’t reliable in telling us the child could infect others, why is everyone quarantined?”
Wouldn’t it be great if we had a reliable test that told us whether someone is infectious? That would be a real game changer. Well, we don’t have one. Sorry. Quarantine duration–recently shortened by the CDC- is based on the vagaries of the incubation period, and best guess as to the period of infectivity.

[BGrahamMD]

Lot to unpack in these emails. I’ll pick one topic:
“BGraham, please explain: For laymen, what we need is an Actionable test. That is, when it says yes, we have it, when it says no, we don’t. With a reasonable level of error, like <5%.”
Accuracy of a test depends on a couple of factors. One, existence of a true “gold standard.” For example MRI scan can now predict with some accuracy the existence of prostate cancer. One looks for certain signs that indicate cancer, then take the prostate out, examine it, and see if you’re right. That path exam is the “gold standard.” But how do you know if you REALLY have/had COVID? What do you measure PCR testing against? You see the difficulty. Secondly, accuracy depends on large part on pre-test probability; Bayes theorem.

Here’s a reference on Bayes theorem in Medicine:
https://blogs.scientificamerican.com/cross-check/bayes-s-theorem-what-s-the-big-deal/
Here’s a (technical) paper on false positive PCRs:
https://www.icd10monitor.com/false-positives-in-pcr-tests-for-covid-19

[BGrahamMD]

So, the question of whether PCR tests are reliable hinges on “reliable for what?” Are they reliable in telling us whether a person can currently infect others? No, they are not. As the Oxford study implies, potential for infectivity is inversely related to number of cycles.
But the author of your posted article says PCR tests are “useless,” and implies that they are not reliable for diagnosing Covid infection. That is simply not true. He talks about “false positives.” The failure to grow the virus on culture is not a false positive. It means the virus is dead.

“Because this latest blow to PCR testing raises the probability that we are not enduring a killer virus pandemic, but a false positive pseudo-epidemic.” No person on the front lines of Covid care would possibly write such a thing. Shame!
Benjamin Graham, MD

[BGrahamMD]

This is a very disappointing and misleading article. The Oxford study correlated the number of PCR cycles and the ability to grow virus on culture. At a cycle threshold of 35 or higher, viral cultures were almost uniformly negative, so the chances of someone being able to infect others was less than 3%. IT DOES NOT MEAN THEY NEVER HAD COVID-19. It is well known that patients can shed viral fragments for weeks after infection. In the early days of the epidemic, repeat negative PCRs were required to leave the hospital, return to work, etc, but no one does that anymore.
I used to enjoy reading Automatic Earth for its insightful economic articles. It’s degenerated into articles like this and US election conspiracy theories.
Benjamin Graham, MD

[Author]

Posts