The PCR test is being used clinically in two major ways. First, people with a positive test are advised to isolate themselves, to avoid infecting others. As I noted in my very first reply, a positive PCR test doesn't necessarily mean that the patient is shedding viable virus (i.e., that the patient could infect others), and that's the kernel of truth behind Mullis's statements. I posted a link above to a paper that attempted to culture virus from PCR-positive samples:
> It can be observed that at Ct = 25, up to 70% of patients remain positive in culture and that at Ct = 30 this value drops to 20%. At Ct = 35, the value we used to report a positive result for PCR, https://academic.oup.com/cid/advance-article/doi/10.1093/cid...
So it's absolutely true that some of the people who test positive by PCR would never have infected anyone else, and there was no need for them to isolate. But it's hard to distinguish who--the viral culture takes too much time to run for each patient, and no one really knows how sensitive that is anyways. So to be conservative, they just advise everyone to isolate.
People have talked about advising isolation only if Ct is less than some lower cutoff, since there's a clear correlation between Ct and positivity in culture. But since a patient early in the infection might have high Ct now but lower Ct later, public health authorities haven't done so.
The second way that PCR tests are being used is to diagnose the cause of illness--a patient comes in sick, and the doctor is trying to find the cause. At this point, we already know that the patient is sick, because they're coughing, complaining that they feel terrible, etc. The PCR test isn't what tells us that they're sick, but rather what tells us that the cause of their sickness is SARS-CoV-2. The alternative hypothesis is that people are dying of some different disease, and by coincidence the excess mortality due to that different disease correlates strongly with SARS-CoV-2 test positivity (which e.g. is why that's high in the USA, and low in Australia; you can also see the correlation within a single region if you look at the time series). That would be a pretty strange coincidence, and no one has proposed any mechanism for that.
So do the above clinical uses seem reasonable to you? Or is there some other clinical use that seems unreasonable to you in light of Mullis's comments?
Perhaps this confusion comes from the definition of "illness" or "sick". In a certain sense, if I don't feel sick, then no test can tell me otherwise. For example, was Typhoid Mary "sick"? Is an HIV patient with no symptoms? Kary Mullis says no:
> A man with gray hair and a goatee raises his hand. He says he's been HIV-positive since 1984, that he took the anti-AIDS drug AZT for a couple years but stopped. Now, he says, his T-cell count -- the number of a kind of white blood cell that is killed by HIV -- has gone way down.
> Mullis interrupts him: "Change doctors!"
> The man continues. His T-cell count is down to 150, which is usually thought of as dangerously low. He asks Mullis for advice.
> "I would say there is no evidence that I can find in the scientific literature that you should worry about HIV or your T-cell count," Mullis tells him. "If you'd stop worrying, maybe you'll be all right. You look pretty healthy to me."
https://www.washingtonpost.com/archive/lifestyle/1998/11/03/...
Do you think he gave that patient good advice?