Hacker News Reader

Comment by andy_ppp | Hacker News Reader

Does anyone know if PCR tests will find out if you've had the virus, after you had the vaccine?

PCR tests don't detect antibodies, they detect active virus. Antibody tests may be what you're thinking of.

See: https://www.fda.gov/consumers/consumer-updates/coronavirus-d...

I don't know the answer to your question, but it would make sense unless the antibody test(s) look for a specific type of antibody to COVID-19 that is produced naturally when infected and not just any antibodies that might be produced when vaccinated.

incorrect, PCR tests detect any virus matter - active or otherwise. This is a very important distinction since we've seem to lost our minds collectively and label PCR tests with a cycle threshold of >40 as a true positive.

"lost our minds collectively" -- no, made a sane tradeoff in the sensitivity vs. specificity ratio when the disease is common. It's not like the sensitivity is wonderful even with a large number of PCR cycles, and the false positive rate was not too bad.

As the disease becomes less common, it makes sense to turn down the sensitivity of the test. Base rate fallacy, yada yada.

here's a study showing above CT 32 not a single live virus was able to be cultured [1] - if this study accurately portrays reality then there's nothing resembling sanity with PCR CTs >40

https://academic.oup.com/cid/advance-article/doi/10.1093/cid...

(image from the study) https://i.imgur.com/AaZZIXM.png

People who are infected will not necessarily have a single live virus in a swab you collect. Yes, if you can't culture it, they probably are not infectious right now but that is a different distinction.

You're also putting a whole lot of weight on a study with what, about 50 total positives? And of them, very few with a Ct of over 35.

do you know of a study that shows the opposite? Because I know of more studies that have similar findings.

here's one I found with a quick google (haven't read the study)

https://pubs.acs.org/doi/10.1021/acsomega.1c00166

Even your studies show a relatively low false positive rate with high numbers of cycles, and it's hoped that it increases sensitivity. No, we have no perfect data, because we have no perfect "infection test".

High sensitivity was very worthwhile early in the pandemic, when infection was common and the predictive value of a test with a somewhat higher false positive rate was still high. Now that COVID is becoming rarer, we should absolutely be reducing cycle counts. The target should be to keep >75% of positives with actual infection even as the underlying rate of the disease decreases. I think we've exceeded this threshold by a large margin thus far. If you have a cogent statistical argument that disagrees instead of handwaving about cycle counts, I'm willing to listen.

1st study says zero active virus beyond CT>32 and 2nd study shows 34ct is the cutoff for infectivity and that 2nd study also cites a 3rd study that ct>35 decreased to 8% active viruses

All of those numbers are below 40ct - if this doesn't convince you that 40ct is the wrong number I don't know what to say - you've provided zero evidence that 40ct is beneficial and try to downplay extremely high false positives shown in three studies when you go to 40ct

"Extremely high false positives"-- no, inability to culture from the specimens, which isn't the same thing as a false positive. Please stop conflating the two.

It's a sensitivity vs. specificity issue, as I've said before, and I'm willing to listen to arguments on those grounds.

As the studies you've cited show, the vast majority of positive samples become so before 30 cycles. Then there's these relatively few "weak positive" results, where the ability to culture live virus is less. There's not enough of these to explain the number of cases, deaths, or overall excess mortality (as you've attempted to say elsewhere here, and others have rightfully corrected you). And these results being considered a positive is arguably a good thing while the virus is very common, as I've attempted to explain to you.

> 2nd study shows 34ct is the cutoff for infectivity and that 2nd study also cites a 3rd study that ct>35 decreased to 8% active viruses

Yes, but "infective at the moment you took the test" is not a useful metric. You want more people to be positive than this, so you can A) contact trace people who were infectious immediately before, and B) get people who were just infected to stay home -before- they're infectious.

Of course, now that we're on a major downslope of cases, this tradeoff is changing...

This is just silly, if you reject the methodology used in these studies to determine effective CT thresholds then it's pretty pointless having a conversation with someone who refuses to do anything other than restate their opinion without providing and evidence that supports their opinions.

As I've repeatedly told you-- you're overstating your evidence here. Infectivity is not the gold standard of what should be a positive test.

My girlfriend got infected about ten days after my second AstraZeneca jab. We didn't isolate. I got tested:

- Zero naturally-occurring antibodies (the kind you would get from having had COVID).

- Strong presence of vaccine-induced antibodies.

However, given the timing, I am still uncertain as to whether I contracted COVID during that post-second-jab 4 weeks when my girlfriend was infected (I gave blood for the test while I was still feeling bad in the weeks after the second injection, and I didn't spring for the 'currently infected' test, since the whole package of tests was pretty expensive).

If I got COVID now, would it not be handled exclusively by vaccine-induced antibodies, which means that any test for having had COVID in the future would still be negative?

I'm no kind of doubter, and I'm pretty sure the AZ double-jab helped me lately, but I'm not sure I'll ever know if I was ever infected, to be honest.

Even if you were, if you didn't have any symptoms the vaccine did it's job. In which case you were not infectious neither.

I hope you better half fully recovered by now.

I got a rapid test after getting the first vaccine -- I was told that there were more than one type of antibodies, the vaccine only provides one, so they can test for the others to see if you had the virus proper.

PCR, no, antibody tests do. E.g. if you give blood with the Red Cross right now, they will do a basic antibody test that will react if you have had COVID or if you have been vaccinated. If you test positive for that, they do a second test that will only react if you have been exposed to the real virus.

PCR tests only detect live virus, not antibodies.

there are several antibody tests for covid. As far as I'm aware, the vaccines only trigger the anti-spike protein antibodies, so if you test positive for other types of covid-19 antibodies, you've probably had it.

PCR test detect genetic material of the virus. It may be live virus it may be just fragments, impossible to tell.

Find a lab that offers Roche Anti-Sars-cov2-N test.

Interesting, is this one just for Pfizer as I've had Astra Zenica which I believe is a different set of spike protein DNA sequences?

This one's for the nucleocapsid protein. If you'd like to test antibody generated by the vaccines, you'd have to take the Anti-sarscov2-S test.

All currently used vaccines use the same version of the spike protein sequence.

Incorrect. Pfizer, Moderna, and J&J use a pre-fusion stabilized sequence. AZ does not.

However, all currently used vaccines are just the spike (S) protein plus possible non-COVID proteins (in the viral vector vaccines), so none will induce antibodies against the COVID N protein.

I imagine it might be possible to guess whether someone was vaccinated with the AZ vaccine based on whether they have antibodies that are specific to the site that’s modified in the stabilized sequences. I don’t think anyone does a test like that.

Reply on news.ycombinator.com