15 ms·

The most baffling thing is not standardising CT values for RT-PCR tests. I have looked and I cannot even find how increasing CT values affect the false positiv

by forcry 5y ago

The most baffling thing is not standardising CT values for RT-PCR tests. I have looked and I cannot even find how increasing CT values affect the false positive rate for the PCR testing. If my understanding is correct a single increment of CT essentially doubles the sensitivity of the test, so difference between CT value of 35 and 40 is 32 fold. CDC is suggesting CT value of 28 for detecting breakthrough infections after vaccinations. And if I am not mistaken, a lot of places was using 35 for the CT value in RT-PCT tests. So that means, a breakthrough infection needs to have 128 times viral load in someone who has been vaccinated to be considered as positive, than it is required to have considered as positive in a non-vaccinated person. That is quite ridiculous.

dataflow 5y ago
Apparently it's even worse: "Different machines can produce different Ct values for the same sample, and the same machine can give different Ct values for different samples from the same person." [1] I'm not sure I even understand the statement, but I guess it means it could be that they (somehow) couldn't report them even if they wanted to. [1] https://medical.mit.edu/covid-19-updates/2020/11/pcr-test-result https://medical.mit.edu/covid-19-updates/2020/11/pcr-test-re...
- killjoywashere 5y ago
  I ran one of the first labs to validate the original CDC assay (we got the controls to work). Ct numbers were a slippery slope we should never have gone down for COVID-19 PCR tests because snot is not like blood, a homogenous substance with well known homeostatic parameters that are under tight physiologic control. Do you know someone went swimming? Do you know if they were crying? Did they just eat some particularly spicy tacos? Do you know if they have Sjogren's disease? What's the humidity? Are they dieting? Add in the presence of long COVID, and just long post infectious shedding, and it's mildly amazing we can get this to work at all. Yes, the test platform produces the data and no, it should not be reported. The only thing we can really do is confirm presence of the virus.
  - a9h74j 5y ago
    I would like to see QC statistics from some of the large test-processing centers which have IIRC been established. Surely they send a blind fraction of known-negative or control samples through, to get good statistics on false-positive rates?
    - killjoywashere 5y ago
      Yes, CMS requires clinical labs to submit to external proficiency testing under CLIA'88. The College of American Pathologists is a designated accrediting body and the one I'm used to using. You are not allowed to do anything out of the ordinary with proficiency testing samples. You can't send them out to someone else and report what they reported to you. You just report the results. If you fail, they test you again. If you fail again, you're on probation. If you fail a third time, you're not allowed to report that analyte any more. Results of proficiency testing are reviewed on site in a bi-annual inspection. Officially there are 3 inspections: Joint Commission, CAP, and FDA, that all inspect the lab. Generally, the Joint Commission folks and FDA folks ask for your CAP certificate, and if it's good, they just skip all those items. CAP does not mess around. They will fly to the ends of the earth, they will cite you, they may decertify you. And then no one in the lab gets paid and they have to find new jobs. So, yeah, rest assured, labs work really hard to make sure they report honestly and accurately. Better to fail an occasional proficiency test and repeat, than get sideways on the inspection.
      - a9h74j 5y ago
        I was working in a hospital when they discovered the blood lab was not prepared for the Joint Comission, and I saw the panic based on the scale of consequences you describe. My question refers to blinded testing meaning the lab in question does not know which, and handles, [in this case unexposed swabs from the factory] and it might range around 1-2% of samples on a continuing basis. To me that is the more industrial and continuous connotation of "QC". And the potential importance of "blind" testing. AFAIK, in any hospital lab there are tests which are still sent out, because they are either more challenging to build and train the right [accurate, low-error] procedures around, or too rarely done for the local investment. What if PCR tests for Covid were also too challenging to ramp up to mass scale in a regional startup lab? There are reports around of horrible controls -- perhaps true or not. That is something sending "blinded" samples through, proportional to mass testing being done, could disclose. Again, in the industrial setting there is the procedure you reported to the last ISO9000 auditor, and there is the actual procedure. The true spirit of that is that you should also build in procedures to detect when the production procedures are not being followed. Again back to ongoing QC.
timr 5y ago
> I have looked and I cannot even find how increasing CT values affect the false positive rate for the PCR testing. If my understanding is correct a single increment of CT essentially doubles the sensitivity of the test, so difference between CT value of 35 and 40 is 32 fold. I have never used this particular set of primers, but have done a lot of PCR. In general, at 30+ cycles, PCR is prone to spurious amplification. It depends on the primers, temperature profile, and other details, but at these cycle counts you need to be skeptical of your results. It's easy to get noise. I've never been able to fathom how a PCR amplification at 30+ cycles with no downstream purification or gel visualization is considered definitive diagnosis of an illness. I strongly suspect that the goal was to cast a wide net (i.e. bias toward false positives) at the expense of accuracy, but then "cases" became some kind of top-line media metric...
- mizzack 5y ago
  > I strongly suspect that the goal was to cast a wide net (i.e. bias toward false positives) at the expense of accuracy, but then "cases" became some kind of top-line media metric... Then in January the WHO updated the diagnostic protocol [1] because of that false positive/low confidence problem. Unsurprisingly, case counts plummeted in the following weeks. [1] https://www.who.int/news/item/20-01-2021-who-information-notice-for-ivd-users-2020-05 https://www.who.int/news/item/20-01-2021-who-information-not...
  - timr 5y ago
    Yep. It was stunning to me that this was rarely mentioned in the press.
  - themark 5y ago
    When I bring this up in conversations with people they make me feel like a kook. I am not sure what to think anymore.
    - wut2doobiedo 5y ago
      Because what to do is highly variable doesn’t mean we need to take unnecessary risk to know literal truth. Urban area population density requires different mitigations than rural, except rural communities rely on urban ones as logistics pipelines. So normalize; mask up, stay home. Prefer all gas, no brakes, based on stats? There may not be enough real people to keep the internet on later. Nothing about the lockdown was for saving you or me specifically, but systems of behavior we rely on.
74d-fe6-2c6 5y ago
this is at the heart of what "Covidiots", "Querdenker" and other tin foil hat people have been saying for more than a year ... just b/c you find a single virus or a part of it somewhere doesn't mean anything. but ... LOCKDOWN! and think of the children!
xyzzy21 5y ago
Well if you want to "control public fear" for political and economic gain, you just adjust the CT value as required to bump up the positive rate or to lower it. Note that the US CDC is using two different CT values: a higher one for unvaccinated and a lower one for vaccinated. That doesn't make any sense until you start thinking Hobbsian/Hegalian/Machiavellian. If I was wanting to exploit the situation, this is exactly what I'd do. Instead I have ethics however.
- briefcomment 5y ago
  Is there any proposed logic behind the difference in thresholds between the two groups? It seems designed to do just one thing - overestimate cases in the unvaccinated, and underestimate cases in the vaccinated. What is the actual rationale the CDC provides?
  - rsfern 5y ago
    Vaccinated people have a strong immune response to the virus, and presumably it would take a much higher viral load to make them clinically ill. To me this seems like a much more plausible explanation for dual thresholds than some kind of conspiracy to inflate numbers for political reasons. Immunology is not my field though.
jonahbenton 5y ago
No. Nearly all EUA PCR COVID tests are qualitative. On saliva/sputum samples, that's really the best that can be done. They can report one or more Ct values, but those reflect specific characteristics of the platform and can not be normalized across platforms and cannot be used even to infer things like viral load.
- blywi 5y ago
  Well apparently it is possible, and it has been done, probably more than once. In his German language podcast Dr. Drosten, a Coronavirus specialist from Charite Berlin, is addressing this exact issue. In case you have never heard of him, his lab was the first to publish a working PCR test protocol for SARS-CoV-2 back in January 2020 [1] Here is an DeepL translated excerpt from the transcript for his latest podcast [2] "The whole thing has a certain complication. The Ct values that we have here are not easily comparable between the individual test manufacturers. Basically, you can say that a high Ct value always indicates a low viral load. And if the Ct value then becomes lower, then that also becomes a higher viral load. But we can only compare them numerically as long as we are in the same test system. The differences there are sometimes considerable. There are test manufacturers where a value of, let's say, 25 is nothing at all worrying, while the same value of 25 in another manufacturer's test shows that this is already a seriously infectious concentration. This is simply because these test manufacturers do not standardize on the Ct value. That would not make sense either. Instead, it makes sense to simply determine what lies behind the Ct values, namely the actual viral load. You can do that, you have to calibrate that." and further "We did that in the fall. All the laboratory work that is necessary for this was done in September and October. I had already explained that to the public in the summer, how that works. We worked in the lab to make this possible. We have also come so far that viral load standards... You really have to imagine it as a small plastic vial with a test solution in it. It contains killed virus of a known, defined concentration. You can order it in two or three defined concentrations from a company that sells such a thing. The purpose of this company is to provide quality assurance for laboratories and to offer the necessary calibration standards. And these calibration standards are produced here in our laboratory, this killed and exactly quantified virus. So we have produced this calibration standard. We have also developed instructions, which are then recommended by the Robert Koch Institute, on how the laboratories can use this calibration standard to convert their Ct values into viral load ranges, which either actually lead to an exact viral load or which - and this is our recommendation - lead to assessment ranges. And that is to an assessment of highly infectious, low infectious, and borderline. So roughly speaking, that is expressed a little bit more genteel and precise. There's even a recommendation on how to express that on the medical findings then. Medical laboratories can do all that. This is also done in practice in the hospital sector. routinely used for discharge decisions. For example, a patient is in the intensive care unit. He is getting better. He should be transferred to a normal ward. Now the question is: Can we do that? Is he still highly infectious? Then a quantitative PCR test is carried out with these findings." [1] https://www.eurosurveillance.org/content/10.2807/1560-7917.ES.2020.25.3.2000045 https://www.eurosurveillance.org/content/10.2807/1560-7917.E... [2] https://www.ndr.de/nachrichten/info/coronaskript306.pdf https://www.ndr.de/nachrichten/info/coronaskript306.pdf
blywi 5y ago
Do you have source for that claim? I tried to find that recommendation, but all I came up with, was this fact check [1]. According to the fact check the CDC did NOT change cycle threshold, and the thresholds used to decide if test result is positive, are not different for vaccinated and unvaccinated. The quoted 28 cycle threshold is apparently only used for deciding if a sample can be submitted for sequencing. [1] https://www.politifact.com/factchecks/2021/jun/03/tweets/cdc-did-not-change-its-criteria-what-considered-br/ https://www.politifact.com/factchecks/2021/jun/03/tweets/cdc...
- jonahbenton 5y ago
  Also, PCR is not one test, one method, one process. There are hundreds of distinct PCR-type molecular nucleic amplification tests. CDC guidance would only apply to the one process they provide/support. For a survey of the ecosystem and the difficulties in standardizing Ct values, see e.g. https://www.aacc.org/science-and-research/covid-19-resources/statements-on-covid-19-testing/aacc-recommendation-for-reporting-sars-cov-2-cycle-threshold-ct-values https://www.aacc.org/science-and-research/covid-19-resources...
- forcry 5y ago
  https://news.ycombinator.com/item?id=27995222 https://news.ycombinator.com/item?id=27995222
ericlavigne 5y ago
Imagine the CDC yelling: "Don't throw that away! I need more details! Please! If you won't finish the job, please let me do that work!" That's my translation of the document I found when looking into this issue. More details below... === "CDC is suggesting CT value of 28 for detecting breakthrough infections after vaccinations... So that means, a breakthrough infection needs to have 128 times viral load in someone who has been vaccinated to be considered as positive, than it is required to have considered as positive in a non-vaccinated person." This sounded strange to me, so I searched for more information about it. It looks like you just misunderstood the CDC's policy. This document regarding breakthrough case investigations was the second result in a Google search for "cdc ct value 28". The relevant text can be found on page 5 of that document. https://stacks.cdc.gov/view/cdc/105217/cdc_105217_DS1.pdf https://stacks.cdc.gov/view/cdc/105217/cdc_105217_DS1.pdf "If SARS-CoV-2 sequencing will not be performed locally and a specimen is available, the state public health laboratory should request the residual clinical respiratory specimen for subsequent shipping to CDC. For cases with a known RT-PCR cycle threshold (Ct) value, submit only specimens with Ct value <=28 to CDC for sequencing." In other words... Imagine some lab just found a breakthrough case. And this breakthrough case had an especially high viral load (Ct 28). And the lab was just going to report a positive and throw away the sample... Imagine the CDC yelling: "Don't throw that away! I need more details! Please! If you won't finish the job, please let me do that work!" That's what the document is saying. There's a rare event that needs extra analysis. CDC is just letting the labs know in advance that if they ever see this event, and didn't have the resources to fully analyze it, please send that sample to the CDC so it gets the attention it deserves. Nothing to do with whether the test is considered positive or a breakthrough - just about whether to put extra effort into gathering more details on that particular case. CDC is volunteering to do this extra effort only for higher Ct values. Whether or not they do this extra work, it's still a positive result either way.
- forcry 5y ago
  I am not sure I like someone vocalising CDCs thought. It is kind of disturbing. I mean, you can be cutting someone throat and saying "be quite my child, it is all for best".. anyway here is what I have responded to a similar comment https://news.ycombinator.com/item?id=27995222 https://news.ycombinator.com/item?id=27995222
ericlavigne 5y ago
Very disappointing that the top comment on this article is COVID misinformation. The forcry account was created 3 weeks ago and specializes in COVID misinformation. No, CDC isn’t messing with the tests like this. And all this fancy talk about CT values is just cover for a false statement about CDC policy.
- forcry 5y ago
  It is not misinformation. This is the actual document that provides guidelines to labs to assess breakthrough infections https://www.cdc.gov/vaccines/covid-19/downloads/Information-for-laboratories-COVID-vaccine-breakthrough-case-investigation.pdf https://www.cdc.gov/vaccines/covid-19/downloads/Information-... but it is no longer available at that location, for some reason, and here is a link to the old version.. https://web.archive.org/web/20210429184157/https://www.cdc.gov/vaccines/covid-19/downloads/Information-for-laboratories-COVID-vaccine-breakthrough-case-investigation.pdf https://web.archive.org/web/20210429184157/https://www.cdc.g... Also, nice try to look up and discredit what I am saying on the basis of the age of my account. I am afraid such tactics won't fly well here.
simpleguitar 5y ago
So forget breakthrough infections for a second. Covid has both pre-sympomatic AND asymptomatic transmission. This has been proven. In many places, a negative PCR test means you do not have/carry the disease and can participate in risky activities. The NFL, and the NBA, did this. Movie studios did this. Australia and New Zealand did it as a whole country. To do that, you need a very sensitive test, not something that only confirms symptomatic infection. Of course, a very sensitive test has more false positives. But you have to weigh that with the possibility of an outbreak among those who tested negative. Imagine what would have happened if NBA had an outbreak after PCR tests--"The Tests Are USELESS!" the media would say. How many false positives? Look at how many players sat out during the NBA bubble. Now back to breakthrough infections. The CDC's thought is that vaccines protect well against existing strains, and their guidelines follow that assumption. Nothing in their guidelines accounted for the Delta, though they are changing their stance now, but slowly.
- forcry 5y ago
  >To do that, you need a very sensitive test, not something that only confirms symptomatic infection. Now, that makes sense. But you have also take into account that these are not just numbers. But these are numbers that can destroy countries and communities by implying perpetual lockdowns. It becomes even more comical that RT-PCR does even when positive, does not imply presence of the virus, but some fragments of a dead virus or even genetic material of some other viruses. I can only be appalled at the indifference of these people to mandate lockdowns that destroy lives and businesses, based on positivity rate of such a test.

by forcry 5y ago

dataflow 5y ago

Apparently it's even worse: "Different machines can produce different Ct values for the same sample, and the same machine can give different Ct values for different samples from the same person." [1] I'm not sure I even understand the statement, but I guess it means it could be that they (somehow) couldn't report them even if they wanted to. [1] https://medical.mit.edu/covid-19-updates/2020/11/pcr-test-result https://medical.mit.edu/covid-19-updates/2020/11/pcr-test-re...

killjoywashere 5y ago
I ran one of the first labs to validate the original CDC assay (we got the controls to work). Ct numbers were a slippery slope we should never have gone down for COVID-19 PCR tests because snot is not like blood, a homogenous substance with well known homeostatic parameters that are under tight physiologic control. Do you know someone went swimming? Do you know if they were crying? Did they just eat some particularly spicy tacos? Do you know if they have Sjogren's disease? What's the humidity? Are they dieting? Add in the presence of long COVID, and just long post infectious shedding, and it's mildly amazing we can get this to work at all. Yes, the test platform produces the data and no, it should not be reported. The only thing we can really do is confirm presence of the virus.
- a9h74j 5y ago
  I would like to see QC statistics from some of the large test-processing centers which have IIRC been established. Surely they send a blind fraction of known-negative or control samples through, to get good statistics on false-positive rates?
  - killjoywashere 5y ago
    Yes, CMS requires clinical labs to submit to external proficiency testing under CLIA'88. The College of American Pathologists is a designated accrediting body and the one I'm used to using. You are not allowed to do anything out of the ordinary with proficiency testing samples. You can't send them out to someone else and report what they reported to you. You just report the results. If you fail, they test you again. If you fail again, you're on probation. If you fail a third time, you're not allowed to report that analyte any more. Results of proficiency testing are reviewed on site in a bi-annual inspection. Officially there are 3 inspections: Joint Commission, CAP, and FDA, that all inspect the lab. Generally, the Joint Commission folks and FDA folks ask for your CAP certificate, and if it's good, they just skip all those items. CAP does not mess around. They will fly to the ends of the earth, they will cite you, they may decertify you. And then no one in the lab gets paid and they have to find new jobs. So, yeah, rest assured, labs work really hard to make sure they report honestly and accurately. Better to fail an occasional proficiency test and repeat, than get sideways on the inspection.
    - a9h74j 5y ago
      I was working in a hospital when they discovered the blood lab was not prepared for the Joint Comission, and I saw the panic based on the scale of consequences you describe. My question refers to blinded testing meaning the lab in question does not know which, and handles, [in this case unexposed swabs from the factory] and it might range around 1-2% of samples on a continuing basis. To me that is the more industrial and continuous connotation of "QC". And the potential importance of "blind" testing. AFAIK, in any hospital lab there are tests which are still sent out, because they are either more challenging to build and train the right [accurate, low-error] procedures around, or too rarely done for the local investment. What if PCR tests for Covid were also too challenging to ramp up to mass scale in a regional startup lab? There are reports around of horrible controls -- perhaps true or not. That is something sending "blinded" samples through, proportional to mass testing being done, could disclose. Again, in the industrial setting there is the procedure you reported to the last ISO9000 auditor, and there is the actual procedure. The true spirit of that is that you should also build in procedures to detect when the production procedures are not being followed. Again back to ongoing QC.

timr 5y ago

> I have looked and I cannot even find how increasing CT values affect the false positive rate for the PCR testing. If my understanding is correct a single increment of CT essentially doubles the sensitivity of the test, so difference between CT value of 35 and 40 is 32 fold. I have never used this particular set of primers, but have done a lot of PCR. In general, at 30+ cycles, PCR is prone to spurious amplification. It depends on the primers, temperature profile, and other details, but at these cycle counts you need to be skeptical of your results. It's easy to get noise. I've never been able to fathom how a PCR amplification at 30+ cycles with no downstream purification or gel visualization is considered definitive diagnosis of an illness. I strongly suspect that the goal was to cast a wide net (i.e. bias toward false positives) at the expense of accuracy, but then "cases" became some kind of top-line media metric...

mizzack 5y ago
> I strongly suspect that the goal was to cast a wide net (i.e. bias toward false positives) at the expense of accuracy, but then "cases" became some kind of top-line media metric... Then in January the WHO updated the diagnostic protocol [1] because of that false positive/low confidence problem. Unsurprisingly, case counts plummeted in the following weeks. [1] https://www.who.int/news/item/20-01-2021-who-information-notice-for-ivd-users-2020-05 https://www.who.int/news/item/20-01-2021-who-information-not...
- timr 5y ago
  Yep. It was stunning to me that this was rarely mentioned in the press.
- themark 5y ago
  When I bring this up in conversations with people they make me feel like a kook. I am not sure what to think anymore.
  - wut2doobiedo 5y ago
    Because what to do is highly variable doesn’t mean we need to take unnecessary risk to know literal truth. Urban area population density requires different mitigations than rural, except rural communities rely on urban ones as logistics pipelines. So normalize; mask up, stay home. Prefer all gas, no brakes, based on stats? There may not be enough real people to keep the internet on later. Nothing about the lockdown was for saving you or me specifically, but systems of behavior we rely on.

74d-fe6-2c6 5y ago

this is at the heart of what "Covidiots", "Querdenker" and other tin foil hat people have been saying for more than a year ... just b/c you find a single virus or a part of it somewhere doesn't mean anything. but ... LOCKDOWN! and think of the children!

xyzzy21 5y ago

Well if you want to "control public fear" for political and economic gain, you just adjust the CT value as required to bump up the positive rate or to lower it. Note that the US CDC is using two different CT values: a higher one for unvaccinated and a lower one for vaccinated. That doesn't make any sense until you start thinking Hobbsian/Hegalian/Machiavellian. If I was wanting to exploit the situation, this is exactly what I'd do. Instead I have ethics however.

briefcomment 5y ago
Is there any proposed logic behind the difference in thresholds between the two groups? It seems designed to do just one thing - overestimate cases in the unvaccinated, and underestimate cases in the vaccinated. What is the actual rationale the CDC provides?
- rsfern 5y ago
  Vaccinated people have a strong immune response to the virus, and presumably it would take a much higher viral load to make them clinically ill. To me this seems like a much more plausible explanation for dual thresholds than some kind of conspiracy to inflate numbers for political reasons. Immunology is not my field though.

jonahbenton 5y ago

No. Nearly all EUA PCR COVID tests are qualitative. On saliva/sputum samples, that's really the best that can be done. They can report one or more Ct values, but those reflect specific characteristics of the platform and can not be normalized across platforms and cannot be used even to infer things like viral load.

blywi 5y ago
Well apparently it is possible, and it has been done, probably more than once. In his German language podcast Dr. Drosten, a Coronavirus specialist from Charite Berlin, is addressing this exact issue. In case you have never heard of him, his lab was the first to publish a working PCR test protocol for SARS-CoV-2 back in January 2020 [1] Here is an DeepL translated excerpt from the transcript for his latest podcast [2] "The whole thing has a certain complication. The Ct values that we have here are not easily comparable between the individual test manufacturers. Basically, you can say that a high Ct value always indicates a low viral load. And if the Ct value then becomes lower, then that also becomes a higher viral load. But we can only compare them numerically as long as we are in the same test system. The differences there are sometimes considerable. There are test manufacturers where a value of, let's say, 25 is nothing at all worrying, while the same value of 25 in another manufacturer's test shows that this is already a seriously infectious concentration. This is simply because these test manufacturers do not standardize on the Ct value. That would not make sense either. Instead, it makes sense to simply determine what lies behind the Ct values, namely the actual viral load. You can do that, you have to calibrate that." and further "We did that in the fall. All the laboratory work that is necessary for this was done in September and October. I had already explained that to the public in the summer, how that works. We worked in the lab to make this possible. We have also come so far that viral load standards... You really have to imagine it as a small plastic vial with a test solution in it. It contains killed virus of a known, defined concentration. You can order it in two or three defined concentrations from a company that sells such a thing. The purpose of this company is to provide quality assurance for laboratories and to offer the necessary calibration standards. And these calibration standards are produced here in our laboratory, this killed and exactly quantified virus. So we have produced this calibration standard. We have also developed instructions, which are then recommended by the Robert Koch Institute, on how the laboratories can use this calibration standard to convert their Ct values into viral load ranges, which either actually lead to an exact viral load or which - and this is our recommendation - lead to assessment ranges. And that is to an assessment of highly infectious, low infectious, and borderline. So roughly speaking, that is expressed a little bit more genteel and precise. There's even a recommendation on how to express that on the medical findings then. Medical laboratories can do all that. This is also done in practice in the hospital sector. routinely used for discharge decisions. For example, a patient is in the intensive care unit. He is getting better. He should be transferred to a normal ward. Now the question is: Can we do that? Is he still highly infectious? Then a quantitative PCR test is carried out with these findings." [1] https://www.eurosurveillance.org/content/10.2807/1560-7917.ES.2020.25.3.2000045 https://www.eurosurveillance.org/content/10.2807/1560-7917.E... [2] https://www.ndr.de/nachrichten/info/coronaskript306.pdf https://www.ndr.de/nachrichten/info/coronaskript306.pdf

blywi 5y ago

Do you have source for that claim? I tried to find that recommendation, but all I came up with, was this fact check [1]. According to the fact check the CDC did NOT change cycle threshold, and the thresholds used to decide if test result is positive, are not different for vaccinated and unvaccinated. The quoted 28 cycle threshold is apparently only used for deciding if a sample can be submitted for sequencing. [1] https://www.politifact.com/factchecks/2021/jun/03/tweets/cdc-did-not-change-its-criteria-what-considered-br/ https://www.politifact.com/factchecks/2021/jun/03/tweets/cdc...

jonahbenton 5y ago
Also, PCR is not one test, one method, one process. There are hundreds of distinct PCR-type molecular nucleic amplification tests. CDC guidance would only apply to the one process they provide/support. For a survey of the ecosystem and the difficulties in standardizing Ct values, see e.g. https://www.aacc.org/science-and-research/covid-19-resources/statements-on-covid-19-testing/aacc-recommendation-for-reporting-sars-cov-2-cycle-threshold-ct-values https://www.aacc.org/science-and-research/covid-19-resources...
forcry 5y ago
https://news.ycombinator.com/item?id=27995222 https://news.ycombinator.com/item?id=27995222

ericlavigne 5y ago

Imagine the CDC yelling: "Don't throw that away! I need more details! Please! If you won't finish the job, please let me do that work!" That's my translation of the document I found when looking into this issue. More details below... === "CDC is suggesting CT value of 28 for detecting breakthrough infections after vaccinations... So that means, a breakthrough infection needs to have 128 times viral load in someone who has been vaccinated to be considered as positive, than it is required to have considered as positive in a non-vaccinated person." This sounded strange to me, so I searched for more information about it. It looks like you just misunderstood the CDC's policy. This document regarding breakthrough case investigations was the second result in a Google search for "cdc ct value 28". The relevant text can be found on page 5 of that document. https://stacks.cdc.gov/view/cdc/105217/cdc_105217_DS1.pdf https://stacks.cdc.gov/view/cdc/105217/cdc_105217_DS1.pdf "If SARS-CoV-2 sequencing will not be performed locally and a specimen is available, the state public health laboratory should request the residual clinical respiratory specimen for subsequent shipping to CDC. For cases with a known RT-PCR cycle threshold (Ct) value, submit only specimens with Ct value <=28 to CDC for sequencing." In other words... Imagine some lab just found a breakthrough case. And this breakthrough case had an especially high viral load (Ct 28). And the lab was just going to report a positive and throw away the sample... Imagine the CDC yelling: "Don't throw that away! I need more details! Please! If you won't finish the job, please let me do that work!" That's what the document is saying. There's a rare event that needs extra analysis. CDC is just letting the labs know in advance that if they ever see this event, and didn't have the resources to fully analyze it, please send that sample to the CDC so it gets the attention it deserves. Nothing to do with whether the test is considered positive or a breakthrough - just about whether to put extra effort into gathering more details on that particular case. CDC is volunteering to do this extra effort only for higher Ct values. Whether or not they do this extra work, it's still a positive result either way.

forcry 5y ago
I am not sure I like someone vocalising CDCs thought. It is kind of disturbing. I mean, you can be cutting someone throat and saying "be quite my child, it is all for best".. anyway here is what I have responded to a similar comment https://news.ycombinator.com/item?id=27995222 https://news.ycombinator.com/item?id=27995222

ericlavigne 5y ago

Very disappointing that the top comment on this article is COVID misinformation. The forcry account was created 3 weeks ago and specializes in COVID misinformation. No, CDC isn’t messing with the tests like this. And all this fancy talk about CT values is just cover for a false statement about CDC policy.

forcry 5y ago
It is not misinformation. This is the actual document that provides guidelines to labs to assess breakthrough infections https://www.cdc.gov/vaccines/covid-19/downloads/Information-for-laboratories-COVID-vaccine-breakthrough-case-investigation.pdf https://www.cdc.gov/vaccines/covid-19/downloads/Information-... but it is no longer available at that location, for some reason, and here is a link to the old version.. https://web.archive.org/web/20210429184157/https://www.cdc.gov/vaccines/covid-19/downloads/Information-for-laboratories-COVID-vaccine-breakthrough-case-investigation.pdf https://web.archive.org/web/20210429184157/https://www.cdc.g... Also, nice try to look up and discredit what I am saying on the basis of the age of my account. I am afraid such tactics won't fly well here.

simpleguitar 5y ago

So forget breakthrough infections for a second. Covid has both pre-sympomatic AND asymptomatic transmission. This has been proven. In many places, a negative PCR test means you do not have/carry the disease and can participate in risky activities. The NFL, and the NBA, did this. Movie studios did this. Australia and New Zealand did it as a whole country. To do that, you need a very sensitive test, not something that only confirms symptomatic infection. Of course, a very sensitive test has more false positives. But you have to weigh that with the possibility of an outbreak among those who tested negative. Imagine what would have happened if NBA had an outbreak after PCR tests--"The Tests Are USELESS!" the media would say. How many false positives? Look at how many players sat out during the NBA bubble. Now back to breakthrough infections. The CDC's thought is that vaccines protect well against existing strains, and their guidelines follow that assumption. Nothing in their guidelines accounted for the Delta, though they are changing their stance now, but slowly.

forcry 5y ago
>To do that, you need a very sensitive test, not something that only confirms symptomatic infection. Now, that makes sense. But you have also take into account that these are not just numbers. But these are numbers that can destroy countries and communities by implying perpetual lockdowns. It becomes even more comical that RT-PCR does even when positive, does not imply presence of the virus, but some fragments of a dead virus or even genetic material of some other viruses. I can only be appalled at the indifference of these people to mandate lockdowns that destroy lives and businesses, based on positivity rate of such a test.