4 ms·

PCR has a 30% false negative rate. You make it sound like they entered a clean room. They did not.

persedes 6y ago
I've been curious about the COVID PCR tests performance metrics, do you have a source for that (to also include sensitivity, specificity etc)?
- wittyreference 6y ago
  Ai et al. https://pubmed.ncbi.nlm.nih.gov/32101510/ https://pubmed.ncbi.nlm.nih.gov/32101510/ analysis of 1014 patients in Wuhan; Yang et al. looked at another 213 in Guangdong. Kucirka et al (https://www.acpjournals.org/doi/10.7326/M20-1495 https://www.acpjournals.org/doi/10.7326/M20-1495) did a pooled analysis of seven other studies and found that "over the 4 days of infection before the typical time of symptom onset (day 5), the probability of a false-negative result in an infected person decreases from 100% (95% CI, 100% to 100%) on day 1 to 67% (CI, 27% to 94%) on day 4. On the day of symptom onset, the median false-negative rate was 38% (CI, 18% to 65%). This decreased to 20% (CI, 12% to 30%) on day 8 (3 days after symptom onset) then began to increase again, from 21% (CI, 13% to 31%) on day 9 to 66% (CI, 54% to 77%) on day 21." A lot of folks are jumping into this thread very confident that my numbers are wrong, but they're pretty well established and uncontroversial.
  - persedes 6y ago
    thanks for those! I wasn't trying to dispute them, just curious in general as I haven't heard much about the PCR-tests metrics. I recently had an IgG antibody test done and there the sensitivity was much higher, but also a function over time.
hdkrgr 6y ago
This makes it sound more dramatic than it is. Let's assume the "real" share of infectious people in the population is 10/100k (That's twice as much as the most recent reported 7-day incidence (5.1/100k) of new infections for Leipzig, the area where the study was performed.) Further let's assume a PCR Test has a 30% FNR, and a 10% FPR (numbers completely made up, I don't have a source). Then out of 1500 People who have tested negative, we'd expect 0.05 to have the virus.
- jariel 6y ago
  Probably more like 14-20 day window and much greater than 2x incidence as loads of people show little or no symptoms. Once the expected value starts to creep over 1, then that's enough to infect quite a lot of people. It's not too much past that to expect that one of the infected has 'underlying condition' and risking death. In British Coloumbia there's been a big outbreak due to a single gathering in the interior, notably (shamefully) attended by Medics and Doctors. It's a little risky. It's just not the time to have major gatherings, but at least this one is in the service of the science of 'stopping it' - which is better than most places are doing.
wesleywt 6y ago
I don't think your percentage is correct. It depends on the PCRs target gene, the quality of the samples and the quality of the machine used. They are using reverse transcriptase quantitative PCR. What is your source for that FNR?
- wittyreference 6y ago
  Ai et al. https://pubmed.ncbi.nlm.nih.gov/32101510/ https://pubmed.ncbi.nlm.nih.gov/32101510/ analysis of 1014 patients in Wuhan; Yang et al. looked at another 213 in Guangdong. Kucirka et al (https://www.acpjournals.org/doi/10.7326/M20-1495 https://www.acpjournals.org/doi/10.7326/M20-1495) did a pooled analysis of seven other studies and found that "over the 4 days of infection before the typical time of symptom onset (day 5), the probability of a false-negative result in an infected person decreases from 100% (95% CI, 100% to 100%) on day 1 to 67% (CI, 27% to 94%) on day 4. On the day of symptom onset, the median false-negative rate was 38% (CI, 18% to 65%). This decreased to 20% (CI, 12% to 30%) on day 8 (3 days after symptom onset) then began to increase again, from 21% (CI, 13% to 31%) on day 9 to 66% (CI, 54% to 77%) on day 21."
tinus_hn 6y ago
> A negative test can tell us: > the person was unlikely to be infectious at the time of the test. https://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-health-advice-general-public/assessment-and-testing-covid-19/covid-19-test-results-and-their-accuracy https://www.health.govt.nz/our-work/diseases-and-conditions/... Of course if you are testing proper samples from the airways and they come back negative, it’s not very likely these people can infect others through droplets from their airways at that moment. The problem is that they might be in the incubation period but that is irrelevant here. And anyway PCR does not nearly have a 30% false negative rate, I’m not quite sure where you got these stats.
- wittyreference 6y ago
  Ai et al. https://pubmed.ncbi.nlm.nih.gov/32101510/ https://pubmed.ncbi.nlm.nih.gov/32101510/ analysis of 1014 patients in Wuhan; Yang et al. looked at another 213 in Guangdong. Kucirka et al (https://www.acpjournals.org/doi/10.7326/M20-1495 https://www.acpjournals.org/doi/10.7326/M20-1495) did a pooled analysis of seven other studies and found that "over the 4 days of infection before the typical time of symptom onset (day 5), the probability of a false-negative result in an infected person decreases from 100% (95% CI, 100% to 100%) on day 1 to 67% (CI, 27% to 94%) on day 4. On the day of symptom onset, the median false-negative rate was 38% (CI, 18% to 65%). This decreased to 20% (CI, 12% to 30%) on day 8 (3 days after symptom onset) then began to increase again, from 21% (CI, 13% to 31%) on day 9 to 66% (CI, 54% to 77%) on day 21." You are mistaken.
  - tinus_hn 6y ago
    Your first study takes rt-pcr testing as the reference and cannot make statements on its accuracy. Your second study is flawed as indicated by the comments on it, it’s a meta study based on poor data. If you choose a point and go looking for studies to support it, regardless of quality and ignoring studies that do not, that’s not science. And by the way, your studies are about something that is not important here. What is important here is if the patient carries virus particles in his airways so they can infect others, not if they are infected themselves.

I've been curious about the COVID PCR tests performance metrics, do you have a source for that (to also include sensitivity, specificity etc)?

wittyreference 6y ago
Ai et al. https://pubmed.ncbi.nlm.nih.gov/32101510/ https://pubmed.ncbi.nlm.nih.gov/32101510/ analysis of 1014 patients in Wuhan; Yang et al. looked at another 213 in Guangdong. Kucirka et al (https://www.acpjournals.org/doi/10.7326/M20-1495 https://www.acpjournals.org/doi/10.7326/M20-1495) did a pooled analysis of seven other studies and found that "over the 4 days of infection before the typical time of symptom onset (day 5), the probability of a false-negative result in an infected person decreases from 100% (95% CI, 100% to 100%) on day 1 to 67% (CI, 27% to 94%) on day 4. On the day of symptom onset, the median false-negative rate was 38% (CI, 18% to 65%). This decreased to 20% (CI, 12% to 30%) on day 8 (3 days after symptom onset) then began to increase again, from 21% (CI, 13% to 31%) on day 9 to 66% (CI, 54% to 77%) on day 21." A lot of folks are jumping into this thread very confident that my numbers are wrong, but they're pretty well established and uncontroversial.
- persedes 6y ago
  thanks for those! I wasn't trying to dispute them, just curious in general as I haven't heard much about the PCR-tests metrics. I recently had an IgG antibody test done and there the sensitivity was much higher, but also a function over time.

hdkrgr 6y ago

This makes it sound more dramatic than it is. Let's assume the "real" share of infectious people in the population is 10/100k (That's twice as much as the most recent reported 7-day incidence (5.1/100k) of new infections for Leipzig, the area where the study was performed.) Further let's assume a PCR Test has a 30% FNR, and a 10% FPR (numbers completely made up, I don't have a source). Then out of 1500 People who have tested negative, we'd expect 0.05 to have the virus.

jariel 6y ago
Probably more like 14-20 day window and much greater than 2x incidence as loads of people show little or no symptoms. Once the expected value starts to creep over 1, then that's enough to infect quite a lot of people. It's not too much past that to expect that one of the infected has 'underlying condition' and risking death. In British Coloumbia there's been a big outbreak due to a single gathering in the interior, notably (shamefully) attended by Medics and Doctors. It's a little risky. It's just not the time to have major gatherings, but at least this one is in the service of the science of 'stopping it' - which is better than most places are doing.

wesleywt 6y ago

I don't think your percentage is correct. It depends on the PCRs target gene, the quality of the samples and the quality of the machine used. They are using reverse transcriptase quantitative PCR. What is your source for that FNR?

wittyreference 6y ago
Ai et al. https://pubmed.ncbi.nlm.nih.gov/32101510/ https://pubmed.ncbi.nlm.nih.gov/32101510/ analysis of 1014 patients in Wuhan; Yang et al. looked at another 213 in Guangdong. Kucirka et al (https://www.acpjournals.org/doi/10.7326/M20-1495 https://www.acpjournals.org/doi/10.7326/M20-1495) did a pooled analysis of seven other studies and found that "over the 4 days of infection before the typical time of symptom onset (day 5), the probability of a false-negative result in an infected person decreases from 100% (95% CI, 100% to 100%) on day 1 to 67% (CI, 27% to 94%) on day 4. On the day of symptom onset, the median false-negative rate was 38% (CI, 18% to 65%). This decreased to 20% (CI, 12% to 30%) on day 8 (3 days after symptom onset) then began to increase again, from 21% (CI, 13% to 31%) on day 9 to 66% (CI, 54% to 77%) on day 21."

tinus_hn 6y ago

> A negative test can tell us: > the person was unlikely to be infectious at the time of the test. https://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-health-advice-general-public/assessment-and-testing-covid-19/covid-19-test-results-and-their-accuracy https://www.health.govt.nz/our-work/diseases-and-conditions/... Of course if you are testing proper samples from the airways and they come back negative, it’s not very likely these people can infect others through droplets from their airways at that moment. The problem is that they might be in the incubation period but that is irrelevant here. And anyway PCR does not nearly have a 30% false negative rate, I’m not quite sure where you got these stats.

wittyreference 6y ago
Ai et al. https://pubmed.ncbi.nlm.nih.gov/32101510/ https://pubmed.ncbi.nlm.nih.gov/32101510/ analysis of 1014 patients in Wuhan; Yang et al. looked at another 213 in Guangdong. Kucirka et al (https://www.acpjournals.org/doi/10.7326/M20-1495 https://www.acpjournals.org/doi/10.7326/M20-1495) did a pooled analysis of seven other studies and found that "over the 4 days of infection before the typical time of symptom onset (day 5), the probability of a false-negative result in an infected person decreases from 100% (95% CI, 100% to 100%) on day 1 to 67% (CI, 27% to 94%) on day 4. On the day of symptom onset, the median false-negative rate was 38% (CI, 18% to 65%). This decreased to 20% (CI, 12% to 30%) on day 8 (3 days after symptom onset) then began to increase again, from 21% (CI, 13% to 31%) on day 9 to 66% (CI, 54% to 77%) on day 21." You are mistaken.
- tinus_hn 6y ago
  Your first study takes rt-pcr testing as the reference and cannot make statements on its accuracy. Your second study is flawed as indicated by the comments on it, it’s a meta study based on poor data. If you choose a point and go looking for studies to support it, regardless of quality and ignoring studies that do not, that’s not science. And by the way, your studies are about something that is not important here. What is important here is if the patient carries virus particles in his airways so they can infect others, not if they are infected themselves.