4 ms·

> On the other hand PCR test are not inaccurate at all, at least regarding false positives. https://www.thelancet.com/journals/lanres/article/PIIS2213-2600(20)

by collyw 6y ago

> On the other hand PCR test are not inaccurate at all, at least regarding false positives. https://www.thelancet.com/journals/lanres/article/PIIS2213-2600(20)30453-7/fulltext https://www.thelancet.com/journals/lanres/article/PIIS2213-2... between 0.8% and 4% false positive rate according to that, making them completely unsuitable for mass testing. On top of that BBC's panorama did a program pointing out the sloppy standards at centers where the tests are processed.

mrtksn 6y ago
How did you decide that this range exactly is unsuitable enough that we are better off we simply assume that the reality is the one that suits our agenda? IMHO that's a pretty good accuracy, significantly better than the "I feel like", "I want to believe that", "what if it's all about" or the "It's suits me better if" methodologies. I'm not being snarky, I simply want to outline the alternatives I'm aware of. I respect the accuracy of the PCR, it can literally tell us why people in India are dying off on the streets right now in thousands and we even cannot see the assailant through optical magnification. It can also tell us that it's the same thing that caused the exact same situation in NY and Italy. It's amazing, much more intriguing than any other explanation that I came across.
- MR4D 6y ago
  I’m not the OP, but the rest of the paragraph that the quote above was linked to reads as follows (note the last 3 words in the quote): * This rate could translate into a significant proportion of false-positive results daily due to the current low prevalence of the virus in the UK population, adversely affecting the positive predictive value of the test.2 Considering that the UK National Health Service employs 1·1 million health-care workers, many of whom have been exposed to COVID-19 at the peak of the first wave, the potential disruption to health and social services due to false positives could be considerable.*
- collyw 6y ago
  At the height of the "second wave" here, it was estimated that around 0.8% of the population was infected at the time. That means that if you were randomly tested there was a greater chance that you would have a false positive than an actual infection.
  - mrtksn 6y ago
    That would be unfortunate to all those prevalence studies where the difference between 1.1 and % 0.3 means the world for them. They may never know if it was %0.5 or %0.8 of the people infected. On the other hand, it's a pretty good accuracy when you have a hospital full of sick people. Sure, you may end up recommending 14 days of staying at home every now and then when it's not necessary. See, the PCR test is just one of the clues that is used to diagnose people. Negative test with clearly Covid-19 symptoms and medical image of the lungs showing spots gets you a bed and treatment in the hospital. Positive test and no symptoms gets you stay at home and wear a mask recommendation.
    - collyw 5y ago
      That's why I said they were completely unsuitable for mass testing. I said nothing about them being used where covid is suspected.
SketchySeaBeast 6y ago
Nevermind the fact that as we find more people positive we see an increase in hospitalizations and mortality shortly afterwards, why in these charts[1] did Alberta get so many sub 0.8% 7 day rolling averages then? https://www.alberta.ca/stats/covid-19-alberta-statistics.htm#laboratory-testing https://www.alberta.ca/stats/covid-19-alberta-statistics.htm...

by collyw 6y ago

mrtksn 6y ago

How did you decide that this range exactly is unsuitable enough that we are better off we simply assume that the reality is the one that suits our agenda? IMHO that's a pretty good accuracy, significantly better than the "I feel like", "I want to believe that", "what if it's all about" or the "It's suits me better if" methodologies. I'm not being snarky, I simply want to outline the alternatives I'm aware of. I respect the accuracy of the PCR, it can literally tell us why people in India are dying off on the streets right now in thousands and we even cannot see the assailant through optical magnification. It can also tell us that it's the same thing that caused the exact same situation in NY and Italy. It's amazing, much more intriguing than any other explanation that I came across.

MR4D 6y ago
I’m not the OP, but the rest of the paragraph that the quote above was linked to reads as follows (note the last 3 words in the quote): * This rate could translate into a significant proportion of false-positive results daily due to the current low prevalence of the virus in the UK population, adversely affecting the positive predictive value of the test.2 Considering that the UK National Health Service employs 1·1 million health-care workers, many of whom have been exposed to COVID-19 at the peak of the first wave, the potential disruption to health and social services due to false positives could be considerable.*
collyw 6y ago
At the height of the "second wave" here, it was estimated that around 0.8% of the population was infected at the time. That means that if you were randomly tested there was a greater chance that you would have a false positive than an actual infection.
- mrtksn 6y ago
  That would be unfortunate to all those prevalence studies where the difference between 1.1 and % 0.3 means the world for them. They may never know if it was %0.5 or %0.8 of the people infected. On the other hand, it's a pretty good accuracy when you have a hospital full of sick people. Sure, you may end up recommending 14 days of staying at home every now and then when it's not necessary. See, the PCR test is just one of the clues that is used to diagnose people. Negative test with clearly Covid-19 symptoms and medical image of the lungs showing spots gets you a bed and treatment in the hospital. Positive test and no symptoms gets you stay at home and wear a mask recommendation.
  - collyw 5y ago
    That's why I said they were completely unsuitable for mass testing. I said nothing about them being used where covid is suspected.

SketchySeaBeast 6y ago

Nevermind the fact that as we find more people positive we see an increase in hospitalizations and mortality shortly afterwards, why in these charts[1] did Alberta get so many sub 0.8% 7 day rolling averages then? https://www.alberta.ca/stats/covid-19-alberta-statistics.htm#laboratory-testing https://www.alberta.ca/stats/covid-19-alberta-statistics.htm...