Page 1 of 1

90% of COVID Tests Might be Wrong

Posted: Wed Sep 02, 2020 3:10 pm
by ragtimejoe1
Man, I missed this the other day when it came out.

https://www.nytimes.com/2020/08/29/heal ... sting.html

I know that technology very well and this is unbelievable. 90% of positive cases are at cycle 37+!! In my experience, things get a little dicey once you pass around 32-34 cycles. The best test I've ever seen using that same type of assay was reliable to 35 cycles. Most of the scientists in that article are purposely avoiding stating how bad that actually is. 37-40 cycles isn't just "too sensitive"; it has a very high chance of being very unreliable.

To me, this is the biggest bombshell I've seen associated with COVID. I mean WOW! Really shocked that this hasn't received more attention (maybe I just missed it).

Re: 90% of COVID Tests Might be Wrong

Posted: Wed Sep 02, 2020 4:25 pm
by laxwyo
There’s so much evidence that points to a HUGE overreaction. This is just more. Just like JA’s false positive. Play ball!

Re: 90% of COVID Tests Might be Wrong

Posted: Thu Sep 03, 2020 8:18 am
by ragtimejoe1
I just forwarded that article to as many reps as I could think of. The validation assays for each lab need to be made publicly available. If the assay isn't validated out to 37 or 40 cycles and if the products of the reaction haven't been quality checked, then that means our actual COVID numbers might be 90% less. This is a big deal because all the lock down bs is predicated on case numbers.

Wondering if the false positives are leading to over inflation of estimations of asymptomatic carriers, spread, and how contagious it really is.

We very much need to get more information on this. If this article's data is correct, then it completely changes everything associated with COVID in the US.

If you don't recall this from your biology class, basically this assay uses specific "primers" that bind to the viral genome. The process uses PCR to amplify that section of DNA. Each "cycle" is a doubling of that section of DNA (1 to 2; 2 to 4; 4 to 8; etc.). During the process a reporter dye is incorporated (various ways; in some cases a fluorescent marker is released) which is read by the machine after each cycle. If you start with a lot of virus, then there is a lot of viral DNA. When you start with a lot of DNA, it doesn't take as long to amplify and detect a signal. The test is fantastic for x number of cycles (you have to run validation assays to see how long that is), but once it gets past that number of cycles, the primers can bind each other or perhaps they bound regions they weren't supposed to, or other situations that give you "background noise". It provides a signal on the machine but is actually meaningless.

It is more than a little suspicious that 90% of the cases show positive at the very last cycle of the assay. Even for yes/no this is unacceptable. There is a very high chance that at the point of the assay, they amplified a non-target molecule which provided a signal but is not COVID.

Re: 90% of COVID Tests Might be Wrong

Posted: Thu Sep 03, 2020 10:24 am
by ZapPoke
I wish they’d get over this whole “follow the science “ The science has turned out to be worse than statistics. You can manipulate both to support whatever narrative you are pushing.

Mark Twain had it right. If you don’t read the news you are uninformed. If you do read the news you are misinformed.

Re: 90% of COVID Tests Might be Wrong

Posted: Thu Sep 03, 2020 10:33 am
by ragtimejoe1
ZapPoke wrote: Thu Sep 03, 2020 10:24 am I wish they’d get over this whole “follow the science “ The science has turned out to be worse than statistics. You can manipulate both to support whatever narrative you are pushing.

Mark Twain had it right. If you don’t read the news you are uninformed. If you do read the news you are misinformed.

It is mind boggling. The other is myocarditis. Big 10 estimates 30 to 35% of their cases had myocarditis and use that as a big component of why they shut down the season. On the other hand, the CDC tells you that if you don't have symptoms or only mild symptoms (the same for most college athletes), you don't need to get tested under any circumstance including if you have had contact with someone who has COVID. By virtue of recommending you to heal at home on your own without being tested, they are telling you the myocarditis issue is not a severe health threat. If it were, the CDC recommendation would be to get tested and monitor heart function.

This poop is unraveling quickly now.

Re: 90% of COVID Tests Might be Wrong

Posted: Thu Sep 03, 2020 10:42 am
by BeachPoke
This is all just a hype for the elections nothing more. I’ve had several friends and was sick myself for a week with a mayor headache and all the other symptoms. I tested myself and it came up negative. Worst headache I’ve ever had, I understand the need to keep safe but wash your hands stay 6 feet apart and get on with life. Just something for information. Google. Plandemic and watch the documentary it is pretty revealing especially with Bill Gates. Pay attention to the last few minutes when he states there will be another break out and laughs.

Re: 90% of COVID Tests Might be Wrong

Posted: Thu Sep 03, 2020 11:00 am
by Adv8RU12
There are two simple metrics for most tests. Most people can understand them. I don't understand why they aren't always listed for a test. These are the numbers (metrics) that you want to know and their definition:

SENSITIVITY: The probability that you HAVE the condition if the test says you have it.
SELECTIVITY: The probability that you DON'T HAVE the condition if the test says you don't have it

Re: 90% of COVID Tests Might be Wrong

Posted: Thu Sep 03, 2020 11:06 am
by Adv8RU12
SORRY -LAST POST NEEDS A CORRECTION

SENSITIVITY should be replaced with SPECIFICITY

Re: 90% of COVID Tests Might be Wrong

Posted: Thu Sep 03, 2020 11:12 am
by Adv8RU12
SORRY AGAIN !!

The two metrics are, in the order first stated

SENSITIVITY and
SPECIFICITY

Re: 90% of COVID Tests Might be Wrong

Posted: Thu Sep 03, 2020 11:13 am
by ragtimejoe1
Adv8RU12 wrote: Thu Sep 03, 2020 11:00 am There are two simple metrics for most tests. Most people can understand them. I don't understand why they aren't always listed for a test. These are the numbers (metrics) that you want to know and their definition:

SENSITIVITY: The probability that you HAVE the condition if the test says you have it.
SELECTIVITY: The probability that you DON'T HAVE the condition if the test says you don't have it
For this test, my bigger worry is the accuracy. They are calling a 37 cycle or 40 cycle detection as positive, but their assay is likely inconclusive past say 34 cycles. Even if they are detecting some odd ball remnant of viral particle (DNA) it is likely not at a level for infection or transmission. This is exactly why people have been testing positive for weeks or retesting as positive after they have cleared the disease.

A positive at say 20 cycles means you are likely positive and can shed the virus. A positive at 37 or 40 cycles is suspect at best.

Re: 90% of COVID Tests Might be Wrong

Posted: Thu Sep 03, 2020 6:34 pm
by Asmodeanreborn
BeachPoke wrote: Thu Sep 03, 2020 10:42 am Google. Plandemic and watch the documentary it is pretty revealing especially with Bill Gates. Pay attention to the last few minutes when he states there will be another break out and laughs.
Plandemic is as much of a documentary as Monty Python's Quest for the Holy Grail.

Seriously.

Re: 90% of COVID Tests Might be Wrong

Posted: Thu Sep 03, 2020 7:09 pm
by 'PokeForLife
Adv8RU12 wrote: Thu Sep 03, 2020 11:00 am There are two simple metrics for most tests. Most people can understand them. I don't understand why they aren't always listed for a test. These are the numbers (metrics) that you want to know and their definition:

SENSITIVITY: The probability that you HAVE the condition if the test says you have it.
SELECTIVITY: The probability that you DON'T HAVE the condition if the test says you don't have it
Apologies, but really I think these should be:

Positive predictive value
and
Negative predictive value

Sensitivity is the percentage of the "true positives" that the test will correctly identify. So if there are 10 people who have the rona and you get 9 positives after testing them all, then the test has a sensitivity of 90%.
Specificity is the percentage of the "true negatives" the test correctly identifies. If you have 10 people who really have flu and not rona, and you get 9 negatives/1 positive rona test, then the specificity would be 90%.

I don't know what the stated sensitivity, specificity, PPV, or NPV on these tests really are, but it sounds like they could have pretty crappy specificity if the above article is true.

Re: 90% of COVID Tests Might be Wrong

Posted: Fri Sep 04, 2020 8:56 am
by ragtimejoe1
'PokeForLife wrote: Thu Sep 03, 2020 7:09 pm
I don't know what the stated sensitivity, specificity, PPV, or NPV on these tests really are, but it sounds like they could have pretty crappy specificity if the above article is true.
I think they are actually pretty good if the results are interpreted in the proper range. In other words, at cycles x through y, the results are likely near 100% PPV. If no signal, the test is probably near 100% on NPV.

I'm am shocked that they are using data that comes up at the absolute end of the assay (last cycle for 90% of the freaking tests!). At best, it is too sensitive as some in the article imply, thus likely not biologically relevant (i.e. the person is not shedding virus). At worst, the test is actually measuring an off-target DNA sequence and is completely worthless (i.e. false positive). They absolutely should at minimum be applying biological relevance to cycle number AND sequencing some reaction products at 37 or 40 cycles to ensure it isn't off-target amplification.

Keep in mind every 3.3 cycles is a 10 fold difference. So, 3.3 cycles = 10x; 6.6 = 100x; 9.9 = 1000x; 13.2 = 10,000x; 16.5 = 100,000x; 19.8 = 1,000,000x; etc. 37 cycles would be approximately 1.4 x 10^11 amplification of each copy of viral DNA that was put in reaction.

I'm very very surprised this isn't getting more play. It might be by those in the know?