Skip to the bottom bolded text for the important stuff.That's what I find strange. They're not updating the IFUs, even in electronic form. That doesn't make sense.
This page says that there's evidence. Now, we all know that this is based solely on manufacturer's claims. It even says so on the page in 2 places.
https://www.tga.gov.au/post-market-review-antigen-and-rapid-antigen-tests
why then are the IFUs not updated to reflect as such? Is it because that doesn't meet standards elsewhere?
https://www.tga.gov.au/covid-19-rapid-antigen-self-tests-are-approved-australiaNow, they have 3 tests registered under the same ARTG entry. The Alltest 502, the Alltest 802 and the My covid test 802. The 802s are identical except for branding. The my covid test IFU is actually linked under the alltest name on the TGA's site.
ICOV-802 (oral). Last mention is Delta and there's a viral copies figure given. That tells me that they haven't completed any kind of real-world omicron testing. Only enough to satisfy the TGA (in silico?).
ICOV-502 (nasal).
I mean, I've said it before but here it is again. Going back to the Germans...
On the TGA's site, they're both listed as ICOV, there's no INCP. However I actually do think that it is the INCP version due to the (nasal swab) in the name. These inconsistencies in the numbering / naming conventions can cause confusion. It's certainly fooled me in the past.
The stats given in the IFU for the oral are 100 / 97.7 / 87.
The stats shown below are 100 / 55 / 0.
The nasal one doesn't even give a breakdown even though they're from the same manufacturer.
There's an even bigger issue here though. Take a look at the copies/ml statistics. Take a look at their claimed levels of detection for any given strain. Some are measured in ng/ml. we'll ignore those for now.
Others are measured in copies per µL, which is 0.001mL. For reference, ADO's LOD is claimed as 2.81 x 10^5 genome copies/mL, which is 281,000/mL or 281/µL.
Back to this test's LOD.
For delta, the tests claim 10^5 µL
For alpha, the tests claim 10^2 µL
For beta, the tests claim 10 µL/mLFirst thing that's easily noticeable here is that both tests claim different sensitivity figures, yet claim to have the exact same levels of detection for particular strains.
The second thing is that the LOD claims vary wildly from strain to strain:
10^5 = 100,000
10^2 = 100
10 = 10So, according to that, the test is 10,000x more sensitive to beta than it is to delta? Yeah, OK... At that rate it should be 1,000,000 less sensitive to Omicron and any sensitivity figures should go out the window.
The third thing is that
10^5 µL/mL= 100,000 copies per µL, or 100,000,000 copies/mL
10^2 µL/mL= 100 copies per µL, or 100,000 copies/mL
10 µL/mL= 10 copies per µL, or 10,000 copies/mL
Going back to ADO's test, it's 281,000 copies/mL claimed. So this test is either much worse, or much better, depending on what strain they're claiming LODs on.
Let's look at their stated overall LOD.
8x10^2 is 800. So 800 TCID50/ml
According to the VIDRL, 1.3 x 10^6 copies/mL is 5.8 x 10^3 TCID50 /mL. 1,300,000 copies/mL is 5,800 TCID50/mL.
Converting the 5800 value into the test's 800 to get the equivalent viral copies/ml.
5800 / 800 = 7.25
1300000 / 7.25 = 1793103
The test's claimed 8x10^2 TCID50/ml is equivalent to 1,793,103 copies/ml unless I'm wrong.
So they are saying that they're good at detecting as little as 1.8 million copies / ml. But their strain specific claims range from 10 thousand to 100 million.
None of that makes sense.
And the TGA are passing this stuff. Are they even reading the manuals? Do any calculations get made? Anyone can see at a glance that 10 for one strain and 10^5 for another makes absolutely 0 sense.
(20min delay)