Have you heard anything about the test being only 65% reliable?????

[LAS14]

Last edited Sat Apr 4, 2020, 03:06 PM - Edit history (1)

My daughter-in-law is an ICU physician. She was home for four or five days after a fever and chills, waiting for test results. It came back negative, so she's back at work. But at home she's still self isolating because her understanding is that the tests can give false negatives 35% of the time. Is that what you've heard????? Sobering.

tia
las

[customerserviceguy]

1. I've not heard that

but with so many variations of the tests out there, it might be useful to know which ones were the most accurate, with the fewest numbers of false positives and false negatives.

[hlthe2b]

2. Antigen tests are dependant on viral load. Very small amounts of virus and that sequestered

in less accessible areas of the respiratory tract or other areas of the body might not be picked up but still be present. Once additional replication occurs, future tests will likely be positive.

People are always impressed with the fact that RT-PCR is so sensitive it can detect very few nucleic acid particles associated with the virus. While that is true, there are still detection limits.

The bottom line this is more likely to be a sampling/timing artifact than lack of reliability of the test.

That is why we SHOULD Have had both an IgM and an IgG antibody test to use alongside the antigen tests. IgM antibody comes up early in infection and could be very useful with testing. Once IgM antibody is gone and IgG is detectable (and antigen is negative) one could reliably assume the infection had cleared in an asymptomatic patient.

[SheltieLover]

9. Thank you for explaining

I have wondered about that too and, in my nonmedical reasoning, concluded that viral load would need to be sufficient to show a positive.

I know that the sailors in Guam have been retested a couple of days apart.

[unblock]

3. I've heard 30%, but that's still high

Most tests have either very low false positives and higher false negatives, or it's the other way around.

The standard Covid-19 test (known as a reverse transcriptase polymerase chain reaction (RT-PCR) test) appears to have very low false positives (meaning if it says you have it, it's almost surely actually do have it), but high false negatives.


https://www.livescience.com/covid19-coronavirus-tests-false-negatives.html

[SheltieLover]

11. Thank you

For explaining & for the link. 😊

[Clash City Rocker]

4. If so, the test would be basically useless

Let’s hope that isn’t the case.

Oh, and just to be pedantic... 65% + 45% is 110%.

[LAS14]

16. :-) my arithmetic deteriorates day by day, hour by hour...

[Hoyt]

5. That's pretty much the case with flu and similar quick tests. Supposedly, if you

test positive, it is highly likely you have it.

But, if you test negative, you still might have it.

Of course, even if a negative were 100% reliable, you might get it from the testing, an hour later, etc. So, negative doesn’t mean you are good to go, even with more sophisticated tests.

[Yonnie3]

6. I've heard some unverified reports of 30%

You say 65% reliable and then get 45% false negative - 100 minus 65 is 35% not 45%. So what you are saying is in line with what I've heard.

Scuttlebutt is saying that the small, one at a time Abbott machines that 'Rump showed off are more likely to have false negatives than the hospital lab tests.

Locally they seem to run the full Respiratory Panel first so that flu can be ruled out. Then the CoViD test is run if the Respiratory Panel is negative. Did she have other testing?

[LakeArenal]

7. Would you post where you heard this?

It seems to be a bit fear mongering.

[LAS14]

18. I heard it from my daughter-in-law, the ICU physician, who was explaining why she still doesn't...

... eat or sleep with family members.

[Yonnie3]

8. Here's an article about this to read if you like.

https://www.livescience.com/covid19-coronavirus-tests-false-negatives.html

[LAS14]

19. Thanks! nt

[LAS14]

20. I'm thinking that some of the false negatives could be associated...

... with carriers who aren't very infections.... Low viral load. Here's hoping.

[Yonnie3]

21. and maybe a poorly taken or mishandled sample. n/t

[Drahthaardogs]

10. Pretty standard for a PCR test

False positives don't really happen.

False negatives are 40%

[ProfessorGAC]

12. WSJ Two Days Ago Reported...

...that the medical community consensus right now is 30% false negatives.
The obvious concern is continued spread. The solution would normally be simple. Test 2x. Now, were only talking 9% false negatives. But, we can't test enough now, so they can't be doing people twice. So much for "normally".

[unblock]

13. Testing twice may not help

If the test randomly fails, it would help. But more than likely, there's a reason why the test gave a false negative and the second test may very well give a false negative for the same reason.

E.g., you're infected, but not enough has accumulated at the back of your nasal cavity yet.

Testing twice, 3-4 days apart might work. But then, you might have acquired it during the interval.

Accuracy aside, some people, such as health care workers in frequent contact with Covid-19 patients, should be tested multiple times (weekly?) regardless.

Again, the shortage of tests is a problem.

[ProfessorGAC]

17. That's A Good Point

That said, we couldn't do it anyway. As we both said, the tests aren't out there.

[femmedem]

14. I have a friend who tested negative, was told she had pneumonia but not covid. But a week later

she went back because she still was feeling bad. They said she still had pneumonia but bloodwork didn't indicate COVID-19. But the next day she got a call saying a radiologist had looked at her x-ray and said her lung damage indicated she had it and she should quarantine.

But after the first test but before her second, she visited her parents because she is their primary caretaker. The doctors hadn't warned her about false negatives.

[Igel]

15. I've seen a few reports.

Some said 30% chance of false negative. Others said 30-50%. Estimates.

Little chance of false positives.

No word on if it's a failure of the test itself or of sampling techniques.

[tblue37]

22. Yes. I think it depends on viral load, like pregnancy tests depend on hormone levels.

Test too early & maybe get a false negative.