Covid-19

[Lloyd Braun]

Wow. You’ve gone off the rails.... and I followed you there. You will continue to spew lies and fail to comprehend simple reading and I will continue to get sucked into your tangled web of “isms” hoping one day you will admit you were MAYBE wrong or misspoke. We’ve reached the pity stage however as there is no chance of either of those things happening ever. Good day, sir.

[D-West & PO-Z]

My example was very specifically the actual COVID test, not the antibody test. It seems surprising to me that a group of that many people, including people over 70 and 2 with Lupus would all be positive but asymptomatic. Then again this is a strange virus. I'm sure we will learn more as time goes on, but right now the randomness of it seems shocking.

Yeah I didnt mean to imply you were confused.

I have known several people asymptomatically positive as well.

[Muskie in dayton]

My example was very specifically the actual COVID test, not the antibody test. It seems surprising to me that a group of that many people, including people over 70 and 2 with Lupus would all be positive but asymptomatic. Then again this is a strange virus. I'm sure we will learn more as time goes on, but right now the randomness of it seems shocking.

The NY Times article I linked above mentioned a newer COVID test with 95% accuracy on the positive and negative sides. That sounds pretty good, but doing the math reveals the problem. With a testing population of 10,000, assume 1% actually have COVID. So 100 have it, 9900 do not. Then doing some basic math with with 95% accuracy, you can expect:
- 5 False Negatives (.05*100)
- 495 False Positives (.05*9900)

So theoretically, this testing method will show 95 true positives and 495 false positives. That means that roughly 4 our of 5 positive tests is not real. That's a problem.

[paulxu]

Georgia Gov. Brian Kemp (R) on Wednesday explicitly banned cities and counties from adopting rules requiring masks or other face coverings.

I'm at a loss to understand how this would help anything. If some city wanted to require masks in public places (government offices, etc.) why would the governor care? Who is it hurting?

[X-man]

I'm at a loss to understand how this would help anything. If some city wanted to require masks in public places (government offices, etc.) why would the governor care? Who is it hurting?

Because he's an idiot who wants to politicize the mask issue.

[boozehound]

The NY Times article I linked above mentioned a newer COVID test with 95% accuracy on the positive and negative sides. That sounds pretty good, but doing the math reveals the problem. With a testing population of 10,000, assume 1% actually have COVID. So 100 have it, 9900 do not. Then doing some basic math with with 95% accuracy, you can expect:
- 5 False Negatives (.05*100)
- 495 False Positives (.05*9900)

So theoretically, this testing method will show 95 true positives and 495 false positives. That means that roughly 4 our of 5 positive tests is not real. That's a problem.

That's an interesting point. It probably doesn't impact the results as much in smaller cohorts (i.e. if you test 5 people in a family and they all test positive, the results are probably valid) but can certainly skew the data in widespread testing, particularly of asymptomatic / healthy individuals.

Because he's an idiot who wants to politicize the mask issue.

It seems to me like a risky battle to pick, but what do I know?

[Lloyd Braun]

The NY Times article I linked above mentioned a newer COVID test with 95% accuracy on the positive and negative sides. That sounds pretty good, but doing the math reveals the problem. With a testing population of 10,000, assume 1% actually have COVID. So 100 have it, 9900 do not. Then doing some basic math with with 95% accuracy, you can expect:
- 5 False Negatives (.05*100)
- 495 False Positives (.05*9900)

So theoretically, this testing method will show 95 true positives and 495 false positives. That means that roughly 4 our of 5 positive tests is not real. That's a problem.

This was just for the antibody test though, correct? The article you linked used the 95% as an example unless my reading comprehension is poor (gulp!). The molecular testing -nasopharyngeal swab that everyone uses for widespread testing- should be highly specific (99%+) thus greatly decreasing false positives. That’s what I have been told anyways unless it’s the BD brand which has been under question for a few months.

[Muskie in dayton]

This was just for the antibody test though, correct? The article you linked used the 95% as an example unless my reading comprehension is poor (gulp!). The molecular testing -nasopharyngeal swab that everyone uses for widespread testing- should be highly specific (99%+) thus greatly decreasing false positives. That’s what I have been told anyways unless it’s the BD brand which has been under question for a few months.

My read was that it was the actual COVID infection test. The article hopped around between the two, but was discussing the older COVID tests missing so many infections, then moved right into the 95% discussion.

[xavierj]

Meanwhile, with the increase use of masks the cases keep Jumping. Also the death rate to number of cases fell below 4%. Should end up around 0.4% about a year from now, if not lower. I think we are still under 10,000 deaths for people under 55.

[Lloyd Braun]

My read was that it was the actual COVID infection test. The article hopped around between the two, but was discussing the older COVID tests missing so many infections, then moved right into the 95% discussion.

Unless you linked 2 articles recently, I didn’t get that at all, just that the 95% was used as an arbitrary number to demonstrate that it sounds like a good accurate test, but really isn’t. I can try to find data later for the PCR tests, which vary by manufacturer, but the specificity is significantly higher with those ~98-99% - IF they are done correctly. FWIW in a small sample size (25-30) I have seen first hand the local CVS minute clinics give patients the swab to do themselves. The patient swabs their own nose 1/4” inside and don’t go all the way up to the nasopharynx, which is more likely to obtain viral nucleic acid. The true lag is in false negatives with the current PCR tests, as they are more specific than sensitive. Antibody tests seem to have more propensity for false positives.