Follow

Covid rapid test psa 

Just a reminder that the rapid test is about 80% effective in diagnosing symptomatic Covid cases. If you’re asymptomatic, it’s about 30% effective — less accurate than a coin toss.

If you’re concerned about a Covid exposure but you don’t have symptoms, get a PCR test, not a rapid test! I don’t know why clinics are offering rapid tests to everyone regardless of symptoms, but I don’t know why anyone is doing anything in 2020 tbh.

(Source of my stats: nytimes.com/2020/11/02/health/)

Covid rapid test psa 

@courtney How is it possible to be less accurate than a coin toss?

Covid rapid test psa 

@waweic because a coin toss is 50% and this is 32% and 50>32

Covid rapid test psa 

@courtney Then you could invert the result and it would be 68%

Covid rapid test psa 

@waweic if it was accurate 68% of the time it would be...more effective than a 50/50 guess. Which is what a coin flip is.

Anyway, have a good night.

Covid rapid test psa 

@waweic @courtney I was wondering this too, like if someone got a 0% on a 100 question true false quiz they probably knew all the answers. These numbers must be relative to 50%? Like maybe 30% translates to 65% absolute? Dunno

Covid rapid test psa 

@waweic OK, I think I get it. The article says it detects 32% of cases detected by PCR. This isn't the same thing as saying that 32% of the time, it is wrong about whether you have COVID. If the test comes back positive, it's very likely you have it, depending on the rate of false positives. If it comes back negative, it doesn't really say much about whether you have it or not.

Covid rapid test psa 

@courtney This is really important info I''m going to share w friends and family, especially in states like Iowa and Wisconsin. 2 spouses w cases at their work were directed to rapid tests then told to return on the negative result.

Genuinely wonder if the misleadingly low numbers are WHY employers direct them to the rapid tests and thus clinics offer them.

Covid rapid test psa 

@courtney this bit of the article really calls that summary into question. The research screened 1500+ asymptomatic people and PCR found *any* covid in 19. Of those 19, 6 were also detected by the rapid test. Of the remaining 13 positive PCR, *12* of the had C.T. >30. That's at the very far end of PCR detection possibilities, I've never seen a C.T. value greater than 25 in previous covid PCR papers.

Lower CT means it was easier to detect (greater quantity of viral genetic material in the sample) and is highly correlated with both symptom severity and infectivity.

The "flip a coin" metaphor is also a problem, it seems to imply that the test will give false positives (say someone is infected when they're not), but the rapid antigen tests have specificity >99.9% according to the CDC, which means essentially no false positive results.

Definitely PCR is more effective, it's more sensitive to later-stage and asymptomatic cases. But there's a place for a cheaper test too.

@courtney When looking at these things, the magic words are sensitivity (what percent of infections does it detect) and specificity (what percent of "detections" are correct). There's no way to combine the two into a single, meaningful number -- any time an article just gives one number, you're not getting the full picture.

@aparrish

Sign in to participate in the conversation
Friend Camp

Hometown is adapted from Mastodon, a decentralized social network with no ads, no corporate surveillance, and ethical design.