Rapid coronavirus test used by White House misses nearly half of positive cases

[RandySF]

A rapid coronavirus diagnostic test manufactured by Abbott may miss nearly half of all positive infections, according to a pre-published study from New York University.

The analysis of Abbott's ID NOW system, which has not been peer-reviewed, found the test to be "unacceptable" in a clinical setting.

But Abbott said it's not clear if the researchers used the samples correctly. A spokesperson said the company's own rate of false negatives that it has shared with the Food and Drug Administration (FDA) is .02 percent.

"In communications with the users of the test, it is performing as expected," Darcy Ross said in a statement to The Hill. "ID NOW is an important tool that delivers information where it’s needed most — taking the fight to the frontlines so that public health officials and healthcare providers can make critical decisions."



https://thehill.com/policy/healthcare/497661-abbotts-rapid-coronavirus-test-misses-nearly-half-of-positive-cases-study

[MFM008]

1. All it takes

Is one to slip through.

[dewsgirl]

3. Promise?😉

[OhioChick]

2. These have been throwing false negs at an alarming rate n/t

[tanyev]

4. So they're relying on fake news.

[dalton99a]

5. Fake news. The tests are beautiful and perfect

[ProfessorGAC]

6. The Third Paragraph Tells Me A Lot

But Abbott said it's not clear if the researchers used the samples correctly. A spokesperson said the company's own rate of false negatives that it has shared with the Food and Drug Administration (FDA) is .02 percent.

This is an obvious indicator that the method is highly technique sensitive, & they know it.

Across multiple operators, one can assume (reasonably) that the spread is +/- 50% of finding.
So, good operators might get a false negative 3 in 8 times, while the less skilled tech would get 5 in 8.

I'm guessing their internal testing involved 2 VERY meticulous operators who got good results because they were involved in method development.

They probably also assumed a "one sample, run twice, with referee" protocol. I don't know we can assume that's how it's being done in the field.

[gibraltar72]

7. Why don't they try the magic 8 ball? Many people say it's very reliable.