...The number one quoted me -- and no idea how accurate it is -- is that rapid tests only identify 1 out of 5 active Covid cases. Basically, they can identify a 'high density' of the virus, so if folks are very sick, they work. Otherwise, they're very unreliable. The workflow our legal counsel suggested were we to implement rapid testing is to conduct the rapid test and to only trust positive results they return. For negative tests, we were instructed to have them sent to Labcorp for a second diagnosis that would take 5-7 days. [/QUOTE]
I can't find any specific results for the Quidel test, just that it has a high degree of specificity, but lower sensitivity than PCR tests:
https://www.bioworld.com/articles/435020-fda-applies-eua-... Thus, you can trust the positive results, but a negative result is less meaningful, and you might want a lab PCR test done as well, pretty much as you suggested.
Here's a different rapid test that does quote specifics:
https://www.thedailybeast.com/fda-approved-veritor-corona... [quote=daily beast]As it turns out, the Veritor machine with a coronavirus assay accurately detects the virus only 84 percent of the time, BD found after a trial involving 226 patients in 21 locations. Laboratory molecular tests, by contrast, are nearly 100-percent sensitive but can take days to complete.