Some SARS-CoV-2 diagnostic tests might be picking up small traces of the virus within some individuals and giving them a positive result when they are unlikely to be contagious, the New York Times reports.
For PCR-based tests, it writes that the issue stems from the cutoff values used to determine if someone has the virus, which is typically a cycle threshold of about 40 — meaning it takes that many rounds of amplification to find the virus in the sample. But some experts tell teh Times that cutoff is too sensitive and a lower one would be more reasonable, however others argue that lower thresholds might miss newly infected people who don't yet have a high viral load.
To the Times, this suggests that other tests, like rapid tests, though they are less accurate, might be better suited as the first-line test as they can be repeated more frequently. Since they can be repeated, those newly infected people could be caught on a re-test. Harvard T.H. Chan School of Public Health's Michael Mina tells the Times that testing in the US at this stage of the pandemic has to be quick, inexpensive, and easy to get so that even if not every infection is caught, the people most likely to transmit the virus are.