Thread by @michaelmina_lab

Why do we need frequent testing over very sensitive testing? It’s easy

The average person is PCR+ for ~30+ days but infectious for ~7.

Most people are asymptomatic. So, infrequent PCR based surveillance will detect + people but ~75% will no longer be transmitting virus

1/...

We know this. The CDC recommends that after someone recovers from infection and quarantines for ten days, they do not test again bc they are likely to be positive but no longer infectious

2/

So we know that testing someone more than ten days or so days after they are infected doesn’t make a lot of sense.

If performing random testing on asymptomatics, most who are Found to be positive will be discovered only after the first week or so of infection...

3/

And these people by and large are not infectious anymore. You can do the math 1-(7/30) = 76% who are + are discovered when they are no longer infectious given random testing of asymptomatic.

So we need to not test randomly for people, we need test frequenty...

4/

If we are to use testing as a useful tool to stop transmission, we have to test individuals frequently enough that we find them in their 1st or 2nd day of transmission, and remove them from interaction with others, rather than swabbing them ten days after they are infectious

5/

So to use testing as a means to stop transmission in asymptomatics, we need need frequent testing with fast results that detect ppl with high viral loads. We don’t need infrequent tests like PCR that detect people long after they are done transmitting virus to their neighbors.

Note this is very different from diagnostic testing of sick people where we need to be able to gather all evidence possible for why someone is sick.

If sick, get a PCR test and discuss with your doctor

If not symptomatic, etc, a one time test provides very little information.