DC's caregiver is visiting family (by car and just for the day I think), which will effectively expose caregiver to 2-3 households as family members / households combine for dinner (outdoors, but in heated tent, which seems the same as indoors to my layperson's understanding . . .). What's an appropriate amount of time to wait to confirm no exposure before caregiver comes back? Should we test beforehand?

have her get a test on the 5th day after exposure, and then come back when you get a negative result.
That's exactly what I was thinking - so glad to hear from someone else that they think this is reasonable vs. a full 14 day quarantine. Thanks for the feedback.
I agree with this. Unless someone in the group subsequently tests positive.
Ugh all of these necessary decisions make me anxious
Simplify. Keep it just yourselves
Here (VT) the guidelines are to wait 7 days after returning for the Covid test. But unfortunately our turnaround times are something like 6 days now.
You don't have rapid testing there?
@Anonymous my understanding is that rapid tests have too many false positives. True cases are confirmed by PCR. I don't think rapid tests are standard of care.
I thought you could get rapid result PCR tests, though?
FWIW, the turn around at Health and Hospital Sites for test results is currently under 36 hours. They are the regular test, but they are processing them locally using something called a "pooling" method which makes results come faster.
I thought pooling meant that they take a bunch of tests together and if there’s one positive, the whole bunch has to come back and get retested. (Only efficient if there’s low disease prevalence.)
I don't think pooling has to work that way. they can test a pool and if the pool is positive, then they test each individual sample in the pool to identify who in the pool was positive and then that person is notified.
@Anonymous I spoke to people who are good at math, and they said pooling is a good approach. I can't speak to the math of it, but it's still faster to use this method. As the second poster wrote, you don't need to retest. Also, even at 3% the prevalence is still low.