OK so here we go.

When I order a test for your child as a pediatrician, I have to ensure that there is enough "pretest probability" or "index of suspicion", so that the test results are meaningful and useful. https://twitter.com/AndreaFeller2/status/1264789668261957633
This isn't just because of lab capacity or costs of tests. That's never a primary consideration - what is best for the patient IS.

So it's because if I order a lab, I HAVE to be prepared to manage misleading results if there wasn't a high pretest probability.
Let's look at negatives - Say there is West Nile Virus (WNV) around in Niagara (happens every year). Say a patient with no symptoms somehow self refers for a WNV test. It's negative, so they think they are "clear" so they stop wearing repellant/being careful. And maybe get WNV.
Let's look at timing - let's say someone wants to know if they have coeliac disease. Say they self refer somehow for blood test - and get a negative result. So they stop their months long gluten free diet. And get sick. (Test was negative due to gluten-free diet.)
Having risk isn't the same as needing a test.

A high risk profession isn't the same as having a decent index of suspicion or pretest probability.

Testing capacity IS important - critically important to have testing available for everyone with even the most mild of symptoms.
Most important, healthcare providers want you to know what it means if you're getting tested for the new coronavirus. If you do not have any symptoms at all, you DESERVE to know that those results may be misleading, as it is different than how testing usually works.
On to positive and negative predictive value another time.

#medtwitter, @somedocs, share your examples of how a negative doesn't mean "clear" or "risk free" when asymptomatic.
You can follow @AndreaFeller2.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: