I keep seeing posts "concerned" about the lack of testing being done/available. Here's a statistical explanation why that's not a good idea. I originally posted this as a comment but decided it needs to be a top level post.

I want to give some perspective on why rampant testing is, in general, not a good idea. It comes from Bayes Rule in statistics.

Generally, medical tests have something called sensitivity and specificity. Sensitivity is the ability of a test to correctly identify that a patient has a condition. It's the probability of a positive test given that the person has the disease p(+|Disease). Specificity is the probability that the test gives a negative result given the person doesn't have the disease p(-|NoDisease).

What we really want to know is if a person who tests positive really has the disease, or p(Disease|+). In order to know that, we need to know the probability of having the disease or p(Disease) in the population we're testing. This is something that's generally not known.

So, let's assume that we have a population that we run the test on that's 100,000. Let's also assume that the population we're testing has a 25% chance of having the disease, or 25,000 people have the disease. For argument's sake, let's say that the sensitivity is 90% and sensitivity is 99%. These numbers are generally pretty good. (I think the nose swab is somewhere around 70-90% for the annual flu).

So, we have 25,000 with the disease. 90% or 22,500 of these people are correctly identified as having the disease while 2500 are false negatives. For the 75,000, who don't have the disease, 99% (74,250) are correctly identified as not having the disease. That leaves 750 false positives. So, the probability of having the disease given that you test positive is 22,500/(22500+750) ~ 97%. That's pretty good.

Now, what if we start testing all willy-nilly with everyone, including the hypochondriacs panicking about the disease. That drops the probability of the people actually having the disease to a much lower number (again, we don't know the actual number). Let's say that the p(Disease) = 1% now. With the same test, the probability of having the disease given a positive test result goes down to 47.6%. Or 52.4% of the people won't actually have the disease. No better than a coin flip.

These are all hypothetical numbers. The above just illustrates why you want to make sure that the population you test has a higher probability of having the disease. That said, I'd imagine that the sensitivity and specificity of the nose swab is lower than 90%/99% for the Wuhan flu because it is a newly developed test, but that's just my speculation. The gold standard for testing flu virus is a virus culture, but this gets cost prohibitive if you have a lot of positive tests.