You're not quite thinking in the right direction. No matter the statistical model a 1/14 result, is absolute evidence against there being no infections at all.
The problem lies with the test itself, which might have an unknown false positive rate. Although in this case we're basically looking at what I understand to be the gold standard in RNA/DNA evidence, combined with matching symptoms.
Also apparently they had 2 separate teams testing the samples using different methodology, so we've got at least a decent amount of confidence that something is going on with that sample, although it doesn't rule out systematic bias.
Yes. It's possible (I can't put a number on how likely) that the sample was mishandled and contamination introduced. Depending at what stage this occurred, this could affect all future tests of the sample. Ideally they are able to follow up with serological tests against the specific patient. While a positive carries no information (might have gotten COVID in the intervening time), a negative would point to some sort of sample contamination.
This isn't a knock against their skill or their competence. Sometimes shit just happens. Certainly if this is true, I'd be very interested in how this changes the nature of our models of the virus.
If it is possible to sequence the sample they would be able to confirm that it came from a very early branch of the genetic tree, rather than one introduced at a much later stage.
Yeah, but for IRB reasons, you may not be able to contact the patient. IMHO I would prefer this (potentially) imperfect short communication being out in the scientific world now than a more perfect communication requiring several more months.
If this turns out to be a false positive from contamination, how many man-month do you estimate will be wasted on making models that don't describe reality before the error is caught?
The problem lies with the test itself, which might have an unknown false positive rate. Although in this case we're basically looking at what I understand to be the gold standard in RNA/DNA evidence, combined with matching symptoms.
Also apparently they had 2 separate teams testing the samples using different methodology, so we've got at least a decent amount of confidence that something is going on with that sample, although it doesn't rule out systematic bias.