DEADLINE open thread. Have a party.

It bothers me. A LOT. Because, at least according to Kelly, only 2 in 10,000 reservoir cases get better. It's a lot higher than, um, 0 in 10,000 but... the fact that they could tell is very, very, very disturbing. It implies an awful lot more indeed.

...You know what gets me about that frequency? It's exactly the frequency of my own chronic illness, Ehlers-Danlos syndrome. So I was possibly less freaked out than I should've been. *g*

I'd say, don't get too freaked out about that. There are often such coincidences in statistical frequency, usually either as artifacts of the information collection mechanism or of the analysis tool, but sometimes because of an underlying truth about the mechanism. An awful lot of recessive traits, for example, seem to occur in 12.5% (+/- 2.5%) of a given population.

Mm. Though the type of EDS that I have is actually autosomal dominant, a 50/50 split if you have an affected parent (or two).

Oh man, I didn't even think of that. How many tens of thousands of people did they need to study to get that accurate?

Probably not nearly as many as you'd think. I'm not familiar with how the "X in Y-thousand" probabilities are reached, but that seems like an extrapolatory figure, not interpolatory - in other words, their sample size was considerably less than 10,000 and they did some mathematical work to figure out what the real number should be.

Of course, It's not impossible that they got that figure by interpolation - plenty of people do Amplify, and the CDC gets blood test results for all of them.