Re: The Basis for Many Medical Tests

[Guest guest]

Hi ,

I haven't actually seen the graphs for the thyroid hormones, but to add

to the melting pot, I think that the TSH and T4 (so probably T3) are

skewed. In practice, this means that the lovely bell-curve that you

would hope to get is squished on one side and fat on the other.

For me, the interesting thing is that 5% of the population wouldn't be

expected to fit within the distribution. So given 20 people in the

general population, 1 will not be within range anyway. I don't know if

these naturally " out-of-range " people are more/less likely to have

thyroid problems, but even if there is no difference to the " in-range "

population, I find it amusing to think of a doctor seeing 20 patients

and trying to fit them all within a small part of the distribution...

(well, academically amusing).

If anyone has some stats papers about all of this, I'd be interested to

read them. Please post references if available, on-line is obviously

easiest but I have access to mainstream stats journals (not medical

ones) otherwise.

Cath