In article <000a01bdeb7f$27ca8680$3b5b95c1@default>, owen dempsey
<[log in to unmask]> writes
>and i was wrong; suprising but true.......knew it didn't sound right
<snip>
>its interesting (to me); but the sensitivity and specificity of a test
>(positivity in disease, and negativity in health) are unaffected by
>prevalence: however the negative predictive value, (the post-test
>probability of being healthy if you have a negative test), actually
>falls if the prevalence rises, so the answer to mary's question is that
>it depends on how high the prevalence is; eg
<snip>
>took me a while to come to grips with this yesterday;
Doesn't this speak volumes about the 'mathematical masturbation' that
Sacket et al have introduced with EBM. I know its useful but I find it
difficult to use in a natural manner. I would think the majority said
nothing because, like me, they skipped the mental gymnastics needed to
follow the argument.
Human nature suggests that the more convoluted it is, the more likely it
won't get used and the more likely its wrong.
Oh for clarity!
Regards
George
--
I'm not in denial.
I'm just selective
about the reality
I choose to accept.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|