Mayer, Dan wrote:
> First of all, hello to Piersante, I hope you had a good autumn and
your winter is not as cold and snowy as mine. I am writing this from Los
Angeles instead of Albany. I am here with five of my grandchildren for a
few weeks of vacation, so pardon my rambling.
It freezes at night somimes, but winter in Tuscany is still milder than
it used to be.
>
> The next level is diagnostic accuracy or diagnostic thinking. In
either of these situations, the question is whether the test performs as
well as the gold standard or will change clinical thinking. The PICO
here becomes; in patients who are suggestive of having the target
disorder (P), will the test (I) have better operating characteristics
(O) than the gold standard (C). This can also be expanded to ask if the
test result will result in different clinical practice. I suppose that
is a different PICO with the same P, I, and C and the O being whether
there would be a change in the therapy for the patients diagnosed with
the test as opposed to the gold standard.
Even in this case, I prefer that my students think "I" as doing the test
and "C" as not doing the test, with "O" as the probability of disease:
In this case, "I" then becomes the post-test probabilities (with a
positive or negative result), and "C" corresponds to the pre-test
probability.
I really insist that pre-test probabilities be part of a clinical
diagnostic question.
The risk with other alternatives is that they concentrate on the test,
forgetting about the patient
cheers,
Piersante Sestini
|