(Apologies if this question has been answered before.)
Does anyone know of a sample size/power calculation for the following
situations:
Situation #1:
-comparison of a new diagnostic test to a reference test
-both tests performed on each patient in a group of patients
-want to detect a 5 or 10% difference in sensitivity between the 2 tests
Situation #2:
-comparison of two diagnostic tests, each compared to an independent
reference standard
-again both tests performed in same group of patients
-want to detect a 5-10% difference in sensitivity or specificity between the
2 tests
This is an attempt to sytematically review and compare the results of
studies that report
comparisons of diagnostic operating characteristics for a specific
situation, but report little more than the sensitivities, specificities for
each test (without confidence intervals in most cases).
Thanks!
Margaret Piper, PhD, MPH
Technology Evaluation Center
BCBSA
Chicago, IL, USA
|