I think you were posting in response to Jay Pershad but I'll chime in
anyway! I agree with you that point prevalence does impact on posterior
probability and not sensitivity, but there are conditions in the
pre-test state that may alter the sensitivity of the diagnostic test.
This concept, called spectrum bias by some methodologists, observes that
diagnostic tests perform differently at different stages of a disease
process. One of the best examples in the ER is the WBC in acute
appendicitis. Early in the disease (when we need the most help!), the
sensitivity of the WBC is poor, whereas later it's sensitivity improves.
CK-MB in rule-out MI is another example. It is likely that the U/A
suffers from spectrum bias as well, and that it's sensitivity on day one
is weaker on day one than day two.
Jay Fisher MD
Las Vegas NV
For more information, send mail to [log in to unmask] with the message: info PED-EM-L
The URL for the PED-EM-L Web Page is: