Although the performance of medical diagnostic tests is usually
described in terms of their sensitivity and specificity, these accuracy
indices by themselves are of little help to clinicians who need to make
decisions about patient care. Estimating the probability that a patient
with a positive or negative test has or lacks a disorder requires the
clinician to include the pretest probability of the disorder in the
interpretation of the test result. This article described several
approaches to the problem of interpreting diagnostic data in actual
clinical contexts.
Abstract Teaser