Predictive value of tests is the
probability of a target condition given by the result of a test,[1] often in regard to
medical tests.
In cases where
binary classification can be applied to the test results, such yes versus no, test target (such as a substance, symptom or sign) being present versus absent, or either a
positive or negative test), then each of the two outcomes has a separate predictive value. For example, for positive or negative test, the predictive values are termed
positive predictive value or
negative predictive value, respectively.
In cases where the test result is of a continuous value, the predictive value generally changes continuously along with the value. For example, for a
pregnancy test that displays the urine concentration of
hCG, the predictive value increases with increasing hCG value.
A
conversion of continuous values into binary values can be performed, such as designating a pregnancy test as "positive" above a certain
cutoff value, but this confers a loss of information and generally results in less accurate predictive values.