 |
As we raise the cut-off level for a positive test, we increase
the proportion of normal individuals who correctly test negative,
thus increasing specificity. On the other hand, we decrease the
proportion of diabetics who correctly test positive, thus
decreasing sensitivity. Since many cancer screening tests,
like a mammogram or a Pap smear, require the interpretation of
a physician or cytotechnologist as opposed to giving an automated
result, the concept of the value of an easily defined cut-point
affecting sensitivity and specificity of the test is often not
applicable.
|
