Review Question - QID 105635

When the cut-off point of a test is increased, this is associated with an increase in specificity (TN/(TN+FP)) and a decrease in sensitivity (TP/(TP+FN)).

Sensitivity and specificity are inversely related. In the context of a diagnostic test, moving a cut-off point to identify more patients with the disease increases sensitivity; however, this comes at the expense of decreased specificity, with more false positives occurring. Conversely, minimizing false positives leads to increased specificity at the expense of decreased sensitivity (more false negatives). Ultimately a cut-off point should be found where a balance is struck between optimal sensitivity and specificity values within a certain clinical context.

Sharp and Lipsky discuss various tools for screening for depression. The cut-off levels for these screening tools have been optimized to strike the optimal balance between obtaining the highest sensitivity and specificity possible for detecting symptoms suggestive of depression. Such screening tests may be manipulated by clinicians through increasing or decreasing cut-off points (and thereby altering sensitivity and specificity of the test) for diagnosis or the need to pursue further work-up.

Ohle et al. discuss the use of the Alvarado score for predicting acute appendicitis (Illustration B). A cut-off value of 5 is used to rule-out appendicitis as a reason for admission, while a cut-off value of 7 is used to rule-in appendicitis and proceed to surgery. When the cut-off moves from 5 to 7, the sensitivity decreases with a goal towards increased specificity.

Illustration A shows a graph illustrating how two different cut-points for creatine kinase levels in MI patients may elicit different sensitivies and specificities; note that there will be more missed diagnoses (false negatives) with a higher cut-off point, while conversely, there will be more misdiagnoses (false positives) with a lower cut-off point. Illustration B is a diagram summarizing the Alvarado criteria; note how different cut-off points may be used to elicit different sensitivities and specificities from the test. Illustration C summarizes the implications of a test with low sensitivity and high specificity. Illustration D graphically summarizes where TP, TN, FP, and FN fall relative to a cut-off point.

Incorrect Answers:
Answers 1,4,5: Increasing the cut-off point is associated with decreased sensitivity and increased specificity.
Answer 2: Lowering the cut-off point is associated with increased sensitivity and decreased specificity.