In a diabetes detection program with test A cutoff 165 mg/dl and test B cutoff 135 mg/dl, which statement is FALSE?

In a diabetes detection program with test A cutoff 165 mg/dl and test B cutoff 135 mg/dl, which statement is FALSE?

• A. The specificity of test A is greater than that of test B
• B. The number of false-negative results is greater with test A than with test B
• C. The number of false-positive results is less with test A than with test B
• D. The sensitivity of test A is less than that of test B

Answer: (C)

When you raise the cutoff for a diagnostic test, you make the test more stringent. That tends to increase specificity (fewer non-diseased people test positive) and decrease sensitivity (more diseased people test negative). It also means fewer false positives and more false negatives. With test A having a higher cutoff (165 mg/dl) than test B (135 mg/dl): - Specificity is higher for the tougher cutoff, so test A has greater specificity than test B. - Sensitivity is lower for the tougher cutoff, so test A has lower sensitivity than test B. - Fewer people without diabetes will test positive with the higher cutoff, so the number of false positives is smaller with the higher cutoff. - More people with diabetes will be missed by the higher cutoff, so false negatives are more numerous with the higher cutoff. Thus, the statement that the number of false-positive results is less with the higher cutoff is true, not false. In this setup, all these relationships hold, so there isn’t a false statement among them; the given key appears to be incorrect.

When you raise the cutoff for a diagnostic test, you make the test more stringent. That tends to increase specificity (fewer non-diseased people test positive) and decrease sensitivity (more diseased people test negative). It also means fewer false positives and more false negatives.

With test A having a higher cutoff (165 mg/dl) than test B (135 mg/dl):

• Specificity is higher for the tougher cutoff, so test A has greater specificity than test B.

• Sensitivity is lower for the tougher cutoff, so test A has lower sensitivity than test B.

• Fewer people without diabetes will test positive with the higher cutoff, so the number of false positives is smaller with the higher cutoff.

• More people with diabetes will be missed by the higher cutoff, so false negatives are more numerous with the higher cutoff.

Thus, the statement that the number of false-positive results is less with the higher cutoff is true, not false. In this setup, all these relationships hold, so there isn’t a false statement among them; the given key appears to be incorrect.