Examkin

Question

Difficulty: MediumDiagnostic and Screening Test Performance

A public health task force is reviewing how changes in diagnostic testing protocols and population demographics affect statistical performance metrics. Match each strategic modification (left) with its corresponding expected effect on the test's performance parameters (right). Assume the physiological spectrum of the disease and the test's mechanics remain identical.

  • The numeric cutoff criterion for a positive test result is made less strict (e.g., lowering the biomarker threshold).Sensitivity increases, specificity decreases, positive predictive value (PPV) decreases, and negative predictive value (NPV) increases.
  • The numeric cutoff criterion for a positive test result is made more strict (e.g., raising the biomarker threshold).Sensitivity decreases, specificity increases, positive predictive value (PPV) increases, and negative predictive value (NPV) decreases.
  • The screening protocol is implemented in a patient cohort with a significantly higher prevalence of the target disease.Sensitivity and specificity remain unchanged; positive predictive value (PPV) increases, and negative predictive value (NPV) decreases.
  • The screening protocol is implemented in a patient cohort with a significantly lower prevalence of the target disease.Sensitivity and specificity remain unchanged; positive predictive value (PPV) decreases, and negative predictive value (NPV) increases.

Answer

Less strict cutoffs match with increased sensitivity/NPV and decreased specificity/PPV. Stricter cutoffs match with increased specificity/PPV and decreased sensitivity/NPV. Higher prevalence matches with increased PPV and decreased NPV (no change to intrinsic parameters). Lower prevalence matches with decreased PPV and increased NPV (no change to intrinsic parameters).
Test performance metrics are predictably altered by strategic modifications. Adjusting a numeric cutoff trades sensitivity for specificity (or vice versa), which cascades to alter predictive values based on the resulting false positive and false negative rates. Conversely, applying the same test to populations with varying disease prevalences affects only the predictive values (PPV and NPV) because the intrinsic accuracy (sensitivity and specificity) of the assay is biologically fixed.

Step-by-Step Solution

1
Analyze the impact of modifying the diagnostic cutoff threshold on intrinsic test parameters.
Lowering the cutoff (less strict) increases sensitivity and decreases specificity. Raising the cutoff (more strict) decreases sensitivity and increases specificity.
Changing the cutoff shifts the decision boundary, inversely affecting the true positive rate and false positive rate.
2
Determine the downstream impact of these cutoff modifications on predictive values.
A less strict cutoff increases false positives, which lowers PPV, and decreases false negatives, which raises NPV. A more strict cutoff decreases false positives, raising PPV, and increases false negatives, lowering NPV.
Predictive values depend mathematically on the proportions of true and false results produced by the altered sensitivity and specificity.
3
Analyze the impact of altering disease prevalence on test parameters and predictive values.
Sensitivity and specificity remain constant. However, higher prevalence increases PPV and decreases NPV, while lower prevalence decreases PPV and increases NPV.
Prevalence changes the pre-test probability of disease, which alters predictive values via Bayes' theorem, but does not change the test's physiological ability to detect the disease.

Key Concept

Shifting test cutoffs inversely alters sensitivity and specificity, subsequently affecting predictive values. Altering population prevalence changes predictive values (PPV and NPV) but does not affect intrinsic test parameters (sensitivity and specificity).
Rate this question

Topics

Diagnostic and Screening Test Performance

More questions from this topic