Rational Diagnostic Testing

On this page

Diagnostic Foundations - Test Truths

  • Purpose: To reduce diagnostic uncertainty, aiding clinical decisions and patient management.
  • Types of Tests:
    • Screening: Detects potential disease/risk in asymptomatic individuals (e.g., Pap smear).
    • Diagnostic: Confirms/excludes disease in symptomatic individuals (e.g., ECG for chest pain).
    • Monitoring: Tracks disease progression or treatment response (e.g., INR for warfarin).
  • Role of Prevalence (Pre-test Probability):
    • Significantly influences a test's predictive values (PPV, NPV).
    • ↑ Prevalence generally → ↑ Positive Predictive Value (PPV).
    • ↓ Prevalence generally → ↑ Negative Predictive Value (NPV).

⭐ No diagnostic test is 100% accurate; all tests have limitations and potential for error.

Test Performance Metrics - Number Crunch

Evaluate diagnostic tests using metrics derived from a 2x2 contingency table:

Disease +Disease -Total
Test +True Pos (TP)False Pos (FP)TP + FP
Test -False Neg (FN)True Neg (TN)FN + TN
TotalTP + FNFP + TNTP+FP+FN+TN
  • Sensitivity (Sn): Proportion of actual positives correctly identified. $Sn = TP / (TP + FN)$
  • Specificity (Sp): Proportion of actual negatives correctly identified. $Sp = TN / (TN + FP)$
  • Positive Predictive Value (PPV): Likelihood that a positive test means disease is present. $PPV = TP / (TP + FP)$
  • Negative Predictive Value (NPV): Likelihood that a negative test means disease is absent. $NPV = TN / (TN + FN)$
  • Likelihood Ratios (LR): Quantify how much a test result changes the likelihood of disease.
    • LR+ (Positive): $Sensitivity / (1 - Specificity)$. How much a positive test increases disease odds.
    • LR- (Negative): $(1 - Sensitivity) / Specificity$. How much a negative test decreases disease odds.

📌 SNOUT: Highly Sensitive test, when Negative, rules OUT disease. 📌 SPIN: Highly Specific test, when Positive, rules IN disease.

⭐ Sensitivity and Specificity are intrinsic test characteristics, unaffected by disease prevalence. PPV and NPV, however, are prevalence-dependent.

Strategic Test Selection - Smart Choices

Effective diagnosis hinges on smart test choices, guided by pre-test probability (PTP) and test characteristics.

  • Bayes' Theorem in Practice:

    • Estimate PTP (clinical findings).
    • Use Likelihood Ratios (LRs) for Post-test Probability (PoTP).
    • Formula: $Post-test\ odds = Pre-test\ odds \times LR$.
      • Pre-test odds = $PTP / (1 - PTP)$.
      • PoTP = $Post-test\ odds / (1 + Post-test\ odds)$.
    • LRs >10 (rule-in) or <0.1 (rule-out) are impactful.
    • Fagan's nomogram visualizes this. Fagan's Nomogram for Diagnostic Test Interpretation
  • Test Selection Factors (📌 "AC AIP"):

    • Accuracy (Sens, Spec, LRs)
    • Cost
    • Availability
    • Invasiveness
    • Patient values

⭐ Positive Predictive Value (PPV) is highly dependent on the prevalence of the disease in the population being tested.

Diagnostic Pitfalls & Screening - Testing Traps

  • Common Diagnostic Biases:
    • Confirmation Bias: Seeking data supporting initial hypothesis, ignoring contradictory evidence.
    • Availability Bias: Over-relying on easily recalled (recent/dramatic) diagnoses.
  • Testing Consequences:
    • Overtesting: ↑ false positives, patient anxiety, unnecessary costs, iatrogenic harm.
    • Undertesting: Can lead to missed/delayed diagnosis, ↑ morbidity & mortality.
  • Principles of Screening:
    • Early detection in asymptomatic individuals for significant health problems.
    • Test must be accurate, acceptable, safe. Effective treatment must exist.

    ⭐ The Wilson-Jungner criteria provide a framework for evaluating the appropriateness of a disease screening program.

High‑Yield Points - ⚡ Biggest Takeaways

  • Always estimate pre-test probability before ordering any diagnostic test.
  • Likelihood Ratios (LRs) are best for assessing how test results change disease probability.
  • Sensitivity and Specificity are fixed test properties, independent of prevalence.
  • Use high Sensitivity tests to rule out disease (SnNOut).
  • Use high Specificity tests to rule in disease (SpPIn).
  • PPV and NPV vary significantly with disease prevalence in the population.
  • Avoid shotgun testing; select tests to confirm or refute specific hypotheses based on clinical reasoning.

Practice Questions: Rational Diagnostic Testing

Test your understanding with these related questions

Calculate the sensitivity of a screening test: True Positives=80, False Negatives=20, True Negatives=90, False Positives=10

1 of 5

Flashcards: Rational Diagnostic Testing

1/5

A patient with treatment-resistant hypocalcemia after massive blood transfusion, also most likely has a co-existing _____.

TAP TO REVEAL ANSWER

A patient with treatment-resistant hypocalcemia after massive blood transfusion, also most likely has a co-existing _____.

hypomagnesemia

browseSpaceflip

Enjoying this lesson?

Get full access to all lessons, practice questions, and more.

Start Your Free Trial