Likelihood Ratios in Diagnostics

Jul 2, 2025

Overview

This lecture explains likelihood ratios, how to calculate them, and how to use them to interpret diagnostic test results in clinical practice.

What Are Likelihood Ratios?

  • Likelihood ratios (LRs) synthesize sensitivity and specificity to assess how much a test result changes the probability of disease.
  • LR+ (positive likelihood ratio) indicates how much the odds of disease increase with a positive test.
  • LR- (negative likelihood ratio) shows how much the odds of disease decrease with a negative test.

Calculating Likelihood Ratios

  • LR+ = Sensitivity / (1 - Specificity)
  • LR- = (1 - Sensitivity) / Specificity
  • A good LR+ is greater than 10; a good LR- is less than 0.1.
  • Likelihood ratios close to 1.0 have little impact on confirming or rejecting a diagnosis.

Practical Application Example (ACL Tear)

  • Estimate pretest probability (e.g., for ACL tear, 60% based on symptoms and prevalence).
  • For the Lachman test: sensitivity = 85%, specificity = 94%.
  • LR+ = 14, LR- = 0.16 for Lachman test.
  • Use a nomogram to graphically determine post-test probability by connecting pretest probability and LR.
  • For a positive Lachman, post-test probability ≈ 95%; for a negative, ≈ 18%.

Calculating Post-Test Probability (Mathematical Method)

  • Convert pretest probability to odds: Odds = Probability / (1 - Probability).
  • Multiply pretest odds by LR to get post-test odds.
  • Convert post-test odds to probability: Probability = Odds / (Odds + 1).
  • Positive test: Pretest odds 1.5 × LR+ 14 = 21; probability = 21/22 = 95%.
  • Negative test: Pretest odds 1.5 × LR- 0.16 = 0.24; probability = 0.24/1.24 = 19%.

Using Tests Effectively

  • Use tests with the highest LR+ to confirm a diagnosis.
  • Use tests with the lowest LR- to rule out a diagnosis.

Key Terms & Definitions

  • Sensitivity — Probability a test correctly detects disease when it is present.
  • Specificity — Probability a test correctly identifies no disease when it is absent.
  • Likelihood Ratio (LR) — Ratio summarizing how much a test result changes disease odds.
  • Pretest Probability — Estimated chance of disease before a test result is known.
  • Pretest Odds — Odds calculated from the pretest probability.
  • Post-test Probability — Revised chance of disease after considering the test result.

Action Items / Next Steps

  • Review sensitivity, specificity, and predictive values if not already done.
  • Practice calculating LR+, LR-, and post-test probability for other tests.
  • Optional: Read the blog post on physiotutors.com for further study.