Interpretation
| Result | Meaning |
|---|---|
| LR > 10 | Large increase in probability (often rules in) |
| LR 5–10 | Moderate increase |
| LR 2–5 | Small increase |
| LR 0.5–2 | Little change |
| LR 0.1–0.5 | Moderate decrease |
| LR < 0.1 | Large decrease (often rules out) |
Formula
Pre-test odds = p ÷ (1 − p) · Post-test odds = pre-test odds × LR · Post-test probability = odds ÷ (1 + odds). LR+ = sens ÷ (1 − spec); LR− = (1 − sens) ÷ spec.
Clinical pearls
- Likelihood ratios from one study may not transfer to a population with a different disease spectrum.
- Applying several tests in sequence by multiplying LRs assumes the tests are independent, which is often untrue.
Examples
post-test probability pretest 30% LR 5pretest 20 percent, negative test with sensitivity 90% specificity 80%bayes prior 10% positive test sens 95 spec 90
FAQ
How do you calculate post-test probability?
Convert the pre-test probability to odds, multiply by the likelihood ratio, then convert back: probability = odds ÷ (1 + odds). Pre-test 30% with LR 5 gives 68%.
What is a good likelihood ratio?
LR+ above 10 or LR− below 0.1 produce large, often conclusive changes in probability; values between 0.5 and 2 rarely change management.