What is post-test probability?
Post-test probability is the updated chance that a patient has a condition after a diagnostic test result is known. It is the heart of evidence-based diagnosis: a test does not simply say "yes" or "no" — it shifts your prior estimate up or down. This calculator applies Bayes' theorem in its convenient odds form, using a likelihood ratio (LR) to update the pre-test probability.
How to use it
Enter the pre-test probability (your estimate of disease before the test, as a percentage) and the likelihood ratio of the test result. Use the positive likelihood ratio (LR+) for a positive result and the negative likelihood ratio (LR−) for a negative result. The tool returns the post-test probability, plus the intermediate pre- and post-test odds.
The formula explained
Because probability and odds describe the same thing, we first convert probability to odds: \(\text{PreOdds} = P / (1 - P)\). Multiplying by the likelihood ratio gives \(\text{PostOdds} = \text{PreOdds} \times \text{LR}\). Finally we convert back to a probability:
$$\text{PostProb} = \frac{\text{PostOdds}}{1 + \text{PostOdds}}$$An LR above 1 raises the probability of disease; an LR below 1 lowers it; an LR of exactly 1 leaves it unchanged.
Worked example
Suppose the pre-test probability is 30% and the test result has an LR of 10. Pre-test odds = \(0.30 / 0.70 = 0.4286\). Post-test odds = \(0.4286 \times 10 = 4.2857\). Post-test probability = \(4.2857 / (1 + 4.2857) = 0.8108\), or about 81%. The positive result has raised the suspicion of disease from 30% to 81%.
Key Terms: Sensitivity, Specificity, Odds and Prior
- Pre-test probability (prior, prevalence): the probability that a person has the condition before the test result is known. It may come from population prevalence, clinical context, or risk scores.
- Post-test probability (posterior): the updated probability after incorporating the test result. After a positive result it equals the positive predictive value (PPV) when the pre-test probability equals the prevalence. After a negative result, the probability of disease is 1 − NPV (negative predictive value).
- Probability vs odds: probability \(p\) is the chance of an event, from 0 to 1. Odds are \(\frac{p}{1-p}\), and converting back gives \(p = \frac{O}{1+O}\). For example, a probability of 0.25 corresponds to odds of 1:3, or 0.333.
- Sensitivity: the proportion of people with the condition who test positive (true positive rate).
- Specificity: the proportion of people without the condition who test negative (true negative rate).
- Positive likelihood ratio (LR+): \(\text{LR+} = \frac{\text{sensitivity}}{1-\text{specificity}}\). It shows how much a positive result raises the odds of disease.
- Negative likelihood ratio (LR−): \(\text{LR−} = \frac{1-\text{sensitivity}}{\text{specificity}}\). It shows how much a negative result lowers the odds of disease.
- Bayes' theorem: the rule for updating a probability with new evidence. In odds form it is simply \(\text{post-test odds} = \text{pre-test odds} \times \text{LR}\).
Unlike PPV and NPV, likelihood ratios do not depend on prevalence. That is why they can be applied to patients with different pre-test probabilities. This tool provides general educational information and is not a substitute for professional medical judgment.
FAQ
Where do likelihood ratios come from? They are derived from a test's sensitivity and specificity: \(\text{LR+} = \text{sensitivity} / (1 - \text{specificity})\), \(\text{LR−} = (1 - \text{sensitivity}) / \text{specificity}\).
What is a "good" LR? As a rule of thumb, an LR+ above 10 or an LR− below 0.1 produces large, often conclusive changes in probability; values near 1 are clinically unhelpful.
Can pre-test probability be 100%? At 0% or 100% the odds become 0 or infinite, and no test can change a certainty — that is why estimates near, but not at, the extremes are used in practice.