likelihood ratio calculator

What is a likelihood ratio?

A likelihood ratio tells you how much a test result changes the odds of disease. In plain language, it answers this question: “Given this test result, should I be more or less convinced that the condition is present?”

Unlike sensitivity and specificity alone, likelihood ratios are designed to work directly with pre-test probability using Bayes’ theorem. That makes them highly practical for real decision-making in medicine, screening, and risk assessment.

The two key numbers: LR+ and LR−

Positive likelihood ratio (LR+)

LR+ describes how much more likely a positive test is in someone with disease compared with someone without disease.

• Formula: LR+ = Sensitivity / (1 − Specificity)
• Higher LR+ values are better for ruling in disease.

Negative likelihood ratio (LR−)

LR− describes how much more likely a negative test is in someone with disease compared with someone without disease.

• Formula: LR− = (1 − Sensitivity) / Specificity
• Lower LR− values are better for ruling out disease.

How this calculator works

The calculator first converts your sensitivity and specificity percentages into decimal form, computes LR+ and LR−, and then (if pre-test probability is provided) performs these steps:

• Convert pre-test probability to pre-test odds.
• Multiply odds by LR+ (for a positive result) or LR− (for a negative result).
• Convert post-test odds back to post-test probability.

This is mathematically equivalent to a Fagan nomogram, but done numerically for precision.

Interpreting likelihood ratio values

Metric	Typical range	Interpretation
LR+	> 10	Large, often clinically important increase in probability (good for ruling in)
LR+	5 to 10	Moderate increase in probability
LR+	2 to 5	Small but sometimes meaningful increase
LR+	1 to 2	Minimal increase
LR−	< 0.1	Large decrease in probability (good for ruling out)
LR−	0.1 to 0.2	Moderate decrease
LR−	0.2 to 0.5	Small decrease
LR−	0.5 to 1	Minimal decrease

Worked example

Suppose a diagnostic test has:

• Sensitivity: 90%
• Specificity: 80%
• Pre-test probability: 25%

Then:

• LR+ = 0.90 / (1 − 0.80) = 4.5
• LR− = (1 − 0.90) / 0.80 = 0.125

Interpretation: a positive test meaningfully increases probability, and a negative test strongly lowers it. This is usually a useful clinical test in both directions, especially for triage and follow-up strategy.

Common pitfalls

1) Ignoring pre-test probability

The same LR can lead to very different post-test probabilities depending on baseline risk. A strong LR+ may still not push probability high enough if pre-test risk was very low.

2) Confusing probability and odds

Bayesian updating uses odds internally. This calculator handles that conversion for you to avoid arithmetic mistakes.

3) Over-trusting point estimates

Real test performance varies by population, spectrum of disease, and quality of study design. Sensitivity/specificity from one paper may not transfer perfectly to your setting.

FAQ

Can LR+ be infinite?

Yes. If specificity is exactly 100%, the false-positive rate is zero, so LR+ approaches infinity. In practice, this indicates very strong rule-in power, but verify sample size and confidence intervals.

Can LR− be zero?

Yes. If sensitivity is exactly 100%, false negatives are zero, and LR− becomes zero. That suggests a negative result is extremely convincing for ruling out disease.

Is this only for medical testing?

No. The same framework applies anywhere you update beliefs from evidence: fraud detection, quality control, forensics, and machine learning threshold analysis.

Bottom line

If you want a clear and quantitative way to move from test characteristics to real-world probability changes, likelihood ratios are one of the best tools available. Use the calculator above to quickly estimate LR+, LR−, and post-test probabilities from your own assumptions.