The number needed to treat (NNT) puts the benefit of a treatment into a number anyone can picture: how many people must be treated for one extra person to benefit. This calculator takes the event rate in the control and treated groups and gives the NNT, the absolute and relative risk reductions, the relative risk and the odds ratio.
How the calculation works
- Absolute risk reduction (ARR) = control rate − treated rate
- Relative risk reduction (RRR) = ARR ÷ control rate
- Relative risk (RR) = treated rate ÷ control rate
- Odds ratio = [treated ÷ (1 − treated)] ÷ [control ÷ (1 − control)]
- NNT = 1 ÷ ARR, rounded up; a negative ARR means a number needed to harm
Example
In a trial, 20 % of the control group and 15 % of the treated group have the event. The ARR is 5 percentage points, the RRR is 25 %, and the relative risk is 0.75. The NNT is 1 ÷ 0.05 = 20: treating 20 people prevents one event, or 50 events prevented per 1,000 treated. The odds ratio is 0.71.
Why it matters
Headlines love relative figures such as “cuts the risk by a quarter”. The NNT shows what that means for a patient, and lets you compare a drug that prevents a heart attack with a vaccine, a screening test or a side effect on the same scale.
Good to know
The NNT applies to the time period and the population of the study. It has no confidence interval here, although published values should have one. Different baseline risks can give a different NNT for the same relative effect. This calculator is for education and does not give medical advice. When you read a study, also look at how long patients were followed, how the outcome was defined and whether the result is statistically significant, because a small trial can produce a large effect by chance.
Related tools
The relative risk calculator adds confidence intervals, the odds calculator converts between odds and probability, and the sensitivity and specificity calculator assesses a diagnostic test.