Skip to main content
v2026.1714 entries · CC-BY 4.0
CASRAI

Clinical research & EBM · Reference

What is the number needed to treat?

The number needed to treat is a measure of an intervention’s effect, defined as the number of people who would need to receive it for one additional person to benefit. It is the reciprocal of the absolute risk reduction and helps express effect sizes intuitively.

How NNT is defined

The number needed to treat is the reciprocal of the absolute risk reduction (ARR), the difference in the rate of an outcome between the control group and the intervention group. If an intervention lowers the outcome rate by an absolute 5 percentage points (an ARR of 0.05), the NNT is 1 ÷ 0.05 = 20: on average 20 people would need to receive it for one additional good outcome. A smaller NNT indicates a larger effect. NNT is presented as a methodological measure of effect size, not as advice about whether any individual should receive an intervention.

Why NNT is useful

NNT is valued because it is more intuitive than relative measures. A relative risk reduction of 50% sounds impressive but says nothing about how common the outcome is; the same relative reduction can correspond to a very large or a very small absolute benefit. By anchoring the effect in absolute terms, the NNT makes the scale of benefit clearer and guards against overstating effects from relative figures alone. It is therefore widely used to communicate the practical size of an effect found in a trial.

Cautions when interpreting NNT

NNT must be read with care. It depends on the baseline risk in the population studied, so an NNT from one trial does not transfer automatically to populations with different underlying risk. It also depends on the time horizon over which the outcome is measured, so an NNT is only meaningful alongside its follow-up period. Like any estimate it carries uncertainty and should be reported with a confidence interval. The mirror-image measure for harms is the number needed to harm (NNH), the reciprocal of the absolute risk increase of an adverse outcome.

Key facts

At a glance

  • Definition: People treated for one additional good outcome
  • Formula: NNT = 1 ÷ absolute risk reduction
  • Smaller NNT: Indicates a larger effect
  • Contrast: More intuitive than relative risk reduction
  • Depends on: Baseline risk and follow-up time
  • Harm version: Number needed to harm (NNH)

Common questions

FAQ

How is the number needed to treat calculated?+

The NNT is the reciprocal of the absolute risk reduction — that is, 1 divided by the difference in outcome rates between the control and intervention groups. For example, an absolute risk reduction of 0.05 gives an NNT of 20.

Why is NNT preferred over relative risk reduction?+

A relative risk reduction does not reveal how common the outcome is, so the same percentage can mean a large or tiny absolute benefit. NNT anchors the effect in absolute terms, which makes the practical size of the benefit clearer and harder to overstate.

Does an NNT from one study apply everywhere?+

No. The NNT depends on the baseline risk of the population and on the follow-up period over which the outcome was measured. An NNT should always be interpreted with its time horizon and its underlying population in mind, and reported with a confidence interval.

The step most authors miss

Doing CRediT right? Don’t stop at the statement.

A CRediT statement credits you inside one paper. The recognition CRediT was built for happens when those roles are tied to you, persistently. Sign in with your ORCID — free — and claim your CRediT contributions on casrai.org, the home of the standard. They become a verified, portable part of your identity, not a line that disappears into one PDF.

Free: claim your contributions, then export a journal-ready CRediT statement, schema.org structured data, JATS XML, CSV or BibTeX — and preview your public profile. A membership publishes that profile publicly and verifies the journals you serve.

Referenced across the research world

University of Cambridge logoColumbia University logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoCrossref logoUniversity of Cambridge logoColumbia University logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoCrossref logo
  • University of Cambridge logo
  • Columbia University logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo
  • Crossref logo

View CASRAI adoption →