Number Needed to Treat (NNT) is the average number of patients who must receive a treatment for one additional patient to benefit compared with the control. An NNT of 10 means you treat 10 patients to prevent one adverse outcome.

Number Needed to Harm (NNH) is the average number of patients who receive a treatment before one additional patient experiences a harmful event. It uses the same formula as NNT but applied to adverse event rates.

What is a good NNT value?

Lower NNT values indicate more effective treatments. An NNT of 1 means every treated patient benefits — the theoretical optimum. NNTs below 10 are generally considered clinically meaningful, though context (disease severity, side effects) matters greatly.

How is ARR different from RRR?

Absolute Risk Reduction (ARR) is the simple difference between control and treatment event rates. Relative Risk Reduction (RRR) expresses ARR as a proportion of the control rate. ARR is more informative for clinical decisions; RRR can look impressive even when baseline risk is very low.

How are the confidence intervals calculated?

The 95% CI for ARR uses the standard error of the difference between two proportions. The CI for NNT is then derived as 1/CI_upper to 1/CI_lower. This method assumes large enough sample sizes for the normal approximation to hold.

What if my event rates are identical?

If control and treatment rates are equal, ARR is 0, meaning the treatment has no effect. NNT is mathematically undefined (infinity) and the calculator will display this clearly.

Number Needed to Treat

Name: Number Needed to Treat
Availability: InStock
Author: Nham Vu

Enter control and experimental event rates to instantly compute NNT, NNH, ARR, RRR, and 95% confidence intervals.

Event Rates

Analysis type

Lower event rate in treatment group = beneficial.

Control group event rate (0 – 1 or 0 – 100%)

CER

Treatment group event rate (0 – 1 or 0 – 100%)

EER

Optional — for 95% confidence intervals

Control n

Treatment n

Quick examples

Enter event rates and click Calculate

Copied!

Summary

Enter control and experimental event rates to instantly compute NNT, NNH, ARR, RRR, and 95% confidence intervals.

How it works

Enter the event rate in the control group (e.g., 0.20 for 20%).
Enter the event rate in the treatment group.
Select whether a lower event rate is beneficial (treatment) or harmful (harm).
The calculator computes ARR, RRR, NNT/NNH, and 95% confidence intervals.
Review the interpretation panel for a plain-English summary of results.

Use cases

Interpret randomized controlled trial (RCT) results.
Compare efficacy across multiple treatments.
Communicate clinical benefit to patients in intuitive terms.
Assess whether a drug benefit outweighs its harm profile.
Calculate NNH from adverse-event data in safety analyses.
Support evidence-based medicine and systematic review work.

Frequently Asked Questions

Last updated: 2026-06-11 · Reviewed by Nham Vu