Medicine & HealthcareEpidemiologyUniversity
AQAIB

Number Needed to Treat

Calculate NNT from absolute risk reduction.

Understand the formulaSee the free derivationOpen the full walkthrough
Open Full WalkthroughTry Calculator

This public page keeps the free explanation visible and leaves premium worked solving, advanced walkthroughs, and saved study tools inside the app.

Core idea

Overview

The Number Needed to Treat (NNT) is an epidemiological measure used to communicate the effectiveness of a medical intervention. It represents the average number of patients who need to receive a treatment to prevent one additional bad outcome compared to a control group.

When to use: Use this formula when evaluating therapeutic efficacy in clinical trials where outcomes are binary, such as recovery or death. It requires a clear definition of the time period over which the events occur and assumes the treatment effect is consistent across the population.

Why it matters: NNT translates abstract risk percentages into a practical figure that helps clinicians weigh the benefits of a drug against its costs or potential side effects. It is a critical tool for evidence-based medicine and facilitates shared decision-making between patients and healthcare providers.

Symbols

Variables

NNT = Number Needed to Treat, CER = Control Event Rate, EER = Experimental Event Rate

NNT
Number Needed to Treat
Variable
CER
Control Event Rate
Variable
EER
Experimental Event Rate
Variable

Walkthrough

Derivation

Understanding Number Needed to Treat (NNT)

NNT is the number of patients who need a treatment for one additional patient to benefit, based on the absolute risk reduction.

  • Event rates are estimated from comparable groups (e.g., a randomised trial).
  • Absolute risk reduction applies over the stated time horizon.
  • Risks are expressed as proportions (0 to 1), not percentages, when taking reciprocals.
1

Define absolute risk reduction (ARR):

ARR is the difference in event risk between control and treatment groups (use the sign that reflects benefit).

2

Define NNT as the reciprocal of ARR:

If ARR = 0.10, then NNT = 10, meaning treat 10 people to prevent one additional event.

Note: NNT is usually rounded *up* to the next whole number and should be reported with the time period and confidence interval.

Result

Source: Oxford Handbook of Clinical Medicine — Evidence & Statistics

Visual intuition

Graph

The graph forms a hyperbola because ARR appears in the denominator of the equation, creating a vertical asymptote at the y-axis where NNT increases toward infinity as ARR approaches zero. For a student of medicine, this shape illustrates that small ARR values result in a very high NNT, meaning a large number of patients must be treated to prevent one additional adverse outcome, whereas larger ARR values correspond to a lower NNT and higher treatment efficiency. The most important feature of this curve is that NNT never reaches zero, which highlights that even with a high absolute risk reduction, there is always a requirement to treat at least one person to achieve a clinical benefit.

Graph type: hyperbolic

Why it behaves this way

Intuition

Imagine two parallel groups of patients, one receiving treatment and one not. The NNT is like counting along the treated group until you find one patient whose outcome was prevented *because of the treatment*, relative

NNT
The average number of patients who must receive a specific medical intervention for one additional patient to benefit (i.e., avoid an adverse outcome) compared to a control group.
A lower NNT indicates a more effective intervention, as fewer patients need treatment to prevent one outcome.
ARR
The absolute difference in the rate of an adverse event between an experimental (treated) group and a control (untreated) group over a specified period.
A larger ARR means the treatment has a greater absolute impact on reducing the risk of the event.
CER
The proportion of individuals in the control group who experience the adverse outcome over a specified period.
Represents the baseline risk of the adverse outcome without the intervention.
EER
The proportion of individuals in the experimental (treated) group who experience the adverse outcome over a specified period.
Represents the risk of the adverse outcome with the intervention.

Signs and relationships

  • 1/ARR: The reciprocal relationship converts a risk difference (ARR) into a count of individuals (NNT). A smaller ARR (less effective treatment) means a larger NNT (more patients needed for one benefit), and vice-versa.
  • CER - EER: This difference quantifies the absolute benefit of the intervention. A positive value (CER > EER) indicates the treatment reduces the event rate, making NNT positive and meaningful for 'number needed to treat'.

Free study cues

Insight

Canonical usage

The Number Needed to Treat (NNT) is a dimensionless quantity representing a count of individuals, typically reported as a whole number.

Common confusion

A common mistake is to use percentages directly for CER and EER in the formula instead of converting them to decimal proportions, leading to incorrect NNT values.

Dimension note

All components of this equation (CER, EER, ARR, and NNT itself) are dimensionless quantities. CER and EER represent event rates as proportions or probabilities, ARR is the difference between these proportions, and NNT is

Unit systems

NNTpatients · Represents the number of patients who need to receive a treatment to prevent one additional adverse outcome. Often rounded up to the nearest whole number.
ARRproportion · Absolute Risk Reduction must be expressed as a decimal proportion (e.g., 0.05 for 5%), not a percentage, when used in the formula.
CERproportion · Control Event Rate must be expressed as a decimal proportion (e.g., 0.10 for 10%), not a percentage.
EERproportion · Experimental Event Rate must be expressed as a decimal proportion (e.g., 0.05 for 5%), not a percentage.

One free problem

Practice Problem

In a clinical trial for a new cholesterol medication, the event rate for heart attacks in the control group was 0.12 (12%), while the experimental group had an event rate of 0.08 (8%). Calculate the Number Needed to Treat (NNT).

Control Event Rate0.12
Experimental Event Rate0.08

Solve for: NNT

Hint: First calculate the Absolute Risk Reduction (ARR) by subtracting the experimental rate from the control rate.

The full worked solution stays in the interactive walkthrough.

Where it shows up

Real-World Context

In a clinical or health-monitoring context involving Number Needed to Treat, Number Needed to Treat is used to calculate NNT from Control Event Rate and Experimental Event Rate. The result matters because it helps estimate likelihood and make a risk or decision statement rather than treating the number as certainty.

Study smarter

Tips

  • Always round NNT up to the nearest whole integer for clinical safety and interpretation.
  • A lower NNT value indicates a more effective treatment, with 1 being the theoretical ideal.
  • Ensure the ARR is calculated as a decimal (e.g., 5% is 0.05) before taking the reciprocal.

Avoid these traps

Common Mistakes

  • Using relative risk instead of ARR.
  • Ignoring confidence intervals.

Common questions

Frequently Asked Questions

NNT is the number of patients who need a treatment for one additional patient to benefit, based on the absolute risk reduction.

Use this formula when evaluating therapeutic efficacy in clinical trials where outcomes are binary, such as recovery or death. It requires a clear definition of the time period over which the events occur and assumes the treatment effect is consistent across the population.

NNT translates abstract risk percentages into a practical figure that helps clinicians weigh the benefits of a drug against its costs or potential side effects. It is a critical tool for evidence-based medicine and facilitates shared decision-making between patients and healthcare providers.

Using relative risk instead of ARR. Ignoring confidence intervals.

In a clinical or health-monitoring context involving Number Needed to Treat, Number Needed to Treat is used to calculate NNT from Control Event Rate and Experimental Event Rate. The result matters because it helps estimate likelihood and make a risk or decision statement rather than treating the number as certainty.

Always round NNT up to the nearest whole integer for clinical safety and interpretation. A lower NNT value indicates a more effective treatment, with 1 being the theoretical ideal. Ensure the ARR is calculated as a decimal (e.g., 5% is 0.05) before taking the reciprocal.

References

Sources

  1. Wikipedia: Number Needed to Treat
  2. Gordis, L. Epidemiology. 5th ed. Elsevier Saunders; 2014.
  3. Sackett, D. L., et al. Clinical Epidemiology: A Basic Science for Clinical Medicine. 2nd ed. Little, Brown and Company; 1991.
  4. Gordis, L. (2014). Epidemiology (5th ed.). Elsevier Saunders.
  5. Fletcher, R. H., Fletcher, S. W., & Fletcher, G. S. (2014). Clinical Epidemiology: The Essentials (5th ed.).
  6. Straus, S. E., Glasziou, P., Sacket, D. L., & Haynes, R. B. (2019). Evidence-based medicine: How to practice and teach EBM (5th ed.).
  7. Oxford Handbook of Clinical Medicine — Evidence & Statistics