Your outcomes:
Control group
Experimental group

NNT Calculator

By Łucja Zaborowska
Last updated: Jun 15, 2020

The NNT calculator is a simple tool that assesses both the number needed to treat, and the absolute risk reduction (ARR, risk difference) of a trial, research or any scientific paper.

Our number needed to treat calculator will also provide you with a detailed explanation of the computed NNT's meaning.

Thanks to the article below, you will also be able to calculate number needed to treat on your own, discover how to calculate the number needed to harm, and the difference between these two concepts.

Read on to find a step-by-step explanation of all the NNT formulas.

How to use the NNT calculator?

  1. The first step required by our NNT calculator is to choose the type of your data.
    • Some research results are given in as percentages - e.g., group A had a 20% complete response to an old treatment, and group B had a 40% response to the new treatment.

    • Some scientists prefer to use exposure data, or patient-years. They would supply you with a number of events in both groups, and the period over which the study happened. - e.g., the follow-up lasted for 5 years. During that period, 5 patients died in group A, and 10 patients died in group B.

💡 The word event can have quite a wide range of meanings - it may be an adverse effect, response to treatment, or the number of deaths, births, or infections.

⌛ To obtain the patient years, multiply the number of patients by the time they were in the study. 5 patients with a follow up of 15 years would give us 75 patient-years.
  1. Enter the data for both the control and experimental groups - the control group is usually the old, well-known way of doing something. In contrast, the experimental group is the novelty that we wish to test.

  2. If you want to see your calculated absolute risk reduction (ARR), click the advanced mode button! 🔘

How to calculate NNT?

Let's say we'd like to calculate the NNT for the health benefits of dark chocolate . 🍫

We conducted a unique study that lasted for 10 years. Our primary outcome was the incidence of strokes among people who eat small amounts of dark chocolate (experimental group), versus those who don't eat any (control group).

Let's assume both the control and experimental group were:

  • Of a similar size;
  • Had the same gender ratios; and
  • The patients were of similar age, ethnicity, and had the same chronic diseases.
  1. Results in percentage The incidence of strokes in the experimental group (chocolate) was 1%, while the rate of strokes in the control group (no chocolate) was 2%. We'll use the following number needed to treat formulas:

ARR = (Control group)−(Experimental group); and

NNT = 1/ARR.

❗ Remember, you need to transform the percentages (2% = 0.02)!

ARR = 0.02 - 0.01 = 0.01; and

NNT = 1/ 0.01 = 100.

Our number needed to treat is equal to 100, which means that out of every 100 people who eat chocolate 1 person will benefit and not have a stroke.

Note, that our NNT is positive - it means that our intervention (eating chocolate) will help avoid a particular event, instead of causing it.

  1. Results in patient-years

We've observed 200 patients for 2 years. During that period, 2 people in the control group and 1 person in the experimental group had strokes. We can calculate our patient years as patient-years = 200 * 2 = 400. We're gonna use the equations presented below:

R₀ = 1 - e(-Control group/Patient-years)

R₁ = 1 - e(-Experimental group/Patient-years)

ARR = R₀ - R₁


💡 e in mathematics is the base of the natural logarithm, has a value of around 2.71828182845904.

R₀ = 0.004988

R₁ = 0.002497

ARR = 0.0024906

NNT = 401.5

This means that for every 401.5 people who eat chocolate, one will avoid a stroke.

Number needed to harm formula

The NNH formula is the same as the number to treat equation: NNT = 1/ARR

So what is the difference?

Number needed to harm describes the amount of side effects, or any kind of harm. NNH calculations look for bad things, while NNT focuses on the positive.

An example: ☢️ We're checking how many people have to be exposed to a radioactive debris in order to develop a given kind of cancer.

💡 The concept of relative risk might also be helpful in this kind of evaluations.

Absolute risk reduction calculation

Absolute risk reduction describes the proportion of patients that benefited from the use of experimental therapy. It's a measure of a patient's gain from a given treatment.

The formulas for ARR are as follows:

ARR = Control event rate - Experimental event rate, ARR = R₀ - R₁

ARR = 1/NNT.

Łucja Zaborowska
People also viewed…

Books vs e-Books

The books vs e-books calculator answers the question: how ecological is your e-book reader? 📚

Framingham risk

Framingham risk calculator allows you to compute risk of a heart attack or coronary death for people aged 30-79


This millionaire calculator will help you determine how long it will take for you to reach a 7-figure saving or any financial goal you have. You can use this calculator even if you are just starting to save or even if you already have savings.