Number Needed To Treat Calculator

Variables used

The number needed to treat method was created by Laupacis et al. in 1988 and is used to measure the relative benefit of an active treatment in the case of a control group, therefore the variables accounted for are:

■ Control group total number;

■ Control group interest cases;

■ Experimental group total number;

■ Experimental group interest cases.

The NNT to aim for in a treatment is 1, where everyone improves with treatment and no one improves with control. The effectiveness of the treatment decreases as the NNT increases.

Algorithm

These are the results returned by the NNT method and the formulas used to get to them:

■ Control event rate - percent of subjects from the control group that presented the study’s outcome of interest.

CER = Control group interest cases/Control group total number  x 100

■ Experimental event rate - percent of subjects from the experimental group that presented the study’s outcome of interest.

EER = Experimental group interest cases/Experimental group total number x 100

■ Absolute risk reduction - absolute value of difference between CER and EER. Evidences the excess risk in relation to the control group action.

ARR = (CER-EER) or ARR= 1/NNT

■ Number needed to treat: the subjects to be given the experimental treatment to prevent the study’s outcome.

NNT = 1/(CER-EER) or NNT = 1/ARR

■ Relative risk reduction: dividing the absolute risk by the control event rate:

RRR = (CER-EER)/CER or RRR = ARR/CER

■ Reported relative risk:

RR = 1-(CER-EER)/CER or RR = 1-RRR

NNT benefits

NNT helps both at an individual case basis (evaluation of the likelihood of benefit of a treatment) and in summarizing the results of therapeutic trials.

It was also found to be a better measure of the relative benefit of an active treatment over a control than other calculations, such as the relative risk, the relative risk reduction (RRR) or the odds ratio.

It is often used in establishing patient decision aids such as: a particular medication reduces the risk of a particular condition.

NNT limitations

Despite their easefulness of application, the interpretation of NNTs can only go up to a certain extent as the measure cannot be used for a meta-analysis because the baseline risk can vary considerably between the trials.

Although they can be used by clinicians to discuss particular benefits of a treatment, the measure itself, the NNT, cannot be used to calculate the risk/benefit ratio for an individual person and requires sample sizes.

Similarly, this is not a measure of degree of benefit but an expression of how many individuals are needed to undergo the treatment, for one to benefit.

Example calculation

By taking a study with the following characteristics:

■ Control group total number: 136;

■ Control group interest cases: 82;

■ Experimental group total number: 97;

■ Experimental group interest cases: 58.

The result from the Number needed to treat calculator is:

■ CER percent (Control event rate) = 82/136 = 0.603 (60.3%);

■ EER percent (Experimental event rate) = 58/97 = 0.598 (59.8%);

■ ARR (Absolute risk reduction) = 0.603-0.598 = 0.005 (0.5%);

■ NNT (Number needed to treat) = 1/0.005 = 200;

■ RRR (Relative risk reduction) = 0.005/0.603 = 0.008 (0.8%);

■ RR (Reported relative risk) = 1-0.008 = 0.992 (99.2%).

Original source

Laupacis A, Sackett DL, Roberts RS. An assessment of clinically useful measures of the consequences of treatment. N Engl J Med. 1988; 318(26):1728-33.

Other reference

Christensen PM, Kristiansen IS. Number-needed-to-treat (NNT)--needs treatment with care. Basic Clin Pharmacol Toxicol. 2006; 99(1):12-6.