About the number needed to treat (NNT)
The number needed to treat is a standard evidence-based-medicine statistic that translates a clinical trial's results into a practical count: how many patients, on average, need to receive a treatment instead of the control for one additional patient to avoid the outcome being measured. It was introduced by Laupacis, Sackett, and Roberts in 1988 and is now a routine way to report the benefit of a therapy alongside its relative-risk figures.
How the calculation works
NNT is built from two event rates measured over the same follow-up period:
- Control event rate (CER): the number of control-group patients who had the outcome, divided by the total number of control-group patients.
- Experimental event rate (EER): the number of treatment-group patients who had the outcome, divided by the total number of treatment-group patients.
- Absolute risk reduction (ARR): CER minus EER — the plain difference in event rates between the two groups.
- NNT: 1 divided by ARR, rounded up to the next whole number, since you cannot treat a fractional patient.
Two related figures usually accompany NNT: relative risk reduction (RRR), which is ARR expressed as a percentage of the control rate, and relative risk (RR), which is EER divided by CER. If the treatment group's event rate turns out higher than the control group's — meaning the treatment is associated with more of the outcome, not less — the absolute risk reduction is negative, and the same 1-divided-by-ARR arithmetic produces the number needed to harm (NNH) instead.
Interpreting the result
A smaller NNT means a smaller number of patients need to be treated for one to benefit, which generally signals a larger treatment effect for that outcome and time period. As a widely used rule of thumb, an NNT in the single digits is often considered a large effect, while NNT values in the tens or higher indicate a smaller absolute benefit — even when the relative risk reduction looks impressive. NNT should always be reported together with the time frame and outcome it refers to, since the same drug can have a very different NNT for different endpoints or durations.
When to consult a professional
This tool performs the standard NNT arithmetic from the counts you provide. It does not evaluate trial quality, statistical significance, or confidence intervals, and it is not a substitute for reading the source study or for professional medical or biostatistical advice when the result informs a treatment decision.