Number Needed to Treat Explainer
The number needed to treat, or NNT, estimates how many people would need a treatment for one additional person to benefit compared with a control over a specified period. It is derived from the absolute risk reduction. If an event falls from 10 percent to 5 percent, the reduction is 5 percentage points, and the NNT is 20.
An NNT is never just a number. It must name the outcome, comparison, population, and timeframe. An NNT of 20 to prevent one event in a year means something different from 20 over ten years, and preventing a mild symptom differs from preventing a stroke.
Lower NNTs generally indicate a larger average benefit, but direct comparisons can be misleading when studies use different patients, follow-up, outcomes, or background care. Confidence intervals may be wide, and the estimate can change with baseline risk. Individual benefit may therefore differ from the trial average.
NNT does not by itself describe side effects, cost, inconvenience, or severity of outcomes. Number needed to harm can summarize an additional adverse event, but it has the same context requirements. Decisions should consider both benefits and harms in absolute terms, plus the person’s preferences.
Subtract the treated event rate from the control event rate to find absolute risk reduction, then take its reciprocal. A 0.05 reduction gives 1 divided by 0.05, or 20.
Rounding conventions and statistical uncertainty matter. Published estimates should ideally include a confidence interval.
The NNT generally falls as follow-up lengthens because more events can occur. It should not be moved to a different timeframe without appropriate analysis.
The outcome must be clinically meaningful. Avoid treating a composite outcome as if every component has equal importance.
Compare it with the treatment burden and the frequency and severity of harms. Ask whether your baseline risk resembles the study population's.
Use NNT as one translation of evidence, not a ranking system that automatically identifies the best treatment.
Explore more "Explainers"
Discover additional explainers across politics, science, business, technology, and other fields. Each explainer breaks down a complex idea into clear, everyday language—helping you better understand how major concepts, systems, and debates shape the world around us.
