Number Needed to Treat (NNT)

NNT is the reciprocal of an absolute beneficial risk difference for a specified outcome, comparator, population, and period. It estimates how many people would receive the intervention per additional beneficial outcome on average under those conditions.

For a binary beneficial outcome, the number needed to treat is the reciprocal of the absolute benefit difference between an intervention and comparator over a specified period. If event risk falls from ten percent to eight percent, the absolute reduction is 0.02 and NNT is fifty. It represents one additional event prevented on average per fifty treated under those conditions.

NNT does not identify which person benefits or show that everyone else receives no benefit. It concerns the specified outcome and comparison, leaving other benefits and harms separate. Baseline risk, follow-up time, and uncertainty change its interpretation; when the estimated difference approaches zero, the reciprocal becomes unstable. A large NNT can still matter for a severe outcome or inexpensive intervention.

When to use it

When translating a supported binary-outcome difference into an incremental count-based measure for a defined comparison.

How it can help

Calculate NNT from absolute differences, include uncertainty, and compare outcome severity, harms, costs, and applicability.