NNH

NNH stands for number needed to harm, a statistic used in evidence-based medicine and pharmacovigilance to express the risk of an adverse effect from a treatment. It represents the number of patients who must be treated for one additional patient to experience a harmful outcome attributable to the treatment, compared with a control.

Calculation and interpretation: NNH is computed as the reciprocal of the absolute risk difference for the adverse event between treated and control groups. That is, NNH = 1 / (risk_harm_treated − risk_harm_control). A higher NNH indicates a lower risk of harm, while a lower NNH indicates a greater risk. NNH is closely related to the number needed to treat (NNT), which applies to beneficial outcomes; together they summarize trade-offs in a trial.

Example: If 6% of patients receiving a drug experience a specific adverse event and 2% in the control group do, the risk difference is 0.04, giving an NNH of 25. This would mean that about 25 patients need to be treated for one additional person to experience the adverse event, under the study conditions.

Limitations: NNH depends on the selected adverse event and the duration of follow-up, and can be unstable when events are rare. It does not capture the severity of harm, and comparisons of NNH across different harms or populations should be done with caution. Confidence intervals for NNH are typically derived from those for the risk difference.

See also: number needed to treat (NNT), absolute risk reduction, risk difference.