Statistics How To

Number Needed to Harm NNH: Definition

Randomized Clinical Trials > Number Needed to Harm

What is Number Needed to Harm?

number needed to harm

NNH rates are usually calculated relative to another active drug or a placebo (“sugar pill”).

Number Needed to Harm (NNH) is a measure of how many people need to be treated (or exposed to a risk factor) in order for one person to have a particular adverse effect. For example, March et. al showed that the antidepressant sertraline had an NNH of 64 for suicidal thoughts when used in children and adolescents. This means that for every 64 patients given sertraline, one experienced suicidal thoughts. The lower the NNH, the more risk of harm; An NNH of 1 would mean that every patient treated is harmed.

A different NNH is calculated for each specific adverse event. For example, one IBIS study showed that the breast cancer drug tamoxifen caused a thromboembolic event (a blood clot that blocks a blood vessel) in about 1 in 138 women, giving it an NNH of 138 for thromboembolic events. On the other hand, tamoxifen’s NNH for endometrial cancer and cardiovascular events was small enough to be labeled “not significant.”


Number needed to harm is similar to number Number needed to treat (NNT); While NNH is a measure of harm or adverse effects, NNT is a measure of how many patients needed to be treated in order for one to benefit. Together, these statistics help physicians decide on courses of treatment. Lower NNT and higher NNH treatments are preferred over those with higher NNT and lower NNH. However, that doesn’t mean you should always choose a drug based on NNT/NNH statistics alone: the decision to use a specific drug depends on how much of a benefit a patient receives vs. the seriousness of side effects. For example, the recalled drug Vioxx had a very good NNT of 2.2, but an NNH for heart attacks of 333.


Number need to harm is calculated in the same way as number needed to treat: divide 1 by the absolute risk increase.

IBIS investigators. First results from the International Breast Cancer Intervention Study (IBIS-I): a randomised prevention trial. Lancet 2002;360:817–24.
Rabbie, et. al. “Ibuprofen with or without an antiemetic for acute migraine headaches in adults.” Cochrane Database Syst Rev. 2010; (10): CD008039.
Published online 2010 Oct 6. doi: 10.1002/14651858.CD008039.pub2
Retrieved 3/21/2017 from:
“Market Withdrawal of Vioxx: Is it Time to Rethink the Use of COX-2 Inhibitors?”

Comments are now closed for this post. Need help or want to post a correction? Please post a comment on our Facebook page and I'll do my best to help!
Number Needed to Harm NNH: Definition was last modified: October 12th, 2017 by Andale