RLO: Numbers needed to treat (NNT) and numbers needed to harm (NNH)



In the example we have used the NNT was calculated as a whole value, but this is rarely the case. Suppose, that in our previous example the experimental event rate had been 60%. The NNT would then be 3.3. In all cases the NNT is rounded up, so in this case it would be 4.

In a perfect world an intervention would have an NNT of 1 meaning that everyone would experience the benefit of the drug. Unfortunately, an NNT of 1 is very rare, but the smaller the NNT, the better the intervention.

So what do you think is an acceptable maximum NNT? Use the sliding scale to register your view.

Unfortunately, there are no hard and fast rules as to what is an acceptable value for an NNT. It depends on a number of factors including the severity of the condition being treated, possible side effects and costs, along with individual values and preferences.


RLO Transcript (Rich Text Format 45kb)