### Jump to comment:

- Response to Dr. MoranShow More
I believe thst clarification is needed regarding the statement that the NNT "is calculated by taking the reciprocal of the absolute risk reduction between 2 treatment options"? I believe that is consistent with the formula I presented in the table footnote of the paper. If you are calculating the NNT from ARR expressed as percentage, then it's 100/ARR. I take it from Dr. Moran's letter that he/she interpreted the writt...

Competing Interests: None declared. - Number needed to treat correction.I enjoyed reading Dr. McAlister's article about the usefulness and limitations of the number needed to treat. I would like to point out that the formula presented to calculate the number needed to treat in Table 1 (100/absolute risk reduction) disagrees with what it was mentioned in the text (1/ARR) and would lead to confusion and erroneous results. The absolute risk reductions presented in the table are incorrect: the absolute...Show MoreCompeting Interests: None declared.
- NNT misses the mark on baseline risksShow More
I very much agree with the points brought up by Finlay McAllister, however I believe there is one missing issue around the use of NNT and in fact for all the evidence-based terms (relative risk reduction and absolute reduction) we are so used to using.

Consider the 2 following scenarios

Scenario 1) We have a drug that reduces the chance of dying from an MI from 3 down to 2% - say over 2-3 years - th...

Competing Interests: None declared.