Stephen J.LurieMD, PhD, Senior EditorIndividualAuthor
In Reply: The objective of our article was
to assess the pattern used in reporting results of RCTs. We demonstrated that
there was a strong preference for reporting relative risk reduction (RRR).
I agree with Drs Wu and Kottke that the NNT must be interpreted with caution.
It is inappropriate to use the NNT to compare the results of a study with
patients at one baseline risk to patients with a different risk and to extrapolate
the NNT to time points not considered in the trial.1,2
Nevertheless, the methods used to display data may influence the interpretation
of the study results,3 and exclusive use
of RRR may mislead the reader.4
Nuovo J. Interpreting the Number Needed to Treat—Reply. JAMA. 2002;288(7):830-832. doi:10.1001/jama.288.7.828