Stephen J.LurieMD, PhD, Contributing EditorIndividualAuthor
In Reply: The different NNTs in the 4 trials
we described were largely related to baseline risks in the respective control
groups. We therefore stressed the strategic importance of comparing trials
carried out under different circumstances and not just with different drugs.
However, the influence of possible differences due to the intrinsic efficacy
of each statin or the dosages used cannot be ruled out.
We agree with Dr Lam that NNTs refer to additional benefits beyond those
associated with "control treatment" and thank him for making this comparison
more explicit. Regarding derivation of the average NNT per year, we stated
that it was extrapolated by assuming that the effects of statins were evenly
distributed over the period of treatment for which the NNT itself was calculated.
The product of the corresponding period and the NNT therefore gave a value
for the average NNT per year.
Fontanarosa PB, Kumana CR, Cheung BMY, Lauder IJ. Assessing Prevention Interventions by "Number Needed to Treat"—Reply. JAMA. 2000;284(3):303-305. doi:10.1001/jama.284.3.303