Abstract
Despite substantial agreement with points made by Andrew C. Leon, PhD, in his article, I am not in complete agreement in a few areas. The definition of noninferiority proposed by Leon allows drugs somewhat less effective than placebo to be characterized as noninferior to placebo, and 2 active drugs may each be simultaneously noninferior to the other. Moreover, including a placebo arm in comparing 2 active drugs is of no use in deciding whether the study is well designed or not, since a significant difference between one of the active arms and the placebo may be due to chance or to a bias in the design. An alternative view of the situation is presented.
Users
Please
log in to take part in the discussion (add own reviews or comments).