Some issues in the design and analysis of equivalence trials

Given the number of approved drugs, it is increasingly the case that the comparison arm for a new drug or combination product is another drug or combination, that is, the trial uses an "active-control." Such active-controlled trials raise issues not seen in placebo-controlled trials. This note reviews and discusses some issues associated with the design and analysis of equivalence trials. Included are discussions of the choice of the equivalence allowance in designing the trial, and the roles of confidence intervals and intent-to-treat analyses.