An equivalence test for the comparison between a test drug and placebo in human abuse potential studies

Statistical methodologies for human abuse potential studies are rarely evaluated. Human abuse potential studies assess whether test drugs produce positive and negative subjective responses on abuse-related measures using volunteers with histories of recreational drug use. These studies typically have a randomized, double-blind, placebo- and positive-controlled crossover design with at least four treatments: placebo, at least two doses of a test drug, and at least one dose of a positive control drug (a drug with known abuse potential). This article proposes a new test for assessing whether a test drug produces data indicative of abuse potential when compared to placebo.