ABSTRACT
Abstract The analysis of the efficacy of a drug in a clinical trial amounts to quantifying the expected response of patients to the drug. The analysis of safety data, though, amounts to characterizing the unexpected, or extreme, response of a minority of patients. As such, the statistical methods used to analyse efficacy data are inappropriate to the analysis of safety data. We illustrate how univariate and multivariate approaches to the analysis of clinical trial safety data can identify and characterize potential toxicities.
