Statistical Analyses and Issues in the Testing of Means and Proportions in Clinical Endpoint Studies for Evaluation of Generic Products
*Pina D'Angelo, Novum Pharmaceutical Research Services 

Keywords: ANOVA, superiority testing, pooling of treatments

This session will provide an overview of the current FDA recommendations for generic submission of clinical endpoint studies for which the primary objective involves testing of means (continuous endpoint) or proportions (dichotomous endpoint) for assessment of superiority of active to placebo products and equivalence of test and reference products. Statistical issues pertinent to the analyses of these studies will be examined. Examples include 1. Method of analysis for superiority: should data for superiority be analyzed in one ANOVA model (continuous endpoint) with pairwise comparisons for each of the active treatments to placebo or two separate ANOVA models for each comparison? 2. Pooling of active treatments: should active treatments be pooled for comparison to placebo as a method to evaluate superiority in order to demonstrate that the study is sufficiently sensitive to identify a clinical effect? These questions will be addressed and findings from statistical analyses will be presented.