Abstract:
Non-inferiority in Clinical Trials: Biometrical and Regulatory Aspects by Prof. Dr. Dieter Hauschke
Objective:It is scientific consensus that the most convincing way to establish efficacy of a new treatment is done by demonstrating its superiority to placebo in a randomized clinical trial. If, however, treatments with proven efficacy are available for an indication, performance of placebo-controlled trials may be unethical. Hence, non-inferiority trials have become popular. In such a clinical trial, it may be sufficient to confirm that the new treatment is not inferior to a reference treatment by more than a pre-specified, clinically irrelevant amount. Placebo-controlled and non-inferiority trials differ fundamentally with respect to methodological issues. These difference will be discussed from a statistical and regulatory point of view.

Abstract: Non-inferiority with Meta-Analyses by Dr. Stefan Witte
Objective:Investigating non-inferiority can be an objective of meta-analyses like in single clinical trials. Some special topics have to be considered. The proposed methods have their origin in the framework of non-inferiority trials or meta-analyses. Two issues are highlighted in the paper: The choice of delta for the meta-analysis as well as how to deal with different analysis sets, such as the full-analysis set with the intention-to-treat (ITT) principle or the per-protocol analysis set. Sensitivity analyses, meta-regression, and graphical presentations may help to analyze the data. The confidence interval approach will be discussed with its limitations in the light of non-inferiority with meta-analyses.