Description
There are some well known challenges to the design and analysis of active controlled trials, especially when it is impractical or unethical to include a placebo arm. Many common statistical approaches need to be reconsidered in this context: it is not obvious what statistical hypotheses should be tested; a sample size reassessment may impact the type 1 error rate; the analysis populations need to be carefully defined. Recently, choice of the non-inferiority margin has received much attention. This meeting aims to contribute to the dialogue between statisticians from regulatory agencies and industry. The presentations in the morning will focus on regulatory issues, including the FDA s draft guidance for industry, whereas the afternoon will be dedicated to case studies.
