ABSTRACT
This chapter discusses the generalizability of clinical results by evaluating the sensitivity index under different models that shift in location parameter is random, shift in scale parameter is random, and shifts in both location and scale parameters are random. It illustrates the issue of extrapolation in biosimilar product development. The chapter introduces the concept of a sensitivity index for measuring the degree of population shift. It outlines several scenarios for assessment of sensitivity index. The chapter provides statistical inferences of the effect size for the three possible scenarios. It includes is an example concerning an asthma clinical trial. A placebo-control clinical trial was conducted to evaluate the efficacy of an investigational drug product for treatment of patients with asthma. Food and Drug Administration indicated that the extrapolation across indication may be determined based on data supporting a demonstration of biosimilarity, including data from clinical studies performed for another condition of use.
