General Departmental Seminar Series
Assessment of Individual and Population Bioequivalence: Confidence Intervals for Linear Combinations of Variance Components
Jun Shao, Ph.D., Departments of Statistics, UW-Madison
Friday, March 7, 2003, 12-1 p.m.
6225 Medical Sciences Center (MSC), 1300 University Avenue
A bioequivalence study is a surrogate for a clinical trial in applying for an approval of a generic drug product. In this talk, we will introduce the concepts of individual and population bioequivalence in the U.S. Food and Drug Administration (FDA) 2001 guidance. The main statistical issue involved in assessing individual or population bioequivalence is the construction of a confidence interval for a linear combination of variance components. The method adopted by the FDA 2001 guidance is based on the so-called modified large sample (MLS) method that has the best empirical performance among several other alternative large sample methods. We will show how to apply the MLS method in several bioequivalence studies, including a situation in which the MLS method is not directly applicable so that an extension has to be made. Using Edgeworth expansions, we also show theoretically why the MLS method or its extension works well.
Back to General Departmental Seminar Series