Comparison of Different Analyses of Data from Sequential Designs with Applications in Drug Interaction Studies

• Activity Number: 305
• Type: Contributed
• Date/Time: Wednesday, August 11, 2004 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Yonghua Wang*+ and Robert A. Smith and Jingdong Xie
• Companies: Bristol-Myers Squibb Company and Bristol-Myers Squibb Company and Bristol-Myers Squibb Company
• Keywords: sequential design ; drug interaction ; mixed-effect model ; standard deviation

Abstract

The issues on how to analyze data from sequential designs arise from drug interaction studies. Assuming that Drug X has interaction with Drug A in the form of reducing its bioavailability, the interest is to find a way of co-administering both drugs without sacrificing the bioavailability of Drug A. This can be accomplished by increasing the dose level of Drug A, or adding pro-drug for Drug A, etc. One design can be as follows: all subjects will take Drug A alone (Treatment A) for the first period, then these subjects will be randomized into different co-administration treatments (Treatments B, C, D, etc.). Bioavailability of Drug A from Treatments B, C, D, etc. will be compared to that of Treatment A. Our research is mainly focused on the comparison of the following analyses: (1) mixed effect model only including subjects who have data from both treatments (Drug A alone and one of co-administration treatments) in a comparison; (2) mixed effect model with sequential effect (sequences defined as AB, AC, etc.) in the model, including all subjects; and (3) mixed effect model without sequential effect in the model, including all subjects.