Abstract:
|
Crossover designs are commonly used in preclinical and clinical PK/PD studies, bioavailability, and bioequivalence studies. The preferred designs for higher order crossover studies should be balanced for the carryover effects, and such preferred designs usually have standard layouts depending on the number of periods and treatments. An automated system is developed for the statistical analysis of balanced 3X3 and 4X4 crossover designs. Visual basic for application (VBA), for Excel, is utilized as front end for graphical user interface (GUI), data I/O and manipulation. SAS, invoked inside VBA, using OLE, serves as an engine for statistical analysis and graphics. SAS/ODS generates results in text format and VBA, then finally reformats and delivers the results to the end users in Excel. This system is highly automated, which requires minimum user intervention to input subject allocation information in GUI. A non-statistician (e.g., a scientist or clinical investigator) can use it at ease with no knowledge on modeling SAS and VBA.
|