THE GOAL OF THIS PHASE I PROJECT IS TO DEVELOP AN INTERACTIVE COMPUTER SYSTEM CAPABLE OF ASSISTING INVESTIGATORS IN THE DESIGN OF STATISTICAL ASPECTS OF CLINICAL ONCOLOGY TRIALS.

THE GOAL OF THIS PHASE I PROJECT IS TO DEVELOP AN INTERACTIVE COMPUTER SYSTEM CAPABLE OF ASSISTING INVESTIGATORS IN THE DESIGN OF STATISTICAL ASPECTS OF CLINICAL ONCOLOGY TRIALS. THE SYSTEM WILL ADVISE THE USER CONCERNING CHOICES OF TYPE I ERROR, POWER, DETECTABLE TREATMENT DIFFERENCES, AND SAMPLE SIZE. WHEN EXACT OR APPROXIMATE FORMULAS ARE NOT AVAILABLE, NUMERICAL RESULTS WILL BE OBTAINED BY COMPUTER SIMULATIONS. THIS EFFORT WILL IDENTIFY: 1. DESIGNS IN CURRENT USE FOR ONCOLOGY TRIALS; 2. THEORY FOR THE SELECTION OF DESIGN PARAMETERS; AND 3. METHODS TO MAKE THE COMPUTER SYSTEM EASY TO USE. RESEARCHERS WILL WRITE AND DEMONSTRATE A PROTOTYPE SYSTEM THAT PERFORMS DESIGN COMPUTATIONS FOR A VARIETY OF STUDIES HAVING BINARY, CONTINUOUS, OR FAILURE-TIME OUTCOMES. BY FACILITATING ACCESS TO SOPHISTICATED METHODS FOR THE DESIGN OF EXPERIMENTAL STUDIES, THE SYSTEM WILL HAVE WIDE APPLICATIONS AMONG CLINICAL RESEARCHERS IN CANCER TREATMENT AND BASIC BIOMEDICINE, AND EXPERIMENTERS IN A VARIETY OF INDUSTRIAL AREAS.