Efficient snd Effective Analysis in the Drug Approval Setting

• Activity Number: 289
• Type: Topic Contributed
• Date/Time: Tuesday, August 2, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistical Consulting
• Author(s): William K. Mountford*+
• Companies: PPD
• Keywords: reproducible research ; consulting statistics ; good practice ; clinical trial

Abstract

Statistical reporting is usually the last step encountered in a research study. Study results and their interpretation drive decision making and cannot afford human error. Therefore, statistical results are often required to be reproducible by numerous individuals in numerous settings. For instance, clinical trial results may be calculated by numerous statisticians within a clinical research organization, pharmaceutical company, and regulatory agency. The present discussion will provide an overview of processes implemented in the clinical research organization industry, which allow for an effective and systematic approach to reproducing study results. The process begins at the data collection stage where explicit written specifications are developed to instruct someone how to map a datapoint from a clinical database into an analysis database. Details will be provided on developing an analysis plan and study result mockups as well as the result validation process. In addition, there will be a discussion of techniques that increase efficiencies of the overall process described. In closing the real world pros and cons of the described process will be discussed.