Have Biostatisticians Ceded the High Ground to Modelers, and Can We Get It Back?

• Activity Number: 564
• Type: Contributed
• Date/Time: Wednesday, August 6, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): David Raunig*+
• Companies: ICON Medical Imaging
• Keywords: dynamic modeling ; statistician role

Abstract

Increasingly drug development includes dynamical modeling that extends further than kinetics/dynamics and well beyond the typical statisticians roles in data analysis. Power analyses and sample size estimation are important inputs to study design but endpoint dynamics have thrust physiological modeling into the lead as promising new methods for disease evaluation. Model-based methods for evaluating Parkinson's Disease and tumor growth make headlines for dynamical modeling and even though statisticians were involved in the model development, their participation has been largely relegated to support. Modeling and random effects are integral parts of all statistical efforts but that high ground has been ceded to others who are answering the question of "What is going on?" rather than "Which hypothesis do we accept?" This paper will describe the parallels behind statistical and physiological modeling and how biostatisticians can regain their place in a new era of drug development.