Fisher/Shafer NONMEM Workshop

Workshop Overview

The Fisher/Shafer NONMEM Workshop covered NONMEM, data handling, model building, and model diagnostics. The workshop includeed an extensive set of example problems ranging from simple didactic examples to demonstrate basic concepts to complex examples from the instructors' own experience. Registrants were provided with a comprehensive set of graphical tools programmed in the R and S Plus statistical languages to create high-quality diagnostic and analysis plots, as well as a powerful graphical user interface for NONMEM to facilitate rapid analysis of clinical trial data.

The workshop ran 3 and 1/2 days, from Sunday through Wednesday.

  • Sunday Morning: Basic Theory: (Shafer) Population Approaches, Fixed vs. Random Effects, Structural and Variance Models
  • Sunday Afternoon: (Fisher) Model Building and Diagnostics (PLT Tools)
  • Monday Morning: (Shafer) Pharmacokinetic Models, ADVAN vs. PRED, Closed Form vs. Differential Equations, Goodness of Fit Metrics
  • Monday Afternoon: (Fisher) Pharmacokinetic Models: Input Functions, ETAs, Time Varying Covariates, NONMEM Performance Enhancement
  • Tuesday Morning: (Shafer) Pharmacodynamic Models
  • Tuesday Afternoon: (Fisher) Simulation and Validation
  • Wednesday Morning: (Fisher) Reporting a NONMEM Analysis
The Fisher / Shafer NONMEM Workshop ran NONMEM using the PLT Tools software suite.

PLT Tools provides a graphical interface for controlling and organizing NONMEM analyses, comprehensive graphics to analyze model results, powerful features including visual predictive check, log likelihood profiling, bootstrap analysis, and jackknife model validation. It also has tools for comprehensive auditability and automated report generation. The fully functional PLT Tools is freely available for download.

In the workshops registrants who brought their own data were typically able to assemble datasets (often using the R code learned in the R workshop), and analyze their data using the PLT Tools diagnostic graphics (e.g., log likelihood, profiling, automatic covariate searching, booststrap, and VPC).

The Bethesda 2013 course was our 17th and last workshop. We thank the hundreds of students who have made this course one of the highlights of our lives for the course duration.