KoreAnesthesia 2019

Workshop 10. NONMEM

Time, Place and Language: Nov 2 (Sat) 9:00-12:30, Room 9, Korean

Moderator and Instructors

Moderator :: Gyu-Jeong Noh (University of Ulsan, Korea)

Instructors :

Byung-Moon Choi (University of Ulsan, Korea)
Seong-Heon Lee (Chonnam National University, Korea)
Yong-Hun Lee (University of Ulsan, Korea)
Kwon-Hui Seo (Hallym University)

Description

NONMEM (nonlinear mixed effects modeling) workshop aims to provide fundamental concepts of population pharmacokinetic and pharmacodynamic analysis, and essential knowledge of how to run NONMEM, how to interpret NONMEM results, and how to validate a selected NONMEM model. The objective function of NONMEM, 2LL (log-likelihood), is discussed. This workshop shows how to describe inter-individual random variability of model parameters using covariates. Various components of NONMEM control stream and output are discussed using examples of NONMEM control and result files, respectively. Finally, model selection and validation are also discussed.

Time Table
9:00 - 9:15	Introduction to nonlinear mixed effects modeling
9:25 - 9:55	Structural model, statistical model, covariate model, residual model)
10:05 - 10:35	NONMEM output, objective function, relative standard error
10:45 - 11:15	Variance-covariance matrix, correlation matrix, condition number
11:25 - 11:55	Model selection and validation

Learning Objectives

The attendees understand NONMEM objective function, 2LL (log likelihood).
Pharmacokinetic and pharmacodynamic parameters split into fixed effect and random inter-individual effect parameters in mixed effects modeling.
The attendees understand statistical and error models.
The attendees understand several types of covariate model.
The attendees are encouraged to make a NONMEM control file and interpret NONMEM result file.
The attendees understand that predictive checks and bootstrap are used to internally validate a selected NONMEM model.

Photos