Simulation and Model-Based Decisions

Use fitted population PK and PK/PD models to perform simulations, explore scenarios, and support model-informed decisions.
Tip

Module goal: Learn how fitted models become useful by generating predictions and exploring what-if scenarios.

Module Overview

Until now, the course focused on:

Data → Model → Evaluation

This module introduces the next step:

Model → Simulation → Decision

Once a model has been estimated and evaluated, it can be used to explore scenarios that were not directly observed.

Examples include:

  • changing dose
  • changing schedule
  • changing patient characteristics
  • predicting future outcomes
  • comparing strategies

Simulation is one of the main reasons pharmacometric models are valuable.

Learning Objectives

By the end of this module, you will be able to:

  • explain why simulation is performed
  • generate predictions from fitted models
  • distinguish deterministic and population simulation
  • simulate alternative dosing scenarios
  • interpret simulation outputs
  • connect simulation to model-informed decisions

Lessons in This Module

Lesson 1: Why Simulate?

Introduce the purpose of simulation and explain how simulation differs from estimation.

Lesson 2: Individual and Population Simulation

Generate simulations for typical individuals and virtual populations.

Introduce variability during simulation.

Lesson 3: Exploring Dose Scenarios

Compare alternative dose levels and dosing schedules.

Evaluate exposure changes across dosing scenarios.

Lesson 4: Simulation Using Fitted Models

Use fitted nlmixr2 models to simulate future observations.

Generate concentration predictions from fitted population PK models.

Lesson 5: Interpreting Simulation Results

Translate simulation outputs into practical conclusions.

Discuss uncertainty and limitations.

Software Used

This module continues using simulation tools already introduced earlier.

library(tidyverse)
library(nlmixr2)
library(rxode2)
library(nlmixr2data)

Simulation uses model structure and parameter estimates.

No additional estimation is performed.

Datasets Used

This module primarily reuses:

data(
  "theo_sd",
  package = "nlmixr2data"
)

We continue using a familiar dataset so the focus remains on simulation concepts.

Module Workflow

Conceptually:

Observed Data

↓

Estimated Model

↓

Scenario Definition

↓

Simulation

↓

Interpretation

This shifts attention from:

What happened?

to:

What could happen?

How This Module Fits in the Course

Earlier modules focused on describing observed data through population PK models.

This module shifts the focus toward prediction.

Once a model has been estimated, it can be used to explore dosing scenarios that were never directly observed.

This predictive capability is one of the main reasons pharmacometric models are valuable.

What This Module Does Not Do

This module does not focus on:

  • full PK/PD simulation workflows
  • optimal design
  • trial simulation
  • Bayesian updating
  • advanced MIDD workflows
  • regulatory applications

Here the focus is practical simulation.

Expected Outputs

By the end of this module, you should have:

  • simulated concentration profiles
  • compared dosing scenarios
  • generated typical and population simulations
  • interpreted variability and uncertainty
  • connected simulations to decisions

Next Step

Start with Lesson 1 to understand why simulation is one of the central applications of pharmacometric models.