Design Optimization for Developing Population Pharmacokinetic Models in Critically Ill Children: Application to...

Area of Science:

• Pharmacokinetics and Pharmacodynamics
• Pediatric Pharmacology
• Clinical Trial Design

Background:

• Population pharmacokinetic (popPK) models are crucial for optimizing pediatric drug dosing.
• Existing study designs often lack efficiency, posing challenges due to limited pediatric patient populations and scarce resources.
• There is a need to enhance the precision of popPK models while minimizing patient burden and costs.

Purpose of the Study:

• To optimize study designs for developing popPK models of teicoplanin, piperacillin, and meropenem in pediatric patients.
• To improve model precision and reduce patient burden and economic costs.
• To evaluate designs for pediatric patients with and without continuous kidney replacement therapy (CKRT).

Main Methods:

• Utilized Fisher Information Matrix (FIM) optimization with the $DESIGN option in NONMEM 7.5.
• Selected previously developed popPK models for each antibiotic.
• Fixed the number of subjects to 28 (14 with and 14 without CKRT) and analyzed different sample types (plasma, prefilter, postfilter, effluent).

Main Results:

• Optimized designs required fewer sampling times per individual: 3 for teicoplanin, 4 for piperacillin, and 6 for meropenem.
• Reduced total samples per patient by 25% (teicoplanin), 51% (piperacillin), and 21% (meropenem) compared to original studies.
• Optimized designs demonstrated robustness across varying proportions of patients with/without CKRT and different covariate values.

Conclusions:

• Optimizing study designs enhances accuracy and precision of popPK model parameters.
• Reduced sampling requirements offer significant advantages, particularly for critically ill pediatric patients.
• Efficient study designs are vital for advancing pediatric pharmacotherapy with limited resources.