Related Experiment Videos

Phenobarbital in intensive care unit pediatric population: predictive performances of population pharmacokinetic

Amélie Marsot1, Fabrice Michel2, Estelle Chasseloup1

  • 1Service de Pharmacologie Clinique et Pharmacovigilance, Pharmacologie Intégrée et Interface Clinique Industrielle, Institut des Neurosciences Timone - AMU-CNRS 7289, Aix-Marseille Université, AP-HM, Hopital Timone - Bâtiment F, 264 rue Saint Pierre, Marseille, 13385, France.

Insights

This study validated a phenobarbital population pharmacokinetic model for pediatric intensive care units. The model accurately predicts drug concentrations, supporting optimized dosing regimens for critically ill children.

Area of Science:

  • Pharmacology
  • Pediatric Critical Care
  • Pharmacokinetics

Background:

  • Phenobarbital dosing requires accurate pharmacokinetic models, especially in pediatric intensive care units (PICUs).
  • External validation of existing models is crucial for confirming their applicability and extending recommendations to broader patient populations.
  • The Marsot et al. model for phenobarbital pharmacokinetics in neonates needs evaluation in a diverse PICU population.

Purpose of the Study:

  • To externally evaluate the phenobarbital population pharmacokinetic model by Marsot et al. in a pediatric intensive care unit setting.
  • To confirm the model's predictive performance and assess its suitability for dose adjustment in critically ill children.
  • To extend the validated dosing recommendations to the entire PICU population.

Main Methods:

  • Retrospective analysis of phenobarbital concentrations from 35 pediatric intensive care unit patients.
  • Implementation of the Marsot et al. population pharmacokinetic model using NONMEM 7.3 software.
  • Assessment of predictive performance using bias, inaccuracy, Normalized Prediction Distribution Errors (NPDE), and Visual Predictive Check (VPC).

Main Results:

  • The model demonstrated reasonable bias and inaccuracy in predicting phenobarbital concentrations.
  • Median prediction error was 3.03% and median absolute prediction error was 26.20%.
  • No significant trends were observed in NPDE and VPC, indicating good model performance.

Conclusions:

  • The phenobarbital population pharmacokinetic model by Marsot et al. is externally validated for IV infusion administration in a PICU setting.
  • The model-based dosing regimen can be extended to optimize phenobarbital treatment in all PICU patients.
  • Therapeutic drug monitoring is recommended alongside the model-based dosing due to inter- and intra-patient variability.

Related Concept Videos