Related Experiment Videos
NONMEM and NPEM2 population modeling: a comparison using tobramycin data in neonates
Matthijs de Hoog1, Rik C Schoemaker, John N van den Anker
1Department of Pediatrics, Erasmus University and University Hospital Rotterdam/Sophia Children's Hospital, Dr. Molewaterplein 60, 3015 GJ Rotterdam, The Netherlands. dehoog@alkg.azr.nl
Therapeutic Drug Monitoring
|May 22, 2002
Summary
Nonlinear mixed effects modeling (NONMEM) and nonparametric expectation maximization (NPEM2) yield different tobramycin pharmacokinetic estimates. Differences are mainly due to how error models are incorporated, impacting population parameter precision and bias.
Area of Science:
- Pharmacokinetics
- Pharmacometrics
- Neonatal Medicine
Background:
- Nonlinear mixed effects modeling (NONMEM) and nonparametric expectation maximization (NPEM2) are used for population modeling of tobramycin.
- Understanding differences in these methods is crucial for accurate drug dosing in neonates.
Purpose of the Study:
- To compare NONMEM and NPEM2 for population pharmacokinetic modeling of tobramycin in neonates.
- To evaluate the impact of error models on parameter estimates and predictive performance.
Main Methods:
- Population pharmacokinetic analysis of tobramycin in 470 neonates using NONMEM and NPEM2 with a one-compartment model.
- Models were constructed to compare error pattern incorporation between NONMEM and NPEM2.
- Predictive performance was assessed in a separate group of 61 patients.
Main Results:
- NONMEM and NPEM2 produced dissimilar population pharmacokinetic parameter estimates (e.g., K(el), V(d)) and variation coefficients (CV).
- Adjusting error models reduced differences in CVs between methods.
- NONMEM demonstrated less bias and comparable precision when error patterns were harmonized.
Conclusions:
- NONMEM and NPEM2 provide different population pharmacokinetic estimates for tobramycin.
- The discrepancies are significantly influenced by the methods' error model incorporation strategies.
- Careful consideration of error modeling is essential when comparing results from NONMEM and NPEM2.