A diagnostic tool for population models using non-compartmental analysis: The ncappc package for R

Area of Science:

• Pharmacokinetics
• Computational Biology
• Statistical Modeling

Background:

• Non-compartmental analysis (NCA) is crucial for calculating pharmacokinetic (PK) metrics like area under the concentration-time curve.
• Population PK (PopPK) models require robust diagnostic tools to assess their predictive performance.

Purpose of the Study:

• To introduce ncappc, a novel R package for performing NCA and simulation-based posterior predictive checks (PPC) for PopPK models.
• To facilitate the evaluation of PopPK model adequacy using NCA metrics.

Main Methods:

• The ncappc package estimates NCA metrics from observed concentration-time data.
• It performs PPC by comparing NCA metrics from simulated data to those from observed data at both population and individual levels.
• Normalized Prediction Distribution Error (NPDE) is calculated for individual-level diagnostics.

Main Results:

• ncappc generates graphical and tabular outputs for NCA and PopPK model diagnosis.
• Outputs include comparisons of observed and simulated NCA metrics, deviations, NPDE, and regression parameters.
• The package aids in assessing the ability of PopPK models to simulate drug concentration-time profiles.

Conclusions:

• ncappc is a versatile R tool for estimating NCA metrics and diagnosing PopPK models.
• It provides easily interpretable outputs for model evaluation, including outlier identification.
• The package is freely available on CRAN and GitHub.