Approaches to handling missing or "problematic" pharmacology data: Pharmacokinetics

Area of Science:

• Pharmacokinetics
• Data Analysis
• Biostatistics

Background:

• Missing or inaccurate data is frequent in pharmacokinetic (PK) analyses.
• Issues include incorrect dosing, sub-quantification concentrations, and missing sample times or covariates.

Purpose of the Study:

• To review existing literature on handling problematic PK data.
• To present simulation results evaluating common data error handling methods.
• To provide recommendations for addressing PK data deficiencies.

Main Methods:

• Literature review of data imputation and error handling techniques.
• Simulation studies to assess methods for common PK data errors.
• Evaluation of approaches for missing or inaccurate dose, concentration, time, and covariate data.

Main Results:

• Identified gaps in current recommendations for PK data error handling.
• Simulation results provide insights into the performance of various methods.
• Demonstrated utility of described approaches for diverse PK datasets.

Conclusions:

• A comprehensive review and simulation-based evaluation of PK data error handling methods.
• Guidance is provided to bridge knowledge gaps for scientists analyzing PK data.
• Applicable to both clinical and nonclinical pharmacokinetic data analysis.