Introduction: Target-controlled infusion (TCI) models, such as Marsh and Schnider are commonly used for the induction and maintenance of propofol total intravenous anaesthesia (TIVA). However, substantial inter-individual variability remains in patient response, including time to loss and return of consciousness. Population pharmacokinetic (popPK) analyses require accurate dosing/input data to reliably characterise drug disposition.
Problem: : During surgery, propofol-TIVA is titrated by the anaesthetist, commonly guided by the Bispectrality Index (BIS). This results in non-constant dosing, with frequent changes in infusion rate. . In addition, the software embedded in the computer-controlled infusion devices records only total dose administered over certain time periods, making it difficult to reconstruct multiple infusion-rate changes. Although key data may be available, including plasma propofol concentrations, total administered dose and patient covariates, popPK analysis cannot be performed reliably without detailed dosing-rate information.
Aims: This project aims to use the Berkeley Madonna software program to estimate a theoretical average rate of change between variable infusion rates and thereby reconstruct the likely infusion input profile.
Proposed Method: Using a constant average infusion rate transitions approach, we will develop, visualise and refine a model capable of estimating infusion-rate transitions. The resulting dosing profile may then be used as input for subsequent popPK analysis.
Expected Outcomes: We anticipate greater variability between the measured and predicted plasma propofol concentrations during the initial induction and early maintenance phases, when infusion rates are changing rapidly. Variability is expected to reduce during periods of relatively stable infusion rates. Specifically, we expect to reconstruct the complete infusion dosing profile administered by the TCI model.
Significance: These findings may provide a practical method for estimating infusion-rate transitions in popPK analyses where variable-rate infusion data are incomplete.