Adaptive pharmacokinetic and pharmacodynamic modelling to predict propofol effect using BIS-guided anesthesia

Area of Science:

• Anesthesiology
• Pharmacology
• Biomedical Engineering

Background:

• Propofol is a key anesthetic for hypnosis and general anesthesia.
• Bispectral Index (BIS) monitors anesthetic depth.
• Existing automated infusion systems lack personalized variability adjustments.

Purpose of the Study:

• To develop and validate a real-time pharmacokinetic/pharmacodynamic (PK/PD) model for propofol.
• To account for intra- and inter-patient variability in propofol administration.
• To improve anesthetic control and patient outcomes.

Main Methods:

• Utilized BIS data and propofol infusion records from 60 adult patients.
• Employed a real-time PK/PD compartmental model.
• Tuned key model parameters (ce50, γ, ke0) online by minimizing a performance function.

Main Results:

• The real-time PK/PD model demonstrated reduced prediction error compared to fixed-parameter models.
• Identified significant intra- and inter-patient variability in ce50, γ, and ke0.
• Provided mean parameter values: ce50 (3.89 mg/l), γ (4.63), and ke0 (0.36 min⁻¹).

Conclusions:

• The real-time PK/PD model offers a more accurate reflection of the patient's current state.
• Enables precise titration of propofol for desired effect and duration.
• Facilitates reduced drug waste and mitigation of post-operative side effects.