Drug interaction: focusing on response surface models

Area of Science:

• Pharmacology
• Anesthesiology
• Mathematical Modeling

Background:

• Anesthesiologists recognize the significance of optimal drug combinations.
• Drug interactions are classified as additive, synergistic, or antagonistic.
• A zero interaction model is essential for classifying drug combinations.

Purpose of the Study:

• To review and evaluate models for classifying anesthetic drug interactions.
• To identify optimal approximating models for anesthetic drug response surfaces.
• To explore methods for robust inference through model averaging.

Main Methods:

• Utilizing the sigmoid Emax (Hill equation) for response surface modeling.
• Evaluating established drug interaction models (Greco, Machado, Plummer, Carter, Minto, Fidler, Kong).
• Applying the Akaike information criterion (AIC) for model selection.
• Investigating parametric models with polynomial interaction functions.

Main Results:

• Several models are adequate for analyzing anesthetic drug interaction data.
• Parametric models with polynomial interaction functions can capture complex interactions.
• AIC is a popular method for selecting the best approximating model.
• Model averaging offers more robust inference than single model selection.

Conclusions:

• Optimal model selection is critical for understanding anesthetic drug interactions.
• Parametric models provide flexibility in describing diverse interaction patterns.
• Model averaging enhances the reliability of conclusions drawn from anesthetic drug interaction studies.