Closed-loop vasopressor control: in-silico study of robustness against pharmacodynamic variability

Area of Science:

• Critical care medicine
• Biomedical engineering
• Pharmacology

Background:

• Vasopressor potency can vary significantly in clinical settings due to factors like drug interactions and organ dysfunction.
• Previous work established the initial feasibility of a novel closed-loop controller for vasopressor infusions.

Purpose of the Study:

• To evaluate the effectiveness and robustness of a novel closed-loop vasopressor controller under conditions of substantial drug potency variation.
• To determine if the controller maintains target blood pressure convergence despite a tenfold range of norepinephrine responsiveness.

Main Methods:

• A physiologic simulator generated 250 randomized simulated septic patients.
• The closed-loop controller managed simulated patients across seven norepinephrine responsiveness conditions (0.1x to 10x expected population response).
• Controller performance was assessed using Varvel's criteria, time-out-of-target, median performance error, and wobble.

Main Results:

• The controller demonstrated effectiveness across all tested norepinephrine responsiveness levels (0.1x to 10x).
• Median performance error and median absolute performance error remained below 5% in all conditions.
• While generally stable, the controller approached instability at the highest response level (10x), and showed slower initial hypotension correction at lower response levels (0.1x, 0.2x).

Conclusions:

• The closed-loop vasopressor controller is robust and effective in simulated patients with drug response variability from 0.1x to 10x the expected population response.
• The controller shows potential for reliable vasopressor management in diverse clinical scenarios with fluctuating drug efficacy.
• Further investigation is warranted regarding controller stability at extreme drug response levels.