Towards Closed-Loop Sedation In Intensive Care Units

Recently, we’ve seen a glut of closed-loop controllers being developed for patient use. Whether it be the progress towards an artificial pancreas or autonomous anesthesia (a la McSleepy), the trend seems that letting a computer handle adjusting some parameters both frees up physicians and nurses and provides better control of the process at hand.
The latest example of that is work being done at the Georgia Institute of Technology regarding sedation in the ICU. This Medgadget editor is currently working in an ICU and has seen first-hand how difficult it can be to walk the tight-rope between needless agitation and over-sedation. Traditionally, the ICU nurse assesses the patient on some sort of sedation scale, then titrates the sedative/analgesic medicines to the desired level of sedation.
“The challenge in developing closed-loop control systems for sedating critically ill patients is finding the appropriate performance variable or variables that measure the level of sedation of a patient, in turn allowing an automated controller to provide adequate sedation without oversedation,” said Behnood Gholami, a postdoc working on the project.
In this project, the inputs included “facial expression, gross motor movement, response to a potentially noxious stimulus, heart rate and blood pressure stability, noncardiac sympathetic stability, and nonverbal pain scale” to determine a level of sedation. They found that the computer model predicted agitation 18% of the time that the traditional models missed (but also had a 5% “false-negative” rate). Currently, this is just a Bayesian manipulation of human-assessed data, but future plans for this technology include using digital imaging to assess patients’ agitation level.
Press release: Automated ICU Sedation: Computer System for Evaluating Sedation Level Shows Strong Agreement with Clinical Assessment