Towards individualised anaesthesia: A comparison between target-controlled infusion and closed-loop control

Abstract: Individualised healthcare is the future of medicine, due to the so called inter-patient variability. The inter-patient variability involves differences in the drug response between different patients. The effect of a drug can usually be divided into pharmacokinetics (PK) and pharmacodynamics (PD). One way to individualise anaesthesia, may be to automate it using closed-loop control systems. This thesis focused on comparing a commonly used method to calculate the dosage of propofol for anaesthesia, target-controlled infusion (TCI), with closedloop control using a proportional-integral-derivative (PID) controller. The TCI method was implemented with quadratic programming in two ways: one to calculate the optimal propofol dosage for a reference patient, and the other to calculate an optimal propofol dosage for all patients in the patient set. The patient model set is based on a model developed by Eleveld et al, which is a PKPD model using six different covariates to cover a broad population. The two TCI methods and the closed-loop control method were simulated on the same set of 100 patients, and the results were compared. The results show that the range of resulting depth of hypnosis after stabilisation for all patients in the set was smaller for closed-loop control than it was for the both TCI-methods. Furthermore, the simulation of closed-loop control resulted in all patients being within the desired interval, whereof the majority reached the desired depth of hypnosis. This indicates the high potential of closed-loop controlled anaesthesia in the future, and a more individualised healthcare.