HYBRID SIMULATION TECHNIQUE FOR PATIENT CONTROLLED ANALGESIA USING REAL PATIENT DATA

摘要

In this paper we assume that the patient's demand in patient controlled analgesia (PCA) is a random process with a unique shape and parameters. In order to find this process we investigated two randomly selected, real data based morphine and fentanyl PCA logs and created patients' behavioral model that approximated real demand data. We used the created patient behavioral models to create 500 virtual PCA logs of both morphine and fentanyl analgesia. These logs allowed pharmacokinetic simulation of the effect compartment concentration. We used quantized state system model to create hybrid aggregate model of PCA. The proposed methodology allows an estimation of frequency and duration of critical episodes during PCA.