Modeling and estimation for patient controlled analgesia of chronic pain

An existing mathematical model of the patient and the pain in patient controlled analgesia is extended to account for the potent metabolite morphine-6/spl beta/-glucuronide and to characterize chronic pain. Nominal population parameters characterizing pain and analgesic effect are tuned to make the model match observed clinical behavior. The model is used to design an on-line estimator generating state estimates which should be suitable for use in a stochastic PCA algorithm delivering a self-adjusting continuous infusion.<
>