Modelling estimation and control in the relief of post-operative pain

• Published in: Automatica (Oxford) Vol. 21; no. 4; pp. 349 - 360
• Main Authors: Jacobs, O.L.R., Bullingham, R.E.S., Lammer, P., McQuay, H.J., O'Sullivan, G., Reasbeck, M.P.
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.07.1985; Elsevier
• Subjects: Bayes methods; Biological and medical sciences; biomedical; computer control; Computerized, statistical medical data processing and models in biomedicine; control engineering computer applications; demand analgesia; drug-delivery systems; estimation; Kalman filters; Medical management aid. Diagnosis aid; Medical sciences; medical systems; nonlinear systems; patient-controlled analgesia; estimation; nonlinear systems; control engineering computer applications; medical systems; computer control; Bayes methods; demand analgesia; drug-delivery systems; Kalman filters; biomedical; patient-controlled analgesia; Postoperative; Evaluation; Non linear estimation; Analgesic; Pain; On line; Perfusion; Automatic monitoring; Mathematical model; Pump; Microcomputer
• ISSN: 0005-1098; 1873-2836
• DOI: 10.1016/0005-1098(85)90072-X

The paper arises from collaborative work to explore contributions which the technology of control engineering can make to pain relief. One particular demand analgesia system was studied in which Fentanyl was intravenously infused to relieve pain during the first day after total hip-replacement. A mathematical model is presented in which pain is modelled as the difference between discomfort and comfort and is measured by the frequency at which the patient presses a button. This is inherently nonlinear because there is no button-pressing when comfort exceeds discomfort. A novel nonlinear estimation scheme to handle the nonlinearity is implemented in an on-line microcomputer to provide model-based stochastic control. Experimental results from trials on 20 real patients and 500 simulated patients validate the model and show that the model-based control gives substantially better pain relief than is achieved by conventional demand analgesia, and is robust against modelling errors. Directions for further work are indicated and it is concluded that there is promise for model-based on-line estimation in demand analgesia and promise for the mathematical model as a basis for further research on mechanisms of pain and its relief.