Predicting optimal anesthesia level from propofol and remifentanil concentration: analysis of covariate factors for individualization

Abstract

General anesthesia involves some targeting effects which aim to prevent the patient from suffering against the therapeutic aggression. These effects are hypnosis, analgesia, amnesia and immobility and to achieve them a combination of drugs is delivered into the patient, from which propofol and remifentanil are highlighted. In the operating room, monitoring systems are used to assess the depth of anesthesia in real time. This monitoring includes basic systems such as arterial blood pressure, oxygenation or electrocardiogram and electroencephalogram derived measures, which are more complex; from this last group, BIS index is a good indicator. Being able to predict the anesthetic depth from a set of input variables could be valuable during the surgery, as it would help the anesthesiologists to prevent adverse effects, and it would help the post-operative recovery. Knowing this, the aim of this project is to predict the probability to be in the optimal level of anesthesia, which is related to the BIS index. This probability is obtained from the input concentration of propofol and remifentanil, a hypnotic and an analgesic drug respectively, and from the demographic variables such as age, height or gender. To do so, a Logistic Regression model will be built with data from patients undergoing general anesthesia in Cirurgia Major Ambulatòria (CMA) in Hospital Clínic.