Auditory evoked responses and learning and awareness during general anesthesia.

Abstract

There is a major distinction between conscious and unconscious learning. Monitoring the mid-latency auditory evoked responses (AER) has been proposed as a measure to ascertain the adequacy of the hypnotic state during surgery. In the present study, we investigated the presence of explicit and implicit memories after anesthesia and examined the relationships of such memories to the AER. We studied 180 patients scheduled for elective surgical procedures. After a thiopental induction, one of four anesthetics were studied: Opioid bolus: 7.5 microg x kg(-1) fentanyl, 70% N2O, with 2.5 microg x kg(-1) supplements as needed (n=100); Opioid infusion: Alfentanil 50 microg x kg(-1) bolus, 1-1.5 microg x kg(-1) x min(-1) infusion, 70% N2O (n=40); Isoflurane 0.3%: Fentanyl 1 microg x kg(-1), 70% N2O, isoflurane 0.3% expired (n=16); Isoflurane 0.7%: Fentanyl 1 microg x kg(-1), 70% N2O, isoflurane 0.7% expired (n=23). AER were recorded before anesthesia, 5 min after surgical incision and then every 30 min until the end of surgery. A tape of either the story of the "Three Little Pigs" or the "Wizard of Oz" was played continuously between the recordings. Explicit memory was assessed postoperatively by tests of recall and recognition, and implicit memory was assessed by the frequency of story-related free associations to target words from the stories, which were solicited twice during a structured interview. Six patients showed explicit recall of intraoperative events: All received the opioid bolus regimen. About 7% of patients reported dreaming during anesthesia. The incidence of picking the correct story that had been presented during anesthesia averaged 49%, i.e., very close to chance level. Overall, priming occurred only at the second association tests for the opioid bolus regimen, for which the frequency of an association to the presented story among those not giving an association to the control story was 26%, which was double the frequency (13%) of an association to the control story among those not giving an association to the presented story. This was significant by McNemar's test, P=0.02. There were significant associations between awareness, priming and AER, e.g., recall was associated with higher Nb amplitudes during anesthesia and priming was associated with shorter wave latencies. The incidence of awareness in patients anesthetized with nitrous oxide and bolus supplementation was 6%. Thus, this anesthetic technique did not reduce the risk of awareness compared with the use of nitrous oxide alone. Implicit memory occurred during nitrous oxide and bolus supplementation. Recording AER during anesthesia may help to predict awareness and implicit memory, particularly the former. The short contents of most of the dreams which were recalled could hamper future studies in this area.