Memory of intraoperative events

BMJ 1994; 309 doi: https://doi.org/10.1136/bmj.309.6960.967 (Published 15 October 1994) Cite this as: BMJ 1994;309:967
  1. J G Jones

    Considerable public interest and anxiety exist about patients waking up during general anaesthesia with explicit memories of painful and terrifying intraoperative events. Using conventional clinical signs, anaesthetists find it almost impossible to recognise conscious awareness in patients with complete neuromuscular blockade.1 In elective surgery such awareness is often due to the anaesthetists not realising that the delivery of anaesthetic has failed. Occasionally there are fictitious claims of conscious awareness, and rarely there are cases with no obvious explanation.

    Fortunately, the incidence of conscious awareness with pain during surgery is only 0.01% during elective general anaesthesia.2 The incidence has fallen considerably since the 1960s, when Hutchinson found that 0.6% of patients anaesthetised with unsupplemented nitrous oxide were awake and in pain.3 The incidence is much higher during operations for major trauma, where anaesthetic concentrations are reduced to preserve cardiovascular function.4

    The psychological consequences of conscious awareness with explicit memory of pain are not known. Moerman et al described the sequelae in 26 patients.1 Eleven had a persistent fear of anaesthesia; seven had sleep disturbances, nightmares, anxiety, or mental distress; and eight had no ill effects. The proportion of such patients who take legal action is unknown, but the cases of the patients who do are well publicised.

    Between 0.2% and 0.4% of patients have explicit memory of some intraoperative events but do not experience pain.2 These memories often come to light only after careful postoperative questioning. This type of conscious awareness is caused by the combination of neuromuscular blockade with light general anaesthesia. Lyons and Macdonald showed that in obstetric anaesthesia a small increase in the dose of anaesthetic could reduce the incidence of this complication from 1.3% to less than 0.4%.5 In their study there were few sequelae and patients rarely sought legal redress.

    Isolating patients' arms from the neuromuscular blocking drugs by means of an inflated blood pressure cuff shows that up to half of patients are awake during some anaesthetic procedures. This can be judged by the fact that they move the isolated arm in response to the anaesthetist's commands.6 These patients have no complaint of pain at the time of surgery, no obvious changes in physical signs, and no explicit memory of any intraoperative event when interviewed postoperatively. This method is, however, unsuitable for prolonged monitoring because of the risk of ischaemia in the isolated arm.

    Patients who seem to be adequately anaesthetised and who have no explicit memory of intraoperative events may show implicit memory of such events when tested postoperatively. In contrast with explicit memory, which entails the conscious recollection of facts and events, implicit memory refers to non-conscious changes in performance or behaviour that are produced by experience.7 There is conflicting evidence that implicit memory of intraoperative events can be registered by the brain.*RF 7-9* Ghoneim and Block list 14 papers showing implicit memory during general anaesthesia, whereas Merikle and Rondi concluded that “there is not a single consistent finding indicating that adequately anaesthetised patients do in fact remember events during anaesthesia”.8,9 In many of the studies that show implicit learning during anaesthesia claims have been made about either the possible advantages or the deleterious effects on patient outcome. The conflicting findings, however, mean that many groups either are unconvinced that implicit memories can be registered during anaesthesia or regard this as a hypothetical possibility.

    No objective measure

    All these studies of implicit memory during anaesthesia are flawed by the lack of an objective measure of the anaesthetic state. Reviews by Schwender and by Thornton and Jones showed that the middle latency (or early cortical) auditory evoked potential in the electroencephalogram had a dose related fall in amplitude and increase in latency with most common general anaesthetics.10,11 Surgical stimulation of the patient while keeping the anaesthetic concentration constant produced a change in the evoked potential similar to that seen by reducing the concentration of anaesthetic. This supports the idea that the middle latency auditory evoked potential is a dynamic measure of the anaesthetic state of the brain.

    By combining psychological and electrophysiological techniques it might be possible to establish whether implicit memory of events during anaesthesia could be registered in apparently anaesthetised patients. Such a study has been reported by Schwender et al in patients having cardiac surgery.12 After sternotomy under general anaesthesia, and before cardiopulmonary bypass grafting, the patients' auditory evoked potentials were recorded and an audiotape of a short version of Robinson Crusoe was played to them. When interviewed postoperatively none of the patients had explicit memories of any intraoperative event. But in an implicit memory test using the code word Friday seven of the 30 patients given the tape associated the word Friday with the story of Robinson Crusoe.

    These patients all had large amplitude middle latency auditory evoked potentials that were similar to those seen in lightly anaesthetised subjects. Those with no implicit memory of the story had the low amplitude potentials as seen in more deeply anaesthetised subjects. This study could resolve the disagreement about the likelihood of registering new implicit memories during anaesthesia because it suggests that the anaesthetised brain needs to be in a particular state of arousal to register these memories.

    Auditory evoked potentials have now also been used to show the graded effects on the brain of many common general anaesthetics. Changes in auditory evoked potentials correlate with changes in explicit memory at low doses of anaesthesia,13,14 and Schwender et al suggest that the technique could indicate the point where implicit memory of intraoperative events is ablated.12

    It remains to be seen whether these results are confirmed in wider studies and whether the auditory evoked potentials will be a useful routine monitor of cognitive function in anaesthetised patients.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    View Abstract

    Sign in

    Log in through your institution