AWARENESS is a rare but potentially harmful incident during general anesthesia. The reported incidence is 0.2% in general surgery and up to 1.5% in obstetric 1and cardiac 2surgery. In a recent review, Ghoneim 3described several potential causes for anesthetic drug doses not being adequate to suppress awareness and offered a number of suggestions for prevention. Unrecognized interruption of anesthetic drug delivery was identified as an important contributing factor, and development of an awareness monitor was given as one proposed solution. Reliance on traditional vital signs has been shown to be ineffective in identifying such unexpected events in time to prevent intraoperative awareness.

The BIS monitor (Aspect Medical Systems, Inc., Newton, MA) provides a drug-independent index for sedation and hypnosis (Bispectral Index [BIS]) 4–6that correlates with both the incidence of explicit (spontaneous recall) and implicit awareness (recall only after priming or hypnosis). 7–9A BIS between 40 and 60 is recommended for adequate depth of anesthesia, 10and at a value higher than 80, patients regain consciousness. We report two cases of explicit awareness of events during anesthesia due to interrupted anesthetic drug delivery.

Case 1

An 80-yr-old woman (American Society of Anesthesiologists physical status II) was scheduled for right shoulder surgery during general anesthesia combined with an interscalene block. The night before surgery, she was given her usual sleeping pill (1 mg temazepam).

At the time of arrival in the operating room at 7:15 am, the patient was fully awake. After institution of the standard monitoring, 1,000 ml lactated Ringer's solution was connected to an intravenous catheter on the back of her left hand. An interscalene block was performed using a nerve stimulator without administration of sedative drugs. Thirty milliliters bupivacaine, 0.375%, was injected when the appropriate motor response was observed.

Zip-Prep electroencephalographic electrodes (Aspect Medical Systems, Inc.) were then fixed at the frontal positions as recommended for BIS monitoring (Aspect-1000, BIS version 3.2; Aspect Medical Corp., Newton, MA). After confirming establishment of the regional block, a target-controlled infusion of propofol (Diprifusor; Astra-Zeneca Ltd., London, United Kingdom) was started at an initial target plasma concentration of 6 μg/ml together with an infusion of remifentanil at a rate of 0.25 μg · kg−1· min−1. After intubation facilitated with 30 mg atracurium, the patient underwent ventilation with 30 vol% oxygen in air, and the target propofol concentration was reduced to 3 and then 2 μg/ml, which was sufficient to maintain the BIS between 40 and 60. No further relaxants were administered thereafter. The remifentanil infusion was adjusted according to the blood pressure.

With the patient in the beach chair position, the anesthetist had no direct access to the patient's head and left hand, and the infusion line was not directly visible. Forty-five minutes after induction of anesthesia, surgery was started. Because 10 min after skin incision the BIS increased from 45 to 65, the propofol target concentration was increased to 3 μg/ml, which decreased the BIS again to 45. However, only a few minutes later, the BIS increased again to 65 and then suddenly to 90 without hemodynamic changes. No movement of the patient was observed at that time, 1 h after the induction dose of atracurium. Although the target propofol concentration was immediately increased from 2.5 to 6 μg/ml, the BIS remained at 90. Only a few minutes later, heart rate suddenly increased from 66 to 110 beats/min, and systolic blood pressure increased from 105 to 170 mmHg. Isoflurane administration was started, and it was recognized that propofol (and remifentanil) were backed up in the intravenous tubing. After repositioning of the left hand, the intravenous catheter flowed freely, the BIS decreased to 35, isoflurane was discontinued, and the target propofol concentration was again reduced.

After completion of the operation (2 h later), the propofol and remifentanil infusions were discontinued. The patient woke up 4 min later and underwent extubation. When asked about her anesthetic experience, she reported that she had woken up during surgery, hearing the surgeons talking about politics. She spontaneously mentioned the name Schroeder, which could be related to the surgeons’ discussion on the German chancellor during the first part of the operation. She did not feel any pain or discomfort but was unable to speak with the tube in her trachea. At the postoperative visit, an explanation for the awareness was given to the patient. She was not worried about the event because she did not feel pain.

Case 2

A 40-yr-old man (American Society of Anesthesiologists physical status I) was scheduled for lower back surgery. Because he was enrolled in an anesthesia research protocol, a Dräger Xenon Cicero EM anesthesia work station (Drägerwerke AG, Lübeck, Germany) was used with a charcoal filter inserted after the inspiratory valve of the circle system to allow for rapid elimination of volatile anesthetics. The filter was normally bypassed by closing a valve manually.

After institution of the standard monitoring and the BIS monitor (Aspect-1000, BIS version 3.2), anesthesia was induced with 2.5 mg/kg propofol, 100 μg remifentanil, and 0.1 mg/kg vecuronium. Immediately after loss of consciousness, the desflurane vaporizer (Dräger Devapor; Drägerwerke AG) was set at 4 vol% (fresh gas flow 3 l/min) and the patient underwent mask ventilation without difficulty. However, no desflurane or expired carbon dioxide was detected by the gas monitor (Datex AS3 Compact; Datex-Ohmeda, Instrumentarium Corp., Helsinki, Finland). The sampling tube was checked for kinking or obstruction, and an improper fit of the water trap at the sampling port of the monitor was detected. End-tidal carbon dioxide was now evident, but still no desflurane was detected, although the vaporizer output had been increased to 18 vol%.

Two minutes after injection of propofol, the BIS reached a nadir of 30 and slowly increased to 64 over the next 5 min. An additional dose of remifentanil (50 μg) was administered before tracheal intubation. We suspected that there was a technical problem with the display of the desflurane concentration, particularly because there were no clinical signs of inadequate anesthesia and because the BIS remained stable between 60 and 70. At the time of tracheal intubation, the BIS was 68. After intubation, systolic blood pressure increased from 119 to 160 mmHg, and heart rate increased from 76 to 113 beats/min. The gas analyzer still did not detect desflurane, although the capnography curve was normal. Only after another 10 min, it was discovered that the charcoal filter was not bypassed. Meanwhile, the BIS had remained stable below 70 and promptly decreased from 66 to 48 after the charcoal filter was bypassed and the desflurane concentration increased.

The rest of anesthesia and surgery were uneventful. Immediately after extubation, the patient spontaneously recalled tracheal intubation as an unpleasant but not a painful event, but he did not recall further events during surgery. The patient was then followed up for 6 months, and there were no psychologic sequels.

Both cases of explicit awareness were caused by an unrecognized interruption of anesthetic drug delivery and illustrate typical clinical situations in which awareness during general anesthesia may occur. Neither patient initially demonstrated hemodynamic or motor reaction as evidence of inadequate anesthesia. However, at least the second patient may have been paralyzed. Laryngoscopy with intubation as illustrated by the second case is a situation especially prone to awareness. 11 

The BIS value in the first patient was at 90 for several minutes. Propofol and remifentanil are drugs with rapid redistribution and elimination. If infusion is accidentally stopped, the patient's state of hypnosis, as in this case, changes rapidly from unconscious to fully awake. Because hemodynamic response to surgical stimulation may occur only after the patient is awake, electroencephalographic monitoring, such as BIS monitoring, is a useful tool for early detection of an unintended decrease of hypnotic drug effect. If an increase of the BIS does not respond to immediate increase in anesthetic drug concentration, the infusion line used for drug delivery should be checked immediately. The injection site should be made visible as much as possible and antireflux valves must be used to allow an uninterrupted drug supply. Unfortunately, an antireflux valve was not available in the first case.

The desflurane remaining undetectable for more than 20 min in the second patient would have prompted intravenous administration of a hypnotic drug if the BIS had increased above 70. Glass et al.  6reported a small probability of recall during propofol sedation at a BIS below 70 (BIS95(absence of recall)= 77; 95% confidence interval, 72–83). In their series of 90 trauma patients, Lubke et al.  9detected an increasing risk of auditory information processing and implicit recall as BIS values increase above a range of 40–60. They failed to define a BIS level with a significantly increased risk of explicit recall, probably because of the low number of patients with a BIS higher than 60. Flaishon et al.  12reported that 50% of their study patients responded to command within 4 min when the BIS increased to 65, although none had an explicit recall of any event before regaining consciousness after a bolus of propofol. Aware of the data mentioned herein, the anesthesiologist accepted a moderately increased BIS (between 60 and 70) for more than 15 min while searching for the technical problem with desflurane measurement. In retrospect, end-tidal monitoring was effective in identifying an interruption in anesthetic delivery, but previous technical problems in obtaining an adequate capnography curve led the anesthesiologist to conclude incorrectly that the agent analyzer was malfunctioning.

The current cases illustrate the results of Flaishon et al. , 12indicating that the probability of awareness and recall is not only related to the absolute BIS level but also to the duration of an increased BIS. The time lag between the increase of the BIS above 60 in the both patients and the event of explicit recall corresponds exactly to the 4 min reported in their article.

In a recent prospective survey, 11the use of end-tidal anesthetic gas analyzers apparently did not reduce the incidence of awareness. In contrast, education of personnel, regardless of monitoring, was more effective. 1,2The use of anesthesia depth monitors, such as BIS 5and midlatency auditory-evoked potentials, 13are thus to be considered in that context. They add important information to clinical signs of inadequate anesthesia and to measurement of anesthetic gas concentrations. To be effective in prevention of awareness and recall, depth-of-anesthesia monitors should be used according to rational practice guidelines when situations with an increased risk for awareness should be identified. One of them is laryngoscopy with intubation, 11which implies that depth-of-anesthesia monitors should already be installed for induction. Another is increase of the BIS above 60 for more than 3 min. If not attributable to an increase in electromyographic activity, this should be treated promptly by increasing the hypnotic drug concentration. 14If an increased BIS does not respond to the increase of the hypnotic drug, an accidental interruption of drug delivery 3has to be suspected, and another hypnotic drug should be administered by another route. We conclude that depth-of-anesthesia monitors may be useful in prevention of awareness but only together with rational practice guidelines and continuous education.

Lyons G, Macdonald R: Awareness during caesarean section. Anaesthesia 1991; 46: 62–4
Ranta S, Jussila J, Hynynen M: Recall of awareness during cardiac anaesthesia: Influence of feedback information to the anaesthesiologist. Acta Anaesthesiol Scand 1996; 40: 554–60
Ghoneim MM: Awareness during anesthesia. A nesthesiology 2000; 92: 597–602
Sigl JC, Chamoun N: An introduction to bispectral analysis for the electroencephalogram. J Clin Monit 1994; 10: 392–404
Rampil IJ: A primer for EEG signal processing in anesthesia. A nesthesiology 1998; 89: 980–1002
Glass PS, Blom M, Kearse LA, Rosow C, Sebel PS, Manberg P, Bloom M, Kearse L, Sebel P: Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers. A nesthesiology 1997; 86: 836–47
Kearse LAJ, Rosow C, Zaslavsky A, Connors P, Dershwitz M, Denman W: Bispectral analysis of the electroencephalogram predicts conscious processing of information during propofol sedation and hypnosis. A nesthesiology 1998; 88: 25–34
Liu J, Singh H, White PF: Electroencephalographic bispectral index correlates with intraoperative recall and depth of propofol-induced sedation. Anesth Analg 1997; 84: 185–9
Lubke GH, Kerssens C, Phaf H, Sebel PS: Dependence of explicit and implicit memory on hypnotic state in trauma patients. A nesthesiology 1999; 90: 670–80
BIS Clinical Reference Manual. Newton, Massachusetts, Aspect Medical Systems, Inc., 1998
Sandin RH, Enlund G, Samuelsson P, Lennmarken C: Awareness during anaesthesia: A prospective case study. Lancet 2000; 355: 707–11
Flaishon R, Windsor A, Sigl J, Sebel PS: Recovery of consciousness after thiopental or propofol. A nesthesiology 1997; 86: 613–9
Dutton RC, Smith WD, Rampil IJ, Chortkoff BS, Eger EI: Forty-hertz midlatency auditory evoked potential activity predicts wakeful response during desflurane and propofol anesthesia in volunteers. A nesthesiology 1999; 91: 1209–20
Bruhn J, Bouillon TW, Shafer SL: Electromyographic activity falsely elevates the bispectral index. A nesthesiology 2000; 92: 1485–7