Introduction

Three large prospective studies of the incidence of intraoperative awareness from Australia, Europe, and North America suggest that the overall rate is in the range of 0.1% to 0.2% or 1 to 2 per 1000 patients. Intraoperative awareness can be a minor or a major complication, depending on the severity and the response of the individual patient; in severe cases post-traumatic stress disorder may occur. In select patient populations the rate of intraoperative awareness may be substantially higher, such as in cardiac surgery patients, in which the rate has been reported to be in the range of 0.4% to 1%. Prospective studies of intraoperative awareness in children found a rate of 0.8% to 1.1%. Conversely, the rate of intraoperative awareness may be lower in a particular setting. A retrospective analysis of quality assurance data from a single medical center suggested that the incidence of intraoperative awareness was 0.0068% or 1 per 14,560 patients. Methodologic criticisms can be made of all of these studies of the incidence of intraoperative awareness. However, as a whole, the literature suggests that intraoperative awareness is a significant problem. Many anesthesiologists find a rate of intraoperative awareness in the vicinity of 0.1% to be unacceptably high. Most patients affected by intraoperative awareness find the experience to be unacceptable, especially if they experience pain and anxiety. Can we prevent recall during anesthesia or, at least, lower the rate substantially?

Options

Some episodes of intraoperative awareness are caused by specific, identifiable errors in anesthetic drug administration. Examples of these errors include the following:

• 1.
  

  Administration of a muscle relaxant instead of a hypnotic during induction of anesthesia, resulting in an awake, paralyzed patient

  
  
• 2.
  

  Unrecognized failure of a pump to deliver an intravenous hypnotic drug such as propofol (see Rowan for a particularly vivid example)

  
  
• 3.
  

  An unrecognized empty vaporizer

Thus prevention of drug administration errors could be useful for reducing intraoperative awareness. Discussion of drug administration errors and strategies for prevention are beyond the scope of this chapter, and readers are referred to previous publications.

Many, if not most, cases of intraoperative awareness occur without a specific error in drug administration and are probably related to an unusually large anesthetic dose requirement, due to either lower than average sensitivity to one or more drugs or faster than average clearance of one or more drugs. Large variation between individuals in anesthetic drug effect or anesthetic drug clearance is well-documented for a variety of anesthetic drugs. Identification of higher risk individuals in advance and administration of larger doses of anesthetic to these individuals might reduce the rate of intraoperative awareness. Unfortunately, a practical clinical method for identifying such individuals does not currently exist.

Patients receiving nondepolarizing muscle relaxants during the maintenance phase of anesthesia may be at greater risk of intraoperative awareness, presumably because they may not be able to move as readily, thereby giving a clue to the anesthesiologist that the anesthetic depth is inadequate. Some anesthesiologists take the approach of using as small a dose of muscle relaxant as possible to provide surgical exposure, with the idea that if patients are too lightly anesthetized they will still be able to move. This practice probably makes sense, although it is clear from case reports that patients may not move during an episode of intraoperative awareness even in the absence of neuromuscular blocking drugs.

Another option could be to give all patients very large doses of anesthetic drugs that would be adequate for even the least sensitive patient. The drawbacks to this approach are numerous, including cost, the potential for slow wake up, and cardiovascular side effects, not to mention that there are no data that show what dose of anesthetic drug would be large enough to prevent intraoperative awareness under every circumstance in every patient.

Likewise, no particular drug has ever been shown to be uniquely reliable for preventing awareness in every circumstance in every patient; intraoperative awareness has been reported in patients receiving apparently adequate doses of almost every possible anesthetic agent. The available evidence suggests that total intravenous anesthesia has the same risk of intraoperative awareness as inhalational anesthesia.

Finally, there is the option to somehow monitor the depth of anesthesia and titrate anesthetic drugs accordingly. Hypothetically, such an approach might prevent intraoperative awareness by identifying the patients who require larger doses of anesthetic drugs. The rest of this chapter will focus on this last approach.

Evidence

Electroencephalography (EEG) has been the most widely applied technology for measuring anesthetic depth. Auditory evoked potentials have also been used either alone or in combination with EEG. For a comprehensive review of the methodology of EEG and auditory evoked potentials to measure anesthetic depth, the reader is referred to previous publications.

Although it may seem reasonable that depth of anesthesia monitoring would reduce the incidence of intraoperative awareness, that outcome was certainly not assured. The opposite hypothesis was entertained by some—that depth of anesthesia monitoring would actually increase the incidence of intraoperative awareness because numerous studies had previously shown that, on average, patients received less anesthetic drug when monitored with an EEG depth of anesthesia monitor.

Four studies have suggested that intraoperative monitoring with EEG (specifically, the bispectral index [BIS] monitor) can significantly reduce the incidence of intraoperative awareness ( Table 43-1 ). The first was a retrospective case-comparison study of 5057 consecutive BIS-monitored patients from two hospitals in Sweden compared with 7826 non–BIS-monitored patients from the same institutions. Two cases of intraoperative awareness occurred in the BIS-monitored series compared with 14 in the non–BIS-monitored case-matched control group. This difference was statistically significant ( p < 0.039).

The second study was a prospective, randomized, international multicenter trial of 2463 patients at high risk of intraoperative awareness (e.g., cardiac, trauma, obstetric patients) assigned randomly to BIS or non-BIS groups (the so-called B-AWARE trial). High-risk patients were chosen for this trial for the purpose of increasing the statistical power of the study. Two cases of intraoperative awareness occurred in the BIS-monitored group compared with 11 in the non–BIS-monitored group. Again, the difference was statistically significant ( p = 0.022).

Avidan’s group has published two prospective, randomized trials that compared two interventions intended to reduce the incidence of awareness: BIS monitoring (target BIS range, 40 to 60) and analysis of targeted end-tidal anesthetic gas concentration (ETAC) (target range, 0.7 to 1.3 minimum alveolar concentration [MAC], with gas analyzers audibly alarmed at these limits). The first of these trials was a single center study involving approximately 2000 patients, and the second was a multicenter study involving approximately 5800 patients. The patients were required to be at “high risk” of intraoperative awareness, estimated to be perhaps 1%, based on a specific set of criteria. Approximately 25% to 30% of the patients underwent cardiac surgery. BIS and ETAC data were collected for both groups, but BIS values were not visible in the operating room for the ETAC group. Patients were assessed for intraoperative awareness three times, at 0 to 24 hours, 24 to 72 hours, and 30 days after extubation. Classification of no awareness, possible awareness, or definite awareness was made by a panel of reviewers unaware of monitoring allocation.

There were 13 cases of definite awareness in the two studies by Avidan et al combined, which yielded an overall incidence of 0.17%. There were nine cases of definite awareness in the BIS-monitored groups and four cases in the ETAC groups. No significant difference was found in the incidence of definite awareness between BIS and ETAC groups in either study. Among the patients with definite awareness in the BIS-monitored groups, four of the patients had some BIS values greater than 60 (BIS values less than 60 are generally considered to be desirable for the purpose of avoiding intraoperative awareness), and five of the patients did not have any BIS values greater than 60. All patients with definite awareness in the ETAC groups had some ETAC values less than 0.7 MAC (although there were patients with “possible awareness” without any ETAC values less than 0.7 MAC). Interpretation of the data is complicated by substantial amounts of missing ETAC and BIS data. Six of the BIS-monitored patients with definite awareness had epochs of missing BIS data lasting as long as 90 minutes. No explanation for the missing data was provided. One cannot help but wonder whether intraoperative awareness may have occurred during an epoch of missing BIS data in the BIS-monitored patients, and whether the availability of BIS data would have enabled the anesthesia providers to prevent awareness in these patients. The argument can be made that no monitoring device is able to provide usable data under all circumstances, and the prevalence of missing data contributes (negatively) to the overall performance and usefulness of any monitor. Nevertheless, it would be very valuable to distinguish intraoperative awareness that occurs with BIS values in the target range (less than 60) from intraoperative awareness that occurs in the absence of usable BIS data.

The B-AWARE trial by Myles and colleagues was a comparison of BIS monitoring with “standard practice” in high-risk patients. The standard practice group had an incidence of awareness of approximately 1%, which was the expected incidence, compared with approximately 0.2% in the BIS-monitored group, which was a statistically significant difference in favor of BIS monitoring. The studies by Avidan and colleagues were not a comparison of BIS monitoring with standard practice; rather, they were a comparison of BIS monitoring with another intervention in which practitioners were instructed to keep ETACs within a particular range with the use of gas monitor audible alarms set to activate when the concentrations were outside the prescribed range. Given that the expected incidence of awareness in the studies by Avidan and colleagues was approximately 1% (as estimated by the authors), and the observed overall incidence of definite awareness was less than 0.2% with BIS monitoring or ETAC, one could conclude that BIS monitoring and targeted ETAC analysis were similarly effective in reducing the expected incidence of intraoperative awareness. Unfortunately, Avidan and colleagues did not have a true standard practice control group for comparison, so one cannot know with certainty what the incidence of intraoperative awareness would have been in their patients without either BIS monitoring or targeted ETAC analysis.

It may be instructive to look more closely at patients who have had intraoperative awareness, despite the use of a BIS monitor. In the Swedish case-control study, two BIS-monitored patients had intraoperative awareness, both of which occurred during intubation, with a BIS value greater than 60. In the first multicenter randomized prospective trial (B-AWARE), two BIS-monitored patients had intraoperative awareness, one during laryngoscopy with a BIS value of 79 to 82 and one during cardiac surgery with a BIS value of 55 to 59. In this later case, intraoperative awareness occurred despite BIS values in the recommended range. In the studies by Avidan et al, two patients with intraoperative awareness had a complete record of BIS data (no missing data) and no BIS values greater than 60. Despite the possibility that intraoperative awareness can occur with a BIS value less than 60, the use of BIS resulted in reduction of the incidence of intraoperative awareness from about 1% (either an actual measured incidence as reported by Myles and colleagues or an expected incidence from Avidan et al ) to about 0.2% in both the Myles and Avidan studies, which suggests that BIS is useful.

Given that intraoperative awareness can occur at a BIS value less than 60 (or ETACs greater than 0.7 MAC), it is important to use the traditional methods of detecting light anesthesia (e.g., movement and vital signs) and give reasonable doses of anesthetic drugs, regardless of the BIS value—those who understand BIS technology have never seriously suggested otherwise. As a general principle, the wise practitioner realizes that no monitoring device, single number, or data point should be used as the sole guide to patient care.

Intraoperative awareness appears to be less likely at depth-of-anesthesia monitoring values in the recommended range (e.g., less than 60 for BIS); however, it is evidently possible for BIS values to exceed the recommended range without the occurrence of intraoperative awareness. In addition, the sufficient conditions to produce intraoperative awareness are not known. The Swedish case-control study reported the distribution of BIS values greater than 60, as found in 5057 consecutive BIS-monitored patients. They found average times with BIS values greater than 60 to be 1.9 minutes during induction of anesthesia (range, 0 to 10 minutes) and 2.0 minutes during maintenance (range, 0 to 178 minutes). As noted previously, only two of these patients had intraoperative awareness.

There have been very few published case reports of individual patients with intraoperative awareness in the presence of BIS values in the recommended range, that is, less than 60. In two published case reports of purported intraoperative awareness with BIS values less than 60, the BIS data were taken retrospectively from an anesthesia record, not from the continuous record stored in the memory of the monitor. Because BIS values are recorded intermittently on a handmade anesthesia record, it is possible that the BIS values pertinent to the episode of intraoperative awareness may not have appeared on the anesthesia record. In the instance of one of the case reports, when the complete record was obtained at a later time from the flash memory of the monitor, there were substantial time periods with BIS values greater than 60 that were not recorded on the anesthesia record.