Accidental awareness during general anesthesia

INTRODUCTION — The term "accidental awareness during general anesthesia" (AAGA) encompasses both intraoperative consciousness and later explicit recall of intraoperative events. Although the incidence of AAGA may be reduced with preventive measures, it may not be eradicated completely. However, AAGA with recall is specifically associated with use of neuromuscular blocking agents (NMBAs), and is virtually unknown in patients who are not paralyzed. Thus, while all patients undergoing general anesthesia should be informed that AAGA is rare but can occur, specific focus should be on those who will receive NMBAs.

This topic will discuss incidence, risk factors, prevention, recognition, and management of AAGA.

INCIDENCE AND CHARACTERISTICS OF AWARENESS — The incidence of AAGA varies widely due to methodologic differences in postoperative assessment of awareness and differences in anesthetic practice [1,2]. The key difference in reported rates depends on whether specific questions were asked in the postoperative interview (eg, "what was the last thing you remember before anesthesia, then the next thing, and do you remember anything in-between?"), rather than relying only on patients' self-reporting of AAGA [3-5]. Most AAGA events occur during induction and emergence, rather than during the maintenance phase [5]. Experiences range from isolated auditory perceptions to being fully awake, immobilized, and in pain.

Large, prospective, multicenter studies of adults undergoing surgery with general anesthesia in North America and Europe have consistently reported an incidence of one to two AAGA occurrences per 1000 patients who were specifically queried [6-11]. Studies relying on retrospective review of patient-initiated reports or surveys of anesthesiologists note a much lower incidence of approximately 1 in 14,000 [5,11-15]. The largest such review relying on patient self-reporting is the fifth National Audit Project (NAP5), which reported an incidence of 1 in 19,000 occurrences of AAGA in an estimated 3 million general anesthetics performed across the United Kingdom during 2013 [5,13,14,16]. Incidence was much higher in obstetric patients (1 in 600). The incidence of self-reported awareness in pediatric patients has been estimated to be approximately 1:40,000 [15]. AAGA events occur more frequently if a neuromuscular blocking agent (NMBA) is used during the anesthetic [5,6,16,17] (see 'Neuromuscular blockade' below). Also, total intravenous anesthesia (TIVA) has been associated with a higher risk compared with techniques based on a volatile inhalation anesthetic agent [5,18,19]. (See 'Total intravenous anesthesia' below.)

RISK FACTORS

Neuromuscular blockade — The most important risk factor for AAGA is the use of a neuromuscular blocking agent (NMBA) [5,6,17]. Findings from the 5^th National Audit Project (NAP5) in the United Kingdom confirmed that the incidence of AAGA was approximately 1 in 8000 if an NMBA was administered, compared with approximately 1 in 136,000 when no NMBA was used [5].

This association is understandable. Muscle paralysis removes one of the physiologic signs of patient awareness (ie, purposeful movement). Since complete paralysis worsens the psychologic trauma of an AAGA experience, even more so than pain, the potential for long-term psychologic sequelae may be increased [10,17,20,21]. Explanations of this include the fact that pain is an experience recognized by all; although unpleasant, the patient has a pre-existing understanding of it. In contrast, whole body paralysis is an extremely rare human experience that is described as panic due to being "buried alive," or "being dead" [22]. In fact, the need for general anesthesia is determined, in part, by the need to prevent awareness of the distress of paralysis.

The risks posed by use of NMBAs are exacerbated by:

●Failure to employ monitoring of neuromuscular blockade. (See "Monitoring neuromuscular blockade".)

●Inadequate reversal of neuromuscular blockade with sugammadex or neostigmine [23]. Full reversal of neuromuscular blockade reduces potential for AAGA, as well as aiding recovery of respiratory function, as discussed in other topics [24,25]. (See "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Reversal of neuromuscular block' and "Clinical use of neuromuscular blocking agents in anesthesia", section on 'Avoidance of residual neuromuscular blockade'.)

Total intravenous anesthesia — Total intravenous anesthesia (TIVA) is associated with a higher risk for AAGA compared with techniques based on a volatile inhalation anesthetic agent [5,18,19]. A survey of 172 anesthesiologists in Japan that included 85,156 anesthetics noted TIVA had been used in 21 of 24 patients who had definite (n = 14) or possible (n = 10) awareness [18]. However, most of the 85,156 patients included in the survey received a volatile inhalation agent rather than TIVA alone.

Increased risk with use of a TIVA technique is likely due to lack of availability of monitors of blood concentration of intravenous (IV) anesthetic agents, which may lead to underdosing. This is in contrast to inhalation anesthetics, for which continuous monitoring of end-tidal (ie, exhaled) anesthetic concentration (ETAC) allows real-time dose adjustments [10,11] (see 'End-tidal anesthetic concentration' below). During TIVA, detection of AAGA has relied in part on patient responses to noxious stimuli to indicate insufficient anesthetic depth. However, if an NMBA has been administered, movement may be prevented.

Furthermore, disconnection of the IV tubing from the IV catheter or subcutaneous infiltration of the IV catheter may occur when the limb containing an IV catheter is tucked at the patient's side or not continuously visible due to surgical draping, with resultant failure to deliver the intended medication. (See "Intravenous infusion devices for perioperative use", section on 'Extravasation of a continuous infusion'.)

Dosing errors are among the most common anesthesia-related medication errors [26].Use of a smart IV infusion pumps can theoretically prevent pump programming errors, although it is still possible for a clinician to administer the wrong medication or wrong concentration (see "Intravenous infusion devices for perioperative use", section on 'Smart pumps' and "Intravenous infusion devices for perioperative use", section on 'Risks for medication errors'). Similarly, use of target-controlled infusion (TCI) devices may reduce risk of dosage miscalculations that can result in inadequate anesthesia (figure 1) [27]. A TCI device was not used to administer TIVA for any of the patients who developed AAGA in a large survey of anesthesiologists [18]. However, no study has conclusively demonstrated that use of TCI lowers risk of AAGA during TIVA. Furthermore, TCI technology is not approved by the US Food and Drug Administration (FDA) for use in the United States. (See "Intravenous infusion devices for perioperative use", section on 'Target-controlled infusion systems'.)

Brain monitoring such as processed electroencephalography (EEG) monitors (eg, bispectral index [BIS]) may improve clinical ability to ensure that general anesthesia has produced unconsciousness during use of TIVA (or other anesthetic techniques), and to avoid oversedation [18,19,28,29]. (See 'Brain monitoring' below.)

Other developing technology may allow prediction of propofol concentrations in serum by determining its concentration in exhaled air, but such methods are not widely available [30]. (See "Intravenous infusion devices for perioperative use", section on 'Likely future developments'.)

Surgery-related risk factors — As noted above, surgical procedures associated with higher risk for AAGA typically involve use of NMBAs (see 'Neuromuscular blockade' above). Patients undergoing certain procedures (eg, trauma and emergency surgical procedures, cardiac surgery with cardiopulmonary bypass [CPB], cesarean delivery) are at particularly high risk [10,17,31].

For trauma or emergency surgery, the time interval between anesthetic induction and surgical incision is necessarily brief (eg, due to the need to control hemorrhage). Also, anesthetic depth may be deliberately reduced to limit adverse consequences of hemodynamic instability. In some institutions, an additional factor for after-hours emergency or trauma surgery is the need for rapid sequence induction and intubation (RSII) that may be performed by a junior anesthesia team member [5,32]. Use of thiopental during anesthetic induction, and not using an opioid (eg, fentanyl) during induction were other factors in such cases. Specific considerations for these cases are discussed in separate topics. (See "Anesthesia for adult trauma patients", section on 'Strategies to minimize risk of awareness' and "Rapid sequence induction and intubation (RSII) for anesthesia".)

During cardiac surgery, use of CPB may alter pharmacokinetics and pharmacodynamics of anesthetic agents, which may lead to lighter depth of anesthesia than anticipated and increased risk for AAGA. (See "Management of cardiopulmonary bypass", section on 'Maintenance of anesthesia and neuromuscular blockade'.)

Cesarean section is another procedure associated with higher risk for AAGA [33-35]. In one study that employed standardized interviews to detect AAGA, the incidence was 1 in 256 patients [35]. Factors that have been associated with AAGA after cesarean section include the need for RSII that may be performed by a junior anesthesia team member, difficult airway due to airway changes associated with pregnancy and/or obesity (see 'Patient-related risk factors' below), the brief time interval between anesthetic induction and surgical incision (due to the need to immediately deliver the neonate), and use of thiopental or ketamine to induce general anesthesia [5,32-36]. (See "Anesthesia for cesarean delivery", section on 'General anesthesia'.)

Patient-related risk factors

●Difficult intubation – Difficulty with intubation in any surgical procedure may lead to AAGA, presumably due to insufficient anesthesia during prolonged intubation attempts [17]. Maintenance of adequate anesthesia during attempts is addressed elsewhere [5,37]. (See "Management of the difficult airway for general anesthesia in adults", section on 'Planning the anesthetic approach' and "Management of the difficult airway for general anesthesia in adults", section on 'Induction of anesthesia'.)

If "awake intubation," is planned for a patient with a potentially difficult airway, it is critically important to explain the planned sedation technique to avoid patient expectation for complete amnesia prior to induction of general anesthesia [5]. (See 'Management of patient expectations' below and "Management of the difficult airway for general anesthesia in adults", section on 'Awake intubation'.)

●Obesity – The association of obesity with a higher incidence of AAGA may be related to difficulty with intubation [36,38]. (See "Anesthesia for the patient with obesity", section on 'Difficulty with airway management'.)

●Resistance or tolerance to anesthetics – Patients with a history of AAGA may be at higher risk for future events. One observational study noted an adjusted fivefold higher incidence in patients with a previous AAGA event [39].

•Genetic resistance to anesthetic agents – The NAP5 study reported that 1 in 20 AAGA patients either had a previous event or genetic relative(s) with such event(s) [5]. Genetic variations may result in resistance to hypnotic or amnesic actions of certain anesthetic agents, although such variations have not been identified in humans [40].

•Acquired tolerance to anesthetic agents may occur:

-Chronic ethanol use – Cytochrome P450 2E1 is induced by alcohol and isoniazid [41]. Since the inhalation agents, benzodiazepines, and opioids are metabolized to varying degrees by the cytochrome P450 category of enzymes, patients with habitual alcohol intake may require higher doses of these agents. However, acute alcohol intoxication is typically associated with decreased anesthesia requirements since alcohol is a central nervous system (CNS) depressant. Details regarding recognition and perioperative management of excessive alcohol use are described separately. (See "Identification and management of unhealthy alcohol use in the perioperative period".)

-Chronic opioid use – Opioid tolerance and/or opioid-induced hyperalgesia may increase opioid dosing requirements in the perioperative period. Inadequate analgesia during a surgical procedure results in a higher level of cortical stimulation due to pain, and may increase risk of AAGA [42,43]. Details regarding anesthetic management of patients with acute or chronic opioid use are discussed separately. (See "Perioperative uses of intravenous opioids in adults: General considerations", section on 'Acute opioid intoxication' and "Perioperative uses of intravenous opioids in adults: General considerations", section on 'Chronic opioid use'.)

-Chronic benzodiazepine use – There is an increased expression of excitatory glutamatergic receptors upon benzodiazepine withdrawal after chronic exposure, which could underlie the symptoms observed in benzodiazepine withdrawal syndrome. This overexpression of receptors could theoretically increase anesthetic requirement and risk of AAGA. (See "Benzodiazepine use disorder", section on 'Physical dependence'.)

-Other prescription and nonprescription medications – Several prescription and nonprescription medications induce cytochrome P450 3A (eg, efavirenz, nevirapine, barbiturates, carbamazepine, glucocorticoids, phenytoin, rifampicin, St. John's wort), which is involved in the metabolism of opioids [41]. Similar to chronic opioid use, chronic administration of these medications may increase opioid dosing requirements [44].

●Pediatric patients – Although some reports have suggested that the incidence of AAGA may be slightly higher in children (between 0.2 to 1.2 percent) [45], the NAP5 report noted a negligible incidence in children [5,46]. Assessing the presence of AAGA in children is particularly challenging due to age-related developmental factors and questionable accuracy of postoperative interviews [47,48].

Technical issues — Malfunction or misuse of the anesthesia machine is an infrequent primary cause of AAGA [6]. Problems with anesthetic machines are generally detected by appropriate alarms and machine checks [49]. (See "Anesthesia machines: Prevention, diagnosis, and management of malfunctions".)

During a TIVA technique, malfunction or misuse of the IV infusion pumps can lead to failure to deliver the intended anesthetic agent or underdosing with possible awareness [27]. Also, disconnection of the IV tubing from the IV catheter or subcutaneous infiltration of the IV catheter may occur when the limb containing an IV catheter is tucked at the patient's side table or not continuously visible due to surgical draping, with resultant failure to deliver the intended medication. In such cases, brain monitoring (eg, with processed EEG [BIS]) may aid in recognition of inadequate anesthetic depth. (See 'Brain monitoring' below and "Intravenous infusion devices for perioperative use", section on 'Risks for medication errors'.)

Human factors are more common than technology failure in errors in anesthetic administration or dosing [50] (see "Prevention of perioperative medication errors", section on 'Types and incidence of errors'). Examples of human error include:

●During administration of IV anesthetic agents, a syringe swap may result in administration of an NMBA before the anesthetic induction agent [50]. Also, it is possible for a clinician to administer the wrong concentration of an IV anesthetic even when a smart IV infusion pump is used due to miscalculation or programming error. (See "Prevention of perioperative medication errors", section on 'Intravenous medication errors' and "Intravenous infusion devices for perioperative use", section on 'Programming errors'.)

●During administration of a volatile inhalation anesthetic agent, it is possible to forget to turn on a vaporizer. However, the low inspired anesthetic agent alarm should sound (table 1). (See "Anesthesia machines: Prevention, diagnosis, and management of malfunctions", section on 'Vaporizer malfunction'.)

PREVENTION

Management of patient expectations — In the preanesthetic consultation, patients should be provided with brief written information as well as a verbal explanation regarding their anesthetic plan and what to expect. In particular, patients who report prior episodes of AAGA may be at higher risk; thus, reassurance regarding additional dosing and fastidious monitoring is appropriate [51]. (See 'Patient-related risk factors' above.)

The majority of AAGA incidents occur during induction and emergence, rather than during the maintenance phase of general anesthesia [5,35]. The concept of "neural inertia," (ie, resistance to changes in consciousness induced by general anesthesia) may explain partial memory during these periods in some patients [52]. During the preanesthetic consultation, explaining that the patient may recall experiences such as "a mask on face," "tube in mouth," or "weakness as you wake up," will help to manage expectations and mitigate distress if any such memories are retained.

Notably, patients frequently do not understand the difference between conscious sedation with monitored anesthesia care versus general anesthesia (table 2 and table 3) [53]. In the fifth National Audit Project (NAP5) conducted in the United Kingdom, a striking feature was that reported AAGA was more common with intended sedation compared with intended general anesthesia [5]. This has also been reported after gastrointestinal procedures [54] and after dental surgery [55]. Thus, when sedation for monitored anesthesia care or awake intubation is planned, the consent process should emphasize the fact that the patient will be somewhat "awake," and "not under deep anesthesia", with an explanation that some recall is to be expected even if deep sedation is planned [56].

Monitoring — Prevention of AAGA involves ensuring continuous delivery of an appropriate anesthetic concentration [57]. Historical explanations suggesting that underdosing of anesthetic agents was the primary problem were oversimplified [21]. Development of risk recognition and mitigation strategies that involve monitoring techniques aid in ensuring adequate anesthetic depth while avoiding overdoses of anesthetic agents that may have life-threatening consequences (figure 2).

Three widely available monitoring modalities directly aid in prevention of AAGA:

●End-tidal anesthetic concentration (ETAC) of inhalation agents (when these agents are in use). (See 'End-tidal anesthetic concentration' below.)

●Neuromuscular blockade monitoring (when a neuromuscular blocking agent [NMBA] is administered). (See "Monitoring neuromuscular blockade".)

●Electroencephalographic (EEG) monitoring, which is typically a processed EEG such as the bispectral index (BIS) monitor, which is especially important when a total intravenous anesthetic (TIVA) technique is used [57]. (See 'Brain monitoring' below.)

Although not widely used or available in the United States, a target-controlled infusion (TCI) device is likely useful to avoid AAGA during administration of a TIVA technique [18,27]. The TCI device is designed to ensure that appropriate plasma or effect-site concentration(s)of anesthetic agents are maintained (see "Intravenous infusion devices for perioperative use", section on 'Target-controlled infusion systems'). Other monitors of awareness that are not widely used include the isolated forearm technique (IFT) and nociception monitors, as noted below. (See 'Other monitors and techniques' below.)

Of course, physiologic parameters such as heart rate (HR), blood pressure (BP), respiratory pattern, muscle tone, and purposeful movement in response to a noxious stimulus are always monitored [58,59]. However, HR and BP are unreliable for determining unconsciousness or the degree of anesthetic depth. Although tachycardia and/or hypertension may occur as a response to pain (indicating insufficient anesthetic depth), patients with hypovolemia or significant beta-adrenergic blockade often have minimal change in HR or BP. Conversely, sympathetic activation due to surgical stimuli may cause hypertension or tachycardia in some patients, even when anesthetic depth is adequate. Furthermore, purposeful movement as a sign of AAGA, as well as respiratory pattern and degree of muscle tone usually cannot be assessed in patients who have received an NMBA. (See 'Neuromuscular blockade' above.)

End-tidal anesthetic concentration — Continuous monitoring of ETAC when inhalation anesthetic(s) are administered may decrease risk for AAGA because this allows assessment of whether the concentration of an inhalation agent is close to the desired minimum alveolar concentration (MAC) value (ie, the exhaled concentration of the agent at which 50 percent of patients do not respond with movement to a noxious stimulus) [60]. The MAC requirement for movement suppression (mediated at the level of the spinal cord) is usually higher than MAC for suppression of consciousness (ie, MAC-awake) or memory (ie, MAC-amnesia), which provides a margin of safety to avoid AAGA [61]. Use of inhalation agents with low blood solubility (eg, sevoflurane and desflurane) allows relatively rapid adjustments in anesthetic concentration to achieve the desired anesthetic depth.

While ETAC <0.7 MAC for the selected anesthetic agent has been proposed as an appropriate alarm threshold level, it is not known whether hypnosis and amnesia are consistently achieved at concentrations ≥0.7 MAC [62-64]. Other important limitations of ETAC as a monitor of appropriate anesthetic depth include:

●MAC is associated with movement responses, which are diminished or absent if an NMBA has been administered. (See 'Neuromuscular blockade' above.)

●MAC varies among and even within individuals, and is dependent on age, pharmacogenetics, temperature, and other factors [65-68]. (See 'Patient-related risk factors' above and "Inhalation anesthetic agents: Clinical effects and uses", section on 'Sedation and anesthesia'.)

●MAC is not relevant for intravenous (IV) agents. Although administration of IV anesthetics such as propofol involves analogous concepts (ie, determining a plasma concentration at which 50 percent of patients respond [Cp50]), direct plasma concentrations are not available in most clinical settings [69]. Thus, TCI devices that use pharmacokinetic and pharmacodynamic mathematical modeling were developed to improve maintenance of the targeted concentration of an IV anesthetic agent at the effect site (typically the brain) [27]. (See 'Total intravenous anesthesia' above and "Intravenous infusion devices for perioperative use", section on 'Target-controlled infusion systems'.)

Brain monitoring — Brain monitoring such as processed EEG monitors (eg, BIS) may improve clinical ability to ensure that general anesthesia has produced unconsciousness during use of TIVA (or other anesthetic techniques) [18,19,28,70].

●Processed EEG – Processed EEG monitors translate the raw EEG signal from the time domain to the frequency domain via fast Fourier transformation [71,72]. The analysis of the contribution of each frequency band results in a frequency spectrum. Characteristics of this spectrum are then analyzed to assess anesthetic depth. Commercially available devices include the BIS, density spectral array, entropy, and cerebral state index monitors, with continuing development of new models (table 4) [72-75]. Threshold alarms for possible AAGA are set according to the device manufacturer's specifications. Evidence that any one monitor performs better than others is lacking, although most research involves use of the BIS monitor.

In one randomized study of 5228 patients receiving a TIVA anesthetic technique, AAGA incidence was lower in patients who received BIS-guided monitoring to a target of 40 to 60 compared with those in which BIS monitoring data was recorded but was not available to the anesthesiologist (0.14 versus 0.65 percent) [19]. A 2019 meta-analysis examined comparisons of BIS-guided anesthetic depth versus monitoring clinical signs only (eg, HR, BP, lacrimal tearing), and/or use of ETAC monitoring [76]. The incidence of AAGA was lower in patients with BIS monitoring (3 per 1000 versus 9 per 1000; odds ratio [OR] 0.36, 95% CI 0.21-0.60; 27 randomized trials with 9765 participants) compared with monitoring clinical signs. Furthermore, in some of these studies, BIS monitoring reduced time (by a few minutes) to reach indices of recovery such as eye opening, time to orientation, and time to discharge from the post-anesthesia care unit (PACU). However, the studies that compared BIS versus ETAC monitoring to guide anesthetic depth noted no differences in incidence of AAGA (five studies; 26,572 participants) [76]. Similar results had been noted in a 2014 meta-analysis comparing BIS monitoring with clinical signs only, or BIS versus ETAC monitoring [77].

Processed EEG monitors do provide information that supplements physiologic parameters to aid in assessing anesthetic depth, and may be particularly important during TIVA techniques since ETAC monitoring is unavailable. Although a general relationship between ETAC and BIS values has been derived using data obtained from a large clinical trial, BIS frequently correlates poorly with ETAC, is insensitive to changes in ETAC, and exhibits wide interindividual variability [78,79]. Furthermore, BIS values are influenced by electrical activity generated by the muscles, which is eliminated in patients with complete neuromuscular blockade. In one study in awake volunteers who received an NMBA but no anesthesia, BIS values declined from baseline values of approximately 100 to values as low as 40 [80]. Another limitation of most processed EEG monitoring studies has been statistical modelling as if the monitor is a "therapy," rather than a diagnostic modality. Studies that model processed EEG as a diagnostic tool that determine sensitivity, specificity, and receiver operating curves reveal that such brain monitoring cannot distinguish between awake versus anesthetized states [81]. Thus, processed EEG monitors do not reliably confirm that a patient is unaware, and are not consistently useful to guide anesthetic titration or assess adequacy of anesthetic depth and subsequent recovery [82,83]. Burst suppression episodes may not be detected by monitoring processed EEG data; assessment of raw EEG data is more accurate to identify this phenomenon [84,85].

Importantly, dosing of anesthetic agents in an attempt to maintain adequate anesthetic depth should not produce processed EEG values that are consistently very low, particularly if such dosing results in hypotension or blood pressure significantly lower than baseline values [86,87]. Maintenance of hemodynamic stability with preservation of vital organ function is ultimately a more important anesthetic goal since hemodynamic instability can lead to immediate harm [31]. (See "Perioperative neurocognitive disorders in adults: Risk factors and mitigation strategies", section on 'Avoid excessive depth during general anesthesia' and "Perioperative neurocognitive disorders in adults: Risk factors and mitigation strategies", section on 'Avoid extremes of blood pressure'.)

●Raw (unprocessed) EEG – Raw (unprocessed) EEG monitoring provides more useful information regarding anesthetic depth compared with processed EEG indices, and provides some evidence of an anesthetic state even after administration of an NMBA [83,88,89]. When a patient is awake, the raw EEG typically has lower-power, higher-frequency beta activity (ie, 20 to 30 Hz), and does not have persistent spindles in theta (4 to 8 Hz) or alpha (8 to 12 Hz) frequencies [58]. The raw EEG waveform during general anesthesia varies depending on the classes and combinations of anesthetic agents administered. With administration of propofol or a volatile inhalation anesthetic (eg, desflurane, sevoflurane, isoflurane) at general anesthetic concentrations, the EEG waveform typically has a slow delta pattern (0 to 4 Hz), often coinciding with persistent spindles in theta or alpha frequencies. These EEG markers of unconsciousness can help distinguish between patient responses to noxious stimuli that produce only autonomic activity versus those that produce both an autonomic response and arousal [58]. (See "Neuromonitoring in surgery and anesthesia", section on 'Electroencephalography' and "Neuromonitoring in surgery and anesthesia", section on 'Anesthetic effects on neuromonitoring'.)

However, use and interpretation of raw rather than processed EEG monitoring is impractical in many settings since the technology is not available in all operating rooms, is unfamiliar to many anesthesiologists, and is susceptible to artifact [77,88].

Other monitors and techniques

Isolated forearm technique — The IFT, although not widely used, is another monitoring technique developed to provide supplemental information to monitor for awareness. One forearm is isolated from the systemic circulation using a cuffed upper arm tourniquet inflated after anesthesia induction, but before administration of any NMBA [90]. Theoretically, an awake patient can move their hand/fingers to alert the caregivers. Interestingly, in a 2015 systematic review that included 1131 patients, 31 percent of the patients moved in response to verbal command, but no patient ever moved spontaneously during surgery, and no patient had AAGA [91]. In another study of 100 consecutive patients who did not receive an NMBA, none moved in response to verbal command, although 24 did move during surgery (and required additional dosing of anesthetic agent) [92]. These results have led to speculation that a specific state termed 'dysanesthesia' may occur, in which sensation is uncoupled from perceptions [93,94]. Patients in such a dissociated state may have no recognition or pain due to the surgery (of which they remain technically aware) but can still attend to their name being called as part of a verbal command. Such studies emphasize the distinction between awareness at the time of surgery versus awareness with postoperative recall of intraoperative events. Thus, although IFT monitoring is the only way in which intraoperative awareness can be contemporaneously detected, such monitoring may not be useful to assess wakefulness in the absence of neuromuscular blockade.

Nociception monitors — Nociception monitors that measure sympathetic activation in response to nociception (eg, surgical plethysmographic index [SPI], nociception level index [NOL], analgesia-nociception index [ANI], pupillary pain index [PPI], and pupillometry [measurements of pupillary diameter]) have been developed to guide administration of analgesics (eg, opioids) and sedatives (eg, propofol), and potentially avoid "too light" anesthetic depth that could result in AAGA [95]. Some studies using these monitors have noted lower analgesic or sedative doses or shorter emergence times in the groups with nociception monitoring compared with control groups monitored with clinical signs only [95-98]. However, nociception monitors are not widely available, and there is no evidence that AAGA is preventable with their use.

Avoidance of complete neuromuscular blockade — Avoiding complete neuromuscular blockade when possible may minimize the risk of AAGA. Notably, severe postoperative distress and long-term psychologic harm are most frequently associated with sensations of paralysis rather than pain. (See 'Neuromuscular blockade' above.)

Complete neuromuscular blockade is not necessary for many surgical procedures [21]. Some degree of muscle relaxation facilitates certain procedures (eg, thoracic and abdominal surgery). However, in many cases, partial skeletal muscle paralysis (frequently assessed with a peripheral nerve stimulator) can be used to improve operating conditions for the surgeon while still allowing some purposeful movement to minimize the risk of AAGA.

Avoidance of misuse of technical equipment — During administration of inhalation anesthetic agents, meticulous adherence to published guidelines for a pre-use anesthesia machine and equipment checkout would prevent most episodes of accidental delivery of inadequate anesthesia, as well as other critical incidents due to anesthesia workstation misuse or failure, including failure to deliver adequate levels of anesthesia. Details are available in another topic. (See "Anesthesia machines: Prevention, diagnosis, and management of malfunctions".)

During administration of bolus doses of IV anesthetics, incorrect dosing or administration of the wrong drug may result in underdosing, overdosing, or failure to administer an appropriate amount of the desired anesthetic agent. Possible solutions include standardization and improvements in technology, as described in detail in a separate topic. (See "Prevention of perioperative medication errors", section on 'Intravenous medication errors' and "Prevention of perioperative medication errors", section on 'Technology solutions'.)

During infusion of IV anesthetic agents, possible awareness can result from failure to deliver the intended agent or anesthetic underdosing, although overdosing can also occur. Although use of smart pumps may reduce risk, safety and efficacy depends primarily on appropriate use of the medication library incorporated within the pump for (rather than manual programming that is possible to bypass these safeguards). Avoidance of errors in anesthetic delivery is discussed in other topics. (See "Intravenous infusion devices for perioperative use", section on 'Risks for medication errors'.)

Administration of adjuvant medications — If an episode of AAGA is suspected, adverse consequence may be limited by administering an adjuvant sedative or analgesic agent as soon as possible, as well as deepening anesthesia if appropriate [57]. Typical adjuvant agents include:

●Benzodiazepines – A benzodiazepine may achieve an amnestic effect. The extent to which a preoperative dose of a benzodiazepine (eg, 2 mg IV midazolam) may prevent AAGA or its consequences is unknown.

●Opioids and other analgesics – Analgesic agents may minimize pain in response to noxious stimuli during and immediately after an AAGA event [99-101]. Thus, it is often appropriate to administer an opioid dose if AAGA is suspected, although large opioid doses are avoided [102].

Professionalism — To reduce risk for extreme distress due to later recall of intraoperative events, the operating room staff should maintain professionalism in all discussions regarding the patient, regardless of whether the patient is assumed to be unconscious.

POSTOPERATIVE RECOGNITION AND MANAGEMENT — Recognizing that AAGA may have occurred is important so that discussion, reassurance, and further counseling can be offered to mitigate distress. Psychologic symptoms occur in 33 to 69 percent of adult patients who experience AAGA, but severity is variable [6,7,22,59,103-109]. Patients may experience significant symptoms similar to other causes of posttraumatic stress such as recurrent dreams or nightmares, sleep disturbance, intrusive thoughts, flashbacks, exaggerated startle response, hypervigilance, or avoidance of stimuli associated with the trauma. However, approximately half of patients reporting AAGA may not experience distress [22].

Early emotional responses after an awareness event are risk factors for later development of more severe post-event psychologic disturbances such as posttraumatic stress disorder (PTSD) [22,103,109]. Distress and long-term psychologic harm were most frequently associated with sensations of paralysis rather than pain (see 'Neuromuscular blockade' above), and the combination was particularly distressing [22]. A subgroup of patients resort to formal complaint or register malpractice claims. In a closed claims analysis of cases of awake paralysis, a judgment of substandard clinical care (eg, administration of a neuromuscular blocking agent [NMBA] before or without general anesthesia due to syringe swaps or labelling error) was rendered in a higher percentage of cases of awake paralysis compared with cases of AAGA without paralysis (94 versus 43 percent) [105].

Importantly, contemporaneous intraoperative recognition by the anesthesiologist that AAGA is possible may be helpful if discussion and reassurance regarding this possibility are offered in the postoperative period to mitigate distress [22]. Regardless of whether intraoperative AAGA is suspected, an explicit query about AAGA in the immediate postoperative period and during the anesthesiologist's follow-up visit is important [3,59,103]. However, signs and symptoms of distress may be delayed, and patients may not talk to their clinicians about AAGA unless specifically asked. Thus, after the patient is discharged from the hospital, the surgeon and primary care clinician should be vigilant in identifying potential early signs of a psychologic disturbance.

Analysis of the psychologic impact of AAGA in the NAP5 study has led to specific recommendations for a support pathway, consisting of three broad steps (figure 3) [56]:

●The "meeting," stage, during which the anesthesiologist involved in the patient's care, along with others, listen carefully to patient’s story to detail and understand their experience. The default position is to accept the patient's story as their genuine experience and express regret that the event has happened (this does not constitute an admission of liability). This meeting is an opportunity to start arrangements for consult with a local clinical psychologist or psychiatrist.

●The "analysis," stage, where details of the case are examined to seek a cause for the patient’s report and determine whether possible or probable AAGA occurred. This may involve review of charts and records, as well as staff interviews. The NAP5 study advises recording the degree of likelihood that the AAGA is genuine based on the report and clinical context, as well as seeking an external expert review.

●The "support," stage seeks early detection of clinical impact in the first 24 hours. The four cardinal signs are flashbacks, nightmares, new anxiety states, and depression. Then, active follow-up occurs at two weeks. If clinical impact is present and persists, formal referral to psychiatric/psychologic services.

Further details regarding assessment and treatment of PTSD are available in separate topics:

●(See "Posttraumatic stress disorder in adults: Epidemiology, pathophysiology, clinical features, assessment, and diagnosis".)

●(See "Posttraumatic stress disorder in adults: Treatment overview".)

●(See "Posttraumatic stress disorder in adults: Psychotherapy and psychosocial interventions".)

PUBLISHED GUIDELINES — Published guidelines regarding AAGA have been developed, although these are primarily based on expert opinion.

The American Society of Anesthesiologists (ASA) 2006 guidelines state that the anesthesiologist should [31]:

●Identify risk factors for AAGA. Patients are informed about this possibility, particularly those with risk factors. (See 'Risk factors' above.)

●Use a checklist protocol for anesthesia machines and equipment (table 5). (See "Anesthesia machines: Prevention, diagnosis, and management of malfunctions", section on 'Standardized anesthesia machine checkout' and "Anesthesia machines: Prevention, diagnosis, and management of malfunctions", section on 'American Society of Anesthesiologists Anesthesia Machine Checkout'.)

●Use brain monitoring such as processed electroencephalography (EEG) in patients with risk factors for AAGA. However, attempts to maintain brain monitoring targets should not conflict with patient safety priorities (eg, maintenance of hemodynamic stability and vital organ protection). (See 'Brain monitoring' above.)

●Administer an intraoperative dose of benzodiazepine as an adjuvant agent if AAGA is suspected, as well as deepening anesthesia as appropriate. (See 'Administration of adjuvant medications' above.)

Guidelines from the 5^th National Audit Project (NAP5) in the United Kingdom differ somewhat [5]. The NAP5 study included made 64 recommendations, including seven for national regulations, 12 for hospitals, and 45 for individual clinicians. These include:

●Ensure that neuromuscular blocking agents (NMBAs) are managed appropriately, including avoiding total neuromuscular blockade when feasible, fastidious monitoring of the degree of blockade, and ensuring that reversal of blockade is complete. (See 'Neuromuscular blockade' above and 'Avoidance of complete neuromuscular blockade' above.)

●Use checklists, communication and alarms to ensure maintenance of anesthesia delivery. (See "Anesthesia machines: Prevention, diagnosis, and management of malfunctions", section on 'Standardized anesthesia machine checkout'.)

●Ensure that the consent process manages patient expectations. (See 'Management of patient expectations' above.)

●Use support pathways for patients who report AAGA [56]. (See 'Postoperative recognition and management' above.)

Guidelines from the Association of Anaesthetists and the Society for Intravenous Anaesthesia for use of total intravenous anesthesia (TIVA) include recommendations to [27] (see 'Total intravenous anesthesia' above):

●Use a target-controlled infusion (TCI) device for TIVA techniques (notably, these devices are not available in the United States). (See "Intravenous infusion devices for perioperative use", section on 'Target-controlled infusion systems'.)

●Use processed EEG or other brain monitoring whenever an NMBA is used with TIVA. (See 'Brain monitoring' above.)

●Maintain visibility of the peripheral intravenous (IV) or central venous catheter through which the IV infusion is being delivered. (See "Intravenous infusion devices for perioperative use", section on 'Extravasation of a continuous infusion'.)

Similar guidelines have been developed by professional societies and national organizations [70,110].

SUMMARY AND RECOMMENDATIONS

●Accidental awareness during anesthesia (AAGA) is rare (with an incidence of 1 to 2 events per 1000 anesthetics when patients are specifically queried regarding perioperative recall). Most AAGA events during general anesthesia occur during induction and emergence, rather than during the maintenance phase. Experiences range from isolated auditory perceptions to being fully awake, immobilized, and in pain. (See 'Incidence and characteristics of awareness' above.)

●Risk factors for AAGA include:

•Use of a neuromuscular blocking agent (NMBA). The incidence of self-reported AAGA is approximately 1 in 8000 when an NMBA is used compared with 1 in 136,000 when no NMBA is used. (See 'Neuromuscular blockade' above.)

•Surgical procedures associated with use of an NMBA, rapid sequence induction, or lighter anesthetic depth (eg, trauma and emergency surgical procedures, cardiac surgery with cardiopulmonary bypass [CPB], cesarean delivery). (See 'Surgery-related risk factors' above.)

•Patient-related risk factors (eg, difficulty with intubation, obesity, genetic resistance or acquired tolerance to anesthetic agents). (See 'Patient-related risk factors' above.)

•Total intravenous anesthesia (TIVA) technique. (See 'Total intravenous anesthesia' above.)

•Technical issues such as malfunction or misuse of the anesthesia machine or intravenous (IV) infusion pumps, which are most commonly related to human error. (See 'Technical issues' above.)

●Prevention of AAGA involves ensuring continuous delivery of an appropriate anesthetic concentration. Monitoring techniques developed to recognize and mitigate risk of inadequate anesthetic depth with AAGA, while avoiding anesthetic overdosing include:

•End-tidal anesthetic concentration (ETAC) monitoring during use of inhalation agents to assess whether the anesthetic dose is close to the desired minimum alveolar concentration (MAC) value for the selected agent, with alarms set to detect ETAC concentrations that are too low (eg, <0.7 MAC). (See 'End-tidal anesthetic concentration' above.)

•Brain monitoring with processed or raw electroencephalography (EEG)and use of a target-controlled infusion (TCI) device (if available) when a TIVA technique is used, particularly if an NMBA is administered. (See 'Brain monitoring' above.)

●Additional prevention strategies to avoid or minimize the consequences of AAGA include:

•Mentioning the small risk of AAGA during the consent process, with an explanation that brief recollections at induction and emergence are not uncommon. If awake intubation or sedation for monitored anesthesia care is planned (rather than general anesthesia), the consent process emphasizes that some recall is to be expected. (See 'Management of patient expectations' above.)

•Avoiding complete neuromuscular blockade when possible to minimize risk of AAGA. Interpret any movement during anesthesia as possible AAGA, and deepen anesthesia accordingly. Notably, severe postoperative distress and long-term psychologic harm are most frequently associated with sensations of paralysis rather than pain. (See 'Neuromuscular blockade' above and 'Avoidance of complete neuromuscular blockade' above.)

•Using of technical equipment to deliver anesthetic agents appropriately, including meticulous adherence to published guidelines for anesthesia machine checkout and use, and appropriate use of IV infusion devices. (See 'Avoidance of misuse of technical equipment' above.)

•Administering adjuvant sedatives or analgesics that may limit adverse consequences as soon as possible if an intraoperative AAGA event is suspected. (See 'Administration of adjuvant medications' above.)

•Maintaining professionalism during all intraoperative discussions even if the patient is assumed to be unconscious.

●Recognizing that AAGA may have occurred is important so that discussion, reassurance, and further counseling can be offered to alleviate any psychological symptoms similar to post traumatic stress (eg, recurrent dreams or nightmares, sleep disturbance, intrusive thoughts, flashbacks, exaggerated startle response, hypervigilance, avoidance of stimuli associated with the trauma). (See 'Postoperative recognition and management' above.)

ACKNOWLEDGMENTS — The editorial staff at UpToDate acknowledge Jeffrey H Silverstein, MD, now deceased, who contributed to an earlier version of this topic review.

The editorial staff at UpToDate also acknowledge George Mashour, MD, PhD, and Michael Avidan, MD, who contributed to an earlier version of this topic review.