Author: Igor Kagan
Institution: University of California, Los Angeles
Date: February 2008
Awareness and recollection of surgical events under general anesthesia is an adverse reaction that can lead to psychological disorders including posttraumatic stress disorder. It is estimated that around 0.1% to 0.2% of patients that undergo general anesthesia in the United States are aware of their surroundings and events at some point during their surgery. Awareness under general anesthesia cannot be fully prevented. A patient may become aware for various reasons including levels of anesthesia, drugs, inadequate monitoring devices, and anesthesiologist error. However, steps may be taken to reduce the risk of awareness. The BIS monitor appears to be the most promising way to reduce awareness. But, more research is needed to quantify this promising technology. Nurses and algorithms, procedural protocols, have been suggested to provide a check and balance in reducing the adverse effects of Awareness and patient stress.
Awareness and recollection of surgical events under general anesthesia is an infrequent but well described, phenomenon that may result in posttraumatic stress disorder (PTSD) (Sebel et al. 2004). The level of anesthesia may be inadequate from the beginning, or wear off during the operation. In such situations, a patient may feel the pain of surgery, pressure, hear conversations, and feel as if he or she cannot breathe. The patient may be unable to communicate any distress because they have been given a paralytic/muscle relaxant. Some patients become aware but do not feel pain or other unpleasant sensations(Sebel et al. 2004).
It is estimated that around 0.1% to 0.2% of patients that undergo general anesthesia in the United States experience Awareness at some point during their surgery (JCAHO 2004). With approximately 21 million adults and children (Davidson et al. 2005) undergoing general anesthesia each year in the U.S., Awareness is observed in 20,000 to 40,000 cases per year (JCAHO 2004). It is not influenced, however, by age and sex (Sebel et al. 2004). Present techniques in administering general anesthesia and traditional clinical monitoring modalities during anesthesia can not fully prevent Awareness (Sebel et al. 2004). This is partly because hypertension, tachycardia or end-tidal anesthetic concentration monitoring does not sufficiently help in predicting Awareness. The purpose of this paper is to analyze the reasons why Awareness occurs, why it may be difficult to detect it during surgery, how Awareness can be identified post-surgery and techniques to prevent Awareness.
A monitor that uses a processed electroencephalogram (EEG) derivative, the Bispectral Index® (BIS®; Aspect Medical Systems, Newton, MA) has been introduced into clinical practice for monitoring anesthetic effects on the brain (Johansen and Sebel 2000). It will be discussed as a possible means to prevent Awareness. After having analyzed the causes of Awareness, implementing strategies to reduce human error and introducing BIS monitoring has a theoretical potential to prevent Awareness, however, more research is needed to determine if a reduction in Awareness will truly result.
We searched PubMed (NCBI, National Library of Medicine) online database Medline using "awareness under general anesthesia", "BIS monitoring", "BIS monitoring prevent awareness under general anesthesia", "Prevent awareness during general anesthesia", and "General Anesthesia complications", as search terms and limiting to only human studies published in English in the last 10 years (1995-2005). Searches were accomplished using single keywords, their combinations, and Boolean operator words, AND, OR, or both. The number of citations acquired through each search keyword(s) was read and reduced using inclusion criteria. These criteria included surgical patients experiencing Awareness under general anesthesia. The reduction process began with a perusal of citation titles. From this filtering process, abstracts were obtained. The acquired abstracts were then read, reducing their numbers to select relevant articles. References, or bibliographies, of these articles were accessed to expand the original searches. Abstracts of these citations were filtered to include relevant information. The final list of reference was based on relevance to the topics covered which included prevalence, detection, and solutions for prevention of awareness under general anesthesia.
The number of Awareness cases may actually be higher than what is reported, since determining a patient's Awareness is complicated, there being no scientific threshold. According to Sebel et al. (2004), the percentage of Awareness cases is increased from 0.13% to 0.36% if "possible awareness" was included. Possible awareness is defined as when a patient claims of being aware during surgery yet the patient is unable to recall any events that will definitely indicate awareness. Also, it is unclear if dreaming during anesthesia and Awareness are related. Sebel reported 6% of patients have dreams during anesthesia. Other studies (Sebel et al. 2004; Sandin, RH et al. 2000) have shown that numerous patient interviews are needed to quantify Awareness. Of 18 cases, Sandin et al. (2000) reported that 17/18 (94%) cases of Awareness were only confirmed by final interview. It was concluded that, if one interview was conducted, only about 11/18 (61%) cases would have been found. Sebel also reported that numerous interviews are needed.
Awareness is an adverse reaction to light anesthesia. This is especially a problem with hemodynamically unstable patients, or patients undergoing cardiac, obstetric, or major traumatic surgery (JCAHO 2004). However, in these cases the anesthesiologist was aiming to provide a low level of anesthesia while still maintain the patient in an unconscious state. In China, Wang et al. (2005) reported that awareness can be as high as 23% during cardiac surgery. His study included 300 patients, 18/300 (6%) of which were aware during surgery. Of these patients, 6/18 (33%) experienced Awareness between induction of anesthesia and before commencement of surgical procedures. Wang hypothesized two alternatives for this finding. Firstly, patients may have been administered lower dosages of anesthesia due to their hemodynamic state. Secondly, anesthesiologists did not monitor adequately the patient's level of consciousness. Similarly in the U.S. (Sebel et al. 2004), one conclusion arrived at was that Awareness increased in medically-compromised patients requiring lower dosages of anesthesia.
Detecting awareness during surgery can be complicated. Among these complications are protocols, drugs, and devices.
Protocol deficiencies are focused on the role of the anesthesiologists. Osborne et al. (2005) analyzed 21 reported cases of Awareness under general anesthesia. They concluded that 34% of cases resulted from failure by anesthesiologists to monitor their anesthetic equipment. From 34 cases of "possible Awareness," major monitoring problems occurred with vaporizers (13/34, or 38% cases), ventilators (4/34, or 12% of cases), circuit leaks (4/34, or 12% cases), and oxygen flush operation (2/34, or 6% cases). Twelve of the 34 cases (35%) occurred because of a lack of equipment monitoring, 11 (32%) failure to conduct equipment checklists. The study recommended that anesthesiologists are responsible and required to thoroughly monitor their equipment prior and during surgical procedures in reducing Awareness. Further, the study looked at 20 "paralyzed while awake" cases. Most cases resulted from "syringe swaps," or anesthesiologist injecting anesthetic drugs in the wrong order. Thus, patients were paralyzed before they were sedated. This resulted in a "significant distress" for patients (Osborne et al. 2005).
Certain drugs often mask the warning signs of consciousness. Paralytic agents, administered as muscle relaxants during surgery, in corporation with beta-blockers may mask fast heart rates, physical movements, or hemodynamic changes (JCAHO 2004). Anti-anxiety drugs, used to relieve anxiety before surgery, may induce anesthesia. One such drug is Benzodiazepine. Sandin et al. (2000), however, did not find any evidence to conclude that benzodiazepines, or scopolamine a similar drug, are better at inducing anesthesia or reducing Awareness than general anesthetics. This finding was also confirmed in children (Davidson et al. 2005).
Certain devices that were equipped to detect Awareness have been inadequate. For example, Narcotrend® (MonitorTechnik, Bad Bramstedt, Germany), a machine used to detect changes in the level of unconsciousness, had a 50% accuracy rate (Schneider et al. 2004). Narcotrend® is an electroencephalogram-based monitor that observes brain activity, producing a probability of Awareness that ranges from 0% to 100%. A probability of 100% is interpreted as a fully aware patient. Probabilities between 40% and 64% are supportive of a patient under general anesthesia with no Awareness. Schneider et al. (2004) concluded that Narcotrend® did not provide an adequate index of values that differentiated between different levels of anesthesia. This occurs because readings may be confounded by anesthetic drug combination, reducing the ability to relate a specific drug to a reliable value. It is also uncertain if differentiations in patient's physiology will result in non-standardized readings.
Reducing Awareness during surgery is achieved using improved monitoring methods. These methods include protocols and devices.
An algorithm is a commonly used visual representation of protocols or procedural steps to assist clinicians, in this case, anesthesiologists in managing Awareness. In 1993, a Core Crisis Management algorithm, or Cover ABCD A Swift Check, was developed to prevent Awareness. It was developed as a systematic approach to managing crises during the administration of anesthesia. Osborne et al. (2005) analyzed the effectiveness of the algorithm. Osborne demonstrated the effectiveness of the algorithm in detecting most cases of awareness. Its effectiveness, however, was demonstrated in cases for which Awareness was evident. In "possible Awareness" cases, the algorithm was ineffective.
Another reported solution (Burlingame 2005) is involving nurses in monitoring operating room conditions. Although intuitive, nurses may ameliorate the distress that patients report from Awareness. Nurses are made responsible for keeping external traffic, noises, and conversations to a minimum, reducing inappropriate conditions and conversations. Wang et al. (2005) demonstrated that effectively monitoring conversations in the operating room reduces auditory memory in surgical patients.
Traditional methods for reducing Awareness include monitoring blood pressure and heart rate. However, these methods do not correlate reliable measures of the level of consciousness in surgical patients (Bowdle TA. 2006). The Bispectral Index (BIS) monitor (Aspect Medical Systems, Natick, MA), on the other hand, has shown promise in reducing Awareness. The BIS monitor uses electrical brain activity (EEG) to assign a value from 100 (fully conscious) to zero (no brain activity). This number helps the anesthesiologist determine the depth of consciousness in a patient. Using the BIS monitor, Ekman et al. (2004) reported 77% effectiveness in reducing Awareness. In a double-blinded clinical trial, Myles et al. (2004) reported an 82% decrease in "possibile Awareness." Further, the BIS monitor has provided reliable data correlated to specific drugs used in anesthesia. For example, Bestas et al. (2004) examined anesthetic drugs, fentanyl-propofol and fentanyl-midazolam, using the BIS monitor. The BIS monitor correlated correctly both drugs with the occurrence of Awareness.
The cost of the use of the BIS monitor is a concern. Myles et al. (2004) estimates that it costs $16 per patient to use the BIS monitor. Thus, it would cost $2200 to prevent one case of Awareness in high-risk patients. O'Connor et al. (2001) concludes that this cost undermines the benefit of the BIS monitor in preventing Awareness. This analysis, however, varies depending on the number of Awareness cases. Since the prevalence of Awareness is relatively low, costs may not support the use of technology over improving traditional monitoring approaches.
Currently, the BIS monitor appears superior to present devices. Even though the BIS monitor is FDA approved (Bowdle TA. 2006), its adoption is impeded because of a lack of exactness in predicting anesthesia without Awareness. Using the BIS monitor, studies (Ekman et al. 2004; Myles et al. 2004) have reported Awareness in the range of 40 to 60. Still others (Myles et al. 2004), have reported an Awareness range between 55 and 59 thus further complicating adoption of the BIS monitoring. Myles et al. (2004) suggests that not only range but also duration of Awareness is important in implementing this type of technology. Thus, more research is needed to determine and provide effective standards in monitoring Awareness during surgery.
Technological advances have the potential to improve the detection of Awareness during surgery; however technology is mediated by the clinician, thus individuals possess a vital role in preventing Awareness as well. Using the BIS monitor, Myles et al. (2004) has shown that Awareness ranges have been reported in the ranges as high as 79 and 82. This finding was attributed to anesthesiologists not effectively implementing anesthesia protocols. Thus, nurses and algorithms, procedural protocols, have been suggested to provide a check and balance in reducing the adverse effects of Awareness and patient stress. Nurses have been suggested as effective resources in reducing noise and inappropriate behaviors in the operating room. In addition, nurses may provide a second interview, revealing past histories of adverse reactions to anesthesia, reducing the prevalence of Awareness.
Awareness under general anesthesia can not be fully prevented. A patient may become aware for various reasons including levels of anesthesia, drugs, inadequate monitoring devices, and anesthesiologist error. However, steps may be taken to reduce the risk of awareness. The BIS monitor appears to be the most promising way to reduce awareness. But, more research is needed to quantify this promising technology.
Bestas, A et al. (2004) The effects of two different anaesthesia techniques on bispectral index values and awareness during off-pump coronary artery bypass grafting. J Clin Monit Comput. 18(5-6):347-51
Bowdle TA. (2006) Depth of anesthesia monitoring. Anesthesiology Clinics of North America 24(4): 793 -822
Burlingame BL. (2005) Anesthesia awareness; formalin handling; standing orders; witnessing consents. Clinical Issues. 81(1): 213-218
Davidson, AJ et al. (2005) Awareness during anesthesia in children: a prospective cohort study. Anesth Analg. 100(3):653-61
Ekman, A et al. (2004) Reduction in the incidence of awareness using BIS monitoring. Acta Anaesthesiol Scand. 48(1):20-6
JCAHO (2004). Joint Commission on Accreditation of Healthcare Organizations, "Preventing and managing the impact of anesthesia awareness," Sentinel Event Alert 32
Johansen, JW and PS Sebel (2000) Development and clinical application of electroencephalographic bispectrum monitoring. Anesthesiology 93:133644
Myles, PS et al. (2004) Bispectral index monitoring to prevent awareness during anaesthesia: the B-Aware randomised controlled trial. Lancet. 363(9423):1757-63
O'Connor, MF et al. (2001) BIS monitoring to prevent awareness during general anesthesia. Anesthesiology. 94(3):520-2
Osborne, GA, et al (2005). Crisis management during anaesthesia: awareness and anaesthesia. Qual Saf Health Care. 14(3):e16.
Sandin, RH et al. (2000) Awareness during anaesthesia: a prospective case study. Lancet 355.9205: 707
Schneider, G et al (2004) Narcotrend does not adequately detect the transition between awareness and unconsciousness in surgical patients. Anesthesiology. 101(5):1105-11
Sebel, PS et al. (2004). The incidence of awareness during anesthesia: a multicenter United States study. Anesth Analg. 99(3):833-9
Wang, Y et al. (2005) Investigation and analysis of incidence of awareness in patients undergoing cardiac surgery in Beijing, China. Chin Med J (Engl). 118(14):1190-4