The bispectral index (BIS) is one of several systems used in anesthesiology as of 2003 to measure the effects of specific anesthetic drugs on the brain and to track changes in the patient's level of sedation or hypnosis. In technical terms, the bispectral index itself is a complex mathematical algorithm that allows a computer inside an anesthesia monitor to analyze data from a patient's electroencephalogram (EEG) during surgery. BIS, which has been in use since 1997, is a type of automated direct measurement of the patient's condition, in comparison to the Glasgow Coma Scale and similar scoring systems, which are indirect assessments of sedation.
A brief discussion of anesthetic depth may be helpful in understanding people's interest in monitoring the brain's responses to anesthesia. Ever since the first modern anesthetics (ether, chloroform, and nitrous oxide) were used in the 1840s, doctors have been searching for a reliable method of measuring the depth of the patient's unconsciousness in order to guarantee the safety as well as the painlessness of surgery. Anesthetic drugs, whether inhaled or given intravenously, are toxic in high doses; too high a dose can stop the patient's breathing. On the other hand, too small a dose can result in the patient's coming to various degrees of awareness during surgery. Events of this type occur frequently enough to be publicized in general medical news sources as well as the professional literature. One Australian medical journal reports that postoperative recall of operations, including the patient's overhearing conversations among members of the surgical team as well as feeling helpless and experiencing physical pain, occurs in one of every 1,000 patients undergoing non-cardiac surgery and three of every 1,000 cardiac patients. An Israeli researcher gives the rate of accidental awareness during surgery as between 0.2% and 1.2% of patients. According to an American news account, "An estimated 40,000 to 200,000 mid-operative awakenings may occur each year in the United States alone." Research has indicated that patients' attitudes toward undergoing surgery are affected by the possibility of awakening during the procedure. A group of Australian researchers found that 56% of a group of 200 patients awaiting surgery had heard about awareness during operations, mostly from the mass media; 42.5% of the group expressed anxiety about it. Post-traumatic stress disorder (PTSD) is a common result of awareness episodes; a 2001 study done at Boston University reported that 56.3% of a group of patients who had awakened during surgery met the diagnostic criteria for PTSD—as late as 17 years after their operation.
There are several reasons for anesthesiologists' difficulty in evaluating dosages of anesthetic agents:
Indirect methods that allow an observer to assess a person's level of awareness have been used since the early 1970s. The earliest and most widely used instrument for evaluating impaired consciousness is the Glasgow Coma Scale (GCS), first published in the Lancet in 1974. The GCS evaluates the patient's responsiveness under three headings: eye response (four levels of responsiveness), verbal response (five levels), and motor (movement) response (six levels). A normally conscious individual would score 15, the maximum score. In practice, however, the total score on the GCS is usually broken down into three subscores for the three types of response measured; thus E2V2M3 would represent a total GCS score of 7. Total scores on the Glasgow Coma Scale are interpreted as follows: 13–14 indicates mild impairment of consciousness; 9–12 indicates moderate impairment; 8 or lower indicates coma.
There are about a dozen other scales that have been devised to measure consciousness in addition to the GCS; the two that are the most important in this context are the Ramsay Sedation Score, first published in 1974 as a measurement of sedation in patients receiving intravenous sedatives prior to surgery; and the Observer's Assessment of Alertness/Sedation Scale (OAA/SS), first used in 1990 for the same purpose as the Ramsay. These two instruments are significant because they are commonly used in research evaluations of the bispectral index and similar monitoring systems. The Ramsay Score is a six-point scale ranging from one ("patient agitated or restless") through six ("patient asleep; has no response to firm nailbed pressure or other noxious stimuli"). A score of 1 indicates inadequate sedation; 2–4, an acceptable level of sedation; and 5–6, oversedation.
The OAA/SS resembles the Glasgow Coma Scale in that it evaluates different categories of responsiveness, although the categories are different. The OAA/SS measures the patient's responsiveness to his or her name, quality of speech, degree of facial relaxation, and ability to focus the eyes.
A variety of different physiological responses have been used in attempts to measure the depth of a patient's unconsciousness under anesthesia. Most anesthesiologists use hemodynamic responses—the patient's blood pressure and heart rate—as basic guidelines for adjusting the amount of anesthetic delivered to the patient during surgery. Other direct measurements have been based on movements of the patient's body during surgery, hormonal responses, sweating, eye movement, and the reactivity of the eyes to light. One difficulty that has emerged from these attempts at direct measurement is that that they are not good predictors of the likelihood of awareness during surgery or recall of the procedure after surgery.
Another measurement that researchers have explored in their attempts to measure depth of anesthesia directly is the electroencephalogram, or EEG. The EEG is a complex recording of the electrical activity of the nerve cells in the brain. The first published paper on the EEG was written in 1929 by Dr. Hans Berger, an Austrian psychiatrist, on 73 recordings of brain waves using his son Klaus as the subject. Berger was the first to distinguish between alpha and beta brain waves, and to use the term EEG to describe the technique of electroencephalography . In 1931 Berger discovered that brain waves change in amplitude and frequency when a person is asleep or anesthetized; they slow down, shift to lower frequencies, and become more closely synchronized with one another. He also noted that such diseases as multiple sclerosis and Alzheimer disease affect a person's EEG.
Several attempts were made in the late 1980s and early 1990s to make use of what is known about changes in the EEG in order to monitor anesthetic depth. One attempt is known as spectral edge frequency, or SEF. SEF is the frequency just above 95% of the total power spectrum of electrical energy recorded on an EEG. It was thought that the spectral edge frequency would be useful in guiding adjustments of anesthetics administered during surgery. Unfortunately, SEF is difficult to use with balanced anesthesia; it is also difficult to correlate with such other measures of anesthetic depth as movement or memory of the procedure. Another method that has been tried is median frequency, which is based on the median frequency of the complex EEG electrical signal at any given moment. This method proved to have the same drawbacks as spectral edge frequency. The bispectral index can be understood historically as a slightly later and more sophisticated attempt to use EEG signals to monitor patients' responses to anesthesia.
The bispectral index was first developed in the early 1990s by applying bispectral analysis to EEG recordings. Bispectral analysis is a method of analyzing the mathematical relationships among the various components of an EEG signal (phase couplings) as well as measuring amplitudes and frequencies. To compile a database for the index, researchers recorded EEGs from several thousand patients and volunteers anesthetized with a range of commonly used anesthetics and anesthetic combinations. Each subject's depth of unconsciousness was evaluated on the basis of a modified version of the OAA/SS described earlier (in the volunteers) or the amount of drug concentration in blood serum (in the patients). Segments of the recorded EEGs were used to draw up a set of EEG features that were then tested for their ability to distinguish between different levels of sedation or unconsciousness. The index that resulted from this process was then tested on different EEG recordings from the researchers' larger database. It is scaled from 100 to 0 so that the BIS value decreases linearly with increasing doses of anesthesia.
It should be noted that the bispectral index is a work in progress. As new anesthetic agents are developed and used, the BIS algorithm is continually retested and refined. In addition, the algorithm is proprietary information, which means that it is kept secret by the company that developed it.
When a patient is brought into the operating room , special BIS sensors are applied to his or her forehead. No additional gels or electrodes are required. The anesthesiologist can attach the sensors to the patient in less than 30 seconds, since preparing the patient's skin requires no more than an alcohol wipe to provide good electrical contact. The BIS system itself is integrated into patient monitoring devices produced by a number of different manufacturers that use enhanced EEG monitors. The BIS system displays both raw data from the EEG and a single number between 100 (indicating an awake patient) and 0 (indicating the absence of brain activity) that represents the patient's degree of sedation. The target number for most anesthetized patients is between 50 and 60.
BIS is presently used in intensive care units (ICUs) and some emergency departments as well as in operating rooms. According to company information, the bispectral index is used in about 26% of all hospital operating rooms in the United States as of late 2002. It is claimed that BIS reduces the risk of patient awareness during surgery as well as lowering hospital costs by speeding patient recovery and reducing the overuse of anesthetic agents.
As of 2003, published studies of the bispectral index and other anesthetic monitoring systems presently available indicate that none of them can be considered a "gold standard" for preventing instances of patient awareness under anesthesia or for predicting the depth of anesthesia in a specific patient. Researchers have noted several specific limitations of the BIS system:
Other anesthesia monitoring systems that are in use as of 2003 include the Patient State Analyzer, or PSA 4000, which is also based on EEG data; and the A-Line (R) monitor, which processes signals derived from auditory stimuli. Current opinion among anesthesiologists appears to be that none of the present monitoring systems are sufficiently sensitive to guarantee that patients will not awaken during surgery while simultaneously preventing undesirable cardiovascular reactions, other stress responses, or overuse of anesthetic agents.
Baars, Bernard J. "The Brain Basis of a 'Consciousness Monitor.' A Breakthrough in Testing Unconsciousness During Anesthesia." Science and Consciousness Review 1 May 2002 [cited May 13, 2003]. http://www.psych.pomona.edu/scr/news/articles/20020401.html .
Blanchard, Amy R. "Sedation and Analgesia in Critical Care." Postgraduate Medicine 111 (February 2002): 59–60, 63–64, 67–70.
Drover, D. R., H. J. Lemmens, E. T. Pierce, et al. "Patient State Index: Titration of Delivery and Recovery from Propofol, Alfentanil, and Nitrous Oxide Anesthesia." Anesthesiology 97 (July 2002): 82–89.
Drummond, John C. "Monitoring Depth of Anesthesia: With Emphasis on the Application of the Bispectral Index and the Middle Latency Auditory Evoked Response to the Prevention of Recall." Anesthesiology 93 (September 2000): 876–882.
Frenzel, D., C. A. Greim, C. Sommer, et al. "Is the Bispectral Index Appropriate for Monitoring the Sedation Level of Mechanically Ventilated ICU Patients?" Intensive Care Medicine 28 (February 2002): 178–183.
Gill, M., S. M. Green, and B. Krauss. "Can the Bispectral Index Monitor Quantify Altered Level of Consciousness in Emergency Department Patients?" Academic Emergency Medicine 10 (February 2003): 175–179.
Hameroff, Stuart, MD. "Anesthesia: The 'Other Side' of Consciousness." Consciousness and Cognition 10 (June 2001): 217–229.
Hoymork, S. C., J. Raeder, B. Grimsmo, and P. A. Steen. "Bispectral Index, Predicted and Measured Drug Levels of Target-Controlled Infusions of Remifentanil and Propofol During Laparoscopic Cholecystectomy and Emergence." Acta Anaesthesiological Scandinavica 44 (October 2000): 1138–1144.
Kalkman, Cor J., and John C. Drummond. "Monitors of Depth of Anesthesia, Quo Vadis?" Anesthesiology 96 (April 2002): 784–787.
Leslie, K., and P. S. Myles. "Awareness During General Anesthesia: Is It Worth Worrying About?" Medical Journal of Australia 174 (March 5, 2001): 212–213.
Leslie, K., L. Lee, P. S. Myles, et al. "Patients' Knowledge of and Attitudes Toward Awareness and Depth of Anesthesia Monitoring." Anaesthesia and Intensive Care 31 (February 2003): 63–68.
Muncaster, A. R., J. W. Sleigh, and M. Williams. "Changes in Consciousness, Conceptual Memory, and Quantitative Electroencephalographical Measures During Recovery from Sevoflurane- and Remifentanil-Based Anesthesia." Anesthesia and Analgesia 96 (March 2003): 720–725.
Osterman, J. E., J. Hopper, W. J. Heran, et al. "Awareness Under Anesthesia and the Development of Posttraumatic Stress Disorder." General Hospital Psychiatry 23 (July-August 2001): 198–204.
Renna, M., J. Handy, and A. Shah. "Low Baseline Bispectral Index of
the Electroencephalogram in Patients with Dementia."
Anesthesia and Analgesia
96 (May 2003):
Riess, M. L., U. A. Graefe, C. Goeters, et al. "Sedation Assessment in Critically Ill Patients with Bispectral Index." European Journal of Anaesthesiology 19 (January 2002): 18–22.
Tassi, P., and A. Muzet. "Defining the States of Consciousness." Neuroscience and Biobehavioral Reviews 25 (March 2001): 175–191.
Welsby, Ian J., J. Mark Ryan, John V. Booth, et al. "The Bispectral Index in the Diagnosis of Perioperative Stroke: A Case Report and Discussion." Anesthesia and Analgesia 96 (February 2003): 435–437.
Wong, C. S. "What's the Matter of Depth of Anesthesia?" Acta Anaesthesiologica Sinica 39 (March 2001): 1–2.
American Academy of Emergency Medicine (AAEM). 611 East Wells Street, Milwaukee, WI 53202. (800) 884-2236. http://www.aaem.org .
American Association of Nurse Anesthetists (AANA). 222 South Prospect Avenue, Park Ridge, IL 60068-4001. (847) 692-7050. http://www.aana.com .
American Society of Anesthesiologists (ASA). 520 N. Northwest Highway, Park Ridge, IL 60068-2573. (847) 825-5586. http://www.asahq.org .
American Society of Perianesthesia Nurses (ASPAN). 10 Melrose Avenue, Suite 110, Cherry Hill, NJ 08003. (877) 737-9696 or (856) 616-9600. http://www.aspan.org .
Center for Emergency Medicine. 230 McKee Place, Suite 500, Pittsburgh, PA 15213. (412) 647-5300. http://www.centerem.com .
Society for Technology in Anesthesia (STA). PMB 300, 223 North Guadalupe, Santa Fe, NM 87501. (505) 983-4923. http://www.anestech.org .
Aspect Medical Systems White Paper. Technology Overview: Bispectral Index . Natick, MA: Aspect Medical Systems, 1997.
Greenwald, Scott D., Ph.D., Charles P. Smith, Jeffrey C. Sigl, Ph.D., et al. Development of the EEG Bispectral Index (R) (BIS(R)) . Abstract presented at the first annual meeting of the Society for Technology in Anesthesia (STA). January 2000 [cited May 12, 2003]. http://www.anestech.org/Publications/Annual_2000/Greenwald.html .
Vuyk, Jaap, MD. Does Bispectral Index Monitoring Optimize Intravenous Anaesthetic Drug Delivery? Paper delivered on April 5, 2002 at the fifth annual meeting of the European Society for Intravenous Anaesthesia (Eurosiva) in Nice, France. http://www.eurosiva.org/Archive/Nice/SpeakerAbstracts/Vuyk.htm .
Rebecca Frey, Ph.D.