Dept Of Anesthesiology & Critical Care, University Of Pennsylvania School Of Medicine

An anesthesia resident discusses BIS

The issue of awareness under anesthesia is of great concern for obvious reasons. The BIS monitor was developed with the intention of ensuring patients are adequately anesthetized. However, given the expense of BIS electrodes and the low incidence of awareness the routine use of the BIS for awareness detection may not be economically efficient. With that in mind I asked the questions: “What else can it do? Can the BIS monitor be used to help fine-tune an anesthetic to achieve the best possible care for an individual patient?”  To answer these questions I began using the BIS monitor on as many cases as possible to see if it changed my typical management. The types of cases included craniotomies for tumors and aneurysms, laminectomies, VP shunts, orthopedics, ENT, and an A-fib ablation. Several observations quickly became apparent. First, the BIS revealed that I was frequently overdosing the volatile agent in response to increased heart rate and blood pressure. This observation was consistent with several studies that have shown that hemodynamic indicators do not always correlate with state of consciousness. In these situations the BIS guided me towards using additional narcotic or anti-hypertensive medications rather than simply turning up the anesthetic vapor.

This change in management had several clinical implications. By using less volatile agent emergence times were consistently reduced and cognitive function returned quicker. This was especially beneficial in craniotomies as patients were able to quickly cooperate with neurologic exams at the end of the case.  It was also helpful in appropriately titrating opioid in surgeries such as laminectomies where post-operative analgesia can be challenging. In addition to these important short-term benefits, decreasing the amount of volatile anesthetic may affect long-term outcomes. As we learn more and more about the potential neurodegenerative changes associated with anesthetic drugs it becomes evident that reducing the amount of volatile agent may be beneficial down the road for the patient.

While the reduction in anesthetic use was notable across a broad spectrum of cases the BIS helped in other aspects of management as well. In two patients where thiopental infusions were utilized to achieve EEG burst suppression the BIS allowed easy medication titration. One patient in particular required a significantly higher rate of infusion than we expected. I also found the BIS to be a valuable aid in weaning off the anesthetic agent towards the end of craniotomies with the Mayfield. I was able to comfortably titrate down the agent while maintaining neuromuscular blockade without the fear of awareness. The BIS may also augment patient satisfaction with their anesthetic in regards to post-operative nausea and vomiting. One study has shown a decreased in PONV when using the BIS to titrate sevoflurane in ambulatory surgery. In these select cases I did not observe much difference in PONV although this may be partially explained by my increased use of opiates when using the monitor.

In summary, the information provided by the BIS allowed me to better tailor anesthetic management to provide the best possible care for individual patients. The ability to directly monitor individual brain effect of the anesthetic drug facilitated therapeutic decision making. Using the BIS in conjunction with clinical judgment I was able to achieve a better balance between anesthetic agents and other adjuvant medications. In addition to monitoring awareness during anesthesia the data provided by the BIS can be utilized to deliver a precise and personalized anesthetic.

Jonathan Anson, M.D.

Dr. Anson is a resident in Anesthesia at the University of Pennylvania

A faculty advocate for routine intraoperative brain electrical activity monitoring

WA Kofke is a member of the faculty of the Department of Anesthesiology and Critical Care at the University of Pennsylvania.  He is a strong advocate for the routine use of brain electrical activity monitors such as BIS for the management of general anesthetics.  He recently published, in the Etherway Blog, a summary of how he uses the BIS for the care of his patients.  His summary starts

"I Use The BIS Monitor For A Lot More Than Just Preventing Awareness!

I like knowing how deeply anesthetized my patients are,  not just that they are unaware. How much a patient is unaware as a continuum is a lot more important that the binary notion of awake vs asleep.  I do this with a BIS monitor (and could do the same with Hospira’s PSA monitor) and make lots of decisions based on it and I think provide better care for the effort. My reasons for saying this follow."

Read the rest of his discussion at http://mkeamy.typepad.com/anesthesiacaucus/2007/11/i-use-the-bis-m.html

I like knowing how deeply anesthetized my patients are,  not just that they are unaware. How much a patient is unaware as a continuum is a lot more important that the binary notion of awake vs asleep.  I do this with a BIS monitor (and could do the same with Hospira’s PSA monitor) and make lots of decisions based on it and I think provide better care for the effort. My reasons for saying this follow."

Read the rest of his discussion at http://mkeamy.typepad.com/anesthesiacaucus/2007/11/i-use-the-bis-m.html

Is there a consensus concerning the routine use of BIS monitoring during general anesthesia?

My personal opinion is as follows: Aspect Medical and some patient advocate groups appear to have “pushed” hard for the adoption of BIS or other brain activity monitor as a standard of care for routine intraoperative monitoring of general anesthesia.  The lack of information about the algorithms used to derive the BIS index and other proprietary indexes, the lack of understanding of the factors that lead to awareness under anesthesia, the fact that many operations do not lend themselves to intraoperative brain activity monitoring, and the several published counter-examples in which information provided by the BIS monitor seems to contradict the claims for the product make many thoughtful investigators and practitioners uncomfortable with the concept that this monitor or monitors of this type should be enshrined as “standard.”  The fact that these monitors do not rise to the level of a standard however does not mean that they are not useful.  Nor does it mean that they should not be used.  The decision of the American Society of Anesthesiologists committee on practice parameters to not endorse brain activity monitoring routinely should not be construed as a lack of interest on the part of anesthesiologists in the problem of intra operative awareness (see ASA guidelines on intraoperative awareness, http://www.asahq.org/publicationsAndServices/AwareAdvisoryFinalOct05.pdf ).

David S. Smith, M.D., Ph.D.

What errors are associated with BIS monitoring?

Earlier versions of the BIS monitor were very sensitive to electrocautery and the noise from electrocauteries would sometimes be incorporated into the BIS index value.  Earlier versions had difficulty recognizing and rejecting muscle activity and on occasion burst suppression on the EEG would be interpretated as a high BIS index value (1).  These problems appear to have been corrected.  There are however recurring reports related to failure of the BIS value to correlate with the clinical situation.  Patient movement despite BIS values in the range accepted as indicating adequate general anesthesia is commonly reported.  At least one case report has described a patient who was awake and responding despite a BIS level in the 50s (2).  Another paper reported recall in a patient with a BIS level in the 40s (3).  Another case report found changes in the BIS level with administration of muscle relaxation to lightly anesthetized patients (4).  Others have found changes in the BIS level with the administration of intravenous adrenergic agonists such as ephedrine (5).  On the other hand clinicians reported cases in which the BIS monitor identified the otherwise unrecognized onset of inadequate cerebral perfusion and in at least one case this allowed correction before brain injury occurred (6, 7, 8).  BIS has also been found to correlate with the severity of traumatic brain injury.  One unsolved problem is a dramatic fall in BIS values as potent inhalational agent concentration drops during emergence from nitrous oxide assisted general anesthesia.  This phenomenon, called paradoxical slowing is seen in the EEG so it is not an artifact.  Its mechanism is unclear but it is clearly not a deepening of the anesthetic.  Yet the current BIS algorithm interprets it as such.  Dr. Kofke has queried the scientific personnel at Aspect Medical and at present they have no solution.

1) Johansen JW: Development and clinical application of electroencephalographic bispectrum monitoring.  Anesthesiology 2000;93:1336-14

2) Kakinohana M et al: Emergence from propofol anesthesia in a nonagenarian at a bispectral index of 52. Anesth Analg 2005;101:169-70

3) Mychaskiw G, et al: Explicit intraoperative recall at a bispectral index of 47. Anesth Analg 2001;92:808-9

4) Liu N et al: The influence of a muscle relaxant bolus on bispectral and datex-ohmeda entropy values during propofol-remifentanil induced loss of consciousness. Anesth Analg 2005;101:1713-8 

5) Andrzejowski J et al: The effect of intravenous epinephrine on the bispectral index and sedation. Anaesthesia 2000;55:761-763

6) England MR: The changes in bispectral index during a hypovolemic cardiac arrest.  Anesthesiology 1999;91:1947-9

7) Rath GP and Singh D: Zero bispectral index during coil embolization of an intracranial aneurysm.  Anesth Analg 2007;105:887-888

8) Morimoto Y et al: The detection of cerebral hypoperfusion with bispectral index monitoring during general anesthesia. Anesth Analg 2005;100:158-61

David S. Smith, M.D., Ph.D.

What is the role of BIS or other brain activity monitor in decreasing the incidence of awareness?

Reviews of cases in which awareness has occurred have identified cases in which, had BIS monitoring been used, the inadequate level of anesthesia might have been recognized (1).  However there appear to be cases in which awareness occurred despite BIS monitoring.  This is true for the b-aware and b-unaware trials as well as case reports (2, 3, 4).

1) Bergman IJ et al: Awareness during general anesthesia: a review of 81 cases from the anaesthesia incident monitoring study. Anaesthesia 2002;57:549-556

2) Myles PS: Bispectral index monitoring to prevent awareness during anaesthesia: the B Aware randomized controlled trial. Lancet 2004;363:1757-63

3) Finkel KJ et al: Sensitivity of BIS and MAC in the B-Unaware Trial.  ASA Annual Meeting, San Francisco A728

4) Mychaskiw G, et al: Explicit intraoperative recall at a bispectral index of 47. Anesth Analg 2001;92:808-9

David S. Smith, M.D., Ph.D.

Does BIS or equivalent monitors have a role in the management of anesthetics?

A number of studies have demonstrated that BIS or equivalent monitors facilitate conduct of general anesthetics that are based on or have a significant component from anesthetic drugs that act via GABAa receptors.  These monitors may have particular benefit in “outliers,” that is patients whose drug requirement may differ significantly from the typical or in whom the hemodynamic effects of anesthetics (a common surrogate for anesthesia depth) is masked by cardiovascular drugs.  BIS monitoring may be the only reliable monitor for total intravenous anesthetic techniques (TIVA) in patients receiving muscle relaxants.  Failure in drug delivery when using TIVA may not be reflected in changes in heart rate or blood pressure and muscle relaxation will prevent patient movement.

David S. Smith, M.D., Ph.D.

What anesthetic types are suitable for BIS or similar monitoring?

BIS and other similar monitors that use processed EEG parameters to derive an estimate of anesthetic effect site activity appear to be most sensitive to anesthetic drugs that have their major effect via the GABA a receptor.  These monitors do not “see” the effects of drugs that act via NMDA pathways, endorphin pathways or alpha 2 adrenergic pathways.  Thus drugs such as isoflurane, desflurane, sevoflurane, propofol and etomidate lend themselves to BIS or similar monitoring.

David S. Smith, M.D., Ph.D.

What is BIS?

The BIS monitor is a processed EEG device manufactured by Aspect Medical Corporation.  Using a proprietary multiple regression algorithm it generates a dimensionless value that ranges from 0 to 100.  According to the manufacturer this index correlates to some of the components of the general anesthesia state including level of consciousness and probability of recollection when anesthetic agents that act though increasing chloride conductance into GABAa neurons are used.  They claim that a value between 40 – 60 represents a level of cortical brain inhibition compatible with adequate hypnotic levels of general anesthesia (although not predicting lack of movement or adequate analgesia).  They also claim that the probability of a patient being aware and having recall with BIS values in this range is extremely low (see Aspect Medical’s Web site for a discussion of their claims, http://www.aspectmedical.com/ ).  There are other brands of processed EEG devices being sold or under development for use in this manner.  However the largest published experience has been with the Aspect Medical device.

David S. Smith, M.D., Ph.D.

Anesthetics and the brain

No one disputes the hypothesis that general anesthetics act via the brain.  Consequently, the use of some measure of brain activity to estimate anesthesia depth has been a long standing goal.  The potential ability to use the EEG or some derivative of the EEG for this purpose has been a subject of considerable study with some of the early work done at UPENN by Don Clark and his associates (1-5).  Though it became clear that anesthetics altered the EEG there was considerable difficulty in interpreting multi channel analogue EEG traces.  In particular the early investigators could not reliably identify the onset of unconsciousness from changes in the EEG trace.  With the advent of computer EEG processing work began anew.  However the first attempts, focusing mainly on power spectrum analysis, also failed at differentiating the transition from awake to asleep and attempts at developing a “depth of anesthesia” monitor based on EEG failed.  In 1996 Aspect Medical Corporation was approved to market a brain monitoring device which they initially attempted to correlate with classical clinical measures of anesthetic depth such as MAC (6).  This failed and they began to reconceptualize their monitor toward other roles in the anesthetized patient.

            1) Clark DL, Rosner BS: Neurophysiologic effects of general anesthetics. I. The electroencephalogram and sensory evoked responses in man. Anesthesiology. 38:564-82, 1973

            2) Clark DL, Hosick EC, Rosner BS: Neurophysiological effects of different anesthetics in unconscious man. J Applied Physiol 31:884-91

            3) Beck C: Cerebral electrical activity during cyclopropane anesthesia in man.  J Applied Physiol 28:802-7, 1970

           4) Hosick EC, Clark DL, Adam N, Rosner BS: Neurophysiological effects of different anesthetics in conscious man. J Applied Physiol 31:892-8, 1971

            5) Kavan EM, Julien RM, Lucero JJ: Electrographic alterations induced in limbic and sensory systems during induction of anaesthesia with halothane, methoxyflurane, diethyl ether, and enflurane (Ethrane). Br J Anaesth 44:1234-9, 1972            

           6) Bowdle TA: Depth of Anesthesia monitoring. Anesthesiology Clin 2006;24:793-822

David S. Smith, M.D., Ph.D.

The calculation of oxygen content

Dr. Wilkey discusses the calculation of oxygen content

Oxygen is poorly soluble in water; therefore without an adjunctive means of transport, it cannot be transported in blood in quantities sufficient to sustain life.  That adjunct comes in the form of hemoglobin.  Oxygen can bind to hemoglobin at any of four active sites on each molecule of hemoglobin.  When measured per gram of hemoglobin, its capacity to carry oxygen is 1.39 ml.  As will be seen below, the term for O₂ per gram Hgb is 1.34 rather than 1.39.  (Some sources may report values ranging from 1.31 to 1.36).  This is due to binding of non-O₂ species to hemoglobin, thereby changing its conformation to one that will not bind oxygen.  The ~0.05 ml/gm is comprised of compounds such as methemoglobin and carboxyhemolobin. 

The equation for the concentration of oxygen in arterial blood is as follows:

CaO₂ = (SaO₂ x Hgb x 1.34) + (0.003 x PaO₂)

The term for SaO₂ is expressed as a fraction of 1.0 rather than a percentage (i.e. 0.98 instead of 98%).  Hemoglobin is entered in grams and PaO₂ is in mmHg.

The term 0.003, derived from oxygen solubility coefficients at different temperatures, assumes a normal body temperature of 37ºC and is expressed as mL O₂/dl/mmHg.  (The solubility of oxygen in blood is 0.03 ml O₂/l/mmHg.  This is divided by ten to bring the units into agreement with others in the equation).

The above equation can then be expressed with units included as:

CaO₂ = (SaO₂ x Hgb gm/dl x 1.34 ml/gm) + (0.003 O₂ ml/dl/mmHg x PaO₂ mmHg) = ml/dl.  (Terms that cancel are lined out).

In a hypothetical patient with SaO₂ 1.0 (100%), Hgb of 15 gm/dl and PaO₂ of 100 mmHg, the oxygen content of arterial blood is:

CaO₂ = (1.0 x 15 gm/dl x 1.34 ml/gm) + (0.003 ml/dl/mmHg x 100 mmHg) = 20.4 mL/dl.

Note that the contribution of dissolved oxygen to the total is only 0.3 mL/dl or less than 1.5% of the total—not nearly enough to sustain life.  An exception to this is with the application of hyperbaric oxygen therapy.  100% O₂ at 3 atmospheres would lead to a PaO₂ as high as 2200mmHg.

Keep in mind that while dissolved O₂ usually contributes little to the total arterial content of oxygen, it can make the difference between two hypothetical patients or patient scenarios.   Recalculating this equation with a SaO₂ of 75% will approximate mixed venous oxygen content.  This equation is one of the most fundamental in anesthesia and must be committed to memory—being able to do the calculation quickly and without the use of a calculator is useful as well.

Also of note is the strong contribution of Hgb to oxygen content.  Though other factors such as tissue perfusion and oxygen uptake and extraction can play significant roles in end organ function or dysfunction, often the addition of more hemoglobin is the only readily modifiable way to increase arterial oxygen capacity and its subsequent delivery to tissues. 

1) Stoelting RK, Pharmacology & Physiology in Anesthetic Practice, Lippincott Williams & Wilkins; 3Rev Ed, 1999.

2) Morgan GE, Mikhail MS, Clinical Anesthesiology, McGraw-Hill Medical; 4 Ed, 2005.

3) Miller RD, Miller’s Anesthesia, 6^th Ed, Churchill Livingstone, 2004.

Andrew Wilkey, M.D. is a Cardiothoracic Anesthesia Fellow at the University of Pennsylvania Health System