What physiologic monitors should we use in the operating room? Though it has been over 173 years since William T. G. Morton first successfully administered a general anesthetic, the question of how we should monitor patients under anesthesia care remains an area of passionate debate and the subject of an article in this issue of Anesthesia & Analgesia.1 Before we delve into this debate and the article’s recommendations, it is worth reviewing several points about the use of monitoring in general and the evidence for improving outcomes.
SPEEDOMETERS AND INTRAOPERATIVE MONITORS: THEY DO NOT ACT BY THEMSELVES
First, the use of a monitor does not improve outcomes or prevent adverse events unless the anesthesiologist interpreting data from the monitor(s) takes timely and therapeutic action(s). Though we lack firm data, we believe it is highly likely that nearly all high-speed motor vehicle collisions in the United States in 2019 involved cars with functioning speedometers. That those speedometers did not prevent these high-speed accidents comes as no surprise. We know that speedometers do not prevent cars from exceeding the speed limit; they simply provide information to drivers. How individual drivers use that information to decide whether to step on the gas or the brakes depends on a complex set of calculations: how much the driver believes in following laws, whether the driver is in a hurry, fears a speeding ticket, thinks that speeding will increase the chance of an accident, etc. Similarly, we are unaware of any intraoperative monitor that itself improves outcomes without a proper response by an anesthesiologist “driver.” Just as drivers use the information from speedometers to make decisions that affect the risk of car accidents, anesthesiologists use information from intraoperative monitors to make decisions and take actions that affect patient outcomes.
Although obvious, this is a key point to remember when considering the literature on electroencephalogram (EEG) monitoring and patient outcomes. The trials discussed by the Perioperative Quality Initiative (POQI) article1 used the bispectral index (BIS) monitor in different ways,2–4 and in some studies, the BIS index was used with raw EEG waveform analysis to change care.2 Because EEG monitoring was used in different ways in these studies, it is difficult to conclude whether EEG monitoring in general or BIS monitoring, in particular, is associated with improved outcomes. Trying to determine whether intraoperative EEG monitoring alters outcomes is like asking whether speedometers reduce car accidents. Unless we know how individual anesthesiologist “drivers” respond to monitored information, we cannot answer this question.
The POQI article provides a series of forest plots that quantitatively measure the effects of (processed) EEG monitoring in recent studies. Yet, in each of these studies, the monitors simply provided data to anesthesia providers, who then acted on this information in different ways. For example, in the Electroencephalography Guidance of Anesthesia to Alleviate Geriatric Syndromes (ENGAGES) trial, clinicians acted on EEG monitor data to reduce anesthetic administration from 0.80 to 0.69 minimum alveolar concentration (MAC) (a reduction of 0.11 MAC),2 while in the cognitive dysfunction after anesthesia (CODA) trial, clinicians acted on EEG monitor data to reduce anesthetic dosage from 0.93 to 0.57 MAC (a 0.36 MAC reduction),3 a >3-fold larger absolute reduction in anesthetic dosage than that seen in ENGAGES. In essence, the results of these 2 studies do not reflect opposing arguments on whether EEG usage prevents delirium; instead, they both support the idea that it is how anesthesiologists use the data from these monitors to alter patient care that changes outcomes.
MONITORS AND PARACHUTES: THEY CAN BE USEFUL
Yet, even this discussion is overly simplistic, as the use of a device or monitor to prevent an adverse event will differ based on the clinical context and patient cohort in which it is used. Parachutes are widely believed to prevent death among people jumping out of airplanes, yet a recent randomized controlled trial showed that using parachutes did not prevent deaths among people jumping out of airplanes—when the planes were already on the ground!5 The extent to which a monitor can prevent an adverse event depends on the population and situation in which it is implemented. Just as parachutes do not prevent mortality in individuals jumping out of airplanes on the ground (in whom the baseline mortality risk is virtually nonexistent), EEG monitoring (and tight anesthetic titration) is unlikely to help anesthesiologists reduce complication rates (ranging from delirium to intraoperative awareness) in patients at low risk for these complications, such as a 20-year-old American Society of Anesthesiologists (ASA) physical status I patient undergoing a laparoscopic appendectomy. Consistent with this idea, a recent study in healthy volunteers undergoing general anesthesia without surgery found no correlation between the duration of burst suppression, an indicator of a brain state far deeper than necessary to prevent awareness, and emergence times or other outcomes.6
THE 4 QUESTIONS OF INTRAOPERATIVE MONITORING
So, where does this leave us when deciding whether to use intraoperative EEG monitoring? We believe there are 4 general considerations that should drive the decision to use any monitor (including EEG) in the operating room (Figure). First, how relevant is the information from the monitor for understanding how the patient’s physiology is being modulated by anesthetic drugs/techniques and surgical procedures? Second, to what extent are anesthesiologists likely to alter a patient’s care based on information from the monitor? Third, how much effort, cost, and risk is associated with use of the monitor? Fourth, how much does using the monitor detract from other urgent priorities or tasks in the operating room?
For the first question, the answer is mixed. On the one hand, the central nervous system (CNS) is the target organ for most, if not all, anesthetic drugs, which would argue that we should always monitor CNS activity to understand what our drugs are doing to the brain. Yet, there is no consensus on what the exact EEG correlates of amnesia or unconsciousness are,7 so it remains unclear what EEG monitors should be used, what aspects of the EEG raw waveform or spectrogram should be monitored, or what processed EEG measures should be utilized.
To help convey our limited knowledge here, it is useful to compare processed EEG to electrocardiogram (ECG) monitoring. If an attending anesthesiologist asks a resident, “how does the ECG look?” an answer of “Good, it is 70” is likely to be viewed as inadequate. A more informative response would be “the ECG shows a normal sinus rhythm at a rate of 70, with normal axes and intervals, narrow QRS complexes, and flat ST segments,” which conveys a sophisticated understanding of the function of the cardiac conduction system and the extent to which the heart is receiving sufficient oxygen to meet metabolic requirements. Unfortunately, we are nowhere close to being able to describe the brain with this level of sophistication based on processed EEG information, even though there are some raw EEG patterns clearly visible on spectrogram plots that predict adverse patient outcomes.8 Because these raw EEG waveform patterns that predict adverse outcomes are not detectable by simple processed EEG index values, this suggests that anesthesiologists should learn how to read and interpret raw EEG waveforms and spectrograms.
Furthermore, it makes little sense to avoid monitoring the target organ of our anesthetic drugs. Dosing anesthetics based on their side effects (such as hypotension) on other organ systems (such as the cardiovascular system) without monitoring their target organ (the CNS) makes about as much sense as if internists dosed antihypertensive drugs like angiotensin-converting enzyme (ACE) inhibitors based entirely on their side effects (such as dizziness) on other organs (ie, the brain) without measuring their effects on their target organ (the cardiovascular system; Mashour G, Department of Anesthesiology, University of Michigan School of Medicine, Ann Arbor, MI, personal communication, 2012). That would be inappropriate for an internist, so why is this same logic acceptable for anesthesiologists? Furthermore, one could argue that it is important to monitor the target organ of anesthetic drugs (the brain) just as we monitor patients’ oxygen saturation during surgery, even though pulse oximetry has not been shown to reduce overall complication rates,9 because both provide highly relevant physiologic information.
For the second question, “to what extent are anesthesiologists likely to alter a patient’s care based on information from the monitor?” the answer is again mixed. Prospective studies utilizing processed EEG monitoring in different ways have shown varying effects on anesthetic titration/administration,2,3,10 and observational studies have shown no decrease (and even a slight increase) in anesthetic dosage among patients who underwent BIS monitoring versus routine care.11 The different effects on anesthetic titration across these studies may partly reflect the fact that BIS index values are not simply a reflection of anesthetic dosage, but rather change as a function of both patient age and preoperative cognitive function.11,12 These results also likely suggest that anesthesiologists provided with the same processed EEG monitor data make different decisions about anesthetic titration and care. This variability in clinical decision making is not surprising—prior studies have also found substantial variability in treatment decisions made by different clinicians in response to the same data from pulmonary artery catheters (PACs).13
For the third question about effort and cost, and risk, frontal EEG monitoring involves little effort and minor cost (relative to other perioperative costs). Unlike other intraoperative monitors such as PACs, whose placement can be associated with life-threatening adverse events such as pulmonary artery rupture, placing a frontal EEG monitor sticker has no significant risk of patient harm. Yet, just as PAC usage has been associated with increased adverse events,14 EEG monitor data (especially processed EEG data) could potentially lead clinicians to make decisions that lead not only to benefits but also to adverse events. For example, a 0.36 MAC reduction in the CODA trial EEG group led to significantly lower delirium rates,3 yet a 0.11 MAC reduction in the ENGAGES trial EEG group was associated with a significant increase in patient movement during surgery,2 which could be harmful if movement occurs during a critical or vulnerable point during surgery. Thus, while the placement of frontal EEG stickers themselves is largely risk-free, intraoperative EEG monitor data could potentially lead to patient harm depending on how clinicians act on these data to alter patient care.
For the fourth question, “how much does using the monitor detract from other urgent priorities or tasks in the operating room?” the answer will likely differ depending on the surgical case, anesthetic technique, patient factors, and provider preference and familiarity with a given EEG monitor.
Thus, aside from the forest plots and statistical calculations provided by this thorough POQI article, we believe that the question of whether to use intraoperative EEG monitoring should be guided by the answers to these 4 questions by individual clinicians on a case-by-case basis. Different clinicians may even disagree about whether EEG monitoring should be used or will be helpful in improving outcomes. Such disagreements are reasonable based on current evidence. Monitors do not alter patient outcomes; it is the decisions made and the actions taken by anesthesiologists (sometimes in response to data from monitors) that affect patient outcomes. Beyond this debate over whether current EEG monitors improve the care of today’s patients, attaining a deeper understanding of how anesthetic drugs modulate brain activity to produce general anesthesia (and how to detect such effects via EEG waveform analysis) are fundamentally important goals for our field. Furthering our understanding of how anesthetic drugs modulate neurophysiology may also help us to understand altered brain states associated with perioperative neurocognitive disorders15 and will hopefully lead to better EEG monitors and outcomes for future patients.
1. Chan MTV, Hedrick TL, Egan TD, et al. American Society for Enhanced Recovery and Perioperative Quality Initiative Joint Consensus Statement on the role of neuromonitoring in perioperative outcomes: electroencephalography. Anesth Analg. 2020;130:1278–1291.
2. Wildes TS, Mickle AM, Ben Abdallah A, et al.; ENGAGES Research Group. Effect of electroencephalography-guided anesthetic administration on postoperative delirium among older adults undergoing major surgery: the ENGAGES randomized clinical trial. JAMA. 2019;321:473–483.
3. Chan MT, Cheng BC, Lee TM, Gin T; CODA Trial Group. BIS-guided anesthesia decreases postoperative delirium and cognitive decline. J Neurosurg Anesthesiol. 2013;25:33–42.
4. Radtke FM, Franck M, Lendner J, Krüger S, Wernecke KD, Spies CD. Monitoring depth of anaesthesia in a randomized trial decreases the rate of postoperative delirium but not postoperative cognitive dysfunction. Br J Anaesth. 2013;110Suppl 1i98–105.
5. Yeh RW, Valsdottir LR, Yeh MW, et al.; PARACHUTE Investigators. Parachute use to prevent death and major trauma when jumping from aircraft: randomized controlled trial. BMJ. 2018;363:k5094.
6. Shortal BP, Hickman LB, Mak-McCully RA, et al.; ReCCognition Study Group. Duration of EEG suppression does not predict recovery time or degree of cognitive impairment after general anaesthesia in human volunteers. Br J Anaesth. 2019;123:206–218.
7. Gaskell AL, Hight DF, Winders J, et al. Frontal alpha-delta EEG does not preclude volitional response during anaesthesia: prospective cohort study of the isolated forearm technique. Br J Anaesth. 2017;119:664–673.
8. Hesse S, Kreuzer M, Hight D, et al. Association of electroencephalogram trajectories during emergence from anaesthesia with delirium in the postanaesthesia care unit: an early sign of postoperative complications. Br J Anaesth. 2019;122:622–634.
10. Avidan MS, Zhang L, Burnside BA, et al. Anesthesia awareness and the bispectral index. N Engl J Med. 2008;358:1097–1108.
11. Ni K, Cooter M, Gupta DK, et al. Paradox of age: older patients receive higher age-adjusted minimum alveolar concentration fractions of volatile anaesthetics yet display higher bispectral index values. Br J Anaesth. 2019;123:288–297.
13. Jain M, Canham M, Upadhyay D, Corbridge T. Variability in interventions with pulmonary artery catheter data. Intensive Care Med. 2003;29:2059–2062.
14. Binanay C, Califf RM, Hasselblad V, et al. Evaluation study of congestive heart failure and pulmonary artery catheterization effectiveness: the ESCAPE trial. JAMA. 2005;294:1625–1633.
15. Browndyke JN, Berger M, Smith PJ, et al.; Duke Neurologic Outcomes Research Group (NORG). Task-related changes in degree centrality and local coherence of the posterior cingulate cortex after major cardiac surgery in older adults. Hum Brain Mapp. 2018;39:985–1003.