EEG-Guided Pediatric Anesthesia: The not so new desert island monitor
Myron Yaster MD and Jerry Chao MD
Back when I was in college, my friends and I would argue endlessly over the question: “If you were stranded on a desert island and could have only one vinyl album (OK for most of you, that’s now a streamed electronic album) what would it be?” Beatles or the Stones? Hendrix or Garcia? Davis or Bird? You get the idea. I think all of you would agree that pulse oximetry and capnography are desert island monitors and one would be hard pressed to pick one over the other. Quantitative neuromuscular blockade monitors and intraoperative EEG monitoring are now readily available and are rapidly becoming a new generation of desert island monitors.
When working with residents and young (well everyone is young compared to me) faculty and talking about the days before pulse oximetry, capnography, and automated non-invasive blood pressure devices, I would always get this look of total disbelief. “How could you possibly have provided anesthesia without those monitors?” I think 5-10 years from now, you will get that same look of disbelief when talking about the provision of anesthesia without EEG and quantitative neuromuscular blockade monitors.
In today’s PAAD1 and accompanying editorial2, Miyasaka et al. and Kurth et al. respectively lay out the case for routine EEG-guided pediatric anesthesia. We’ve discussed this in several previous PAADs (August 3, 2022, August 10, 2022, August 16, 2022, and August 25, 2022) many posted before most of you became PAAD readers. (https://ronlitman.substack.com/p/eeg-part-1 https://ronlitman.substack.com/p/eeg-part-2 https://ronlitman.substack.com/p/eeg-part-3 https://ronlitman.substack.com/p/eeg-part-4 ) I’d recommend downloading these if you need to get up to speed. Myron Yaster MD
Editorial
VanBuren JM, French B. Type I Error Control-Avoiding an Upset. JAMA Pediatr. Published online April 21, 2025. doi:10.1001/jamapediatrics.2025.0528. PMID: 40257760
Kurth CD, Gabrielsen DA, Yuan I. EEG-Guided Pediatric Anesthesia-A Quality Innovation? JAMA Pediatr. 2025 Apr 21. doi: 10.1001/jamapediatrics.2025.0514. Epub ahead of print. PMID: 40257772.
Original article
Miyasaka KW, Suzuki Y, Brown EN, Nagasaka Y. EEG-Guided Titration of Sevoflurane and Pediatric Anesthesia Emergence Delirium: A Randomized Clinical Trial. JAMA Pediatr. 2025 Apr 21. doi: 10.1001/jamapediatrics.2025.0517. Epub ahead of print. PMID: 40257811.
“Pediatric anesthesia emergence delirium (PAED) is a frequent complication in pediatric anesthesia. PAED is characterized and scored by behaviors such as lack of eye contact, non-purposeful movements, unawareness of surroundings, restlessness, and inconsolability.”1 Although mostly benign and time limited, it can lead to accidental removal of catheters and drains and injury. “Its etiology remains unknown, although it is associated with exposure to general anesthesia. Volatile anesthetics, particularly sevoflurane, increase propensity for PAED.3 Miyasaka et al. “hypothesized that titration of anesthesia based on interpretation of observed EEG waveforms and density spectral array spectrograms may allow more substantial reductions in exposure to sevoflurane, possibly decreasing the incidence of PAED.”1
In this single center, randomized controlled study, half of the patients were anesthetized by a single research anesthesiologist with training in EEG-guided titration of anesthesia targeting dominant slow-delta and alpha oscillatory patterns on the EEG shown to represent an unconsciousness state4-6 (EEG-guided group) vs standard 1.0–minimum alveolar concentration (MAC) sevoflurane anesthesia (control group). “A total of 30 (35%) in the control group and 19 (21%) in the EEG-guided group developed PAED (difference, 14%; 96.65% CI, −0.0019% to 28%; 95% CI, 0.92% to 27%; P = .04).”
Children in the EEG group were maintained on less sevoflurane with the mean (SD) sevoflurane concentration after intubation 0.9% (0.2) in the EEG-titration group compared to 2.5% (0.0) in the standard management group. Children in the EEG group were also exposed to less sevoflurane: 0.8 (0.5) MAC-hours compared to 2.5 (0.0) in the standard management group.
“Children in the EEG-guided group emerged a mean of 21.4 minutes (96.65% CI, 15.4 to 27.4 minutes) earlier from general anesthesia and spent a mean of 16.5 minutes less (96.65% CI, 10.8 to 22.3 minutes less) in the postanesthesia care unit.”1 In simple English, the EEG guided patients had less PAED, faster wake ups, and shorter PACU stays.
Sounds great, doesn’t it? But hold the phone. For the primary outcome of reducing PAED, the observed effect size was a 14% reduction, an impact that may or may not be clinically meaningful to pediatric anesthesiologists, PACU nurses, and parents/caretakers). The incidence of PAED (defined as a score >10) was high in both groups, even in the EEG guided study group (19/91 [21%] compared to 30/86 [35%] in the non-EEG guided group). “The reduction in incidence of PAED was marginal compared with the substantial change in exposure. In fact, using a PAED incidence cutoff of 12 or higher, considered equivalent in diagnostic accuracy, renders the difference nonsignificant. Furthermore, since an interim analysis was performed, the P value presented (P = .04) might not control for the overall type I error rate of the study at 5%.” A type I error rate, or alpha, refers to the “probability of incorrectly declaring an intervention different from its comparator, superior or inferior, when there is no true effect.” For a more in-depth biostatistical discussion, see the Editorial by VanBuren and French7 discussing type I error rate, or alpha, in clinical trials. Also, “[a]ccidental awareness during general anesthesia may be a concern at reduced levels of sevoflurane. In the absence of surgical stimuli and other anesthetic drugs, the MAC-awake, or the minimum concentration of inhaled anesthetic at which 50% of patients may be aroused by a standard stimulus, was reported to be 0.66% for sevoflurane in children aged 2 to younger than 5 years. The mean (SD) sevoflurane concentration during surgery in the EEG-guided group of 0.9% (0.2%) is approximately 1.4 times MAC-awake, which in conjunction with robust intraoperative antinociception provided an adequate level of unconsciousness to prevent awareness. We believe the dominant slow-delta and alpha EEG pattern represents a profoundly unconscious state and are confident that all participants in the EEG-guided group were unconscious for the duration of anesthesia management in this study. Reports of awareness under anesthesia in children are limited. While there were no spontaneous reports of awareness in this study, we did not perform direct questioning, and this report cannot account for delayed reporting, which may occur years later.”
Taking all these considerations into account, this study may not convince everyone to jump on the EEG bandwagon.
Of note, the Editorial by Kurth, et al.2 lays out a concise contextual background review with recommendations for the use and increased uptake of depth of anesthesia monitoring. Indeed, I (MY) would recommend using this as teaching tool in the ORs for all learners and would emphasize the following points:
“General anesthesia is a state comprising 4 components produced by drug inhibition of synaptic transmission within or between areas of the central nervous system (CNS). The 4 components are hypnosis (no consciousness), antinociception (no pain), immobility (no movement), and no awareness. Hypnosis results from inhibition of transmission between the frontal neocortex (executive center) and thalamus (sensory gateway) or to a lesser extent within the brainstem (reticular activating system). Antinociception and immobility (MAC) result from inhibition of transmission within the spinal cord. Depth of hypnosis can be monitored by processed EEG. Depth of antinociception can be monitored by heart rate and arterial pressure, reflecting sympathetic activation from the spinal cord to the sympathetic chain. Immobility can be assessed by observation or train-of-4 monitoring if a paralytic agent is administered. There is no monitoring of awareness, although the probability of awareness is related to the depth of hypnosis.”2
“To have a reliable EEG monitor for pediatric anesthesia, the device must display several channels of raw waveform, density spectral array (DSA), and several index numbers.8 When an EEG displays these features, it is termed a processed EEG.”2 It is useful guide to the depth of anesthesia for both inhalational agents like sevoflurane and IV general anesthetic agents like propofol.
Finally, Kurth et al ask: What will it take for EEG-guided anesthesia to become mainstream and a desert island monitor? “Three barriers need to be addressed to achieve mainstream adoption and standard for the specialty. First, hospitals should equip operating rooms with processed EEG, recognizing the superior quality and lower cost that this technology can offer. Second, pediatric anesthesiologists must be convinced that EEG-guided anesthesia can improve safety and outcomes. Clinical research, experience, and professional society standards will be necessary for this to occur. Third, pediatric anesthesiologists must be educated and trained on the technology. Of note, all patients in the EEG-guided anesthesia group in the study by Miyasaka et al were cared for by 1 study anesthesiologist, likely because the department lacked expertise in processed EEG–guided anesthesia. In our experience, this is common worldwide. For this reason, a Pediatric Anesthesia Learning Network (PALNET) was formed in 2022, composed of many departments in the United States and Canada, to train pediatric anesthesiologists in EEG-guided anesthesia. The PALNET deploys an effective educational curriculum9 and sends experts to departments to train a few anesthesiologists in the operating room. These trained anesthesiologists become experts to train their colleagues in the department.”2
Are you ready to get on the EEG bandwagon? Have you overcome the obstacles described above? Were there others? Send your thoughts and comments to Myron who will post in a Friday reader response.
References
1. Miyasaka KW, Suzuki Y, Brown EN, Nagasaka Y. EEG-Guided Titration of Sevoflurane and Pediatric Anesthesia Emergence Delirium: A Randomized Clinical Trial. JAMA pediatrics 2025 (In eng). DOI: 10.1001/jamapediatrics.2025.0517.
2. Kurth CD, Gabrielsen DA, Yuan I. EEG-Guided Pediatric Anesthesia-A Quality Innovation? JAMA pediatrics 2025 (In eng). DOI: 10.1001/jamapediatrics.2025.0514.
3. Mason KP. Paediatric emergence delirium: a comprehensive review and interpretation of the literature. British journal of anaesthesia 2017;118(3):335–343. DOI: 10.1093/bja/aew477.
4. Brown EN, Purdon PL, Akeju O, An J. Using EEG markers to make inferences about anaesthetic-induced altered states of arousal. British journal of anaesthesia 2018;121(1):325–327. (In eng). DOI: 10.1016/j.bja.2017.12.034.
5. Adam E, Kwon O, Montejo KA, Brown EN. Modulatory dynamics mark the transition between anesthetic states of unconsciousness. Proc Natl Acad Sci U S A 2023;120(30):e2300058120. (In eng). DOI: 10.1073/pnas.2300058120.
6. Bastos AM, Donoghue JA, Brincat SL, et al. Neural effects of propofol-induced unconsciousness and its reversal using thalamic stimulation. Elife 2021;10 (In eng). DOI: 10.7554/eLife.60824.
7. VanBuren JM, French B. Type I Error Control-Avoiding an Upset. JAMA pediatrics 2025 (In eng). DOI: 10.1001/jamapediatrics.2025.0528.
8. Purdon PL, Sampson A, Pavone KJ, Brown EN. Clinical Electroencephalography for Anesthesiologists: Part I: Background and Basic Signatures. Anesthesiology 2015;123(4):937–60. (In eng). DOI: 10.1097/aln.0000000000000841.
9. Yuan I, Missett RM, Jones-Oguh S, et al. Implementation of an electroencephalogram-guided propofol anesthesia education program in an academic pediatric anesthesia practice. Paediatric anaesthesia 2022 (In eng). DOI: 10.1111/pan.14520.