Authors: Bruzzone, Maria J. MD, MSCR et al

Anesthesia & Analgesia 140(5):p 1127-1139, May 2025. |

Abstract

Postoperative delirium (POD) is frequent in older adults and is associated with adverse cognitive and functional outcomes. In the last several decades, there has been an increased interest in exploring tools that easily allow the early recognition of patients at risk of developing POD. The electroencephalogram (EEG) is a widely available tool used to understand delirium pathophysiology, and its use in the perioperative setting has grown exponentially, particularly to predict and detect POD. We performed a systematic review to investigate the use of EEG in the pre-, intra-, and postoperative settings. We identified 371 studies, and 56 met the inclusion criteria. A range of techniques was used to obtain EEG data, from limited 1-4 channel setups to complex 256-channel systems. Power spectra were often measured preoperatively, yet the outcomes were inconsistent. During surgery, the emphasis was primarily on burst suppression (BS) metrics and power spectra, with a link between the frequency and timing of BS, and POD. The EEG patterns observed in POD aligned with those noted in delirium in different contexts, suggesting a reduction in EEG activity. Further research is required to investigate preoperative EEG indicators that may predict susceptibility to delirium.

Delirium, a neuropsychiatric syndrome secondary to a general medical condition, is characterized by an acute onset of attention deficits and altered cognition with a fluctuating course.¹^,² Postoperative delirium (POD) usually develops between 1 and 3 days after surgery, and its incidence is estimated between 11% and 51% in older adults. In this population, POD’s associated adverse outcomes include accelerated postoperative cognitive decline, loss of independence, and mortality.³ Older individuals with preoperative cognitive vulnerabilities and neurodegenerative disorders who elect surgery are at significantly greater risk of developing POD.^3–5 As the number of older adults with cognitive vulnerabilities and neurodegenerative disorders, as well as those electing for surgical procedures, rapidly increases, techniques aimed at predicting and preventing POD have become an important research focus.

The electroencephalogram (EEG), which measures and records the electrical activity of the cerebral cortex using scalp electrodes, is gaining attention as a noninvasive tool to assess the neurophysiology of delirium.²^,⁶^,⁷ Signals recorded from the scalp are traditionally grouped in frequency bands known as delta (1–3 Hz), theta (4–7 Hz), alpha (8–13 Hz), and beta (13–20 Hz).⁸ The alpha rhythm is the most prominent rhythm in an awake patient during a resting state with eyes closed. During delirium, slower frequencies (in the theta and delta range) are commonly described.⁹ EEG signals are interpreted, and postprocessing techniques can be used to calculate frequency bands’ power (absolute or relative). Furthermore, measurements of functional connectivity, which reflects the degree of synchronization of the activities of different brain regions, can also be obtained using postprocessing techniques.

Several studies have investigated the use of intraoperative limited EEG monitoring guided anesthesia delivery for preventing POD.^10–14 Limited EEG monitoring using processed EEG information from up to 4 channels is a method commonly used to assess anesthetic state¹⁵—with bispectral index (BIS) and SedLine monitors the most widely cited Food and Drug Administration (FDA)-approved devices in the United States. From these tools, power spectra measurements are acquired to help evaluate EEG trajectories during anesthesia induction and emergence.¹⁶ Limited and multichannel EEG devices have been used in pre- and intraoperative settings to try and predict POD—and in the postoperative setting to try and diagnose POD.

Figure 1.:

PRISMA flow chart representing the selection process. PRISMA indicates Preferred Reporting Items for Systematic Reviews and Meta-analyses.

Table 1. – EEG in the Preoperative Setting and POD

Author	Participants (POD, POD%)	Type of EEG used	Delta power	Theta power	Alpha power	Beta power	Gamma power	SEF
Numan et al⁷⁵	159 (29/18.24%)	4-channel EEG (Fp2-Pz and T8-Pz)	No significant differences among groups	No significant differences among groups	Not tested	Not tested	Not tested	Not tested
Gutierrez et al⁴⁰	30 (2/6.67%)	16-channel EEG	No significant differences among groups	No significant differences among groups	No significant differences among groups	No significant differences among groups	No significant differences among groups	Not tested
Tanabe et al³⁹	70 (22/31.43%)	256-channel EEG	No significant differences among groups	No significant differences among groups	Preoperative increase in POD^a	No significant differences among groups	No significant differences among groups	Not tested
Koch et al³⁶	237 (41/17.30%)	4-channel EEG (Fp1-Fp2, F7, F8)	No significant differences among groups	No significant differences among groups	No significant differences among groups	No significant differences among groups	Preoperative decrease in POD OR, 0.568, 95% CI, 0.342–0.944 P = .009	Preoperative decrease in POD Cutoff value for preop SEF of 17.75 Hz. Sensitivity of 0.944, specificity of 0.426, AUC of 0.718; 95% CI, 0.596–0.839, P = .004
Schussler et al⁶⁴	89 (28, 31%)	10 electrodes, analysis from F1 to F2	No significant differences among groups	No significant differences among groups	No significant differences among groups	Preoperative decrease in PACU-D^a	Preoperative decrease in PACU-D^a	Preoperative lower values in PACU-D 23.68 in PACU-D vs 28.81 in non PACU-D P = .001 AUC, 0.72 [0.61–0.83]

The table shows the most common EEG measurements investigated in the preoperative setting and results in subjects that develop POD compared to subjects that did not develop POD. Measures of association included when provided in the study.

Abbreviations: AUC, area under the curve; CI, confidence interval; EEG, electroencephalogram; OR, odds ratio; POD, postoperative delirium; SEF, spectral edge frequency.

^aEven though this result was described as significant, effect size measurements or values are not provided in the text or supplemental material.

The aim of this systematic review was to summarize the literature on EEG use in perioperative settings to predict or diagnose POD and describe common findings and potential clinical applications.

METHODS

Study Selection

The literature review included articles published until November 2023 and used the following search terms: EEG and POD (full search query details contained Supplemental Digital Content 1, Supplemental Information, https://links.lww.com/AA/E879). The search was conducted in PubMed, Embase, and Cochrane Library. Only articles in English were included. Two reviewers completed the initial abstract search and utilized the following inclusion criteria: human subjects older than 18 years, peer-reviewed articles written in English, EEG measured in the pre-, intra-, or postoperative settings, measurement of POD, and the use of an operationalized definition of delirium. Covidence was used to perform the abstract review and full-text review (M.B. and B.C.) once the appropriate articles were selected. Articles that analyzed EEG findings relative to POD were included. Conflicts were resolved by consensus. Once articles were chosen, 3 independent investigators (M.B., J.W., and M.S.) extracted publication information. The information extracted for analysis included (1) sample characteristics (number of participants, number of participants that developed POD, mean age, sex, country of study, type of surgery), (2) delirium outcome and measurement, (3) type of EEG used, (4) time of EEG obtained, (5) EEG measures used, and (6) EEG findings in individuals with and without POD (Supplemental Digital Content 2–4, Supplemental Tables 1, https://links.lww.com/AA/E880 2, https://links.lww.com/AA/E881, and 3, https://links.lww.com/AA/E882). The review followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines (Figure 1).¹⁷ Meta-analysis was not performed due to variability of EEG measurements at different time points.

Quality Assessment

We used the Newcastle-Ottawa Scale (NOS) for cohort studies to assess study quality. This scale focuses on 3 categories in each study: selection, comparability of cohorts, and outcome. Each category is composed of subcategories, and each one of the subcategories is rated with 1 star, with 2 to 4 stars maximum per category. The fewer stars a category receives, the bigger the risk of bias in the study. The selection category is divided into 4 subcategories: (1) representativeness of the exposed cohort: in our case, we checked if the cohort undergoing EEG measurement is a typical representation of patients in the perioperative setting, (2) selection of nonexposed cohort: this category was excluded in our analysis as most of the studies do not include a nonexposed cohort, (3) ascertainment of exposure: we ensured that EEG measurement methods were clearly described and consistently applied, and (4) demonstration that outcome of interest was not present at the start of the study: in our case, we checked for preoperative delirium screening. For the comparability category, 2 subcategories are described: (1) control for the most crucial factor: in our case, type of surgery and (2) control for additional factors: history of cognitive impairment before surgery and age not limited to older than 50 years old. Finally, for the outcome category, 3 subcategories are described: (1) assessment of outcome: if POD was checked using valid and reliable methods¹⁸ and (2) was follow-up long enough for outcomes to occur: in our case, if follow-up for POD was screened for at least 3 days after surgery (as POD has been reported to be more likely to happen in the first 3 days after surgery¹⁹) and postanesthesia care unit (PACU) delirium during the first 12 hours after surgery. The third subcategory, adequacy of follow-up cohorts, was deemed nonapplicable for the study and had no follow-up periods described after discharge.

RESULTS

Studies Retrieved

A total of 487 studies were imported for screening. 116 duplicates were removed, and 371 studies were screened. From those, 166 studies did not meet the inclusion criteria on abstract review, and 190 complete text studies were assessed for eligibility (Figure 1.). After a full-text review, 55 studies were included.

SAMPLES AND METHODOLOGICAL CHARACTERISTICS OF INCLUDED STUDIES

Sample Demographics

Studies were conducted in the United States, Germany, China, Japan, Australia, Chile, Belgium, Deutschland, and the Netherlands. The total number of participants studied was 13,046 of which 2706 developed POD (21%). The largest study recruited 1504 patients,²⁰ and the median (interquartile range, IQR [Q1–Q3]) sample size was 112 patients (56–256). The mean age of all participants was 69 (standard deviation [SD] = 6). Of the 55 studies, 17 included patients older than 18 years of age,¹⁶^,^20–35 with a mean age of 64 years, while the other 37 focused on patients older than 50.

EEG Measurements

Of the 55 studies included (Supplemental Digital Content 2–4, Supplemental Tables 1, https://links.lww.com/AA/E880 2, https://links.lww.com/AA/E881 and 3, https://links.lww.com/AA/E882), all used postprocessed EEG data, and 11²³^,^36–45 investigated EEG measurements in multiple settings. Twenty-four studies used data derived from 1 to 2 channel EEG monitoring devices such as BIS and SedLine,¹¹^,¹³^,¹⁶^,²⁰^,^24–27^,²⁹^,³⁶^,⁴¹^,⁴³^,^45–56 including BIS and power state index (PSI) values and suppression ratios, and 7 of them obtained spectral power EEG measures.¹⁶^,³⁶^,⁴³^,⁴⁵^,⁴⁷^,⁴⁸^,⁵⁶

Delirium Assessment

Delirium was assessed using the Confusion Assessment Method (CAM) or one of its versions in 43 studies: 17 used only the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU),¹⁶^,²⁵^,²⁹^,^33–35^,³⁸^,⁴¹^,^46–48^,^57–60 13 used only CAM²¹^,²⁴^,²⁶^,³⁰^,⁴⁰^,⁴³^,⁴⁹^,⁵¹^,^61–65 and 3 used only 3D CAM.²²^,⁴²^,⁵⁹ Eight studies used a combination of CAM and CAM-ICU³⁶^,⁴⁴^,⁵²^,⁵⁴^,⁵⁶^,^66–68 and 1 used a combination of the 3.¹³ Other methods used to assess POD were the Diagnostic and Statistical Manual of Mental Disorders (DSM) fourth or fifth edition in 5 studies,¹¹^,²⁷^,³⁶^,⁵³^,⁶⁹ the Delirium Rating Scale (DSR) in 4 studies,⁵⁵^,^70–72 the Saskatoon Delirium Checklist (SDC) in 3 studies,²³^,³²^,⁴⁵ and the delirium screening scale (NuDESC) in 1 study.⁷³ Most studies measured delirium twice daily from postoperative day 1 until discharge. Five studies focused on the PACU delirium.¹⁶^,³⁰^,³³^,³⁴^,³⁷ Of note, only 16 studies measured preoperative delirium.¹³^,³⁴^,⁴⁰^,⁴²^,⁵⁰^,^52–55^,⁵⁹^,⁶¹^,⁶³^,⁶⁷^,⁶⁸^,^72–74 In 28 studies, authors controlled for the presence of preoperative cognitive impairement.¹¹^,¹³^,¹⁶^,²⁰^,²²^,²⁴^,²⁶^,³⁶^,⁴⁰^,^42–44^,^48–53^,⁵⁵^,⁵⁸^,⁵⁹^,^61–63^,⁶⁵^,⁶⁸^,⁷⁰^,⁷²^,⁷³^,⁷⁵

Risk of Bias

The cohort studies included in the analyses were evaluated for risk of bias using the NOS for cohort studies (Supplemental Digital Content 5, Supplemental Figure 1, https://links.lww.com/AA/E883). The maximum number of stars for the selection category was 3, for the comparability category was 2, and for the outcome category was 2, with 7 stars being the maximum a study could receive. From the 40 cohort studies included in this analysis, 6 studies received 7 stars.⁴⁰^,⁵²^,⁵⁴^,⁶¹^,⁶³^,⁶⁸ Twenty-nine studies lost a star in the selection category for not controlling for the presence of preoperative delirium. Of those, 10 received 6 stars,^43–45^,⁵¹^,⁵⁸^,⁶⁴^,⁶⁵^,⁶⁷^,⁶⁹^,⁷⁵ 11 received 5 stars due to additionally not considering the type of surgery or age of patients included in the analysis²⁰^,²⁶^,³⁰^,³²^,³⁶^,^47–49^,⁵⁶ or not checking for preoperative cognitive disorder,³¹^,⁴¹ 7 studies obtained 4 stars because, limited delirium screening to less than 3 days,²¹^,²⁵^,²⁷^,²⁹^,³⁵^,³⁸^,⁷² and 1 obtained 3 points due to lack of detailed description of EEG methods utilized.⁶⁰

QUALITATIVE SYNTHESIS

EEG in the Preoperative Phase (16 Studies)

From the 16 studies included, 2 had a cross-sectional design,²³^,⁴² 12 were done prospectively,³⁶^,⁴⁰^,⁴¹^,^43–45^,⁵²^,⁶¹^,⁶³^,⁶⁴^,⁶⁶^,⁷² 1 was a randomized clinical trial (RCT),³⁷ and 1 was performed retrospectively³⁸ (Supplemental Digital Content 2, Supplemental Table 1, https://links.lww.com/AA/E880). Seven out of 16 studies used limited coverage EEG, most commonly bifrontal 4 channel EEG.³⁶^,⁴¹^,⁴³^,⁴⁵^,⁵²^,⁶¹^,⁷⁵ 8 studies used broader scalp coverage, from 10 or more channels³⁷^,⁴⁰^,⁴⁴^,⁶⁴^,⁷² to high-density 256 channels.³⁹ 7 studies performed the recording right before anesthesia induction,^36–38^,⁴⁰^,⁴²^,⁴⁵^,⁶⁴^,⁷⁵ and the other 9 recorded at least the day before surgery.²³^,³⁹^,⁴¹^,⁴³^,⁴⁴^,⁵²^,⁶¹^,⁶³^,⁷² In all studies, the samples were taken with the patients awake, eyes closed, and in a resting state. Only 1 study included sleep as well.⁶¹ The following EEG features were tested for associations with POD in the preoperative setting: (1) power spectrum, the most-used EEG measure tested in 9 studies,²³^,³⁶^,³⁷^,³⁹^,⁴⁵^,⁶³^,⁶⁴^,⁷²^,⁷⁵ (2) EEG asymmetry, reviewed in 1 study,⁴¹ (3) Direct current (DC) shift values analyzed in 1 study,⁴⁴ (4) BIS values studied in 2 studies,⁴⁵^,⁵² and (5) multiscale entropy analysis in 1 study.⁴² Preoperative EEG findings correlated with POD included: (1) lower preoperative spectral edge frequency (SEF) and reduced preoperative γ-band power,³⁶^,⁶⁴ (2) preoperative longer median total sleep, non-rapid eye movement (REM) stage 2 sleep duration and higher sleep efficiency,⁶¹ (3) higher alpha power and increased alpha band connectivity,³⁹ and (4) lower preoperative median dominant frequency (MDF) values,.⁴³ Three studies did not find significant changes in the power spectrum preoperatively,²³^,²⁸^,⁴⁰ and multiscale entropy analysis did not show preoperative differences between the patients that developed POD and those that did not.⁴² Other negative findings include no difference in preoperative DC shifts between patients with and without POD.⁴⁴ Two studies showed contradictory findings regarding BIS values: Bao et al found that preoperative average BIS values were significantly lower in the POD group, while Kinoshita et al⁴⁵ did not report a significant difference in preoperative BIS values between POD and non-POD groups. The most commonly used EEG parameters and findings in POD are shown in Table 1.

EEG in the Intraoperative Phase (38 Studies)

Four studies had a cross-sectional design,¹⁶^,²²^,²³^,⁴² 16 were performed prospectively,²⁰^,²⁶^,³⁰^,³¹^,³³^,⁴⁰^,⁴⁴^,⁴⁵^,^47–49^,⁵⁴^,⁵⁹^,⁶⁷^,⁶⁸^,⁷⁶ 8 were RCTs,¹¹^,¹³^,²⁴^,⁴⁶^,⁵⁰^,⁵³^,⁶²^,⁷³ and 9 had a retrospective design²¹^,²⁵^,²⁷^,³²^,³⁴^,³⁵^,³⁸^,⁵¹^,⁷⁷ (Supplemental Digital Content 3, Supplemental Table 2, https://links.lww.com/AA/E881). All the RCTs evaluated BIS-guided anesthesia (intervention) versus BIS-blinded anesthesia (control arm). Most studies do not specify from which intraoperative phase (induction, maintenance, or emergence) their data were obtained. BIS was the most-used device¹¹^,¹³^,¹⁶^,^24–26^,⁴¹^,^45–47^,⁵³ (11 studies), followed by multichannel (>16 leads) EEG²³^,³⁰^,³⁸^,⁴⁰^,⁴²^,⁵⁹^,⁶⁸^,⁷³ (8 studies), and the SedLine brain function monitoring system³⁶^,⁴⁸^,⁴⁹^,⁶²^,⁶⁷ (5 studies). The primary outcome for most studies was the presence or absence of POD, though 5 studies also focused on PACU delirium.¹⁶^,³⁰^,³¹^,³³^,³⁴ Delirium was predominately measured using the CAM or one of its variants. BIS values and the presence and duration of burst suppression were the most commonly used EEG measures,¹¹^,¹³^,^24–27^,⁴¹^,⁴⁶^,⁴⁷^,⁵³^,⁵⁴^,⁵⁹^,⁶⁵ followed by power spectra measurements,²³^,³⁰^,³¹^,³³^,³⁴^,³⁶^,⁴⁰^,⁴⁵^,⁴⁸^,⁶⁸ seeing more often in the studies published in the last 2 years. Two studies analyzed raw EEG data.³²^,⁶⁷ Three studies also incorporated cerebral oximetry measurements.²⁰^,³⁰^,⁶⁸ Cooter et al⁴⁷ developed a novel measure called the Duke Anesthesia Resistance Scale (DARS), a processed EEG measurement (using BIS) of brain resistance to volatile anesthesia. Other novel measures include DC shift amplitudes,⁴⁴ lateral interconnection ratio,³⁵ and EEG microstates.⁷³ Acker et al⁴² used multiscale entropy analysis.

Table 2. – Intraoperative Burst Suppression and POD

Study	Participants total/with POD, (POD%)	Type of EEG used	BS (incidence/duration)
Lele et al²¹	112/10 (8.93%)	C3′-Fz, Cz′-Fz, C4′-Fz, Fz-mastoid	Higher incidence of BS in POD 100% in POD vs 66.7% in non-POD RR, 1.5, 95% CI, 1.3–1.7, P = .03
Jung et al²²	80/13 (16.25%)	Narcotrend-Compact M monitor (MonitorTechnik)	Longer duration of BS/suppression in POD 27 min in POD vs 5 min in non-POD P = .03
Evered et al¹³	515/131 (25.44%)	BIS	Longer time in BS in POD 7 min (0–28) in POD vs 2 min (0–10) in non-POD P = .028
Fritz et al²⁶	618/162 (26.21%)	BIS Quatro (Medtronic)	Suppression at relatively lower end-tidal concentrations of volatile anesthetics in POD 35% of patients with BS at lower end-tidal concentration of anesthetics developed POD vs 17% of patients requiring higher end-tidal concentrations to develop BS OR, 2.13, 95% CI, 1.24–3.65
Soehle et al⁴¹	81/26 (32.1%)	BIS	Longer time in BS in POD 131 min (50–183) in POD vs 48 min (13–127) in non-POD AUC = 0.73, P = .001
Momeni et al²⁰	1504/303 (20.15%)	NeuroSENSE monitor	Higher magnitudes of EEG suppression in POD OR, 2.247; 95% CI, 1.414–3.571 P = .001^a
Muhlhofer et al⁴⁹	41/7 (17.7%)	SedLine VR brain monitor (SEDLine Inc), 4-lead strip	Longer time in suppression P = .039^a
Xu et al⁵⁰	255/47 (18.43%)	SedLine brain function monitor (SEDLine Inc)	Longer time in suppression in POD 1.7 min in POD vs 0 min in non-POD P = .04
Pedemonte et al⁵¹	159/23 (14.47%)	SedLine monitor (SEDLine Inc), 4 channels (Fp1, Fp2, F7, and F8)	Incidence of BS higher in POD OR:4.1 95% CI: 1.5–13.7, P = .012^a
Ren et al⁶⁵	85/10 (11.76%)	BIS	Incidence of POD higher in BS 6 patients with POD in BS group vs 4 patients with POD in non BS group OR, 4.954, 95% CI, 1.034–23.736, P = .045
Koch et al⁵³	1058/198 (18.71%)	BIS (Fp1, Fp2, F7, and F8)	BS duration longer in POD 27.5 min (± 21.3) in POD vs 21.4 min (± 16.2) in non-POD OR, 1.011, 95% CI, 1.000–1.022, P = .046
Tang et al⁶²	201/34 (16.92%)	SedLine brain function monitor (SEDLine, Inc)	No relation between time on suppression and POD 2.5 min in POD vs 1.4 min in non-POD P = 0.53
Reese et al⁵⁹	83/12 (14.46%)	32-channel EEG	Longer EEG suppression time in POD OR, 1.34, 95% CI, 1.01–1.78, P = .043^a

This table summarize studies that looked at BS incidence and or duration and POD. Measures of association in included when provided in the study.

Abbreviations: AUC, area under the curve; BIS, bispectral index; BS, burst suppression; CI, confidence interval; EEG, electroencephalogram; OR, odds ratio; POD, postoperative delirium.

^aEven though this result was described as significant, effect size values are not provided in the text or supplemental material.

Table 3. – Postoperative Delta Power and POD

Study	Participants	Type of EEG	Delta activity
Evans et al⁷⁰	12/3 (25.00%)	Frontopolar, temporal, central, and occipital EEG	Greater delirium severity associated with greater post operatory day 1 waking delta power r = 0.84, P = 0.02
Numan et al⁷⁵	156/29 (18.59%)	Fp2-Pz and T8-Pz	Increase in delta power in POD group No delirium 0.44 (median, IQR, 0.32–058) vs delirium 0.67 (median, IQR, 0.55–0.74) P ≤ .0001
Ditzel et al⁵⁵	145/47 (32.41%)	Single channel Fp2-Pz	Increase in polymorphic delta activity (PDA) score (measured by algorithm) in POD group PDA score correlated with likelihood of delirium Rs =.37, 95% CI, 0.25–0.47
Xue et al⁶⁰	46/11 (23.91%)	Wearable headset device from Nicolet (Natus Medical Inc), NFS	Increase in relative delta power in POD group 80.29 in POD vs 71.02 in non-POD P = .0417

Summary of studies investigating delta activity in the postoperative setting and POD. Measures of association included when provided in the study.

Abbreviations: CI, confidence interval; EEG, electroencephalogram; IQR, interquartile range; NFS, no further specified; POD, postoperative delirium.

Regarding the significance of intraoperative BIS values, researchers reported contradictory findings for POD. Radtke et al found low BIS values (<20) and Chan et al found higher times with low BIS values (31–49) in patients who developed POD. However, Whitlock et al did not find a longer intraoperative time with BIS <20 in patients who developed POD, and Thudium et al did not find a correlation between intraoperative BIS values and POD. Most studies that looked at the burst suppression ratio or time in burst suppression reported an association between greater intraoperative burst suppression, longer duration of a single episode of suppression on EEG, and longer cumulative duration of EEG suppression and the development of POD.¹³^,¹⁶^,^20–22^,²⁶^,³⁸^,⁴¹^,⁴⁹^,⁵³^,⁵⁴^,⁶⁵ Only 1 study, by Tang et al,⁶² concluded that time in burst suppression did not predict the risk of POD (Table 2). Reese et al⁵⁹ investigated the total time spent in preburst suppression and burst suppression and did not find a significant relationship between POD and the percentage spent in preburst suppression or the total time spent in either preburst suppression or burst suppression. The studies that looked at power spectra showed a decrease in absolute alpha band power after induction³⁶ and lower absolute and relative alpha power during anesthesia,³¹^,³³^,⁴⁰^,⁴⁵^,⁶⁸ lower power in higher EEG frequencies,³¹^,³³^,⁴⁸ lack of spindle power¹⁶ and lack of negative slope for total power across all bands³⁴ during anesthesia emergence among patients that developed POD. During induction of anesthesia, POD patients showed significantly greater DC shifts than non-POD patients.⁴⁴ Connectivity measurements, particularly in the fronto-parietal theta connectivity, were positively associated with POD in 1 study. The same study found no differences in alpha power measurements in the posterior channels among POD and non-POD groups.³⁰ Regarding EEG complexity, Acker et al described no significant differences in this measurement between patients who develop and those who do not develop POD. Cooter et al showed that a low DARS value predicts the development of POD. Studies that checked raw EEG demonstrated a higher incidence of epileptiform discharges,⁶⁷ and >50% decrease in fast frequency amplitude or a >50% increase in theta or delta activity after intubation³² in patients with POD. When looking at EEG microstates, patients with POD had decreased duration of microstate D and increase duration of microstate C.⁷³ Patients that developed POD showed decreased lateral interconnection ratio (LIR) at the end of the surgery.³⁵

EEG IN THE POSTOPERATIVE PHASE (13 STUDIES)

Two studies had a cross-sectional design,⁵⁷^,⁷⁰ 11 were conducted prospectively,²⁹^,³⁶^,⁴³^,⁵⁵^,⁵⁶^,⁵⁸^,⁶⁰^,⁶⁶^,⁶⁹^,⁷⁵ and 1 was an RCT³⁷ (Supplemental Digital Content 4, Supplemental Table 3, https://links.lww.com/AA/E882). Only 1 study performed the EEG before testing for delirium.⁶⁰ Multichannel EEG³⁷^,³⁹^,⁵⁷^,⁵⁸^,^69–71 was the most commonly studied measure, with 2 authors using BIS²⁹ and SedLine³⁶ systems. The primary outcome of all studies was the presence or absence of delirium, and CAM/CAM-ICU were the most-used tools for delirium diagnosis. Spectra and power analysis were the most common EEG measurements,²⁸^,³⁹^,⁴³^,⁵⁶^,⁶⁰^,^69–71^,⁷⁵^,⁷⁶ while functional connectivity was used by van Dellen et al, BIS index was used by Plaschke et al, and visual analysis of the raw EEG was used by Tschernatsch et al⁵⁸ and Ditzel et al.⁵⁵ The main findings in patients that developed POD included an overall increase in slow wave activity and delta power (Table 3),⁵⁵^,⁶⁰^,⁷⁰^,⁷¹^,⁷⁵ reduced beta band power,³⁶ and loss in functional connectivity.⁵⁷^,⁶⁹ Decrease alpha relative power (eyes closed state) and increase theta relative power (eye open state) correlated with CAM delirium severity in 1 study.⁵⁶ White et al⁷¹ also found an independent association between postoperative EEG slowing and delirium severity. Patients with delirium had lower MDF.⁴³ Plascheke et al²⁹ showed reduced bilateral BIS index in delirious patients compared to nondelirious patients. Interestingly, Tschernatsch et al⁵⁸ described a high incidence of postoperative seizures in patients who underwent open chamber cardiac surgery, and the presence of electrographic seizures was associated with the presence of POD.

DISCUSSION

Engel and Romano first described diffuse EEG slowing in patients with delirium in 1944, though this predates diagnostic criteria for delirium.⁷⁸ Since then, EEG changes have usually been described in patients that develop delirium,⁷⁹ with increased dominance of slower delta (<4 Hz) and theta (4–8 Hz) frequencies being the most prominent changes seen on raw EEGs. It has been proposed that delirium emerges when there is a failure in the integration and processing of sensory information and motor responses and that vulnerability could be determined by impairment in baseline network connectivity and an increase in the inhibitory tone within the brain.⁸⁰ As a result, the interest in using EEG to identify patients at risk of developing delirium is growing as it can provide an objective measure of both. The perioperative setting offers a unique opportunity to investigate EEG characteristics in patients at risk of developing delirium in the postoperative setting, as elective procedures often allow a preoperative evaluation to assess delirium risk.

Over the past 3 years, attention to EEG’s value for POD has surged dramatically. Figure 2 shows the annual publication count from the years 2010 to 2023. The last 3 years show the most publications (34, with half of the publications from 2023) on this topic, more than in the preceding decade (22) highlighting the increased interest in exploring this technique for delirium prediction and detection in the perioperative setting.

Figure 2.:

Use of EEG in the perioperative setting. Number of articles published per year on the use of EEG for perioperative delirium prediction and diagnosis from 2010 to 2023. EEG indicates electroencephalogram.

We identified 55 studies that address the use of EEG in the perioperative setting to predict or diagnose POD. While most of the included studies had a low risk of bias, there was tremendous heterogeneity in several aspects, including sample sizes, study designs, type of surgeries, and mainly on the types of devices used to assess EEG signals, as well as in the EEG measures and level of postprocessing used to interpret the data in the pre-, intra-, and postoperative periods, at times generating conflicting results. Heterogeneous measurements of association were used for reporting the data as well, with most studies using area under the curve (AUC) and odds ratio (OR), as described in Tables 1–3.

Preoperative Data

Cognitive decline is a known risk factor for developing POD,⁸¹^,⁸² and recent studies highlight the association among the Alzheimer’s disease (AD) biomarkers (particularly those in the ATN framework) and the development of POD.⁸³^,⁸⁴ However, despite guidelines, preoperative evaluations typically do not include routine cognitive screenings, potentially leaving this risk unidentified before surgery. Previous EEG studies in patients with mild cognitive impairment (MCI) and dementia have led various researchers to propose that the EEG might serve as an effective instrument for identifying individuals with neurocognitive disorders. These patients could then receive a more comprehensive neuropsychological assessment before undergoing surgery⁸⁵ with appropriate recommendations for intraoperative considerations and follow-up.⁸⁶

In our review, we found 5 studies that conducted analyses on the preoperative power of frequencies during resting state. The majority of these studies did not observe significant preoperative differences in the power of delta, theta, or alpha frequencies when comparing patients who developed POD with those who did not. (Table 1).³⁶^,³⁹^,⁴⁰^,⁶⁴^,⁷⁵ When looking at the faster frequencies, only 2 studies demonstrated a decrease in power on beta and gamma bands in patients with POD,³⁶^,⁶⁴ while the other 2 studies did not see such a difference between groups.³⁹^,⁴⁰ These 2 studies included a lower number of participants

It should be noted that frequent EEG observations in individuals with AD and MCI, as reported in existing studies, are marked by a notable rise in the power of slower frequencies alongside a considerable decline in the power of faster frequencies.⁸⁷ This pattern reflects a slowing down of EEG rhythms, aligning with results from some of the previously mentioned research.³⁶^,⁶⁴

Other measures that have been shown to discriminate between patients with AD and controls and have been tested preoperatively are lower mean frequency and, to a lesser extent, lower 95% SEF (a measurement commonly used to monitor depth of anesthesia).⁸⁸ Two studies in our review revealed lower preoperative SEF in patients who developed POD.³⁶^,⁶⁴ This is not surprising as prior studies have demonstrated reduced EEG SEF activity in older adults.^89–91

Some of the limitations we found while analyzing the data included the variability in the metrics used in preoperative studies, the different types of devices used, heterogeneity in sample sizes, demographics, and type of surgeries which complicate the process of deriving clear interpretations and extrapolating results from current studies. While the concept of utilizing EEG in preoperative scenarios is appealing due to its potential for rapidly and inexpensively identifying patients at risk for delirium, further research is imperative. This research should concentrate on validated measurements and tools and meticulous selection of patient populations. An optimal study would involve a large cohort of older patients undergoing similar surgical procedures, using a standardized EEG device, and focusing on easily replicable measurements in everyday clinical practice. Collaborative efforts among neuropsychologists, anesthesiologists, and neurophysiologists are crucial for effectively addressing a complex issue like predicting POD using EEG.

Intraoperative Data

Tracking anesthesia state with limited-channel intraoperative monitoring devices allows analysis of specific EEG changes occurring intraoperatively with the development of POD. The main intraoperative EEG features associated with POD include increased burst suppression time and decreased alpha power.

In the present review, most of the studies looking at intraoperative power spectra showed that patients who exhibit (1) intraoperative slowing of brain activity, represented by decreased in absolute alpha band power after induction³⁶ and lower absolute and relative alpha power during anesthesia³¹^,³³^,⁴⁰^,⁴⁵^,⁶⁸; (2) lower power in higher EEG frequencies in patients that received ketamine⁴⁸; and (3) lack of spindle power during anesthesia emergence¹⁶ have higher odds of developing POD. Prior studies describe that preoperative cognitive function correlates with intraoperative frontal alpha power, with lower preoperative frontal alpha power seen in association with lower preoperative cognitive function, a known risk factor for POD.⁹²

Cortical EEG recordings reflect the underlying metabolic state of the brain. EEG suppression is seen in the setting of severe reduction of brain metabolic activity. Several studies reviewed here (Table 2) have correlated more time in EEG suppression and burst suppression during surgery under general anesthesia with higher incidences of POD,¹³^,¹⁶^,^20–22^,²⁶^,³⁸^,⁴¹^,⁴⁹^,⁵³^,⁵⁴^,⁶⁵ but this finding does not necessarily represent a causal relationship between these 2 variables. Moreover, 1 study focused on anesthetic sensitivity, defined as the increased propensity for burst suppression at low anesthetic doses, and found that increased anesthetic sensitivity and not burst suppression alone was associated with the development of POD.²⁶ Most studies do not account for variations in anesthetic dosage, and inclusion of this factor in future research is appropriate.⁸⁵

Several groups have tried to determine, with conflicting results, whether the use of processed EEG monitoring devices helps to decrease the incidence of POD.

The Electroencephalography Guidance of Anesthesia to Alleviate Geriatric Syndromes RCT was published in 2019. This RCT involved 1232 patients older than 60 years undergoing major surgery. Patients were randomized to receive EEG-guided anesthesia versus usual anesthetic care. In the trial, patients who received EEG-guided anesthetic administration had no significant differences in the incidence of POD compared to patients randomized to usual care.⁹³ A meta-analysis published after this study, based on 5 studies and 3612 patients, revealed no significant effect of EEG-based monitoring on POD.⁹⁴ Conversely, a more recent multicenter RCT by Evered et al,¹³ where 655 at-risk patients undergoing major surgery were assigned to light (BIS 50) or deep (BIS 35) anesthesia, demonstrated that the development of POD was significantly lower in patients randomized to light anesthesia. Shortly thereafter, an updated meta-analysis published conflicting results by concluding that large, well-conducted RCTs are needed to answer the question of whether the use of processed EEG devices to control anesthesia delivery intraoperatively is associated with an incidence reduction of POD.⁹⁵

Postoperative Data

Consistent with prior literature, most of the studies that recorded EEG during POD showed slowing of the frequencies, loss of faster frequencies, or both.²⁹^,³⁶^,³⁹^,⁵⁵^,⁵⁶^,^69–71 (Table 3). The synthesis of these findings with existing literature emphasizes the multifactorial nature of delirium pathophysiology, which integrates metabolic, inflammatory, and neurotransmitter imbalances with impaired neuronal network connectivity.¹ Both hypoxia and hypoglycemia can impair brain function and are present in some patients with delirium.^96–101 Experimental hypoxia and hypoglycemia in humans also produce EEG slowing.¹⁰¹^,¹⁰² It is established that anticholinergic medications increase delirium risk, and it has been shown that acetylcholine (ACh) receptor antagonists can trigger generalized slowing as well as decrease alpha frequencies on EEG.¹⁰³

Few studies reviewed focused on connectivity measurements that were also disrupted during delirium.⁵⁷^,⁷⁵ This loss of functional connectivity⁵⁷^,⁶⁹ demonstrated in POD is believed to be indicative of the failure of network connectivity. Several studies have shown that decreased connectivity strength and efficiency may characterize structural brain networks of patients at risk for delirium, and it is suggested that functional network impairment could be the final common pathway for the syndrome.¹⁰⁴

Even though the use of EEG during delirium is increasingly popular and can help identify patients suffering from acute delirium, its diagnostic role is unestablished. Delirium is a clinical syndrome and its diagnosis is still based on clinical criteria.

LIMITATIONS

This systematic review has several limitations. Comparison of EEG techniques across studies is limited by heterogeneity of study designs, study population, type of surgery, and anesthesia approach, different devices used to record EEG, different postprocessing techniques, and different methods used to evaluate POD. Different limited EEG monitors have different algorithms to process the information. Several studies do not include measurements of association, or when included they use different measurements (AUC and OR) which limited the possibility to perform a more focused meta-analysis of the data. Information on preoperative cognitive profiles was available in less than half of the studies, so the effect of preoperative cognitive deficits on EEG cannot be adequately studied. Sample sizes were very variable, from studies including 12 patients to studies including 1504. Only full-text articles published in English were included.

CONCLUSIONS

EEG’s role in preoperatively predicting POD requires detailed evaluation. Uniform study designs are essential for reliable data comparison. Analyses should focus on processed EEG measurements that have previously been associated with POD risk factors, like cognition, and that have the potential to be easily reproducible. This could lead to more precise POD prediction algorithms. Additionally, the relationship between preoperative EEG and neurodegeneration (ATN) biomarkers within the perioperative setting remains unexplored and warrants attention in future research. While intraoperative EEG shows potential in identifying at-risk patients, optimal usage for POD prevention or reduction is unclear. Systematic preoperative, intraoperative, and postoperative EEG data collection and accurate POD patient identification are key to developing better risk identification algorithms. Integrating this with preoperative EEG and clinical data, including biomarkers of neurodegeneration, promises enhanced predictive models for POD.

REFERENCES

1. Wilson JE, Mart MF, Cunningham C, et al. Delirium. Nat Rev Dis Primers. 2020;6:90.

2. Boord MS, Moezzi B, Davis D, et al. Investigating how electroencephalogram measures associate with delirium: a systematic review. Clin Neurophysiol. 2021;132:246–257.

3. Inouye SK, Westendorp RG, Saczynski JS. Delirium in elderly people. Lancet. 2014;383:911–922.

4. Koirala B, Hansen BR, Hosie A, et al. Delirium point prevalence studies in inpatient settings: a systematic review and meta-analysis. J Clin Nurs. 2020;29:2083–2092.

5. Fick DM, Steis MR, Waller JL, Inouye SK. Delirium superimposed on dementia is associated with prolonged length of stay and poor outcomes in hospitalized older adults. J Hosp Med. 2013;8:500–505.

6. van der Kooi AW, Rots ML, Huiskamp G, et al. Delirium detection based on monitoring of blinks and eye movements. Am J Geriatr Psychiatry. 2014;22:1575–1582.

7. van der Kooi AW, Zaal IJ, Klijn FA, et al. Delirium detection using EEG: what and how to measure. Chest. 2015;147:94–101.

8. Shafi MM, Santarnecchi E, Fong TG, et al. Advancing the neurophysiological understanding of delirium. J Am Geriatr Soc. 2017;65:1114–1118.

9. Jacobson S, Jerrier H. EEG in delirium. Semin Clin Neuropsychiatry. 2000;5:86–92.

10. Sieber FE, Zakriya KJ, Gottschalk A, et al. Sedation depth during spinal anesthesia and the development of postoperative delirium in elderly patients undergoing hip fracture repair. Mayo Clin Proc. 2010;85:18–26.

11. 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;110(suppl 1):i98–105.

12. Yang Y, Song C, Song C, Li C. Comparison of bispectral index-guided individualized anesthesia with standard general anesthesia on inadequate emergence and postoperative delirium in elderly patients undergoing esophagectomy: a retrospective study at a single center. Med Sci Monit. 2020;26:e925314-1.

13. Evered LA, Chan MTV, Han R, et al. Anaesthetic depth and delirium after major surgery: a randomised clinical trial. Br J Anaesth. 2021;127:704–712.

14. Fritz BA, King CR, Mickle AM, et al.; ENGAGES Research Group. Effect of electroencephalogram-guided anaesthesia administration on 1-yr mortality: follow-up of a randomised clinical trial. Br J Anaesth. 2021;127:386–395.

15. Bowdle TA. Depth of anesthesia monitoring. Anesthesiol Clin. 2006;24:793–822.

16. 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.

17. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62:e1–34.

18. Jones RN, Cizginer S, Pavlech L, et al.; Better Assessment of Illness (BASIL) Study Group. Assessment of instruments for measurement of delirium severity: a systematic review. JAMA Intern Med. 2019;179:231–239.

19. Ansaloni L, Catena F, Chattat R, et al. Risk factors and incidence of postoperative delirium in elderly patients after elective and emergency surgery. Br J Surg. 2010;97:273–280.

20. Momeni M, Meyer S, Docquier MA, et al. Predicting postoperative delirium and postoperative cognitive decline with combined intraoperative electroencephalogram monitoring and cerebral near-infrared spectroscopy in patients undergoing cardiac interventions. J Clin Monit Comput. 2019;33:999–1009.

21. Lele AV, Furman M, Myers J, Kinney G, Sharma D, Hecker J. Inadvertent burst suppression during total intravenous anesthesia in 112 consecutive patients undergoing spinal instrumentation surgery: a retrospective observational quality improvement project. J Neurosurg Anesthesiol. 2021;34:300–305.

22. Jung C, Hinken L, Fischer-Kumbruch M, et al. Intraoperative monitoring parameters and postoperative delirium: results of a prospective cross-sectional trial. Medicine (Baltim). 2021;100:e24160.

23. Hofsté WJ, Linssen CA, Boezeman EH, Hengeveld JS, Leusink JA, de-Boer A. Delirium and cognitive disorders after cardiac operations: relationship to pre- and intraoperative quantitative electroencephalogram. Int J Clin Monit Comput. 1997;14:29–36.

24. Chan MTV, Cheng BCP, Lee TMC, Gin T; CODA Trial Group. BIS-guided anesthesia decreases postoperative delirium and cognitive decline. J Neurosurg Anesthesiol. 2013;25:33–42.

25. Thudium M, Heinze I, Ellerkmann RK, Hilbert T. Cerebral function and perfusion during cardiopulmonary bypass: a plea for a multimodal monitoring approach. Heart Surgery Forum. 2018;21:028–E35.

26. Fritz BA, Maybrier HR, Avidan MS. Intraoperative electroencephalogram suppression at lower volatile anaesthetic concentrations predicts postoperative delirium occurring in the intensive care unit. Br J Anaesth. 2018;121:241–248.

27. Santarpino G, Sirch J, Ackermann B, Fischlein TJ, Pfeiffer S. Impact of bispectral index monitoring on postoperative delirium in patients undergoing aortic surgery. Interact Cardiovasc Thorac Surg. 2010;10:S114.

28. Kim J, Lee HC, Byun SH, et al. Frontal electroencephalogram activity during emergence from general anaesthesia in children with and without emergence delirium. Br J Anaesth. 2021;126:293–303.

29. Plaschke K, Fichtenkamm P, Schramm C, et al. Early postoperative delirium after open-heart cardiac surgery is associated with decreased bispectral EEG and increased cortisol and interleukin-6. Intensive Care Med. 2010;36:2081–2089.

30. Vlisides PE, Li D, Maywood M, et al. Electroencephalographic biomarkers, cerebral oximetry, and postoperative cognitive function in adult noncardiac surgical patients: a prospective cohort study. Anesthesiology. 2023;139:568–579.

31. Ostertag J, Engelhard A, Nuttall R, et al. Development of postanesthesia care unit delirium is associated with differences in aperiodic and periodic alpha parameters of the electroencephalogram during emergence from general anesthesia: results from a prospective observational cohort study. Anesthesiology. 2024;140:73–84.

32. Al-Qudah AM, Ta’ani OA, Thirumala PD, et al. Role of intraoperative neuromonitoring to predict postoperative delirium in cardiovascular surgery. J Cardiothorac Vasc Anesth. 2023;38:526–533.

33. Lutz R, Muller C, Dragovic S, et al. The absence of dominant alpha-oscillatory EEG activity during emergence from delta-dominant anesthesia predicts neurocognitive impairment- results from a prospective observational trial. J Clin Anesth. 2022;82:110949.

34. Dragovic S, Schneider G, Garcia PS, et al. Predictors of low risk for delirium during anesthesia emergence. Anesthesiology. 2023;139:757–768.

35. Baron Shahaf D, Hight D, Kaiser H, Shahaf G. Association between risk of stroke and delirium after cardiac surgery and a new electroencephalogram index of interhemispheric similarity. J Cardiothorac Vasc Anesth. 2023;37:1691–1699.

36. Koch S, Windmann V, Chakravarty S, et al.; BioCog Study Group. Perioperative electroencephalogram spectral dynamics related to postoperative delirium in older patients. Anesth Analg. 2021;133:1598–1607.

37. Kim H, McKinney A, Brooks J, Mashour GA, Lee U, Vlisides PE. Delirium, caffeine, and perioperative cortical dynamics. Front Hum Neurosci. 2021;15:744054.

38. Ma O, Crepeau AZ, Dutta A, Bliss DW. Anticipating postoperative delirium during burst suppression using electroencephalography. IEEE Trans Biomed Eng. 2020;67:2659–2668.

39. Tanabe S, Mohanty R, Lindroth H, et al. Cohort study into the neural correlates of postoperative delirium: the role of connectivity and slow-wave activity. Br J Anaesth. 2020;125:55–66.

40. Gutierrez R, Egaña JI, Saez I, et al. Intraoperative low alpha power in the electroencephalogram is associated with postoperative subsyndromal delirium. Front Syst Neurosci. 2019;13:56–65.

41. Soehle M, Dittmann A, Ellerkmann RK, Baumgarten G, Putensen C, Guenther U. Intraoperative burst suppression is associated with postoperative delirium following cardiac surgery: a prospective, observational study. BMC Anesthesiol. 2015;15:1–8.

42. Acker L, Ha C, Zhou J, et al. Electroencephalogram-based complexity measures as predictors of post-operative neurocognitive dysfunction. Front Syst Neurosci. 2021;15:1–11.

43. Kim J, Park S, Kim KN, et al. Resting-state prefrontal EEG biomarker in correlation with postoperative delirium in elderly patients. Front Aging Neurosci. 2023;15:1224264.

44. Windmann V, Dreier JP, Major S, Spies C, Lachmann G, Koch S. Increased direct current-electroencephalography shifts during induction of anesthesia in elderly patients developing postoperative delirium. Front Aging Neurosci. 2022;14:921139.

45. Kinoshita H, Saito J, Kushikata T, et al. The perioperative frontal relative ratio of the alpha power of electroencephalography for predicting postoperative delirium after highly invasive surgery: a prospective observational study. Anesth Analg. 2023;137:1279–1288.

46. Whitlock EL, Torres BA, Lin N, et al. Postoperative delirium in a substudy of cardiothoracic surgical patients in the BAG-RECALL clinical trial. Anesth Analg. 2014;118:809–817.

47. Cooter Wright M, Bunning T, Eleswarpu SS, et al. A processed electroencephalogram-based brain anesthetic resistance index is associated with postoperative delirium in older adults: a dual center study. Anesth Analg. 2022;134:149–158.

48. Barreto Chang OL, Kreuzer M, Morgen DF, Possin KL, García PS. Ketamine associated intraoperative electroencephalographic signatures of elderly patients with and without preoperative cognitive impairment. Anesth Analg. 2022;135:683–692.

49. Muhlhofer WG, Zak R, Kamal T, et al. Burst-suppression ratio underestimates absolute duration of electroencephalogram suppression compared with visual analysis of intraoperative electroencephalogram. Br J Anaesth. 2017;118:755–761.

50. Xu N, Li LX, Wang TL, et al. Processed multiparameter electroencephalogram-guided general anesthesia management can reduce postoperative delirium following carotid endarterectomy: a randomized clinical trial. Front Neurol. 2021;12:666814.

51. Pedemonte JC, Plummer GS, Chamadia S, et al. Electroencephalogram burst-suppression during cardiopulmonary bypass in elderly patients mediates postoperative delirium. Anesthesiology. 2020;133:280–292.

52. Bao L, Liu T, Zhang Z, et al. The prediction of postoperative delirium with the preoperative bispectral index in older aged patients: a cohort study. Aging Clin Exp Res. 2023;35:1531–1539.

53. Koch S, Blankertz B, Windmann V, Spies C, Radtke FM, Rohr V. Desflurane is risk factor for postoperative delirium in older patients’ independent from intraoperative burst suppression duration. Front Aging Neurosci. 2023;15:1067268.

54. Fang PP, Shang ZX, Xu J, et al. Contribution of intraoperative electroencephalogram suppression to frailty-associated postoperative delirium: mediation analysis of a prospective surgical cohort. Br J Anaesth. 2023;130:e263–e271.

55. Ditzel FL, Hut SC, Dijkstra-Kersten SM, et al. An automated electroencephalography algorithm to detect polymorphic delta activity in acute encephalopathy presenting as postoperative delirium. Psychiatry Clin Neurosci. 2022;76:676–678.

56. Guay CS, Kafashan M, Huels ER, et al. Postoperative delirium severity and recovery correlate with electroencephalogram spectral features. Anesth Analg. 2023;136:140–151.

57. van Dellen E, van der Kooi AW, Numan T, et al. Decreased functional connectivity and disturbed directionality of information flow in the electroencephalography of intensive care unit patients with delirium after cardiac surgery. Anesthesiology. 2014;121:328–335.

58. Tschernatsch M, Juenemann M, Alhaidar F, et al. Epileptic seizure discharges in patients after open chamber cardiac surgery: a prospective prevalence pilot study using continuous electroencephalography. Intensive Care Med. 2020;46:1418–1424.

59. Reese M, Christensen S, Anolick H, et al.; MADCO-PC and INTUIT Investigators. EEG pre-burst suppression: characterization and inverse association with preoperative cognitive function in older adults. Front Aging Neurosci. 2023;15:1229081.

60. Xue Y, Liu W, Su L, He H, Chen H, Long Y. Quantitative electroencephalography predicts postoperative delirium in cardiac surgical patients after cardiopulmonary bypass: a prospective observational study. Front Med (Lausanne). 2023;10:1163247.

61. Ibala R, Mekonnen J, Gitlin J, et al. A polysomnography study examining the association between sleep and postoperative delirium in older hospitalized cardiac surgical patients. J Sleep Res. 2021;30:e13322.

62. Tang CJ, Jin Z, Sands LP, et al. ADAPT-2: a randomized clinical trial to reduce intraoperative EEG suppression in older surgical patients undergoing major noncardiac surgery. Anesth Analg. 2020;131:1228–1236.

63. Ross JM, Santarnecchi E, Lian SJ, et al.; SAGES II Study Group. Neurophysiologic predictors of individual risk for post-operative delirium after elective surgery. J Am Geriatr Soc. 2023;71:235–244.

64. Schussler J, Ostertag J, Georgii MT, et al. Preoperative characterization of baseline EEG recordings for risk stratification of post-anesthesia care unit delirium. J Clin Anesth. 2023;86:111058.

65. Ren S, Zang C, Yuan F, et al. Correlation between burst suppression and postoperative delirium in elderly patients: a prospective study. Aging Clin Exp Res. 2023;35:1873–1879.

66. Tanabe S, Parker M, Lennertz R, Pearce RA, Banks MI, Sanders RD. Reduced electroencephalogram complexity in postoperative delirium. J Gerontol A Biol Sci Med Sci. 2022;77:502–506.

67. Li N, Liu X, Gao Y, et al. Association of electroencephalogram epileptiform discharges during cardiac surgery with postoperative delirium: an observational study. Front Surg. 2022;9:900122.

68. Khalifa C, Lenoir C, Robert A, et al. Intra-operative electroencephalogram frontal alpha-band spectral analysis and postoperative delirium in cardiac surgery: a prospective cohort study. Eur J Anaesthesiol. 2023;40:777–787.

69. Numan T, Slooter AJC, van der Kooi AW, et al. Functional connectivity and network analysis during hypoactive delirium and recovery from anesthesia. Clin Neurophysiol. 2017;128:914–924.

70. Evans JL, Nadler JW, Preud’homme XA, et al. Pilot prospective study of post-surgery sleep and EEG predictors of post-operative delirium. Clin Neurophysiol. 2017;128:1421–1425.

71. White MF, Tanabe S, Casey C, et al. Relationships between preoperative cortical thickness, postoperative electroencephalogram slowing, and postoperative delirium. Br J Anaesth. 2021;127:236–244.

72. Arai N, Miyazaki T, Nakajima S, et al. The association between electroencephalography with auditory steady-state response and postoperative delirium. J Pers Med. 2022;13:1–15.

73. Neuner B, Wolter S, McCarthy WJ, et al. EEG microstate quantifiers and state space descriptors during anaesthesia in patients with postoperative delirium: a descriptive analysis. Brain Commun. 2023;5:fcad270.

74. Lutz NW, Viola A, Malikova I, et al. Inflammatory multiple-sclerosis plaques generate characteristic metabolic profiles in cerebrospinal fluid. PLoS One. 2007;2:e595.

75. Numan T, van den Boogaard M, Kamper AM, Rood PJT, Peelen LM, Slooter AJC; Dutch Delirium Detection Study Group. Delirium detection using relative delta power based on 1-minute single-channel EEG: a multicentre study. Br J Anaesth. 2019;122:60–68.

76. Koch S, Feinkohl I, Chakravarty S, et al.; BioCog Study Group. cognitive impairment is associated with absolute intraoperative frontal α-band power but not with baseline α-band power: a pilot study. Dement Geriatr Cogn Disord. 2019;48:83–92.

77. Ren B, Guan W, Zhou Q, Wang Z. EEG-based driver fatigue monitoring within a human-ship-environment system: implications for ship braking safety. Sensors. 2023;23:1–17.

78. Romano J, Engel GL. Delirium: I. Electroencephalographic data. Arch Neurol Psychiat. 1944;51:356–377.

79. Palanca BJA, Wildes TS, Ju YS, Ching S, Avidan MS. Electroencephalography and delirium in the postoperative period. Br J Anaesth. 2017;119:294–307.

80. Sanders RD. Hypothesis for the pathophysiology of delirium: Role of baseline brain network connectivity and changes in inhibitory tone. Med Hypotheses. 2011;77:140–143.

81. Rong X, Ding ZC, Yu HD, Yao SY, Zhou ZK. Risk factors of postoperative delirium in the knee and hip replacement patients: a systematic review and meta-analysis. J Orthop Surg Res. 2021;16:76.

82. Visser L, Prent A, Banning LBD, van Leeuwen BL, Zeebregts CJ, Pol RA. Risk factors for delirium after vascular surgery: a systematic review and meta-analysis. Ann Vasc Surg. 2021;76:500–513.

83. Payne T, Taylor J, Casey C, et al. Prospective analysis of plasma amyloid beta and postoperative delirium in the interventions for postoperative delirium: biomarker-3 study. Br J Anaesth. 2023;130:546–556.

84. Fong TG, Vasunilashorn SM, Gou Y, et al.; RISE Study Group. Association of CSF Alzheimer’s disease biomarkers with postoperative delirium in older adults. Alzheimers Dement (N Y). 2021;7:e12125.

85. Berger M, Ryu D, Reese M, et al. A Real-time neurophysiologic stress test for the aging brain: novel perioperative and ICU applications of EEG in older surgical patients. Neurotherapeutics. 2023;20:975–1000.

86. Price CC. The new frontier of perioperative cognitive medicine for alzheimer’s disease and related dementias. Neurotherapeutics. 2022;19:132–142.

87. Meghdadi AH, Stevanovic Karic M, McConnell M, et al. Resting state EEG biomarkers of cognitive decline associated with Alzheimer’s disease and mild cognitive impairment. PLoS One. 2021;16:e0244180.

88. Poza J, Hornero R, Abasolo D, Fernandez A, Garcia M. Extraction of spectral based measures from MEG background oscillations in Alzheimer’s disease. Med Eng Phys. 2007;29:1073–1083.

89. Scally B, Burke MR, Bunce D, Delvenne J-F. Resting-state EEG power and connectivity are associated with alpha peak frequency slowing in healthy aging. Neurobiol Aging. 2018;71:149–155.

90. Musaeus CS, Nielsen MS, Østerbye NN, Høgh P. Decreased parietal beta power as a sign of disease progression in patients with mild cognitive impairment. J Alzheimers Dis. 2018;65:475–487.

91. Hughes LE, Rittman T, Robbins TW, Rowe JB. Reorganization of cortical oscillatory dynamics underlying disinhibition in frontotemporal dementia. Brain. 2018;141:2486–2499.

92. Giattino CM, Gardner JE, Sbahi FM, et al.; MADCO-PC Investigators. Intraoperative frontal alpha-band power correlates with preoperative neurocognitive function in older adults. Front Syst Neurosci. 2017;11:24.

93. 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.

94. Sun Y, Ye F, Wang J, et al. Electroencephalography-guided anesthetic delivery for preventing postoperative delirium in adults: an updated meta-analysis. Anesth Analg. 2020;131:712–719.

95. Sumner M, Deng C, Evered L, et al. Processed electroencephalography-guided general anaesthesia to reduce postoperative delirium: a systematic review and meta-analysis. Br J Anaesth. 2022;130:e243–e253.

96. Yokota H, Ogawa S, Kurokawa A, Yamamoto Y. Regional cerebral blood flow in delirium patients. Psychiatry Clin Neurosci. 2003;57:337–339.

97. Pfister D, Siegemund M, Dell-Kuster S, et al. Cerebral perfusion in sepsis-associated delirium. Crit Care. 2008;12:R63.

98. Caplan GA, Kvelde T, Lai C, Yap SL, Lin C, Hill MA. Cerebrospinal fluid in long-lasting delirium compared with Alzheimer’s dementia. J Gerontol A Biol Sci Med Sci. 2010;65:1130–1136.

99. Kealy J, Murray C, Griffin EW, et al. Acute inflammation alters brain energy metabolism in mice and humans: role in suppressed spontaneous activity, impaired cognition, and delirium. J Neurosci. 2020;40:5681–5696.

100. Bendahan N, Neal O, Ross-White A, Muscedere J, Boyd JG. Relationship between near-infrared spectroscopy-derived cerebral oxygenation and delirium in critically ill patients: a systematic review. J Intensive Care Med. 2019;34:514–520.

101. Sejling AS, Kjær TW, Pedersen-Bjergaard U, et al. Hypoglycemia-associated changes in the electroencephalogram in patients with type 1 diabetes and normal hypoglycemia awareness or unawareness. Diabetes. 2015;64:1760–1769.

102. Wang D, Yee BJ, Wong KK, et al. Comparing the effect of hypercapnia and hypoxia on the electroencephalogram during wakefulness. Clin Neurophysiol. 2015;126:103–109.

103. Ebert U, Kirch W. Scopolamine model of dementia: electroencephalogram findings and cognitive performance. Eur J Clin Invest. 1998;28:944–949.

104. van Montfort SJT, van Dellen E, Stam CJ, et al. Brain network disintegration as a final common pathway for delirium: a systematic review and qualitative meta-analysis. Neuroimage Clin. 2019;23:101809.

Electroencephalographic Measures of Delirium in the Perioperative Setting