CLINICAL TRIAL

published: 12 July 2021

doi: 10.3389/fneur.2021.666814

Processed Multiparameter
Electroencephalogram-Guided
General Anesthesia Management
Can Reduce Postoperative Delirium
Following Carotid Endarterectomy: A
Randomized Clinical Trial
Na Xu 1 , Li-Xia Li 1 , Tian-Long Wang 1*, Li-Qun Jiao 2 , Yang Hua 3 , Dong-Xu Yao 1 , Jie Wu 1 ,
Yan-Hui Ma 1 , Tian Tian 1 and Xue-Li Sun 1
1
Department of Anesthesiology, Xuanwu Hospital, Capital Medical University, Beijing, China, 2 Department of Neurosurgery,
Xuanwu Hospital, Capital Medical University, Beijing, China, 3 Department of Vascular Ultrasound, Xuanwu Hospital, Capital
Medical University, Beijing, China

Edited by:
Jacqueline Leung, Background: Patients undergoing carotid endarterectomy (CEA) for severe
University of California, San Francisco,
United States
carotid stenosis are vulnerable to postoperative delirium, a complication frequently
Reviewed by:
associated with poor outcome. This study investigated the impact of processed
Wolfgang Muhlhofer, electroencephalogram (EEG)-guided anesthesia management on the incidence of
University of Alabama at Birmingham,
postoperative delirium in patients undergoing CEA.
United States
Rochelle Zak, Methods: This single-center, prospective, randomized clinical trial on 255 patients
University of California, San Francisco,
United States
receiving CEA under general anesthesia compared the outcomes of patient state index
*Correspondence:
(PSI) monitoring [SEDLine Brain Function Monitor (Masimo, Inc, Irvine, CA)] (standard
Tian-Long Wang group, n = 128) with PSI combined with density spectral array(DSA) -guided monitoring
[email protected]
(intervention group, n = 127) to reduce the risk of intraoperative EEG burst suppression.
Specialty section:
All patients were monitored by continuous transcranial Doppler ultrasound (TCD) and
This article was submitted to near-infrared spectroscopy (NIRS) to avoid perioperative cerebral hypoperfusion or
Neurocritical and Neurohospitalist hyperperfusion. According to the surgical process, EEG suppression time was calculated
Care,
a section of the journal separately for three stages: S1 (from anesthesia induction to carotid artery clamping), S2
Frontiers in Neurology (from clamping to declamping), and S3 (from declamping to the end of surgery). The
Received: 16 February 2021 primary outcome was incidence of postoperative delirium according to the Confusion
Accepted: 09 June 2021
Assessment Method algorithm during the first 3 days post-surgery, and secondary
Published: 12 July 2021
outcomes were other neurologic complications and length of hospital stay.
Citation:
Xu N, Li L-X, Wang T-L, Jiao L-Q, Results: There were no episodes of cerebral hypoperfusion or hyperperfusion
Hua Y, Yao D-X, Wu J, Ma Y-H, Tian T
and Sun X-L (2021) Processed
according to TCD and NIRS monitoring in either group during surgery. The incidence
Multiparameter of postoperative delirium within 3 days post-surgery was significantly lower in the
Electroencephalogram-Guided
intervention group than the standard group (7.87 vs. 28.91%, P < 0.01). In the
General Anesthesia Management Can
Reduce Postoperative Delirium intervention group, the total EEG suppression time and the EEG suppression time during
Following Carotid Endarterectomy: A S2 and S3 were shorter (Total, 0 “0” vs. 0 “1.17” min, P = 0.04; S2 , 0 “0” vs. 0 “0.1”
Randomized Clinical Trial.
Front. Neurol. 12:666814.
min, P < 0.01; S3 , 0 “0” vs. 0 “0” min, P = 0.02). There were no group differences in
doi: 10.3389/fneur.2021.666814 incidence of neurologic complications and length of postoperative hospital stay.

Frontiers in Neurology | www.frontiersin.org 1 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

Conclusion: Processed electroencephalogram-guided general anesthesia

management, consisting of PSI combined with DSA monitoring, can significantly
reduce the risk of postoperative delirium in patients undergoing CEA. Patients,
especially those exhibiting hemodynamic fluctuations or receiving surgical procedures
that disrupt cerebral perfusion, may benefit from the monitoring of multiple EEG
parameters during surgery.
Clinical Trial Registration: www.ClinicalTrials.gov, identifier: NCT03622515.

Keywords: delirium, carotid endarterectomy, electroencephalogram-guided, general anesthesia, cerebral

perfusion, monitoring

KEY POINTS disorganized thinking (11, 12). Postoperative delirium (POD)

is a common yet serious geriatric syndrome that afflicts 10–
• Question: Can EEG-guided general anesthesia management 60% of patients after major surgery (13–16) and up to 91%
reduce the risk of postoperative delirium in high-risk patients of the critically ill (17). Further, POD is associated with worse
undergoing CEA? early and long-term prognosis as well as greater healthcare
• Findings: Processed electroencephalography-guided costs (11, 14, 17–19). Therefore, the United Kingdom National
anesthesia management, including PSI combined with Institute for Health and Care Excellence, the American Geriatric
DSA monitoring, can significantly reduce postoperative Society, the American College of Surgeons, and the American
delirium incidence in patients undergoing CEA. Society of Anesthesiologists have all identified the prevention of
• Meaning: This randomized clinical trial suggests that slight postoperative delirium as a public health priority (20–23).
changes in anesthetic management may play a major role in Risk factors for delirium are usually divided into predisposing
modulating postoperative delirium risk in patients undergoing and precipitating (24). Predisposing factors include preoperative
CEA under general anesthesia. Vulnerable patients, especially vulnerabilities, while precipitating factors are potentially
those exhibiting hemodynamic fluctuations or undergoing reversible events occurring throughout the perioperative period.
surgical procedures that disrupt cerebral perfusion, may POD incidence increases with longer cumulative intraoperative
benefit from the monitoring of multiple EEG parameters electroencephalogram (EEG) suppression duration (25, 26),
during surgery. and processed EEG monitoring during surgery may lower
POD rate and prevent or minimize EEG suppression by using
INTRODUCTION minimal anesthetic doses (27–31). A burst suppression pattern
on the EEG indicates severe inhibition of neuronal activity and
Carotid endarterectomy (CEA) is the gold standard treatment metabolic rate. Patients with severe CS are prone to vascular
for patients with severe carotid stenosis (CS) to reduce the risk cerebral injury, and may therefore be particularly vulnerable
of stroke (1–4). The safety and long-term efficacy of CEA for to EEG suppression (32, 33). It is not clear, however, whether
the management of carotid artery disease has been demonstrated processed EEG-guided anesthesia management can reduce POD
in large randomized controlled trials (5–8). However, cerebral incidence in patients undergoing CEA and if this effect depends
blood supply may be severely disrupted by anesthesia and on decreasing EEG suppression.
surgical manipulations during CEA, and cerebral function is The objectives of this study were to investigate the effect of
highly vulnerable to even brief changes in oxygen and blood processed EEG-guided anesthesia management on the incidence
supply. Cerebral vascular diseases may further increase the of POD in patients receiving CEA under general anesthesia. The
risk of perioperative neurological dysfunction. For instance, primary hypothesis was that processed EEG-guided anesthesia
there is a significant association between carotid artery stenosis management can effectively reduce the likelihood of the
and postoperative delirium (9, 10), an acute state of mental incidence of POD during the first 3 days following CEA.
confusion defined by alterations in attention, consciousness, and

Abbreviations: CEA, carotid endarterectomy; EEG, electroencephalogram; TCD,

METHODS
transcranial Doppler ultrasound; NIRS, near-infrared spectroscopy; CS, carotid
stenosis; POD, postoperative delirium; CAM, Confusion Assessment Method;
Research Design
MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment; This is a single-center, prospective, randomized clinical trial
SDS, Self-rating Depression Scale; SAS, Self-rating Anxiety Scale; ASA, American with two parallel arms. The research protocol was approved
Society of Anesthesiologists; rSO2 , regional cerebral oxygenation; MFVMCA , mean by the Ethics Committee of Xuanwu Hospital of Capital
flow velocity of the middle cerebral artery; BSR, burst suppression ratio; DSA, Medical University (LYS[2018]053) and written informed
density spectral array; PSI, patient state index; MAP, mean arterial blood pressure;
PET CO2 , end-expiratory carbon dioxide partial pressure; SD, standard deviation;
consent was obtained from all subjects participating in the
IQR, interquartile range; OR, odds ratio; ACA, anterior communicating artery; trial. The trial was registered prior to patient enrollment
CBF, cerebral blood flow; CAS, carotid artery stenting; EMG, electromyogram. at clinicaltrials.gov (NCT03622515, Principal investigator: Na

Frontiers in Neurology | www.frontiersin.org 2 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

Xu, Date of registration: August 8, 2018). All surgeries were a 1:1 ratio. A seed was not specified, and blocks were not
conducted at Xuanwu Hospital (Beijing, China). Results are used in randomization. Randomization was conducted after
reported according to CONSORT guidelines. patient consent for research participation during the preoperative
interview. Both patients and research associates conducting
Participants preoperative testing and postoperative outcome assessments
Study candidates were screened by investigators the day before were also blinded to group assignment.
surgery (or on Friday for those scheduled for surgery the
following Monday). We recruited patients at our institution from Anesthesia and Surgery
August 2018 to December 2019. All surgeries were performed by General anesthesia was induced in all patients by intravenous
the same surgical and anesthesia team. Inclusion criteria included etomidate 0.15 mg·kg−1 , and sufentanil 0.2 mg·kg−1 , and
patients scheduled to undergo elective CEA under general maintained by continuous propofol infusion. In the standard
anesthesia, fluency in Chinese Mandarin, an anticipated length of monitoring group, propofol dosage was 50–80 mcg·kg−1 ·min−1 ,
hospital stay over 3 days after surgery, and willingness to comply while in the intervention group, dosage was adjusted according
with the research protocol. Exclusion criteria were as follows: to sedative depth as described in the next section (Intervention).
(1) declining preoperative condition (such as unstable angina, All patients received remifentanil and dexmedetomidine as well
acute myocardial infarction, and heart function NYHA III to IV) as rocuronium (0.6 mg·kg−1 ) or cisatracurium (0.15 mg·kg−1 )
within 4–6 weeks before surgery, (2) severe hepatic dysfunction injection for muscle relaxation. Patients were mechanically
(Child–Pugh grade C) or renal failure (requirement for renal ventilated with FiO2 50%, and SaO2 was maintained at more
replacement therapy), (3) history of schizophrenia, Parkinson’s than 95%. Goal-directed fluid and vasoconstrictive drug therapies
disease, or traumatic brain injury, (4) inability to perform were conducted to maintain stroke volume variability below
neurocognitive testing, (5) preoperative delirium according to 13% and regulated patients’ blood pressure within 180 mmHg
the Confusion Assessment Method (CAM) algorithm (34), (6) according to transcranial Doppler ultrasound (TCD, China
preoperative cognitive impairment according to the Chinese Shenzhen Delica Medical Equipment, ShenZhen, China) and
Mini-Mental State Examination (MMSE) corrected for education near-infrared spectroscopy (NIRS, Cas Medical Systems, Inc,
level (illiterate ≤ 19 points, 1–6 years of primary school ≤ Branford, Connecticut, USA). Intraoperative warming devices
22 points, middle school or above ≤ 26 points) and Montreal were used to maintain nasopharyngeal temperature between
Cognitive Assessment (MoCA) score (illiterate ≤ 13 points, 36◦ C and 37◦ C. Perioperative care was standardized according
primary school ≤ 19 points, middle school or above ≤ 24 to institutional routines for all patients.
points) (35, 36), (7) preoperative depression according the Self- All CEA procedures were performed by the same team
rating Depression Scale (SDS, score > 41 points) or preoperative of 4 experienced neurosurgeons who have worked for more
anxiety according to the Self-rating Anxiety Scale (SAS, score than 15 years. As there is a correlation between intraoperative
> 41 points) (37, 38), (8) change in surgical procedure after hypotension and incidence of delirium (45, 46), all patients were
anesthesia, (9) return to the intensive care unit following surgery, monitored by TCD and NIRS during CEA to minimize the risks
(10) conditions that caused severe hemodynamic fluctuations of perioperative cerebral hypoperfusion and hyperperfusion.
(such as severe allergic reactions or major bleeding), and (11) Regional cerebral oxygenation (rSO2 ) was also monitored
accidental discharge. by NIRS and mean flow velocity of the middle cerebral
artery (MFVMCA ) by TCD. The NIRS probes were placed
Baseline Data Collection on the bilateral forehead. Cerebral ischemia was deemed to
Baseline data included demographics, co-morbidities, and have occurred if the ipsilateral MFVMCA was reduced by
relevant physical and laboratory findings. Preoperative physical >50% compared to baseline or rSO2 decreased by 20% from
condition was evaluated using the American Society of baseline during the clamping process (47–50). An increase
Anesthesiologists (ASA) Physical Status Classification System in MFVMCA of 100% after carotid declamping compared to
(39). Activities of daily living were assessed using the Barthel baseline was considered indicative of cerebral hyperperfusion
Index (score range 0–100, with higher score indicating better (51). Anesthesiologists managed cerebral hypoperfusion or
independent function) (40). Cognitive functions were assessed hyperperfusion during CEA by regulating the patient’s blood
using the MMSE and MoCA. Anxiety and depression were pressure, which was elevated or reduced by 10% before
assessed using the SAS and SDS because several studies have and after declamping the carotid artery, respectively. Carotid
identified preoperative depression (41–43) and anxiety (44) as shunting was conducted based on TCD monitoring and the
risk factors for postoperative delirium incidence or duration. surgeon’s discretion.
Delirium status was assessed with the CAM. Pre-surgical tests
of baseline general cognition, delirium, anxiety, and depression Intervention
were conducted by a neuropsychologist blinded to group During surgery, patients from both the groups were monitored
allocation (see below). using the SEDLine Brain Function Monitor (Masimo, Inc, Irvine
CA), which uses symmetrical bifrontal electrodes to measure
Randomization and Blinding four channels of raw EEG data with separate displays for
Patients were assigned to intervention and control arms before electromyogram (EMG), artifacts (e.g., patient motion), burst
surgery using a computerized random number generator at suppression ratio (BSR), and density spectral array (DSA). The

Frontiers in Neurology | www.frontiersin.org 3 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

SEDLine monitor also estimates sedative depth from digital training process was repeated at 6-month intervals during
EEG waves using a proprietary algorithm and displays it as the study.
a dimensionless parameter called the patient state index (PSI,
ranging 0–100, with 100 indicating wakefulness and 0 isoelectric Magnetic Resonance Image Data Acquisition
EEG) (52). Even though there are some raw EEG patterns Cerebral microembolization is a significant contributor to
clearly visible on the spectrogram plots, we are nowhere close postoperative delirium (57, 58). Therefore, magnetic resonance
to being able to describe the cerebrum depending on the imaging examinations were conducted before and 24 h following
information obtained from the processed EEG. Therefore, in surgery using a Clinical 3-Tesla whole-body MR imager (Verio;
the intervention group, we visually analyzed and inspected Siemens Medical Solutions, Erlangen, Germany). In accordance
DSA as an intervention indicator for EEG burst suppression with previous studies (59–62), new ischemic cerebral lesions were
rather than EEG suppression waveforms (53, 54). Whenever defined as hyperintense regions on post-intervention diffusion-
the DSA indicated burst suppression, propofol dosage was weighted images that were not present on pretreatment images.
reduced 5 mcg·kg−1 ·min−1 , as clinically permitted. In order Ischemic lesions (number and total volume of hyperintense
to prevent intraoperative awareness, if lowering propofol was regions) were evaluated by a radiologist and neurologist both
found to cause PSI to reach 60, but DSA still showed burst blinded to the research protocol, and disagreements were
suppression, we stopped adjusting propofol and adjusted blood resolved by consensus.
pressure according to the stage of surgery to ensure cerebral
perfusion. For instance, blood pressure was elevated by 5% Measurement of Anesthetic Doses
before declamping the carotid artery or decreased by 5% after To determine whether processed EEG-guidance resulted in
declamping. In the standard group, the monitor screen was propofol dosage reduction, we calculated the cumulative
masked, all EEG data and spectrograms were blinded, and doses of all anesthetics according to intraoperative electronic
only PSI values were displayed. Anesthesiologists performed medical records.
anesthesia by conventional methods and maintained the PSI at
25–50 in the standard group. Measurement of Vital Signs
Mean arterial blood pressure (MAP), PSI, end-expiratory carbon
Outcome Assessments dioxide partial pressure (PET CO2 ), bilateral rSO2 , and bilateral
Measurement of Delirium MFVMCA were recorded at three time-points, immediately
Delirium was assessed by trained research team members after general anesthesia induction and before carotid artery
preoperatively and daily on the first three postoperative days clamping as baseline (T1 ), immediately after clamping (T2 ),
using the Chinese version of CAM algorithm, which has and after declamping and subsequent stabilization of cerebral
demonstrated good reliability and validity among the Chinese perfusion (T3 ).
population (55). Postoperative visits were conducted between
10:00 A.M. and 16:00 P.M. at the patient’s bedside. Delirium Measurement of Secondary Outcomes
was defined by acute onset with fluctuating course, inattention, Physical and neurological examinations were conducted by a
disorganized thinking, and (or) altered level of consciousness. neurologist blinded to group allocation before and for the
first 3 days after surgery. Examinations included evaluation
Measurement of EEG Suppression
of neurological deficits according to the National Institutes of
The raw EEG data, PSI, and BSR were obtained from all patients
Health Stroke Scale. From the beginning of anesthesia to 3
using the SEDLine monitor. The EEG recordings were initiated
days after surgery, we also monitored adverse events such as
from anesthesia induction, and ended at the completion of
intraoperative movement or awareness. Finally, length of hospital
surgical manipulation. All EEG data were then edited and saved.
stay was recorded.
The experienced neurophysiologists independently reviewed the
intraoperative EEG traces acquired by the SEDLine R monitoring
system for the entire duration of the operation, recorded Statistical Analysis
whether or not burst suppression was present, and calculated Postoperative Outcome Analysis
the cumulative duration of EEG suppression in minutes. The The Kolmogorov–Smirnov test was used to check for the
neurologists recorded an epoch as having a burst suppression normality of all continuous variables. Continuous datasets
pattern if there was at least 5 s of suppression of the EEG with a normal distribution were compared by independent-
tracing present in a given 30 s epoch (56). We recorded the samples t-test, and continuous datasets with non-normal
cumulative duration of total EEG suppression and the duration distributions by independent-samples Mann–Whitney U-tests.
of EEG suppression for three surgical stages: S1 (from anesthesia Categorical data were compared by χ 2 test or Fisher exact
induction to carotid artery clamping), S2 (from clamping to test as indicated. Measurement data at each time point were
declamping), and S3 (from declamping to the end of surgery). compared between the two groups by analysis of variance
The anesthesiologists received training by neurophysiologists with repeated measures (RT-ANOVA). Normally distributed
in reading and interpreting the DSA before the study was continuous variables are reported as mean ± standard deviation
initiated to ensure that all anesthesiologists could reach 100% (s.d.) and non-normally distributed continuous variables as
consensus on the interpretation of the DSA. Additionally, the median (interquartile range, i.q.r.).

Frontiers in Neurology | www.frontiersin.org 4 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

FIGURE 1 | Flow diagram depicting patient recruitment for this clinical trial. EEG, electroencephalography.

Statistical analyses were conducted using SPSS R Version RESULTS

21(IBM Inc. Chicago, IL, USA). All tests were two-tailed, and a
two-sided P < 0.050 was considered statistically significant. Patient Recruitment
This clinical trial was conducted from August 2018 to December
Sample Size Calculation 2019. Overall, 366 patients were deemed eligible and 272
We calculated the sample size based on the expected incidence patients were included, with patients randomly and equally
of the primary outcome (postoperative delirium). In a pilot divided into processed EEG-guided anesthetic management
study, 20 patients meeting the inclusion criteria were included (intervention) and standard anesthesia care groups (Figure 1).
and randomly divided into an intervention group and standard Anesthesiologists adhered to the protocol for all 136 patients
care group, with 10 patients in each group. One patient assigned to the intervention group. During the study period,
in the standard care group had postoperative delirium, five surgeries were changed to carotid stenting (four cases in
whereas patients in the intervention group did not have the intervention group and one in the standard care group,
any postoperative neurological complications. Therefore, and they did not receive any of the study intervention in
postoperative neurological complications rates where 0% in either group), 10 cases did not complete the postoperative
the interventional group (to avoid errors, the incidence was CAM test (five cases in each group), and two cases in the
increased of 0.5% for calculating the sample size) and 10% in standard care group were transferred to the intensive care unit
the standard group. We set type I error α = 0.05 and the type after surgery. Therefore, 255 patients were included in the final
II error β = 0.1. From this, the sample size of each group was analysis, 127 in the intervention group and 128 in the standard
estimated at 91 cases using the sample size software PASS11 care group (Figure 1). Preoperative demographic, surgical, and
(NCSS, Caseville, Utah, United States). Assuming a 20% loss intraoperative variables for each group are summarized in
to follow-up, we had planned to enroll at least 220 patients Tables 1, 2. Baseline variables, such as age, sex ratio, preoperative
undergoing CEA. co-morbidity, and cognitive and mood assessments, among

Frontiers in Neurology | www.frontiersin.org 5 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

TABLE 1 | Patient demographic and clinical information. TABLE 2 | Intraoperative variables.

Variable Intervention group Standard group P-value Variable Intervention Standard P-values
(n = 127) (n = 128) group care group
(n = 127) (n = 128)
Age (y)†† 62.31 ± 8.02 63.16 ± 7.17 0.37
Sex ratio (F:M)§§ 18:109 15:113 0.56 Surgery duration (min)*## 135 (73.5) 139.5 (100.25) 0.50

Height (cm)†† 168.74 ± 6.98 169.55 ± 6.31 0.33 Anesthesia duration (min)*## 213 (71.5) 216 (98.75) 0.75

Body weight (kg)†† 71.61 ± 10.03 71.88 ± 10.16 0.83 Clamping duration(min)*## 38 (24.75) 35.5 (32) 0.94

BMI (kg/m2 )†† 25.11 ± 2.80 24.95 ± 2.81 0.65 Type of surgery (CEA: CEA+CAS)§§ 91:36 87:41 0.52

Preoperative co-morbidity Total fluid infusion(ml)*## 1,100 (400) 1,100 (500) 0.73

Hypertension## 92 (72.44) 86 (67.19) 0.36 Crystal liquid (ml)*## 900 (400) 1,000 (400) 0.66

Diabetes ##
42 (33.07) 42 (32.81) 0.96 Urine output(ml)*## 700 (600) 700 (600) 0.86

Coronary artery disease## 27 (21.26) 28 (21.88) 0.90 Estimated blood loss(ml)*## 20 (20) 20 (20) 0.24

Previous stroke ##
72 (56.69) 69 (53.91) 0.66 Intraoperative drugs

ASA fitness grade (III: IV)§§ 37:90 31:97 0.38 Propofol (mcg/kg/min)*## 52.13 (14.11) 52.45 (15.98) 0.58

Preoperative level of function and mood Remifentanil (mcg/kg/min)*## 0.20 (0.08) 0.21 (0.06) 0.32

Barthel Index score* ††

98.94 ± 3.79 99.35 ± 3.18 0.45 Dexmedetomidine (mcg/kg/h)*## 0.29 (0.12) 0.27 (0.11) 0.48
††
Education (y) 9.93 ± 3.96 9.44 ±4.41 0.35 Values in parentheses are percentages unless indicated otherwise; values are mean ±
MMSE score††† 27.41 ± 2.06 27.09 ± 2.54 0.26 s.d. and *median (i.q.r.). CEA, carotid endarterectomy; CAS, carotid artery stenting.
§§ The P-value is from the χ2 test, except. ## The P-value is from the independent-samples
MoCA score††† 23.98 ± 2.53 23.63 ± 3.31 0.34
Mann–Whitney U-test.
SDS score‡†† 23.65 ± 4.99 24.45 ± 5.60 0.23
SAS score‡†† 23.56 ±4.37 23.80 ± 4.92 0.67

Values in parentheses are percentages unless indicated otherwise; values are expressed
TABLE 3 | Postoperative outcomes and related intraoperative variables.
as mean ± standard deviation (s.d.) or median (interquartile range, i.q.r.). *Score ranges
†
from 0 to 100, with higher score indicating better function. Score ranges from 0 to
Variable Intervention Standard P-value
30, with higher score indicating better function. ‡ Score ranges from 20 to 80, with
group group
higher score indicating worse mood. BMI, body mass index; MMSE, Mini-Mental State
(n = 127) (n = 128)
Examination; MoCA, Montreal Cognitive Assessment; SDS, Self-rating Depression Scale;
SAS, Self-rating Anxiety Scale.
§§ The P-value is from the χ2 test, except. †† The P-value is from the independent-samples Postoperative outcome
t-test. ## The P-value is from the independent-samples Mann–Whitney U-test. Incidence of delirium within 3 days 10 (7.87) 37 (28.91) 0.000
New cerebral infarctions (symptomatic)§§ 1 (0.79) 3 (2.34) 0.32
New cerebral infarctions (MRI)§§ 33 (25.98) 43 (33.59) 0.18
Intracerebral hemorrhage (MRI) 1 (0.79) 2 (1.56) 0.57
others (Table 1), and intraoperative variables, such as surgery
Duration of hospital stay after surgery 3.99 ± 1.80 4.26 ± 2.00 0.27
duration, anesthesia duration, and clamping duration, among (days)
††

others, were well-matched between the groups (all P > 0.05) Intraoperative EEG Suppression
(Table 2). In addition, intraoperative anesthesia dose did not Total EEG suppression time (min)*## 0 (0) 0 (1.17) 0.04
differ between groups. Time of EEG suppression in S1 (min)*## 0 (0) 0 (0.25) 0.13
Time of EEG suppression in S2 (min)*## 0 (0) 0 (0.1) 0.000
Time of EEG suppression in S3 (min)*## 0 (0) 0 (0) 0.02
Incident Delirium
The incidence of postoperative delirium was significantly lower The values in parentheses are expressed as percentages unless indicated otherwise.
Values are expressed as *median (i.q.r.). S1 is the stage from induction of anesthesia to
among patients receiving processed EEG-guided anesthesia clamping the carotid artery, S2 from clamping to declamping, and S3 from the declamping
management compared with that in the standard care group [10 to the end of surgery.
††
of 127 (7.87%) vs. 37 of 128 (28.91%), P < 0.01] (Table 3). §§ The P value is from the χ 2 test, except The P value is from the independent-samples
t test and ## The P value is from the independent-samples Mann–Whitney U test.

Intraoperative EEG Suppression

Seven patients were excluded from EEG analysis because of
corrupted data. The total duration of EEG suppression was Other Outcomes and Safety of the
significantly shorter in the intervention group than the standard Intervention
care group [0 (0) vs. 0 (1.17), P = 0.04] (Table 3). The effect There were no significant differences in incidence of new cerebral
of the intervention on EEG suppression differed according to infarctions according to symptoms and neuroimaging, incidence
surgical stage, with significantly shorter durations during S2 of intracerebral hemorrhage according to neuroimaging,
and S3 among the intervention group patients compared to the and hospital stay duration between groups (Table 3). No
standard group patients, while there was no significant difference intraoperative movement or awareness was observed in any
in duration during S1 . patient during surgery and no deaths occurred after surgery.

Frontiers in Neurology | www.frontiersin.org 6 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

FIGURE 2 | MAP over time. MAP, mean arterial blood pressure. The FIGURE 4 | Ipsilateral rSO2 over time. rSO2 , regional cerebral oxygenation.
measurement time interval after general anesthesia but before clamping the The measurement time interval after general anesthesia but before clamping
carotid artery was recorded as baseline reference T1, the time interval after the carotid artery was recorded as baseline reference T1, the time interval after
clamping but before declamping was recorded as T2, and the time interval clamping but before declamping was recorded as T2, and the time interval
after the declamping and stabilization of cerebral perfusion but before after the declamping and stabilization of cerebral perfusion but before
completion of surgery was recorded as T3. completion of surgery was recorded as T3.

FIGURE 3 | PET CO2 over time. PET CO2 , end-expiratory carbon dioxide partial FIGURE 5 | Contralateral rSO2 over time. rSO2 , regional cerebral oxygenation.
pressure. The measurement time interval after general anesthesia but before The measurement time interval after general anesthesia but before clamping
clamping the carotid artery was recorded as baseline reference T1, the time the carotid artery was recorded as baseline reference T1, the time interval after
interval after clamping but before declamping was recorded as T2, and the clamping but before declamping was recorded as T2, and the time interval
time interval after the declamping and stabilization of cerebral perfusion but after the declamping and stabilization of cerebral perfusion but before
before completion of surgery was recorded as T3. completion of surgery was recorded as T3.

Intraoperative Monitoring Values at alone. Based on previous studies on POD (11, 14, 17–19), we
suggest that not only the index from a processed EEG but also
Different Time-Points
an index with additional values (in this case, DSA) to guide
There were also no significant differences in MAP, PET CO2 ,
anesthesia management protocol may improve outcomes among
bilateral MFVMCA and rSO2 between groups at corresponding
CEA patients.
time-points (Figures 2–7), while PSI was higher in the
The SEDLine monitor provides computed quantitative
intervention group at all measurement time-points (P <
EEG indices based on retrospective analysis of a diagnostic
0.05) (Figure 8). There were differences in MAP, PET CO2 , PSI,
EEG database of sedated patients. Since regulatory approval
ipsilateral MFVMCA , and bilateral rSO2 at all time-points within
in 2002, these indices have been shown to independently
the two groups. MAP, PET CO2 and contralateral rSO2 at T2
predict deep sedation as assessed by other clinical metrics
were higher than at T1 and T3 (P < 0.05); PSI, ipsilateral rSO2
such as the Ramsay Sedation Score and Modified Observer’s
and MFVMCA at T2 were lower than at T1 and T3 (P < 0.05).
Assessment of Alertness/Sedation Scale (52, 63–65). Our
While contralateral MFVMAC was not significantly different at
findings of reduced POD incidence using processed EEG-guided
all time-points in the two groups (Table 4).
anesthesia management are in accord with previous studies
on other surgical populations (27–29, 66, 67). In contrast, the
DISCUSSION Electroencephalography Guidance of Anesthesia (ENGAGES)
trial of older adults undergoing cardiac or non-cardiac surgery
We demonstrated that in patients undergoing CEA, processed found no reduction in delirium incidence using this EEG-based
EEG-guided anesthesia management with PSI combined with intervention (68). Possible explanations for this discrepancy
DSA reduces the incidence of POD compared to using PSI include differences in baseline conditions between cohorts and

Frontiers in Neurology | www.frontiersin.org 7 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

FIGURE 6 | Ipsilateral MFVMAC over time. MFVMAC , mean flow velocity of

middle cerebral artery. The measurement time interval after general anesthesia FIGURE 8 | PSI at corresponding time points. PSI, patient state index, *P <
but before clamping the carotid artery was recorded as baseline reference T1, 0.05 from standard group (statistically significant). The measurement time
the time interval after clamping but before declamping was recorded as T2, interval after general anesthesia but before clamping the carotid artery was
and the time interval after the declamping and stabilization of cerebral recorded as baseline reference T1, the time interval after clamping but before
perfusion but before completion of surgery was recorded as T3. MFVMAC , declamping was recorded as T2, and the time interval after the declamping
bilateral mean flow velocity of middle cerebral artery. and stabilization of cerebral perfusion but before completion of surgery was
recorded as T3.

suppression is also a function of cerebral perfusion. The patients

recruited in our study all had severe carotid artery stenosis and
exhibited baseline characteristics predictive of greater sensitivity
to anesthesia depth and POD compared to the two cohort.
Thus, patient vulnerability is an important consideration when
identifying patients most suitable for processed EEG monitoring.
A slight change in anesthetic management may play a major
role in modulating postoperative delirium risk in patients with
cerebral ischemia (or patients at a risk of cerebral ischemia).
FIGURE 7 | Contralateral MFVMAC over time. MFVMAC , mean flow velocity of In this study, both groups of patients used the index from a
middle cerebral artery. The measurement time interval after general anesthesia processed EEG, and the intervention group combined the index
but before clamping the carotid artery was recorded as baseline reference T1, with additional values (in this case the DSA) to guide anesthesia.
the time interval after clamping but before declamping was recorded as T2,
It was observed that using the PSI alone did not result in as good
and the time interval after the declamping and stabilization of cerebral
perfusion but before completion of surgery was recorded as T3. MFVMAC ,
outcomes and BS avoidance, possibly due to underestimation of
bilateral mean flow velocity of middle cerebral artery. EEG suppression using the monitoring (56). Patients may have
benefited from BSR and PSI monitoring, visual tracking of the
EEG, and the individualized real-time feedback of an anesthetic
cerebral state provided by the EEG spectrogram or raw waveform
the unique effects of cardiopulmonary bypass on the cerebrum. patterns (70–72), thereby achieving superior anesthesia depth
A comprehensive description of predisposing factors is critical monitoring and reducing POD risk. Anesthesiologists may be
for investigations on POD, especially those including patients able to manage patients’ anesthetic cerebral state better by
with multiple etiologies. Leung and colleagues recently reported monitoring multiple EEG parameters. The occurrence of EEG
the results of their clinical trial in older adults undergoing burst suppression during maintenance may be predictive of
non-cardiac surgery, and reported similar rates of postoperative POD (73). Lower EEG suppression duration may have directly
delirium between patients randomized to receive EEG-guided contributed to the reduced POD incidence in the intervention
anesthetic management vs. those without the monitor (66). group; hence, vulnerable patients may have derived a greater
Unlike our study, most of the patients included in this study benefit. The PSI of the intervention group is higher than that of
were undergoing spinal surgery, and the surgical process had the standard care group, however, if only a single PSI index is
little effect on the cerebral circulation. In contrast to the two observed and maintained at a high level, EEG suppression may
trials, our patient groups were well-matched demographically not be minimized, which may not be enough to significantly
and clinically at baseline, and received similar surgical procedures reduce the risk of POD.
with comparable impacts on cerebral circulation, permitting The EEG suppression is also a function of cerebral perfusion,
similar anesthesia management (except for the EEG-based in other words, the lower the perfusion, the more suppressed
intervention). Another possible explanation for this discrepancy the EEG. Our patients had severe carotid artery stenosis and
is that anesthesia requirements may be highly patient-specific relatively insufficient cerebral perfusion. The severe fluctuations
due to individual sensitivity to anesthetic depth (69). EEG of cerebral blood flow (CBF) during CEA may have further

Frontiers in Neurology | www.frontiersin.org 8 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

TABLE 4 | Intraoperative monitoring values at different time-points.

Variable Intervention group Standard group P-values

n = 127 n = 128
Between groups In the group

MAP (mmHg) T1 97.41 ± 9.01 97.36 ± 8.56 0.74 0.000

T2 105.65 ± 12.36 106.01 ± 11.31
T3 96.61 ± 13.09 97.45 ± 11.43
PET CO2 (mmHg) T1 35.37 ± 1.92 35.16 ± 2.00 0.65 0.000
T2 36.83 ± 2.09 36.56 ± 2.28
T3 35.79 ± 1.96 35.63 ± 2.24
PSI* T1 47.50 (17.50) 39.00 (13.00) 0.000 0.02
T2 45 (15) 36 (11)
T3 46.00 (17.00) 36.00 (14.50)
Ipsilateral rSO*2 T1 68 (5) 66 (6) 0.30 0.000
T2 65 (7) 65 (7)
T3 69 (7) 69 (8)
Contralateral rSO*2 T1 68 (5) 67 (6) 0.67 0.000
T2 69 (6) 68 (6.5)
T3 69 (6) 68 (6)
Ipsilateral MFVMAC (cm/s)* T1 42.00 (15.00) 39.00 (18.00) 0.12 0.000
T2 30.00 (19.00) 26.00 (18.00)
T3 58.00 (30.00) 52.00 (24.00)
Contralateral MFVMAC (cm/s)* T1 50.00 (19.00) 47.00 (18.00) 0.11 0.48
T2 50.0 (22.50) 48.00 (20.00)
T3 50.00 (21.00) 45.00 (19.00)

Values are expressed as mean ± s.d. or *median (i.q.r.). The measurement time interval after general anesthesia but before clamping the carotid artery was recorded as baseline
reference T1, the time interval after clamping but before declamping was recorded as T2, and the time interval after the declamping and stabilization of cerebral perfusion but before
completion of surgery was recorded as T3. MAP, mean arterial blood pressure; PSI, patient state index; PET CO2 , end-expiratory carbon dioxide partial pressure; rSO2 , bilateral regional
cerebral oxygenation; MFVMAC , bilateral mean flow velocity of middle cerebral artery.
The P-value is from the RT-ANOVA. The P-value is from the RT-ANOVA.

aggravated hypoperfusion or led to cerebral hyperperfusion. risk factors for the development of delirium. In addition to
In our study, the combined monitoring of TCD and NIRS adjusting the anesthetic dose during the perioperative period,
may minimize the influence of the patient’s hemodynamics on we also adjusted blood pressure according to the operation
cerebral perfusion. However, the largest changes in CBF occurred stage to avoid potential cerebral perfusion problems that may
during S2 and S3 , and the duration of EEG suppression was lead to burst suppression. Moreover, we conducted multimodal
significantly lower during these stages in the intervention group. monitoring on all patients, which helped us obtain more
In the intervention group, while adjusting the anesthesia dose comprehensive information to identify the extent of individual
to reduce EEG burst suppression, we also paid close attention cerebral perfusion. We adjusted the physiological functions
to the effect on CBF. When adjusting the anesthetic dose, of patients according to anesthetics/techniques and surgical
burst suppression could not be minimized if CBF changes procedures, and strictly controlled arterial blood pressure (77)
were unfavorable. Therefore, even if the TCD and NIRS to avoid insufficient or excessive cerebral perfusion risk related
monitors did not show hypoperfusion or hyperperfusion, we still to POD. All the monitoring we used is non-invasive and has
slightly regulated blood pressure according to the surgical stage. no significant risk of injury to patient. Therefore, we conclude
Although we observed no difference in MAP at each time point that patients with carotid artery stenosis can benefit from
between groups, small adjustments may have slightly improved multimodal monitoring during surgery to reduce the onset of
CBF, reduce EEG suppression, and thereby reduce POD. EEG suppression.
Anesthetic dosage can be reduced using processed EEG (27, The difference in MAP, PET CO2 , ipsilateral MFVMCA , and
74), and previous studies have shown potentially neurotoxic bilateral rSO2 at all time-points is due to differences in blood
effects of general anesthetics, including propofol (75, 76). pressure management measures taken at different surgical stages
Therefore, it is conceivable that POD risk was reduced by lower to ensure that the cerebral perfusion is within desired limits.
anesthetic levels. However, the present results do not support this Despite the difference in contralateral rSO2 , there was no
premise, at least for the anesthetic doses used and for this specific difference in MFVMCA , likely because TCD monitoring reflects
population. All patients had severe CS and were prone to vascular MFVMCA , while NIRS monitors the oxygenation of the frontal
events which had been identified as one of the postoperative cortex mainly supplied by the anterior communicating artery

Frontiers in Neurology | www.frontiersin.org 9 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

(ACA). Since the contralateral MFVMCA is not affected by risk of cerebral ischemia). In turn, reducing POD incidence may
the surgery, there is no difference at all time-points. The improve outcome.
NIRS monitoring value may have been influenced by oxygen
metabolism of extracranial origin as well as by changes in blood DATA AVAILABILITY STATEMENT
pressure and arterial oxygen saturation. The differences in PSI
within group at corresponding time-points are likely because The original contributions presented in the study are included
the largest changes in CBF during clamping and declamping in the article/supplementary material, further inquiries can be
of the carotid artery may also affect EEG, leading to changes directed to the corresponding author/s.
in PSI.
ETHICS STATEMENT
Potential Limitations
The study has several limitations. First, as a single-center trial, The studies involving human participants were reviewed
the generalizability of the results may be limited. Second, and approved by the Ethics Committee of Xuanwu
postoperative delirium has no objective biomarker and so may Hospital of Capital Medical University (LYS[2018]053). The
be difficult to diagnose in certain cases (78). Third, POD was patients/participants provided their written informed consent to
assessed only over 3 days, so transient and later incidences participate in this study.
could have been missed. However, the preponderance of evidence
suggests that most cases occur within the first 3 days after
AUTHOR CONTRIBUTIONS
anesthesia and surgery (14, 79). Fourth, the anesthesiologists
were not blinded to group assignment; therefore, when caring for NX was responsible for the implementation of the project,
patients in the intervention group, they might have been more patient recruitment, patient interviews, data collection, data
cautious in the management of the patient’s overall hemodynamic entry, statistical analyses, and article writing. L-XL, D-XY, JW,
parameters. However, the following measures were adopted Y-HM, TT, and X-LS performed patient recruitment, patient
to avoid bias. The combined monitoring of TCD and NIRS interviews, and data collection. L-QJ and YH reviewed and edited
may minimize the influence of the patient’s hemodynamics on the final manuscript. T-LW managed the project and provided
cerebral perfusion. There was no difference in MAP, bilateral funding. All authors contributed to manuscript revision, read,
MFVMCA and rSO2 at different time-points between the two and approved the submitted version.
groups. The anesthetists did not participate in postoperative
follow-up, and the investigators responsible for postoperative
follow-up and delirium assessments were blinded to group
FUNDING
assignment and did not participate in anesthesia or perioperative This work was supported by Beijing Municipal Administration
care. The anesthetists and investigators also did not communicate of Hospitals Clinical Medicine Development of Special
patient information. Fifth, these findings may not apply to Funding Support (ZYLX201818), Beijing Municipal Health
patients receiving inhalational anesthetics. Commission (Jing2019-2), and Xuanwu Hospital Funding
Support (XWJL-2018017) of Capital Medical University.
SUMMARY
ACKNOWLEDGMENTS
This randomized clinical trial suggests that
electroencephalography-guided anesthesia management The authors sincerely thank all colleagues in the Departments
with both a quantitative EEG index and DSA can reduce of Neurosurgery, Vascular Ultrasound, and Anesthesiology at
postoperative delirium incidence in patients undergoing CEA, Xuanwu Hospital of Capital Medical University, Beijing, for their
especially those vulnerable to disruption of cerebral perfusion cooperation. The authors wish to express special appreciation to
and EEG suppression. A slight change in anesthetic management Ying Han (Department of Neurology, Xuanwu Hospital, Capital
may play a significant role in modulating postoperative delirium Medical University, Beijing) for neuropsychological consultation
risk in patients with ischemic cerebral disease (or patients at and personnel training.

REFERENCES 3. Pulli R, Dorigo W, Barbanti E, Azas L, Russo D, Matticari S, et al.

Carotid endarterectomy with contralateral carotid artery occlusion: is
1. Malcharek MJ, Ulkatan S, Marinò V, Geyer M, Lladò-Carbò E, Perez- this a higher risk subgroup? Eur J Vasc Endovasc Surg. (2002) 24:63–
Fajardo G, et al. Intraoperative monitoring of carotid endarterectomy by 8. doi: 10.1053/ejvs.2002.1612
transcranial motor evoked potential: a multicenter study of 600 patients. Clin 4. Brott TG, Halperin JL, Abbara S, Bacharach JM, Barr JD, Bush RL, et al. 2011
Neurophysiol. (2013) 124:1025–30. doi: 10.1016/j.clinph.2012.10.014 ASA/ACCF/AHA/AANN/AANS/ACR/ASNR/CNS/SAIP/SCAI/SIR/SNIS/
2. Pennekamp CW, Moll FL, de Borst GJ. The potential benefits and the role SVM/SVS guideline on the management of patients with extracranial
of cerebral monitoring in carotid endarterectomy. Curr Opin Anaesthesiol. carotid and vertebral artery disease: executive summary. Stroke. (2011)
(2011) 24:693–7. doi: 10.1097/ACO.0b013e32834c7aa1 42:e420–63. doi: 10.1161/STR.0b013e3182112d08

Frontiers in Neurology | www.frontiersin.org 10 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

5. North American Symptomatic Carotid Endarterectomy Trial Collaborators, 26. Soehle M, Dittmann A, Ellerkmann RK, Baumgarten G, Putensen C,
Barnett HJM, Taylor DW, Haynes RB, Sackett DL, Peerless SJ, Guenther U. Intraoperative burst suppression is associated with postoperative
et al. Beneficial effect of carotid endarterectomy in symptomatic delirium following cardiac surgery: a prospective, observational study. BMC
patients with high-grade carotid stenosis. N Engl J Med. (1991) Anesthesiol. (2015) 15:61. doi: 10.1186/s12871-015-0051-7
325:445–53. doi: 10.1056/NEJM199108153250701 27. Chan MT, Cheng BC, Lee TM, Gin T. BIS-guided anesthesia decreases
6. Endarterectomy for asymptomatic carotid artery stenosis. Executive postoperative delirium and cognitive decline. J Neurosurg Anesthesiol. (2013)
committee for the asymptomatic carotid atherosclerosis study. JAMA. 25:33–42. doi: 10.1097/ANA.0b013e3182712fba
(1995) 273:1421–8. 28. Radtke FM, Franck M, Lendner J, Krüger S, Wernecke KD, Spies CD.
7. Randomised trial of endarterectomy for recently symptomatic carotid Monitoring depth of anaesthesia in a randomized trial decreases the rate
stenosis: final results of the MRC European Carotid Surgery Trial (ECST). of postoperative delirium but not postoperative cognitive dysfunction. Br J
Lancet. (1998) 351:1379–87. Anaesthesia. (2013) 1:i98–105. doi: 10.1093/bja/aet055
8. Barnett HJ, Taylor DW, Eliasziw M, Fox AJ, Ferguson GG, Haynes 29. MacKenzie KK, Britt-Spells AM, Sands LP, Leung JM. Processed
RB, et al. Benefit of carotid endarterectomy in patients with electroencephalogram monitoring and postoperative delirium:
symptomatic moderate or severe stenosis. North American symptomatic a systematic review and meta-analysis. Anesthesiology. (2018)
carotid endarterectomy trial collaborators. N Engl J Med. (1998) 129:417–27. doi: 10.1097/ALN.0000000000002323
339:1415–25. doi: 10.1056/NEJM199811123392002 30. Sieber FE, Zakriya KJ, Gottschalk A, Blute MR, Lee HB, Rosenberg PB, et al.
9. Monk TG, Weldon BC, Garvan CW, Dede DE, van der Aa MT, Heilman Sedation depth during spinal anesthesia and the development of postoperative
KM, et al. Predictors of cognitive dysfunction after major non-cardiac surgery. delirium in elderly patients undergoing hip fracture repair. Mayo Clin Proc.
Anesthesiology. (2008) 108:18–30. doi: 10.1097/01.anes.0000296071.19434.1e (2010) 85:18–26. doi: 10.4065/mcp.2009.0469
10. Miyazaki S, Yoshitani K, Miura N, Irie T, Inatomi Y, Ohnishi Y, 31. Whitlock EL, Torres BA, Lin N, Helsten DL, Nadelson MR, Mashour GA, et
et al. Risk factors of stroke and delirium after off-pump coronary al. Postoperative delirium in a substudy of cardiothoracic surgical patients
artery bypass surgery. Interactive Cardiovasc Thorac Surg. (2011) 12:379– in the BAG-RECALL clinical trial. Anesthesia Analgesia. (2014) 118:809–
83. doi: 10.1510/icvts.2010.248872 17. doi: 10.1213/ANE.0000000000000028
11. Inouye SK. The dilemma of delirium: clinical and research controversies 32. Seder DB, Fraser GL, Robbins T, Libby L, Riker RR. The bispectral index
regarding diagnosis and evaluation of delirium in hospitalized elderly medical and suppression ratio are very early predictors of neurological outcome
patients. Am J Med. (1994) 97:278–88. doi: 10.1016/0002-9343(94)90011-6 during therapeutic hypothermia after cardiac arrest. Inten Care Med. (2010)
12. Lipowski ZL. Delirium (acute confusional states). JAMA. (1987) 258:1789–92. 36:281–8. doi: 10.1007/s00134-009-1691-1
13. Lipowski ZL. Delirium in the elderly patient. N Engl J Med. (1989) 320:578– 33. Umegaki N, Hirota K, Kitayama M, Yatsu Y, Ishihara H, Mtasuki A. A marked
82. doi: 10.1056/NEJM198903023200907 decrease in bispectral index with elevation of suppression ratio by cervical
14. Khadka J, McAlinden C, Pesudovs K. Cognitive trajectories after postoperative haematoma reducing cerebral perfusion pressure. J Clin Neurosci. (2003)
delirium. N Engl J Med. (2012) 367:30–9. doi: 10.1056/NEJMoa1112923 10:694–6. doi: 10.1016/j.jocn.2002.11.001
15. Inouye SK, Westendorp RG, Saczynski JS. Delirium in elderly people. Lancet. 34. Inouye SK, van Dyck CH, Alessi CA, Balkin S, Siegal AP, et al.
(2014) 383:911–22. doi: 10.1016/S0140-6736(13)60688-1 Clarifying confusion: the confusion assessment method. A new
16. Bin Abd Razak HR, Yung WY. Postoperative delirium in patients undergoing method for detection of delirium. Ann Internal Med. (1990)
total joint arthroplasty: a systematic review. J Arthroplasty. (2015) 30:1414– 113:941–8. doi: 10.7326/0003-4819-113-12-941
7. doi: 10.1016/j.arth.2015.03.012 35. Katzman R, Zhang MY, Ouang-Ya-Qu N, Wang ZY, Liu WT, Yu E, et
17. Salluh JI, Wang H, Schneider EB, Nagaraja N, Yenokyan G, Damluji A, al. A Chinese version of the Mini-Mental State Examination; impact of
et al. Outcome of delirium in critically ill patients: systematic review and illiteracy in a Shanghai dementia survey. J Clin Epidemiol. (1988) 41:971–
meta-analysis. BMJ. (2015) 350:h2538. doi: 10.1136/bmj.h2538 8. doi: 10.1016/0895-4356(88)90034-0
18. Crocker E, Beggs T, Hassan A, Denault A, Lamarche Y, Bagshaw S, et 36. Roalf DR, Moberg PJ, Xie SX, Wolk DA, Moelter ST, Arnold SE.
al. Long-term effects of postoperative delirium in patients undergoing The Montreal Cognitive Assessment, MoCA: a brief screening tool
cardiac operation: a systematic review. Ann Thoracic Surg. (2016) 102:1391– for mild cognitive impairment. J Am Geriatr Soc. (2005) 53:695–
9. doi: 10.1016/j.athoracsur.2016.04.071 9. doi: 10.1111/j.1532-5415.2005.53221.x
19. Mashour GA, Woodrum DT, Avidan MS. Neurological complications 37. Zung WW. A self-rating depression scale. Arch Gene Psychiatry. (1965)
of surgery and anaesthesia. Br J Anaesthesia. (2015) 114:194– 12:63–70. doi: 10.1001/archpsyc.1965.01720310065008
203. doi: 10.1093/bja/aeu296 38. Zung WW. A rating instrument for anxiety disorders. Psychosomatics. (1971)
20. American Society of Anesthesiologists. Perioperative Brain Health 12:371–9. doi: 10.1016/S0033-3182(71)71479-0
Initiative Website. (2018). Available online at: https://www.asahq.org/ 39. American Society of Anesthesiologists. New classification of physical status.
brainhealthinitiative (accessed September 16, 2018). Anesthesiology. (1963) 24:111.
21. Mohanty S, Rosenthal RA, Russell MM, Neuman MD, Ko CY, Esnaola 40. Collin C, Wade DT, Davies S, Horne V. The Barthel ADL index: a reliability
NF. Optimal perioperative management of the geriatric patient: a best study. Int Disabil Stud. (1988) 10:61–3. doi: 10.3109/09638288809164103
practices guideline from the American College of Surgeons NSQIP and 41. Rudolph JL, Marcantonio ER. Review articles: postoperative delirium: acute
the American Geriatrics Society. J Am College Surgeons. (2016) 222:930– change with long-term implications. Anesthesia Analgesia. (2011) 112:1202–
47. doi: 10.1016/j.jamcollsurg.2015.12.026 11. doi: 10.1213/ANE.0b013e3182147f6d
22. O’Mahony R, Murthy L, Akunne A, Young J. Synopsis of the 42. Kazmierski J, Kowman M, Banach M, Fendler W, Okonski P, Banys
National Institute for Health and Clinical Excellence guideline A, et al. Incidence and predictors of delirium after cardiac surgery:
for prevention of delirium. Ann Internal Med. (2011) 154:746– Results from The IPDACS Study. J Psychosomat Res. (2010) 69:179–
51. doi: 10.7326/0003-4819-154-11-201106070-00006 85. doi: 10.1016/j.jpsychores.2010.02.009
23. National Institute for Health and Care Excellence. Delirium in Adults. 43. Leung JM. Postoperative delirium: are there modifiable risk factors? Eur J
London: National Institute for Health and Care Excellence (2014). Anaesthesiol. (2010) 27:403–5. doi: 10.1097/EJA.0b013e3283340a99
24. Inouye SK, Charpentier PA. Precipitating factors for delirium in hospitalized 44. Wada S, Inoguchi H, Sadahiro R, Matsuoka YJ, Uchitomi Y, Sato T, et
elderly persons. Predictive model and interrelationship with baseline al. Preoperative anxiety as a predictor of delirium in cancer patients:
vulnerability. JAMA. (1996) 275:852–7. a prospective observational cohort study. World J Surg. (2019) 43:134–
25. Fritz BA, Kalarickal PL, Maybrier HR, Muench MR, Dearth D, 42. doi: 10.1007/s00268-018-4761-0
Chen Y, et al. Intraoperative electroencephalogram suppression 45. Wang J, Li Z, Yu Y, Li B, Shao G, Wang Q. Risk factors contributing to
predicts postoperative delirium. Anesthesia Analgesia. (2016) postoperative delirium in geriatric patients postorthopedic surgery. Asia-
122:234–42. doi: 10.1213/ANE.0000000000000989 Pacific Psychiatry. (2015) 7:375–82. doi: 10.1111/appy.12193

Frontiers in Neurology | www.frontiersin.org 11 July 2021 | Volume 12 | Article 666814

Xu et al. EEG-Guided Anesthesia Reduces Delirium

46. Siepe M, Pfeiffer T, Gieringer A, Zemann S, Benk C, Schlensak C, et al. effects of propofol. Br J Anaesthesia. (2010) 105:172–8. doi: 10.1093/bja/ae
Increased systemic perfusion pressure during cardiopulmonary bypass is q155
associated with less early postoperative cognitive dysfunction and delirium. 66. Tang CJ, Jin Z, Sands LP, Pleasants D, Tabatabai S, Hong Y, et al. ADAPT-2:
Eur J Cardio-Thoracic Surg. (2011) 40:200–7. doi: 10.1016/j.ejcts.2010.11.024 A randomized clinical trial to reduce intraoperative EEG suppression in older
47. Samra SK, Dy EA, Welch K, Dorje P, Zelenock GB, Stanley surgical patients undergoing major noncardiac surgery. Anesthesia Analgesia.
JC. Evaluation of a cerebral oximeter as a monitor of cerebral (2020) 131:1228–36. doi: 10.1213/ANE.0000000000004713
ischemia during carotid endarterectomy. Anesthesiology. (2000) 67. Punjasawadwong Y, Chau-In W, Laopaiboon M, Punjasawadwong S, Pin-
93:964–70. doi: 10.1097/00000542-200010000-00015 On P. Processed electroencephalogram and evoked potential techniques for
48. Rowed DW, Houlden DA, Burkholder LM, Taylor AB. Comparison of amelioration of postoperative delirium and cognitive dysfunction following
monitoring techniques for intraoperative cerebral ischemia. Can J Neurol Sci. non-cardiac and non-neurosurgical procedures in adults. Cochrane Datab
(2004) 31:347–56. doi: 10.1017/s0317167100003437 Systematic Rev. (2018) 5:CD011283. doi: 10.1002/14651858.CD011283.pub2
49. Grubhofer G, Plöchl W, Skolka M, Czerny M, Ehrlich M, Lassnigg A. 68. Wildes TS, Mickle AM, Ben Abdallah A, Maybrier HR, Oberhaus J,
Comparing Doppler ultrasonography and cerebral oximetry as indicators for Budelier TP, et al. Effect of electroencephalography-guided anesthetic
shunting in carotid endarterectomy. Anesthesia Analgesia. (2000) 91:1339– administration on postoperative delirium among older adults undergoing
44. doi: 10.1097/00000539-200012000-00006 major surgery: the ENGAGES randomized clinical trial. JAMA. (2019)
50. Ali AM, Green D, Zayed H, Halawa M, El-Sakka K, Rashid HI. Cerebral 321:473–483. doi: 10.1001/jama.2018.22005
monitoring in patients undergoing carotid endarterectomy using a triple 69. Iwakiri H, Nishihara N, Nagata O, Matsukawa T, Ozaki M, Sessler DI.
assessment technique. Interact Cardiovasc Thoracic Surg. (2011) 12:454– Individual effect-site concentrations of propofol are similar at loss of
7. doi: 10.1510/icvts.2010.235598 consciousness and at awakening. Anesthesia Analgesia. (2005) 100:107–
51. Sundt TM Jr, Sharbrough FW, Anderson RE, Michenfelder JD. Correlation 10. doi: 10.1213/01.ANE.0000139358.15909.EA
of cerebral blood flow and electroencephalographic changes during carotid 70. Purdon PL, Pavone KJ, Akeju O, Smith AC, Sampson AL, Lee J, et al.
endarterectomy: with results of surgery and hemodynamics of cerebral The ageing brain: Age-dependent changes in the electroencephalogram
ischemia. Mayo Clin Proc. (1981) 56:533–43. during propofol and sevoflurane general anaesthesia. Br J Anaesthesia. (2015)
52. Drover D, Ortega HR. Patient state index. Best practice & research. Clin 115(Suppl. 1):i46–57. doi: 10.1093/bja/aev213
Anaesthesiol. (2006) 20:121–8. doi: 10.1016/j.bpa.2005.07.008 71. Purdon PL, Zhou DW, Akeju O, Brown EN. In reply. Anesthesiology. (2015)
53. Ching S, Purdon PL, Vijayan S, Kopell NJ, Brown EN. A neurophysiological- 123:725–8. doi: 10.1097/ALN.0000000000000794
metabolic model for burst suppression. Proc Natl Acad Sci USA. (2012) 72. Purdon PL, Sampson A, Pavone KJ, Brown EN. Clinical
109:3095–100. doi: 10.1073/pnas.1121461109 electroencephalography for anesthesiologists: part i: background
54. Lewis LD, Ching S, Weiner VS, Peterfreund RA, Eskandar EN, Cash SS, et al. and basic signatures. Anesthesiology. (2015) 123:937–
Local cortical dynamics of burst suppression in the anaesthetized brain. Brain. 60. doi: 10.1097/ALN.0000000000000841
(2013) 136:2727–37 doi: 10.1093/brain/awt174 73. Hesse S, Kreuzer M, Hight D, Gaskell A, Devari P, Singh D, et al.
55. Leung, J. l., Leung, V. c., Leung CM, Pan PC. Clinical utility and Association of electroencephalogram trajectories during emergence from
validation of two instruments (the Confusion Assessment Method Algorithm anaesthesia with delirium in the postanaesthesia care unit: an early
and the Chinese version of Nursing Delirium Screening Scale) to detect sign of postoperative complications. Br J Anaesthesia. (2019) 122:622–
delirium in geriatric inpatients. Gene Hospital Psychiatry. (2008) 30:171– 34. doi: 10.1016/j.bja.2018.09.016
6. doi: 10.1016/j.genhosppsych.2007.12.007 74. Luo C, Zou W. Cerebral monitoring of anaesthesia on reducing cognitive
56. Muhlhofer WG, Zak R, Kamal T, Rizvi B, Sands LP, Yuan M, dysfunction and postoperative delirium: a systematic review. J Int Med Res.
et al. Burst-suppression ratio underestimates absolute duration of (2018) 46:4100–10. doi: 10.1177/0300060518786406
electroencephalogram suppression compared with visual analysis 75. Berger M, Nadler JW, Friedman A, McDonagh DL, Bennett ER, Cooter
of intraoperative electroencephalogram. Br J Anaesthesia. (2017) M, et al. The effect of propofol versus isoflurane anesthesia on human
118:755–61. doi: 10.1093/bja/aex054 cerebrospinal fluid markers of Alzheimer’s disease: results of a randomized
57. Bokeriia LA, Golukhova EZ, Polunina AG. Postoperative delirium in cardiac trial. J Alzheimer’s Dis. (2016) 52:1299–310. doi: 10.3233/JAD-151190
operations: microembolic load is an important factor. Ann Thoracic Surg. 76. Tang JX, Baranov D, Hammond M, Shaw LM, Eckenhoff MF, Eckenhoff
(2009) 88:349–50. doi: 10.1016/j.athoracsur.2009.02.031 RG. Human Alzheimer and inflammation biomarkers after anesthesia and
58. Donald W, Reza M. Intraoperative cerebral high-intensity transient signals surgery. Anesthesiology. (2011) 115:727–32. doi: 10.1097/ALN.0b013e31822e
and postoperative cognitive function: a systematic review. Am J Surg. (2009) 9306
198:295–7. doi: 10.1016/j.amjsurg.2008.11.019 77. Joshi B, Ono M, Brown C, Brady K, Easley RB, Yenokyan G, et al. Predicting
59. Flach HZ, Ouhlous M, Hendriks JM, Van Sambeek MR, Veenland JF, the limits of cerebral autoregulation during cardiopulmonary bypass.
Koudstaal PJ, et al. Cerebral ischemia after carotid intervention. J Endovasc Anesthesia Analgesia. (2012) 114:503–10. doi: 10.1213/ANE.0b013e31823d
Ther. (2004) 11:251–7. doi: 10.1583/03-1128.1 292a
60. Henry M, Amor M, Masson I, Henry I, Tzvetanov K, Chati Z, et al. Ischemic 78. Li T, Wieland LS, Oh E, Neufeld KJ, Wang NY, Dickersin K, et al.
cerebral lesions after carotid surgery and carotid stenting. Eur J Vasc Endovasc Design considerations of a randomized controlled trial of sedation level
Surg. (2007) 33:430–5. doi: 10.1016/j.ejvs.2006.11.012 during hip fracture repair surgery: a strategy to reduce the incidence
61. Bonati LH, Jongen LM, Haller S, Flach HZ, Dobson J, Nederkoorn of postoperative delirium in elderly patients. Clin Trials. (2017) 14:299–
PJ, et al. New ischaemic brain lesions on MRI after stenting or 307. doi: 10.1177/1740774516687253
endarterectomy for symptomatic carotid stenosis: a substudy of the 79. Lee H, Oh SY, Yu JH, Kim J, Yoon S, Ryu HG. Risk factors of postoperative
International Carotid Stenting Study (ICSS). Neurology. (2010) 9:353– delirium in the intensive care unit after liver transplantation. World J Surg.
62. doi: 10.1016/S1474-4422(10)70057-0 (2018) 42:2992–9. doi: 10.1007/s00268-018-4563-4
62. Schnaudigel S, Gröschel K, Pilgram SM, Kastrup A. New brain lesions after
carotid stenting versus carotid endarterectomy: a systematic review of the Conflict of Interest: The authors declare that the research was conducted in the
literature. Stroke. (2008) 39:1911–9. doi: 10.1161/STROKEAHA.107.500603 absence of any commercial or financial relationships that could be construed as a
63. Soehle M, Ellerkmann RK, Grube M, Kuech M, Wirz S, Hoeft A, et al. potential conflict of interest.
Comparison between bispectral index and patient state index as measures
of the electroencephalographic effects of sevoflurane. Anesthesiology. (2008) Copyright © 2021 Xu, Li, Wang, Jiao, Hua, Yao, Wu, Ma, Tian and Sun. This is an
109:799–805. doi: 10.1097/ALN.0b013e3181895fd0 open-access article distributed under the terms of the Creative Commons Attribution
64. Schneider G, Heglmeier S, Schneider J, Tempel G, Kochs EF. Patient State License (CC BY). The use, distribution or reproduction in other forums is permitted,
Index (PSI) measures depth of sedation in intensive care patients. Inten Care provided the original author(s) and the copyright owner(s) are credited and that the
Med. (2004) 30:213–6. doi: 10.1007/s00134-003-2092-5 original publication in this journal is cited, in accordance with accepted academic
65. Soehle M, Kuech M, Grube M, Wirz S, Kreuer S, Hoeft A, et al. Patient practice. No use, distribution or reproduction is permitted which does not comply
state index vs bispectral index as measures of the electroencephalographic with these terms.

Frontiers in Neurology | www.frontiersin.org 12 July 2021 | Volume 12 | Article 666814