Ketamine-Associated Intraoperative Electroencephalographic Signatures of Elderly Patients With and Without Preoperative Cognitive Impairment

KEY POINTS

Question: How does ketamine affect the electroencephalogram and cognitive outcomes in older patients with and without preoperative cognitive impairment undergoing spine surgery? Findings: An increase in the electroencephalogram (EEG) power in the moderate frequency range (10–20 Hz) associated with ketamine can be observed in patients considered to be cognitively normal (preoperatively) but not in those considered to be impaired, and ketamine was not associated with increased odds of developing postoperative delirium in patients that are cognitively normal but increases odds of postoperative delirium in patients determined to be cognitively impaired. Meaning: Cognitively vulnerable patients might be able to be identified intraoperatively by the lack of a neurophysiologic response to ketamine.

See Article, page 679

As the population ages, more elderly patients require surgeries.1 Compared to younger patients, elderly patients are at a higher risk of developing brain function impairment and delirium after surgery and anesthesia.2 Given the high rate of complications in this vulnerable population, there has been a move toward improved intraoperative monitoring of cognitive vulnerability during the perioperative period.3,4 At the same time, the rising opioid epidemic has shifted some of the medications we use intraoperatively to decrease opioid consumption, decrease diversion, and prevent opioid dose escalation.5 It is critical to clarify how these alternative anesthetic approaches impact perioperative brain function in older adults.

One of the most common medications to replace opioids is ketamine (Ket), which antagonizes N-methyl-D-aspartate (NMDA) receptors in the central nervous system and also increases α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor current.6 In addition to its effects on glutamatergic neurotransmission, Ket has anti-inflammatory properties, including reduction of microglia activation and a decrease in the cytokines tumor necrosis factor-alpha (TNF-α) and interlukin-6 (IL-6).7 It has been suggested that these mechanisms play a role in reducing the incidence of postoperative delirium in adults8; however, larger studies have found contradictory results.9 Ketamine administration has been associated with decreasing agitation during emergence in children10 and hastening emergence from general anesthesia in rodents.11 Furthermore, Ket can increase the incidence of postoperative nightmares and hallucinations.9 Ketamine’s enigmatic effects reveal that this drug is multidimensional with a wide range of benefits and unintended effects.12

As we attempt to better understand the association between the combined effects of pharmacology and baseline cognitive impairment on postoperative outcomes including delirium, there is recognition that the monitoring of brain function either before surgery with preoperative cognitive assessment or during the surgery (eg, with electroencephalogram [EEG]) could guide the perioperative care of patients with cognitive vulnerability.13 Most frontal EEG-based indices used for titrating anesthetic drugs are limited in their applicability due to problems with accuracy when Ket is administered14 and in patients with advanced age.15 Although higher frequencies are attenuated somewhat by the skull, a wide range of frequencies can be observed even with the abbreviated frontal montage typical for the commonly available devices. Although intraoperative monitoring of brain neurophysiology is not mandatory as an American Society of Anesthesiologists (ASA) standard of care, many practitioners elect to place frontal EEG electrodes for routine cases to inform their intraoperative pharmacologic decisions regarding analgesics,16 and its use is recommended by expert groups.17,18

In this single-center observational study of 98 spine surgery patients ≥65 years, we prospectively evaluated the association of Ket and neurocognitive evaluation with and with specific spectral EEG features that could allow us to identify vulnerable patients who may benefit from perioperative delirium care pathways.

METHODS Participants

This protocol was approved by the University of California, San Francisco institutional review board (UCSF IRB 18-26716). All participants provided written informed consent. These data are an interim analysis of the EEG records from a subset of patients enrolled in an ongoing prospective observational cohort study focused on the long-term effects of surgery with general anesthesia. Data collected from a total of 155 participants are included in this study. No formal a priori power analysis was performed, and the analysis is based only on the data available. Inclusion criteria included age ≥65 years, surgery scheduled for ≥3 hours, any gender, race, or ethnicity. Exclusion criteria included the inability to read, understand, or speak English. All participants completed a preoperative cognitive assessment and postoperative delirium assessment, and 98 participants had an analyzable frontal EEG recording. This article adheres to the applicable Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

Cognitive Assessments

Preoperative cognition was measured by the UCSF Brain Health Assessment (BHA), administered using the TabCAT software platform, developed at the UCSF.19,20 The BHA is a 10-minute digital test administered on an iPad. It consists of 4 subtests: (1) favorites (associative memory), (2) match (executive function and processing speed), (3) line orientation (visuospatial skills), and (4) animal fluency (language). Performance on the BHA is summarized by the BHA-Cognitive Score (BHA-CS), which is the weighted demographically adjusted standardized scores of the BHA component tests. The weights were selected to optimally discriminate neurologically healthy (N = 451) from cognitively impaired (CogImp) (MCI or dementia participants; N = 399). The area under the receiver operating characteristic curve (AUC) for this classification was 0.92.19 In the present study, we conservatively selected a standard score of BHA-CS ≤–1.5z to identify cognitive impairment. In the validation study, <6% of neurologically healthy participants had a BHA-CS score in this range.20 Participants with scores ≤–1.5z were classified as CogImp, and participants with scores >–1.5z were classified as CogNml.

Evaluation for postoperative delirium was done by trained nursing staff using the confusion assessment method for intensive care unit (CAM-ICU) while patients were in the ICU and the Nursing Delirium Screening Scale (NuDESC) while patients were on the hospital floor, using standard procedures.21,22

EEG Recording and Processing

Frontal EEG was recorded with the SedLine Root monitor (Masimo, Corp) at a sample rate of either fs = 178 or fs = 89 Hz using the SedLine electrode strip and stored in the .edf file format. We also denoted the time points of anesthesia induction, start and stop of surgery, extubation, and start and stop of Ket infusion. To correct for the different sample rates, we downsampled the recordings with fs = 178 to 89 Hz after bandpass filtering the signal to 0.5 to 30 Hz using the filtfilt function. We applied the same filter to the data recorded at an 89-Hz sampling rate. We performed EEG analysis on an artifact-free 60-second segment (without burst suppression) that was extracted during maintenance. This segment was visually identified by authors with EEG expertise (O.L.B.C. and M.K.). The power spectral density (PSD) was calculated for each of the 60-second segments with the pwelch function using the default settings, and a frequency resolution set to 0.35 Hz. Because the recorded EEG amplitude is dependent on the display setting of the SedLine,23 we only present the normalized PSD. The normalized PSD was derived by dividing the PSD vector of the 50-second EEG segment by the cumulative power from 0.5 to 30 Hz within this vector. Because coadministration of Ket leads to beta-frequency activation in the EEG,24 we also calculated the EEG beta-band power, which is the cumulative power in the 12- to 25-Hz range. We also calculated the spectral entropy (SpEnt) as a parameter that reflects the shape of the PSD. The SpEnt presents the Shannon entropy applied to the normalized PSD,25 in our case, to the 0.5- to 30-Hz range.

EEG Epochs Selection

The epochs were selected based on the midpoint of surgery. This specific time was calculated by dividing the total time of surgery by 2. Visual inspection was used to identify the 60-second segment closest to the midpoint that was free from electrocautery or movement artifact. These EEG epochs were chosen before archiving the data from drug administration or cognitive testing to minimize potential bias of the investigators. If an appropriate segment within 5 minutes before or 5 minutes after the midpoint of surgery could not be identified, the data from that patient were excluded from the EEG analysis.

Statistical Analysis

For the evaluation of a Ket-induced effect on the EEG, we separated the observational data into 2 main groups: patients who received Ket administration versus no Ket administration. We also conducted analyses to compare the spectral features of EEG and the effect of Ket among patients who were CogNml or CogImp preoperatively. We also compared these features between patients with and without postoperative delirium as a subanalysis.

To evaluate if there were differences between the groups in our study, we analyzed independent variables using a t test for normally distributed continuous variables (mean ± standard deviation [SD]), and for categorical variables, we used the χ2 test. Patients were divided into 2 groups: baseline CogNml or CogImp group. We used a Fisher exact test to evaluate associations between the development of postoperative delirium after intraoperative Ket in the CogNml and CogImp groups and the odds ratio (OR) with 95% confidence intervals (CIs). In addition, we performed analysis to evaluate the relationship between delirium and see if there was an interaction between Ket and baseline cognitive impairment. Stata version 15.1 (StataCorp LLC) was used for the analysis, and a threshold of P < .05 was considered statistically significant.

To statistically compare the PSD between 2 groups, we calculated the area under the receiver-operating characteristic curve (AUC) with 10k-fold bootstrapped 95% CIs. If the 95% CI does not include 0.5, the difference can be considered significant on a P < .05 level.26 In our analyses, we only report results as significant, if a significant difference could be observed in at least 2 neighboring frequencies. This approach has been used with the AUC as well as with t tests previously.27,28 To compare the beta-band power between the groups, we used the Mann-Whitney U test supplemented by the AUC as effect size. According to the traditional point system, an AUC > 0.7 (or AUC < 0.3) reflects a “fair,” an AUC > 0.8 (or AUC < 0.2) reflects a “good,” and an AUC > 0.9 (or AUC < 0.1) reflects an “excellent” effect. Furthermore, AUC > 0.7 (or AUC < 0.3) could indicate a “clinically relevant effect.”28 These classifications are estimations only and should be interpreted with caution. To get a coarse impression of possible relationships between the relative beta power, SpEnt, and Ket concentration, we created linear models that could provide preliminary insights into these relationships.

RESULTS Patient Characteristics

The mean age for the cohort was 72 years (SD, 5.4), and 43% of the cohort was women. Patients were predominantly American Society of Anesthesiologists (ASA) classes II and III. Based on medical record review, a diagnosis of a cognitive disorder was documented for 2%, prior stroke 8%, anxiety 17%, and depression 20%. Baseline impairment by BHA-Cognitive Score (BHA-CS) was 39%. A high percentage of the patients were prefrail (41%) or frail (40%) before surgery. We found that 26.5% of the cohort developed postoperative delirium (Supplemental Digital Content 1, Table S1, https://links.lww.com/AA/D801).

Subgroups

Of the 155 patients undergoing spine surgery, 57 patients were excluded from EEG analysis because the recording was unsuitable for spectral quantification due to artifact or nonstationary signal.

EEG data from the remaining 98 patients were included in this analysis. Figure 1 presents a flow chart of the patients and groups further divided into subgroups (Ket versus no Ket and CogNml versus CogImp). Of the 98 patients, 60 patients received Ket. The rate of baseline cognitive impairment was similar in the Ket (38.3%) and non-Ket (39.5%) groups. There was no significant difference in patient characteristics among the no Ket or Ket groups we used for EEG analysis, as shown in Table 1.

Table 1. - Characteristics of Patients With or Without Intraoperative Ket Variable No Ket (n = 38) Ket (n = 60) P value Age (y)a 71 (67.8–77.5) 70 (67.0–74.8) .156 Sex  Female 15 (39.5) 27 (45) .595  Male 23 (60.5) 33 (55) .595 ASA physical statusb  I 0 (0) 1 (1.67) .698  II 15 (39.5) 25 (41.7)  III 23 (60.5) 34 (56.7) Cognitive  impairment diagnosis 1 (2.6) 1 (1.7) .745  Stroke 5 (13.2) 3 (5) .154  Anxiety 7 (18) 10 (17) .825  Depression 8 (21) 12 (20) .933  Sleep disorder 12 (32) 17 (28) .735  Baseline  impairment by BHA-CS 15 (39.5) 23 (38.3) .911 Preoperative frailty scoreb,c  Robust 7 (19) 11 (18.3) .513  Prefrail 17 (45.9) 23 (38.3)  Frail 13 (35.1) 26 (43.3) Perioperative  Length of  Surgerya 209 (165–374) 306 (232–429) .002  ICU stay 5 (13) 18 (30) .056  Length of  hospital stay (d)a 4 (3–6) 4 (2–6) .013  Delirium after  surgery 7 (18.4) 19 (31.7) .151

Data are median (25th, 75th percentile) or n (%).

Abbreviations: ASA, American Society of Anesthesiologists; BHA-CS, Brain Health Assessment-Cognitive Score; ICU, intensive care unit; Ket, ketamine; No Ket, no ketamine.

aMann-Whitney U test.

bχ2.

cOne patient in the No Ket group is missing frailty score (n = 37).


F1Figure 1.:

Flow chart of the study and groups defined for EEG analysis. EEG indicates electroencephalogram.

Table 2. - Characteristics of Patients Cognitively Normal or Cognitively Impaired at Baseline Variable CogNml (n = 60) CogImp (n = 38) P value Age (y)a 71 (67–75) 70 (67–75) .758 Sex  Female 22 (37) 20 (53) .122  Male 38 (63) 18 (47) .122 ASA physical statusb  I 1 (2) 0 (0) .383  II 27 (45) 13 (34)  III 32 (53) 25 (66) Cognitive Impairment  diagnosis 1 (1.7) 1 (2.6) .745  Stroke 5 (8.3) 3 (7.9) .939  Anxiety 12 (20) 5 (13) .389  Depression 11 (18) 9 (24) .527  Sleep disorder 17 (28) 12 (32) .735 Preoperative frailty scoreb,c  Robust 13 (22) 5 (13) .407  Prefrail 25 (42) 15 (41)  Frail 22 (37) 17 (46) Perioperative  Length of surgery  (min)a 271 (200–418) 286 (175–409) .709  ICU stay 13 (21.7) 10 (26.3) .601  Length of hospital  stay (d)a 5 (3–6) 5 (3–8) .3827  Delirium after  surgery 11 (18.3) 15 (39.5) .021

Data are median (25th, 75th percentile) or n (%).

Bold text represents P < .05.

Abbreviations: ASA, American Society of Anesthesiologists; CogImp, cognitively impaired; CogNml, cognitively normal; ICU, intensive care unit.

aMann-Whitney U test.

bχ2.

cOne patient in the CogImp group is missing frailty score (n = 37).

Table 2 shows the patients divided by baseline cognitive status. There was no significant difference in preoperative patient characteristics. However, more patients in the CogImp group had postoperative delirium when compared to CogNml patients (39.5% versus 18.3%; P = .021).

Ketamine and Risk of Postoperative Delirium

Prior studies had suggested that Ket can decrease the risk of postoperative delirium.7 However, we found that in CogNml patients, the rate of postoperative delirium was not significantly different with or without Ket (19% in Ket group versus 17% in non-Ket group; odds ratio [OR], 1.10; CI, 0.30–4.04; P = 583). Furthermore, we found that among those who received Ket the subgroup that was also CogImp at baseline had a statistically significant increase in postoperative delirium risk (52% postoperative delirium with Ket versus 20% postoperative delirium non-Ket group; OR, 4.36; CI, 1.02–18.22; P = .048). Patients who did not receive Ket had a similar percentage of postoperative delirium in both the CogImp and CogNml groups (20% versus 17%) (Supplemental Digital Content 2, Table S2, https://links.lww.com/AA/D802).

Impact of Ketamine on the EEG

We found a significant increase in the EEG power in the moderate and higher spectral ranges (>10 Hz) in patients who received Ket (n = 60, any Ket dose), when compared to the patients without Ket (n = 38), as depicted in Figure 2A. Additionally, Figure 2B displays the density spectral array (DSA) of an exemplary patient who received a number of Ket boluses. As already indicated in these PSD plots, this activation led to a significant increase in beta-band power in patients with Ket, as depicted in Figure 2A. The detailed information regarding the relative beta power and SpEnt is presented in the scatter plots in the Supplemental Digital Content 3, Figures S1A and S1C, https://links.lww.com/AA/D803. When fitting a linear model to describe the relationship between the EEG parameters and the Ket concentration, we found a significant increase of relative beta power and SpEnt with increasing concentration as presented in Supplemental Digital Content 3, Figures S1B and S1D, https://links.lww.com/AA/D803. The model parameters for the linear fits were as follows:

F2Figure 2.:

Comparison of the normalized PSD plots of patients with (n = 60, blue) or without Ket (n = 38, black). A, Patients who received Ket (n = 60). The patients from the Ket group had significantly higher power in the frequencies above 10 Hz. B, DSA and raw EEG episodes of an exemplary patient who received multiple Ket boluses. The yellow bars indicate Ket delivery. The red line indicates the center of the maintenance period. The black dots in the AUC plot indicate significance, ie, a 95% confidence interval exclusive 0.5. The confidence interval boundaries are depicted by an “x.” AUC indicates area under the receiver-operating characteristic curve; DSA, density spectral array; EEG, electroencephalogram; Ket, ketamine; no Ket, no ketamine; PSD, power spectral density.

rel beta band = 0.11 + 0.0007 * [email protected] (t-stat:2.11; P = .004: R2 = 0.06);

SpEnt = 3.90 + 0.003 * [email protected] (t-stat:2.37; P = .021: R2 = 0.07); the variable [email protected] is the Ket concentration at the time of EEG extraction.

F3Figure 3.:

Comparison of the normalized PSD plots of patients who were cognitively impaired (orange) or cognitively normal (black) during a preoperative assessment. The left plot presents the PSD for all patients, the center plot presents the PSD for all patients who did not receive Ket, and the right plot presents the PSD for the patients who received Ket. A, When all patients are considered, the cognitively normal patients (n = 60) expressed higher normalized power in moderate and higher frequency ranges (10–23 Hz) than the cognitively impaired patients (n = 38). B, With only the patients who did not receive Ket included, the cognitively impaired patients (n = 15) expressed higher normalized power in the beta range and less power in the alpha range than the cognitively normal patients (n = 23). C, With only the patients who received Ket included, there was no clear difference in the spectral information between the cognitively impaired patients (n = 23) and the cognitively normal patients (n = 37). D, For patients who were cognitively normal preoperatively, the patients with Ket (blue) showed a significantly higher relative EEG power (>10 Hz). E, For preoperatively cognitively impaired patients, there was no significant difference in the spectral composition of the EEG between the patients receiving Ket (blue) or not (orange). AUC indicates area under the ROC curve; EEG, electroencephalogram; Ket, ketamine; no Ket, no ketamine; PSD, power spectral density.

F4Figure 4.:

Comparison of the normalized PSD plots of patients with (purple) and without delirium. The left plot presents the PSD for all patients, the center plot presents the PSD for all patients who did not receive Ket, and the right plot presents the PSD for the patients who received Ket. A, Patients without delirium showed significantly elevated power in the range around 20 Hz. B, There was no significant difference in patients that either had delirium or not that did not receive Ket. C, With only the patients who received Ket included, the patients without delirium showed significantly elevated power in the 10- to 25-Hz range. AUC indicates area under the ROC curve; Ket, ketamine; PSD, power spectral density.

For completeness, we present the results for the other EEG frequency bands in Supplemental Digital Content 4, Figure S2, https://links.lww.com/AA/D804.

Association of Preoperative Cognitive Status on the EEG Comparisons Between CogNml and CogImp Patients

When comparing all patients who were CogNml (n = 60) with CogImp (n = 38) at baseline, we found significantly higher power in the CogNml patients in the 10- to 20-Hz range (data include all patients with or without Ket) (Figure 3A). The patients in both groups did not differ significantly in preoperative characteristics (Table 2). For the patient group who did not receive Ket, we found significantly higher power in the range (>20 Hz) in the CogImp patients (n = 15) compared to CogNml patients (n = 23) as presented in Figure 3B. Furthermore, the power in the alpha range was higher in the patients who were CogNml (Figure 3B). Of patients who received Ket, no further increase in beta power was seen in the CogImp patients (Figure 3C). There was no significant difference in the relative power in the beta range between the patients who were CogNml (n = 37) and CogImp (n = 23) when receiving Ket.

Effect of Ketamine on the EEG of CogNml and CogImp

To analyze the independent effects of Ket on beta power, we separated the subjects who were CogNml into groups with or without Ket. The largest EEG differences caused by Ket are in the group who was CogNml at baseline (Figure 3D). There is no distinctive Ket effect on EEG in patients who were CogImp at baseline (Figure 3E).

EEG and Postoperative Delirium

In a subanalysis regarding the association of intraoperative EEG features and postoperative delirium, we found that patients without delirium showed higher power in faster EEG frequencies. This observation was mainly driven by the difference in the EEG between the groups in patients who received Ket. There, patients without delirium had significantly higher power in the 10- to 25-Hz range. The corresponding PSD plots are presented in Figure 4.

DISCUSSION

We prospectively evaluated the association of Ket administration and baseline cognitive impairment (as determined by BHA-CS), with intraoperative EEG during anesthetic maintenance for spine surgery. We observed an increase in spectral EEG power among moderate and higher frequencies (>10 Hz) in patients who received Ket when compared to the patients without Ket. This is in line with the described effects of Ket on the EEG.29,30 Furthermore, we found significant differences in the spectral composition of the EEG between patients who were CogNml versus those who were CogImp preoperatively. The neurophysiologic consequences of Ket administration may interfere with the identification of EEG differences associated with preoperative cognition, as spectral differences associated with preoperative neurocognitive status were only preserved in subgroup analysis with patients that did not receive Ket. The CogImp patients who did not receive Ket had markedly increased EEG power in the frequencies >20 Hz when compared to the CogNml patients without Ket. We cautiously conclude that a Ket-induced decrease in power among the 12- to 25-Hz frequency range (or more accurately the lack of an increase) could potentially be a biomarker for delirium (Figure 4C). This contrasts with our observation that demonstrated no changes in power in this frequency range with the administration of Ket associated with preoperative cognitive impairment (Figure 3C). This emphasizes the fact that not all postoperative delirium is predicted by preoperative cognitive impairment. This finding is consistent with age-related vulnerabilities previously reported using EEG.15,27,31 It is interesting that some overlap exists regarding the differences in spectral EEG associated with Ket administration and the differences associated with preoperative cognitive impairment, highlighting the fact that scalp EEG recordings can reflect neurophysiologic changes due to pharmacology or neurological impairment, but observations of an accelerated EEG do not always mean that a person has either been given Ket or has neurological impairment.

Previous research has associated spectral EEG features with both preoperative cognitive impairment32 and postoperative neurocognitive disorders (eg, delirium).33,34 Although our results are somewhat consistent with these other reports, we did not observe a relationship between delirium and power in the classical alpha band (8–12 Hz) by comparison. This may be due to the smaller age distribution in our sample, the relatively high rate of Ket administration, spine surgery, and/or subtle differences in neurocognitive assessment. In our study, only CogImp patients who did not receive Ket showed a reduction in alpha-band power. However, in the patients that received Ket, an increased power in the moderate and higher frequencies (>10 Hz) was associated with reduced odds of developing postoperative delirium (irrespective of their preoperative cognitive status). Patients who tested positive for delirium did not show the acceleration of the EEG caused by Ket.

Limitations

This is a single-center prospective observational study where patients were not randomly assigned to the Ket or no Ket groups. Our testing is a short global measurement of cognitive impairment. A full test battery could be used to explore if there are specific brain regions/networks responsible for the EEG findings. Because we had to limit our sample rate to 89 Hz, we could not evaluate Ket-induced effects in the higher frequencies of the EEG as previously examined.30 Furthermore, because of the known issues regarding the SedLine EEG export,23 we only present findings regarding the relative power because of possible differences in amplitude scaling. We also focused only on 1 minute of intraoperative EEG. This approach allowed us to obtain a general overview of the Ket-induced effects on the EEG, but further research is necessary to describe the effect for the entire perioperative period. Another limitation is that the cohort of patients is relatively older (age 72 years and SD 5.3) indicating that the results may be not generalizable to younger populations.

It remains possible that our observed dynamic EEG changes associated with Ket administration are not exclusive due to Ket pharmacology but possibly confounded by other factors (eg, Ket concentration, comorbid diagnoses, surgical stimulation, or other unaccounted for variable). In this prospective observational study, the anesthetic regimen for each patient was determined by the anesthesiologist without input from study staff. Patients may have received Ket via intermittent boluses, infusion, or both. For this study, we calculated the total dose of intraoperative Ket administered at the center time of maintenance (see Methods section) and grouped the patients by those who had received Ket and those who did not receive Ket at that time. A larger study that can control dose as well as the multitude of factors that go into deciding who does and who does not receive Ket will be necessary to more completely characterize the relationship of Ket and EEG signals in specific surgical populations.

CONCLUSIONS

The distinction between EEG changes that are age-induced versus those that result from preoperative cognitive impairment presents a challenge for automated EEG interpretation. To individualize anesthesia monitoring in the future, a better estimation of the neurocognitive vulnerability of the patient will be necessary since it may not correspond to the chronologic age. For example, younger patients with a history of drug abuse may express EEG features more typical of a geriatric patient.35 Moreover, we found that Ket did not protect against delirium in CogNml patients, and it negatively affected those who were CogImp at baseline. This goes along with our findings that patients who were CogImp did not have the accelerated EEG (ie, higher power in higher frequencies) that we found in CogNml patients, and furthermore, that failure to observe an effect of Ket administration on the EEG was associated with postoperative delirium. Conversely, seeing an EEG change in response to Ket may be reassuring for postoperative recovery.

DISCLOSURES

Name: Odmara L. Barreto Chang, MD, PhD.

Contribution: This author helped conceive, design, analyze, and interpret the data; draft the manuscript; and critically revise the manuscript for content.

Conflicts of Interest: O. L. Barreto Chang is an investigator for the clinical trial OLIVER from Medtronic and participated as an investigator on the clinical trial from TASLY pharmaceuticals.

Name: Matthias Kreuzer, PhD.

Contribution: This author helped conceive, design, analyze, and interpret the data; draft the manuscript; and critically revise the manuscript for content.

Conflicts of Interest: M. Kreuzer is listed as a contributor to filed patents regarding the visualization of EEG alpha-band activity and the influence of age on the EEG.

Name: Danielle F. Morgen.

Contribution: This author helped analyze the data, draft the manuscript, and critically revise the manuscript for content.

Conflicts of Interest: None.

Name: Katherine L. Possin, PhD.

Contribution: This author helped interpret the data, draft the manuscript, and critically revise the manuscript for content.

Conflicts of Interest: K. L. Possin has received research funding from the NIH, Quest Diagnostics, the Global Brain Health Institute, the Merck Foundation, and the Rainwater Charitable Foundation, consulting fees from ClearView Healthcare Partners and Vanguard, and a speaking fee from Swedish Medical Center.

Name: Paul S. García, MD, PhD.

Contribution: This author helped conceive, design, analyze, and interpret the data; draft the manuscript; and critically revise the manuscript for content.

Conflicts of Interest: P. S. García is listed as a contributor to filed patents regarding the visualization of EEG alpha-band activity and the influence of age on the EEG.

This manuscript was handled by: Oluwaseun Johnson-Akeju, MD, MMSc.

REFERENCES 1. Etzioni DA, Liu JH, Maggard MA, Ko CY. The aging population and its impact on the surgery workforce. Ann Surg. 2003;238:170–177. 2. Inouye SK, Westendorp RG, Saczynski JS. Delirium in elderly people. Lancet. 2014;383:911–922. 3. Gleason LJ, Schmitt EM, Kosar CM, et al. Effect of delirium and other major complications on outcomes after elective surgery in older adults. JAMA Surg. 2015;150:1134–1140.

留言 (0)

沒有登入
gif