Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 1.
Published in final edited form as: J Clin Anesth. 2021 Jun 17;74:110395. doi: 10.1016/j.jclinane.2021.110395

The Association of Bispectral Index Values and Metrics of Cerebral Perfusion During Cardiopulmonary Bypass

Xiuyun Liu 1,*,#, Mitsunori Nakano 1,2,*, Atsushi Yamaguchi 2, Brian Bush 1, Kei Akiyoshi 1, Jennifer K Lee 1, Raymond C Koehler 1, Charles W Hogue 3, Charles H Brown IV 1,#
PMCID: PMC8530850  NIHMSID: NIHMS1713325  PMID: 34147015

Abstract

Study Objective.

Low bispectral index (BIS) values have been associated with adverse postoperative outcomes. However, trials of optimizing BIS by titrating anesthetic administration have reported conflicting results. One potential explanation is that cerebral perfusion may also affect BIS, but the extent of this relationship is not clear. Therefore, we examined whether BIS would be associated with cerebral perfusion during cardiopulmonary bypass, when anesthetic concentration was constant.

Design.

Observational cohort study

Setting.

Cardiac operating room.

Patients.

Seventy-nine patients with cardiopulmonary bypass surgery were included.

Measurements.

Continuous BIS, mean arterial blood pressure (MAP), cerebral blood flow velocity (CBFV), and regional cerebral oxygen saturation (rSO2) were monitored, with analysis during a period of constant anesthetic. Mean flow index (Mx) was calculated as Pearson correlation between MAP and CBFV. The lower limit of autoregulation (LLA) was identified as the MAP value at which Mx increased > 0.4 with decreasing blood pressure. Postoperative delirium was assessed using the 3D-Confusion Assessment Method.

Results.

Mean BIS was lower during periods of MAP < LLA compared with BIS when MAP>LLA (mean 49.35 ± 10.40 vs. 50.72 ± 10.04, p=0.002, mean difference =1.38 [standard error: 0.42]). There was a dose response effect, with the BIS proportionately decreasing as MAP decreased below LLA (β = 0.15, 95% CI for the average slope across all patients 0.07 to 0.23, p<0.001). In contrast, BIS was relatively unchanged when MAP was above LLA (β = 0.03, 95% CI for the average slope across all patients −0.02 to 0.09, p=0.22). Additionally, increasing CBFV and rSO2 were associated with increasing BIS. Patients with postoperative delirium had lower mean BIS and higher percentage of time duration with BIS < 45 compared to patients without delirium.

Conclusions.

There was an association of BIS and metrics of cerebral perfusion during a period of constant anesthetic administration, but the absolute magnitude of change in BIS as MAP decreased below the LLA was small.

Keywords: Bispectral Index, Cardiopulmonary Bypass, Cerebral Perfusion, Delirium, Arterial Blood Pressure, Regional Cerebral Oxygen Saturation

1. Introduction

The bispectral index (BIS™, Medtronics, Inc, Minneapolis, MN) is a widely used processed electroencephalogram (EEG) monitor for estimating hypnotic level during anesthesia[1,2]. Several observational studies and secondary data analyses have demonstrated an association between low BIS values (i.e. <40–45) during surgery and patient morbidity and mortality [3–5], although a recent large clinical trial reported no mortality benefit at 1-year in patients randomized to different BIS targets (50 vs. 35) during general anesthesia [6]. Several early studies also reported that BIS-guided anesthesia reduced anesthetic exposure and decreased the risk of postoperative delirium[7–9], and even cognitive dysfunction [9]. However, the results of subsequent trials have not shown the same benefit [10,11]. Thus, it is unclear if titration of anesthetic level based on BIS values is an effective strategy to reduce postoperative delirium and improve other outcomes.

One potential explanation for these discrepant findings is the unclear role that cerebral perfusion may play on BIS values[12]. Although anesthetic drug concentrations are highly related to underlying EEG changes[13,14], the adequacy of cerebral perfusion is also known to affect the EEG[15,16], and thus might be an unmeasured confounding variable or a target for intervention. Indeed, observational studies which include arterial blood pressure data have suggested a strong relationship between deep anesthesia and complications when the blood pressure was low[17]. Cerebral perfusion to the brain is exquisitely regulated to maintain constant cerebral blood flow across a wide range of blood pressure through the process of cerebral autoregulation[18]. However, below the lower limit of autoregulation (LLA) [19], compensatory mechanisms are inadequate and cerebral blood flow decreases as mean arterial blood pressure (MAP) decreases, potentially leading to cerebral metabolic oxygen imbalance and EEG slowing.

We hypothesized that compromised cerebral perfusion due to MAP below the LLA would result in decreased BIS values during cardiopulmonary bypass (with constant anesthetic) in an observational cohort of patients having cardiac surgery. We also examined the relationship between BIS value and other measures of cerebral perfusion, including cerebral blood flow velocity (CBFV) and regional cerebral oxygen saturation (rSO2). We chose the bypass period because anesthetic administration is generally constant during this period, and so changes in BIS values would not be confounded by varying anesthetic concentrations. If there were indeed an association of BIS with metrics of cerebral perfusion, it would suggest that low BIS during surgery may not only be due to anesthetic drugs but also be related to inadequate cerebral perfusion and potential cerebral ischemia. An exploratory aim was to confirm prior observations that low BIS values are associated with postoperative delirium[7,8], but in particular to determine whether long duration of MAP below the LLA interacts with low BIS values to increase the risk of delirium.

2. Materials and Methods

The Johns Hopkins University School of Medicine Institutional Review Board approved this study, and written informed consent was obtained from all patients.

2.1. Patients

Data from an ongoing observational study with the primary goal of examining ICU-based characteristics of cerebral autoregulation in cardiac surgery patients was used for this analysis. Data from a subset of patients in this analysis has been published previously to address ICU-based characteristics of cerebral autoregulation[20]. The main research questions and analysis plan for this manuscript were agreed upon prior to examining the data. An exploratory analysis regarding an interaction between BIS values, MAP<LLA, and delirium was refined during data analysis. Patients were enrolled between June 20, 2017, and July 12, 2019. Inclusion criteria were age >18 years and having isolated or combined cardiac artery bypass graft, valve, aortic or myectomy surgery. Exclusion criteria were lung or heart transplant or insertion of a ventricular assist device. For the purposes of this analysis, patients without transcranial Doppler (TCD) insonating windows or excessive TCD artifacts or patients in whom an LLA could not be identified were excluded. At Johns Hopkins, anesthetic concentrations during cardiopulmonary bypass are generally kept constant (isoflurane concentration at 0.7% to 1%), but patients who did not have a stable concentration of volatile anesthetic during bypass were also excluded from analysis.

2.2. Data Collection

Monitors were placed at the start of surgery, but only data during cardiopulmonary bypass was used in the analysis. One BIS sensor (BIS™ Complete 2-Channel Monitor, COVIDIEN, Boulder, CO, USA) was placed on the patient’s forehead at anesthesia induction in accordance with manufacturer guidelines. Two near infrared spectroscopy sensors (INVOS, Somenetics, Inc., Boulder, CO, USA) were placed over the patient’s forehead to monitor relative total tissue hemoglobin (rTHb) and rSO2. Direct arterial pressure was measured via femoral or radial artery placed for clinical purpose. CBFV were measured with bilateral transcranial Doppler (Doppler Box, DWL; Compumedics) targeted at the middle cerebral arteries using 2.5-MHz transducers. All hemodynamic, rSO2 and CBFV signals were sampled at 128 Hz and recorded synchronously using ICM+ software (University of Cambridge, Cambridge Enterprise, Cambridge, UK, http://www.neurosurg.cam.ac.uk/icmplus) through an A/D converter (DT9801, Data Translation, Marlboro, MA) or digitally directly from GE Solar monitors.

2.3. Clinical Management

Perioperative care was provided according to usual clinical practice. General anesthesia was generally induced and maintained with fentanyl (5–20 μg kg-1), propofol (0.5–2.0 mg kg-1), and isoflurane with a non-depolarizing muscle relaxant. Dexmedetomidine and/or ketamine infusions were used at the discretion of the attending anesthesiologist and were constant during surgery, including during cardiopulmonary bypass. At the start of cardiopulmonary bypass, the concentration of isoflurane feeding into the membrane oxygenator was set to 0.7% to 1.0% and was not changed for the duration of bypass. Cardiopulmonary bypass was performed with a nonocclusive roller pump with an in-line arterial line filter ≤40 μm. Non-pulsatile flow was maintained between 2.0 and 2.4 L/m2/min, with α-stat pH management. Partial pressure of carbon dioxide was maintained between 35 and 45 mm Hg. A mild hypothermia strategy was used during cardiopulmonary bypass for most patients based on surgeon preference. Rewarming was based on institutional standards, with a goal of maintaining pharyngeal temperature less than 37°C. No intravenous anesthetic agents were given as bolus administrations during bypass.

2.4. Assessment of Delirium

A trained research assistant conducted delirium assessments once daily during three of the first four postoperative days using the 3D-confusion assessment method (3D-CAM) instrument. The sensitivity and specificity of the 3D-CAM for detecting delirium has been reported to be >94%[21]. For intubated patients, the CAM-ICU as performed by bedside nursing staff was used[22,23]. Patients were considered to be delirious if they had any positive 3D-CAM or CAM-ICU assessment during hospitalization. Training on delirium assessment included written materials on administration, videos depicting standard assessments, and supervised administration. Delirium assessors were masked to intraoperative data. Delirium assessments were started after 22 patients had been enrolled in this cohort, resulting in 28% of patients (22 out of 79) without delirium assessment results.

2.5. Analysis of Cerebral Autoregulation Data

The pre-analysis, including artifacts removal, data trend calculation, and all the cerebral autoregulation parameters were calculated through ICM+ software. Artifacts introduced by tracheal suctioning, arterial line flushing, or transducer malfunction were removed manually.

Arterial blood pressure and Doppler signals were time-integrated and resampled as 10-second mean values, to remove pulse, rollerhead, and respiratory frequency variations and preserve low-frequency waveforms associated with autoregulatory vascular reactivity. Next, a continuous, moving Pearson correlation coefficient between 30 consecutive, paired MAP and CBFV values was calculated to generate the mean velocity index (Mx) [24–26]. Mx was updated every 60 seconds from an overlapping, moving 300-second window and paired with the mean arterial pressure value from the same 300-second window. Blood pressure in the autoregulation range is indicated by a Mx value that approaches zero (there is no correlation between flow velocity and MAP), whereas an Mx approaching 1 indicates dysregulated cerebral blood flow (flow velocity and MAP are correlated) [27].

To define the LLA using CA correlation-based parameters, we plotted Mx against MAP in 5-mm Hg bins and applied a “U-shape” curve fitting algorithm [28]. The cutoff of Mx=0.4 [29,30] was used to identify LLA by drawing a straight horizontal line at the cutoff value using ICM+ software[28]. The x coordinate of the point at which the straight line meets the U-shaped curve was defined as the LLA [29,30] or treated as missing in the absence of an intersection.

2.6. Statistical Analysis

Statistical analyses were performed using Matlab software (ver. R2012A, MathWorks, Inc.) and SPSS (version 25.0, IBM, NY, USA). The sample size was based on available date from patients who met inclusion criteria.

Association of BIS Values with Metrics of Cerebral Perfusion

The mean values of BIS during periods of MAP>LLA and MAP<LLA were calculated for each patient and a paired comparison between the two mean BIS values was conducted. In order to analyze the within-participant correlations between metrics of cerebral perfusion (BIS, rSO2, and CBFV) and BIS, linear mixed models were used. Mean values of blood pressure, BIS, rSO2, and CBFV were calculated every 1 minute, resulting in multiple measurements of these indices for each patient. Then, linear mixed-effects models with random intercepts and random slopes were used to assess the within-participant correlations of repeated measurements between BIS vs rSO2, BIS vs CBFV, and BIS vs. MAP[31]. For the latter model, we introduced a spline term at the LLA so that the association of BIS vs. MAP could be modelled separately for the period of MAP<LLA and the period of MAP>LLA [31,32]. We also normalized each patient’s BIS values to the BIS when MAP=LLA, in order to graphically depict % changes in BIS as MAP varied from the LLA.

Interaction of BIS Values, Metrics of Cerebral Perfusion, and Delirium

Student’s t-test and Mann-Whitney test were used to compare average BIS values and time percentage of BIS<45 during cardiopulmonary bypass between patients who did and did not develop delirium. Logistic regression models, both unadjusted and adjusted for age, bypass duration and logEuroScore (determined a priori to be potentially confounding variables), were used to examine the association of BIS values and postoperative delirium. An interaction term was added to the models to assess the interaction of mean BIS value and percent of time with MAP<LLA (categorized by the median value). For all analyses, p  < 0.05 was considered to be significant.

3. Results

3.1. Patient and Perioperative Characteristics

Seventy-nine patients met the inclusion criteria and were analyzed in this study (Fig 1). Mean age was 64.1 ± 11.1 years old (mean ± SD), and 75.9% were male. The mean BIS during cardiopulmonary bypass was 49.9 ± 9.8, and mean MAP at the LLA was 66.5 ± 10.2 mmHg. Patient and perioperative characteristics are further described in Table 1. Delirium assessments were performed after 22 patients had been enrolled and so are available on 57 patients. Patient and perioperative characteristics by delirium status are shown in Table 2.

Fig 1. Patient flow chart.

Fig 1.

Open in a new tab

228 patients were screened, among whom, 61 patients did not have BIS monitoring, 4 patients did not have stable anesthetic administration during bypass (i.e. no change in isoflurane concentration, no change in intravenous anesthetic infusions [if administered], and no anesthetic intravenous bolus) and LLA could not be identified in 84 patients. Finally, seventy-nine patients met the inclusion criteria and were analyzed in this study, with 57 patients having delirium assessment. BIS: bispectral index; CPB: Cardiopulmonary bypass surgery; LLA: lower limit of autoregulation.

Table 1.

Patient Demographics (n=79)

Age (years), Mean (SD) 64.1 (11.1)
Male, n (%) 60 (75.9%)
Race
 White n (%) 65 (82.3 %)
 Black n (%) 9 (11.4 %)
 Other n (%) 5 (6.3 %)
ASA score, Median [IQR] 4 [3–4]
LogEuroSCORE (%), Median [IQR] 3.46 [1.94–5.68]
Surgery
 CABG only, n (%) 47 (59.5%)
 CABG + Valve, n (%) 13 (16.5%)
 Valve only, n (%) 17 (21.5%)
 Other, n (%) 2 (2.5%)
Pre-bypass Medications
 Dexmedetomidine administration, n (%) 26 (32.9%)
 Dose of dexmedetomidine (among recipients of dexmedetomidine), (mcg), mean (SD) 55.7 (41.3)
 Fentanyl administration, n (%) 72 (91.1%)
 Dose of fentanyl (among recipients of fentanyl), (mcg), mean (SD) 404.2 (268.9)
 Ketamine administration n (%) 54 (68.4%)
 Pre-bypass dose of ketamine (among recipients of ketamine), (mg), mean (SD) 55.2 (37.3)
 Methadone administration, n (%) 4 (5.1%)
 Isoflurane concentration (%), median [IQR] 0.68 [0.58–0.77]
 Age-adjusted MAC, median [IQR] 0.58 [0.50–0.67]
Duration of cardiopulmonary bypass (min), Mean (SD) 105.53 (34.27)
Isoflurane concentration during bypass (%), Median [IQR] 1 [1–1]
Lowest temperature (°C), Mean (SD) 33.7 (1.9)
Rewarming time from start of bypass (min), median [IQR] 58.5 [46.0,79.3]
Lower limit of autoregulation (mmHg), Mean (SD) 66.52 (10.17)
Mean arterial pressure (mmHg), Mean (SD) 68.35 (6.51)
Bispectral index , Mean (SD) 49.92 (9.82)
Time percentage of Bispectral index <45, %, Mean (SD) 37.9 (36.0)
Open in a new tab

IQR: Interquartile Range; MAP: mean arterial blood pressure; ASA: The American Society of Anesthesiologists (ASA) Physical Status Classification System. CABG: coronary artery bypass grafting; Valve: heart valve repair or replacement.

Table 2.

Characteristics of patients with and without Delirium

Patient without Delirium (n=46) Patients with Delirium (n=11) P value
Age (years), Mean (SD) 64.2 (11.4) 66.3 (7.9) 0.58
Male, n (%) 37 (80%) 5 (46%) 0.02
Race 0.15
 White n(%) 40 (87 %) 8 (73%)
 Black n(%) 4 (9 %) 3 (27%)
 Other n(%) 2 (4 %) 0
ASA score, Median [IQR] 4 [3–4] 4 [3–4] 0.85
LogEuroSCORE (%), Median [IQR] 3.65 [IQR:1.95–6.12] 3.17 [IQR: 1.82–4.65] 0.56
Surgery 0.72
 CABG only, n (%) 26 (56.5) 7 (63.6%)
 CABG + Valve, n (%) 7 (15.2%) 1 (9.1%)
 Valve only, n (%) 12 (26.1%) 3 (27.3%)
 Other, n (%) 1 (2.2%) 0 (0%)
Pre-bypass Medicine
 Dexmedetomidine administration, n (%) 19 (41.3%) 6 (54.5%) 0.51
 Dose of dexmedetomidine (among recipients of dexmedetomidine), (mcg), mean (SD) 50.8 (35.5) 67.7 (60.5) 0.93
 Fentanyl administration, n (%) 41 (89.1%) 9 (81.8%) 0.61
 Dose of fentanyl (among recipients of fentanyl), (mcg), mean (SD) 320.7 (172.8) 316.7 (330.1) 0.32
 Ketamine administration n (%) 36 (78.3%) 8 (72.7%) 0.70
 Pre-bypass dose of ketamine (among recipients of ketamine), (mg), mean (SD) 49.0 (25.1) 55.6 (37.0) 0.94
 Methadone administration, n (%) 1 (2.2%) 0 (0%) 1.00
 Isoflurane concentration (%), median [IQR] 0.68 [0.58,0.77] 0.67 [0.64,0.75] 0.53
 Age-adjusted MAC, median [IQR] 0.58 [0.50,0.69] 0.60 [0.53,0.66] 0.37
Duration of cardiopulmonary bypass (min), Mean (SD) 108.12 (37.05) 117.45 (32.24) 0.41
Isoflurane concentration during bypass, Median [IQR] 1 [1–1] 1 [1–1] 0.23
Lowest temperature (°C), Mean (SD) 33.3 (1.98) 33.1 (2.45) 0.96
Rewarming time from start of bypass (min), Median [IQR] 57.5 [43.8,78.5][ 64.0 [50.5,82.0] 0.56
Lower Limit of Autoregulation (mmHg), Mean (SD) 67.78 (9.06) 65.64 (11.28) 0.41
Mean arterial pressure (mmHg), Mean (SD) 68.80 (6.06) 69.67 (8.59) 0.97
Bispectral Index , Mean (SD) 51.23 (9.10) 41.32 (9.76) 0.002
Time percentage of Bispectral Index <45, %, Mean (SD) 32.5 (33.0) 71.4 (36.6) 0.003
Open in a new tab

IQR: Interquartile Range; MAP: mean arterial blood pressure; MAC: age - adjusted minimum alveolar concentration; ASA: The American Society of Anesthesiologists (ASA) Physical Status Classification System. CABG: coronary artery bypass grafting; Valve: heart valve repair or replacement.

3.2. The Association of BIS Values with MAP Above and Below the LLA

As shown in Fig 2A, mean BIS values were lower for periods of MAP<LLA compared to periods of MAP>LLA (mean 49.35 ± 10.40 vs. 50.72 ± 10.04, p=0.002, mean difference =1.38 [standard error: 0.42], Fig 2A) during a period of constant anesthetic administration. Fig 2B depicts the continuous relationship between BIS and MAP, demonstrating an association of BIS with MAP, only when the MAP was below the LLA. Specifically, using mixed models with a spline at the LLA, there was a significant association of MAP and BIS when MAP was below the LLA (β = 0.15, 95% CI for the average slope across all patients 0.07 to 0.23, p<0.001), with 95% of the patient-specific slopes estimated to range between −0.42 to 0.72. However, there was no association of MAP and BIS when MAP was above the LLA (β = 0.03, 95% CI for the average slope across all patients −0.02 to 0.09, p=0.22), with 95% CI of the patient-specific slopes estimated to range between −0.36 to 0.42.

Fig 2. The association of BIS values with MAP above and below the LLA.

Fig 2.

Open in a new tab

(A) Mean value of BIS while MAP is above and below LLA of this cohort (n=79). In this panel, mean BIS value were calculated for each patient while MAP<LLA and MAP>LLA, thus each patient had two BIS values. A paired comparison was used to compare the BIS values. (B) Relationship between normalized BIS (normalized to the BIS value when MAP=LLA for each patient) and each ΔMAP (MAP – LLA) level. As shown, there is an association of BIS with MAP when MAP is below the LLA but not when MAP is above the LLA. In this figure, each bin of data represents all patients with BIS values at a particular blood pressure with respect to the LLA. LLA: lower limit of autoregulation. BIS: Bispectral Index. MAP: mean arterial blood pressure.

3.3. The Association of BIS Values with Other Measures of Cerebral Perfusion

BIS values also correlated with both cerebral oximetry and cerebral blood flow velocity in linear mixed models. On average, a one unit increase in RSO2 was associated with a mean 0.29 unit increase in BIS (95% CI for the estimate of the average slope across all patients 0.12–0.45, p<0.001), with 95% of the patient-specific slopes estimated to be −1.17 to 1.73. Similarly, on average, a one unit increase in CBFV was associated with a mean 0.22 unit increase in BIS (95% CI for the estimate of the average slope across all patients 0.15–0.30, p<0.001), with 95% of the patient-specific slopes estimated to be −0.41 to 0.85. When we examined these associations based on whether the MAP was above or below the LLA, the association of BIS and rSO2 was significant during periods of MAP<LLA (β=0.38, 95% CI for average slope 0.20–0.56, p<0.001) but not during periods of MAP>LLA (β=0.08, 95% CI for average slope ‒0.09–0.24, p=0.35). For CBFV, the association of BIS and CBFV was significant during periods of MAP<LLA (β=0.27, 95% CI for average slope 0.16–0.37, p<0.001) and significant, but less strong, during periods of MAP>LLA (β=0.18, 95% CI for average slope 0.09–0.27, p<0.001).

3.4. The interaction of BIS values, MAP below the LLA, and postoperative delirium

Delirium was assessed in 57 patients of whom 11 (19.3%) were positive for delirium. Out of 171 eligible assessments, data were not available for 10 assessments (5 staff unavailable, 4 patient refused, 1 patient unavailable). CAM-ICU was used for 5 assessments, with 3D-CAM information available for the remainder. Patients with delirium compared to without delirium had lower mean BIS values during bypass (mean ± SD, 41.32 ± 9.76 vs. 51.23 ± 9.10, p=0.002; Fig 3A) and a higher percentage of time with BIS<45 (mean ± SD, 71.4% ± 36.6% vs. 32.5% ± 33.0%, p=0.003; Fig 3B). The MAP at the LLA was similar between delirious and non-delirious patients (67.8 ± 9.1 mmHg vs. 65.6 ± 11.3 mmHg, p=0.41). In unadjusted regression models, both lower mean BIS values (odds ratio: 1.14, 95% CI: 1.03–1.26, p=0.010) and longer duration of time of BIS<45 (odds ratio: 1.03, 95% CI: 1.01–1.06, p=0.004) were associated with delirium. Similarly, in models adjusted for age, bypass duration and logEuroScore, both lower BIS values (odds ratio: 1.13, 95% CI: 1.03–1.24, p=0.014) and longer duration of BIS<45 (odd ratio: 1.03, 95% CI: 1.01–1.06, p=0.006) were associated with delirium. However, there was no interaction between BIS values (either mean BIS or % time <45) and duration of MAP<LLA with respect to delirium (all p-interactions >0.05 in unadjusted and adjusted models). In other words, the association of low BIS values and delirium was not significantly different in patients with a long duration of time of MAP<LLA (which might contribute to lower BIS values) compared to a short duration of time of MAP<LLA.

Fig 3. The comparison of BIS between patients with and without postoperative delirium.

Fig 3.

Open in a new tab

(A) Mean BIS and (B) time percentage of BIS < 45 of patients with and without delirium. The patients with delirium had significantly lower BIS (student’s t test) and longer time percentage of BIS<45 (Mann-Whitney test) during bypass compared with that of patients without delirium. BIS: Bispectral Index; MAP: mean arterial blood pressure. Error bar: Standard Deviation.

4. Discussion

In this study, we found that MAP below the LLA was associated with a small dose-dependent reduction in BIS, even while anesthetic concentrations were held constant. BIS values were also related to several other measures of cerebral perfusion, including rSO2 and CBFV, particularly when the MAP was below the LLA. Both lower BIS and longer duration of time with BIS<45 in cardiac surgery were associated with post-operative delirium but there was no interaction with duration of MAP<LLA.

BIS is a processed EEG that is used to monitor and titrate depth of sedation during surgery. Several observational studies or secondary data analyses or trials have demonstrated an association of low BIS values (i.e. <40 – 45) during surgery and patient mortality and delirium [3–5,7–9]. Based on these findings, many practitioners titrate anesthetic agents according to BIS values in high-risk patients to avoid excessive depth of sedation. However, BIS values often correlate poorly with end-tidal anesthetic concentrations, have non-linear relationships with minimum alveolar concentration of anesthetic gases, and may be biased in older adults [33,34]. Additionally, besides anesthetic drugs, other types of medications or physiologic changes may also affect BIS values, such as neuromuscular blocking agents, shock, cerebral ischemia, etc [16,35,36]. Understanding the factors which affect BIS values and may relate to postoperative outcomes is clinically important.

Randomized trials using BIS to guide anesthetic titration have not shown consistent results[7–9]. Early studies demonstrated that BIS-guided anesthesia reduced anesthetic exposure and decreased the risk of postoperative delirium. However, two recent studies reported no reduction in delirium using a strategy to avoid burst suppression and excessively deep anesthesia in high-risk patients under general anesthesia[10] or a strategy targeting light sedation during surgical repair of a hip fracture under spinal anesthesia[11]. Further, a large pragmatic trial in 6644 patients showed no reduction in 1-year mortality or other secondary outcomes (delirium was not reported) in patients randomized to deep anesthesia group (BIS target =35) compared with light anesthesia (BIS target =50)[6]. Therefore, it is unclear if titration of anesthetic to avoid low BIS values is an effective intervention to reduce postoperative delirium and other complications.

The current study was conducted to investigate one possible hypothesis to explain the conflicting findings of prior studies—that low BIS values may result from cerebral hypoperfusion, and thus BIS-guided interventions may need to account for both anesthetic concentration and adequacy of cerebral perfusion. There is strong biologic rationale for this hypothesis. With a decrease in cerebral perfusion, faster frequency components of the EEG decrease while slow wave EEG components increase. If cerebral perfusion continues to decrease, the EEG becomes silent[15,37,38]. This may result in low BIS values as cerebral perfusion declines, because the BIS value is derived from several key EEG features, i.e. the relative BetaRatio, Synch-FastSlow and Burst Suppression[39,40]. Therefore, it is plausible that BIS values seen during surgery may reflect an element of cerebral perfusion. Indeed, the effect of cerebral perfusion on BIS values has been confirmed in case reports of critically ill patients with severely reduced cerebral perfusion, including profound cerebral ischemia and cardiac arrest[41–43].

The present research confirms an association between several markers of cerebral perfusion and BIS values at clinically relevant states of cerebral perfusion during cardiopulmonary bypass. However, the absolute magnitude of the association was small. Compared to periods when the MAP was at the LLA, there was estimated to be less than a 10% reduction in BIS values even when the MAP decreased to 20 mm Hg below the LLA. For a patient with BIS values in the 40–60 range, this would correlate to a reduction in BIS value of only 5, even with a MAP substantially below the LLA. Thus, the clinical relevance of any perfusion contribution to BIS values is limited and likely does not explain conflicting results of prior trials. Nevertheless, clinician should be aware of the potential for adequacy of cerebral perfusion to be reflected in BIS values.

On the other hand, the results of the current study support the biologic relevance of the LLA as determined by correlation methods. Similar results have been reported by Sekhon et al., who demonstrated a positive relationship between partial pressure of oxygen in brain tissue (Pbto2) and MAP while MAP is below optimal value, however, once the MAP is above the optimal value, the Pbto2 tends to be stable[44]. The results suggest that incorporation of BIS values or even EEG waveforms might add precision to LLA-identification algorithms, and this is a future research direction.

Finally, these observations in a small number of patients confirm prior findings that low BIS values are associated with postoperative delirium (9). One possible explanation is that the processed EEG reveals underlying brain vulnerability (such as structural atrophy or decreased cognition) that predispose patients to delirium. Indeed, several studies have reported that baseline cognition is associated with several intraoperative EEG parameters, such as frontal alpha power[45–47]. However, the risk for delirium based on BIS values did not appear to be different by amount of time below the lower limit of autoregulation. In other words, we did not see evidence that low BIS values (either mean BIS or % time <45) and cerebral hypoperfusion (MAP<LLA) interacted to increase the risk of delirium. This may reflect our observation that the absolute change in BIS values was small, even as MAP decreased below the LLA. However, the study sample was small and delirium was assessed using a screening instrument and so further research is needed in this area.

Strengths and Limitations

Strengths of this study include the use of multi-modal monitoring (BIS, arterial blood pressure, NIRS and TCD) during a period of a stable anesthetic level. We identified LLA for each patient using established methods. The findings were examined with respect to a clinically relevant outcome. However, there are several important limitations to consider. First, we used the actual BIS value, instead of the raw EEG. A further research question is to examine the effects of changing cerebral perfusion on EEG indicators of ischemia. Second, delirium was assessed using a screening measure once daily in only a subset of patients, and thus the incidence was lower than has been reported in some studies. Nevertheless, a validated instrument was administered by trained staff for all assessments, and any misclassification should not be different by BIS status. Additionally, because the number of patients in the study was small, some observed relationships with respect to delirium may be due to imprecision. For instance, female sex was associated with delirium in this study, although sex has not been a consistent risk factor for delirium in the literature[48]. Delirium can be multi-factorial and although clinical care is highly standardized at Johns Hopkins, there are always sources of heterogeneity and there is a potential for unmeasured confounding. Cognition is one particularly important factor to measure in future studies as impaired cognition is an important risk factor for delirium, although we would not expect it to confound the association of cerebral perfusion metrics and BIS values. Additionally, the analysis examining an interaction of depth of sedation and cerebral perfusion may be underpowered, and hence this was as an exploratory aim. Third, we did not account for depth of sedation or cerebral perfusion in the post-operative period. A mild hypothermia strategy was also used during the surgery which might affect anesthetic requirements and the LLA, and future studies could examine patients with less temperature changes. Finally, the findings were observed during a period of cardiopulmonary bypass, but the results should be applicable to non-bypass periods with pulsatile perfusion.

Conclusion

BIS values are associated with metrics of cerebral perfusion. However, the absolute change in BIS is small across clinically relevant ranges of cerebral perfusion.

Supplementary Material

1

Highlights.

  • The cerebrovascular contribution to bispectral index (BIS) values during usual clinical care is not well established.

  • The results demonstrated that BIS values were related to several metrics of cerebral perfusion (including mean arterial blood pressure (MAP), regional cerebral oximetry, and cerebral blood flow velocity) while mean blood pressure is below the lower limit of cerebral autoregulation (LLA), although the magnitude of the contributions are small.

  • These results suggest a small cerebrovascular contribution to BIS values at clinically-relevant MAP values that may be relative to an individual patient’s LLA, but no interaction of BIS values, markers of cerebral perfusion, and delirium.

Funding Statement

This study was funded by the NIH K76 AG057020 and the MAGIC that Matters Award (Dr Brown).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Data Availability

The data may be available on reasonable request to corresponding author, Dr. Charles Brown (cbrownv@jhmi.edu), with the appropriate institutional agreements in place.

Conflict of Interest

Dr Brown reported receiving grants from the National Institutes of Health (NIH) during the conduct of the study, and consulted for and participates in a data share with Medtronic. Dr. Hogue reported receiving grants and personal fees for being a consultant and providing lectures for Medtronic/Covidien, Inc., being a consultant to Edwards Lifesciences, Merck, Inc., and receiving grants from the NIH outside of the submitted work. Dr. Lee has received grants and consulted for Medtronic/Covidien. Dr. Lee is also a paid consultant for Edwards Life Sciences and has funding from the NIH outside the submitted work. No authors have patents or financial interests related to bispectral index monitoring, although several companies for which authors have consulted sell processed EEG monitors (Medtronic, Edwards). The entirety of this study was conducted without any participation of these companies, and the authors have no financial interests in the outcome of this study.

References

  • [1].Avidan MS, Zhang L, Burnside BA, Finkel KJ, Searleman AC, Selvidge JA, et al. Anesthesia awareness and the bispectral index. N Engl J Med 2008;358:1097–108. 10.1056/NEJMoa0707361. [DOI] [PubMed] [Google Scholar]
  • [2].Ramsay MAE. Role of brain function monitoring in the critical care and perioperative settings. Semin Anesth Perioper Med Pain 2005. 10.1053/j.sane.2005.10.002. [DOI] [Google Scholar]
  • [3].Monk TG, Saini V, Weldon BC, Sigl JC. Anesthetic management and one-year mortality after noncardiac surgery. Anesth Analg 2005;100:4–10. 10.1213/01.ANE.0000147519.82841.5E. [DOI] [PubMed] [Google Scholar]
  • [4].Lindholm ML, Träff S, Granath F, Greenwald SD, Ekbom A, Lennmarken C, et al. Mortality within 2 years after surgery in relation to low intraoperative bispectral index values and preexisting malignant disease. Anesth Analg 2009;108:508–12. 10.1213/ane.0b013e31818f603c. [DOI] [PubMed] [Google Scholar]
  • [5].Leslie K, Myles PS, Forbes A, Chan MTV. The effect of bispectral index monitoring on long-term survival in the B-aware trial. Anesth Analg 2010;110:816–22. 10.1213/ANE.0b013e3181c3bfb2. [DOI] [PubMed] [Google Scholar]
  • [6].Short TG, Campbell D, Frampton C, Chan MTV, Myles PS, Corcoran TB, et al. Anaesthetic depth and complications after major surgery: an international, randomised controlled trial. Lancet 2019;394:1907–14. 10.1016/S0140-6736(19)32315-3. [DOI] [PubMed] [Google Scholar]
  • [7].Sieber FE, Zakriya KJ, Gottschalk A, Blute MR, Lee HB, Rosenberg PB, et al. Sedation depth during spinal anesthesia and the development of postoperative delirium in elderly patients undergoing hip fracture repair. Mayo Clin Proc 2010;85:18–26. 10.4065/mcp.2009.0469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [8].Radtke FM, Franck M, Lendner J, Krüger S, Wernecke KD, Spies CD. Monitoring depth of anaesthesia in a randomized trial decreases the rate of postoperative delirium but not postoperative cognitive dysfunction. Br J Anaesth 2013;110:i98–105. 10.1093/bja/aet055. [DOI] [PubMed] [Google Scholar]
  • [9].Chan MTV, Cheng BCP, Lee TMC, Gin T. BIS-guided anesthesia decreases postoperative delirium and cognitive decline. J Neurosurg Anesthesiol 2013;25:33–42. 10.1097/ANA.0b013e3182712fba. [DOI] [PubMed] [Google Scholar]
  • [10].Wildes TS, Mickle AM, Abdallah Ben A, Maybrier HR, Oberhaus J, Budelier TP, et al. Effect of electroencephalography-guided anesthetic administration on postoperative delirium among older adults undergoing major surgery the engages randomized clinical trial. JAMA - J Am Med Assoc 2019;321:473–83. 10.1001/jama.2018.22005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [11].Sieber FE, Neufeld KJ, Gottschalk A, Bigelow GE, Oh ES, Rosenberg PB, et al. Effect of Depth of Sedation in Older Patients Undergoing Hip Fracture Repair on Postoperative Delirium: The STRIDE Randomized Clinical Trial. JAMA Surg 2018;153:987–95. 10.1001/jamasurg.2018.2602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [12].Hayashida M, Chinzei M, Komatsu K, Yamamoto H, Tamai H, Orii R, et al. Detection of cerebral hypoperfusion with bispectral index during paediatric cardiac surgery. Br J Anaesth 2003. 10.1093/bja/aeg107. [DOI] [PubMed] [Google Scholar]
  • [13].Hagihira S. Changes in the electroencephalogram during anaesthesia and their physiological basis. Br J Anaesth 2015;115:i27–31. 10.1093/bja/aev212. [DOI] [PubMed] [Google Scholar]
  • [14].Bewernitz M, Derendorf H. Electroencephalogram-based pharmacodynamic measures: A review. Int J Clin Pharmacol Ther 2012;50:162–84. 10.5414/CP201484. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Foreman B, Claassen J. Quantitative EEG for the detection of brain ischemia. Crit Care 2012;16:216. 10.1186/cc11230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [16].Cavus E, Meybohm P, Doerges V, Hoecker J, Betz M, Hanss R, et al. Effects of cerebral hypoperfusion on bispectral index: A randomised, controlled animal experiment during haemorrhagic shock. Resuscitation 2010;81:1183–9. 10.1016/j.resuscitation.2010.05.018. [DOI] [PubMed] [Google Scholar]
  • [17].Sessler DI, Sigl JC, Kelley SD, Chamoun NG, Manberg PJ, Saager L, et al. Hospital stay and mortality are increased in patients having a triple low of low blood pressure, low bispectral index, and low minimum alveolar concentration of volatile anesthesia. Anesthesiology 2012;116:1195–203. 10.1097/ALN.0b013e31825683dc. [DOI] [PubMed] [Google Scholar]
  • [18].Panerai RB. Cerebral autoregulation: from models to clinical applications. CardiovascEng 2008;8:42–59. [DOI] [PubMed] [Google Scholar]
  • [19].Lassen N. Cerebral blood flow and oxygen consumption in man. Physiol Rev 1959;39:183–238. [DOI] [PubMed] [Google Scholar]
  • [20].Nakano M, Nomura Y, Whitman G, Sussman M, Schena S, Kilic A, et al. Cerebral autoregulation in the operating room and intensive care unit after cardiac surgery. Br J Anaesth 2021;126:967–74. 10.1016/j.bja.2020.12.043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [21].Marcantonio ER, Ngo LH, O’Connor M, Jones RN, Crane PK, Metzger ED, et al. 3D-CAM: Derivation and validation of a 3-minute diagnostic interview for CAM-defined delirium: A cross-sectional diagnostic test study. Ann Intern Med 2014;161:554–61. 10.7326/M14-0865. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [22].Ely EW, Bernard GR, Speroff T, Gautam S, Dittus R, May L, et al. Delirium in mechanically ventilated patients: Validity and reliability of the Confusion Assessment Method for the intensive care unit (CAM-ICU). J Am Med Assoc 2001;286:2703–10. 10.1001/jama.286.21.2703. [DOI] [PubMed] [Google Scholar]
  • [23].Ely EW, Margolin R, Francis J, May L, Truman B, Dittus R, et al. Evaluation of delirium in critically ill patients: Validation of the Confusion Assessment Method for the intensive care unit (CAM-ICU). Crit Care Med 2001;29:1370–9. 10.1097/00003246-200107000-00012. [DOI] [PubMed] [Google Scholar]
  • [24].Czosnyka M, Smielewski P, Kirkpatrick P, Laing RJ, Menon D, Pickard JD. Continuous assessment of the cerebral vasomotor reactivity in head injury. Neurosurgery 1997;41:11–9. 10.1097/00006123-199707000-00005. [DOI] [PubMed] [Google Scholar]
  • [25].Czosnyka M, Smielewski P, Kirkpatrick P, Piechnik S, Laing R, Pickard JD. Continuous Monitoring of Cerebrovascular Pressure-Reactivity in Head Injury. Acta Neurochir Suppl 1998;71:74–7. 10.1007/978-3-7091-6475-4_23. [DOI] [PubMed] [Google Scholar]
  • [26].Czosnyka M, Smielewski P, Lavinio A, Pickard JD, Panerai R. An assessment of dynamic autoregulation from spontaneous fluctuations of cerebral blood flow velocity: a comparison of two models, index of autoregulation and mean flow index. Anesth Analg 2008;106:234–9, table of contents. 10.1213/01.ane.0000295802.89962.13. [DOI] [PubMed] [Google Scholar]
  • [27].Lee JK, Kibler KK, Benni PB, Easley RB, Czosnyka M, Smielewski P, et al. Cerebrovascular reactivity measured by near-infrared spectroscopy. Stroke 2009;40:1820–6. [DOI] [PubMed] [Google Scholar]
  • [28].Donnelly J, Czosnyka M, Adams H, Robba C, Steiner LA, Cardim D, et al. Individualizing Thresholds of Cerebral Perfusion Pressure Using Estimated Limits of Autoregulation. Crit Care Med 2017;45:1464–71. 10.1097/CCM.0000000000002575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [29].Ono M, Joshi B, Brady K, Easley RB, Zheng Y, Brown C, et al. Risks for impaired cerebral autoregulation during cardiopulmonary bypass and postoperative stroke. Br J Anaesth 2012;103:391–8. 10.1093/bja/aes148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [30].Brown CH, Neufeld KJ, Tian J, Probert J, Laflam A, Max L, et al. Effect of Targeting Mean Arterial Pressure during Cardiopulmonary Bypass by Monitoring Cerebral Autoregulation on Postsurgical Delirium among Older Patients: A Nested Randomized Clinical Trial. JAMA Surg 2019;154:819–26. 10.1001/jamasurg.2019.1163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [31].Minalu G, Aerts M, Coenen S, Versporten A, Muller A, Adriaenssens N, et al. Application of mixed-effects models to study the country-specific outpatient antibiotic use in Europe: A tutorial on longitudinal data analysis. J Antimicrob Chemother 2011;66:vi97–87. 10.1093/jac/dkr460. [DOI] [PubMed] [Google Scholar]
  • [32].Landau S, Everitt B. Chapter 08. Analysis of Repeated Measures II Linear Mixed Effects Models Computer Delivery of Cognitive Behavioral Therapy. A Handb Stat Anal Using SPSS 2004. [Google Scholar]
  • [33].Whitlock EL, Villafranca AJ, Lin N, Palanca BJ, Jacobsohn E, Finkel KJ, et al. Relationship between bispectral index values and volatile anesthetic concentrations during the maintenance phase of anesthesia in the B-Unaware trial. Anesthesiology 2011. 10.1097/ALN.0b013e3182395dcb. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Ni K, Cooter M, Gupta DK, Thomas J, Hopkins TJ, Miller TE, et al. Paradox of age: older patients receive higher age-adjusted minimum alveolar concentration fractions of volatile anaesthetics yet display higher bispectral index values. Br J Anaesth 2019. 10.1016/j.bja.2019.05.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [35].Messner M, Beese U, Romstöck J, Dinkel M, Tschaikowsky K. The bispectral index declines during neuromuscular block in fully awake persons. Anesth Analg 2003;97:488–91. 10.1213/01.ANE.0000072741.78244.C0. [DOI] [PubMed] [Google Scholar]
  • [36].Myles PS. Bispectral index monitoring in ischemic-hypoxic brain injury. J Extra Corpor Technol 2009;41:p15–9. [PMC free article] [PubMed] [Google Scholar]
  • [37].Vespa PM, Nuwer MR, Juhász C, Alexander M, Nenov V, Martin N, et al. Early detection of vasospasm after acute subarachnoid hemorrhage using continuous EEG ICU monitoring. Electroencephalogr Clin Neurophysiol 1997;103:607–15. 10.1016/S0013-4694(97)00071-0. [DOI] [PubMed] [Google Scholar]
  • [38].Liu X, Pu Y, Wu D, Zhang Z, Hu X, Liu L. Cross-Frequency Coupling Between Cerebral Blood Flow Velocity and EEG in Ischemic Stroke Patients With Large Vessel Occlusion. Front Neurol 2019;10:194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [39].Kelley SD. Monitoring Consciousness Using the Bisectral Index During Anesthesia 2nd edit. Asp Med Syst 2007. [Google Scholar]
  • [40].Dahaba AA. Different conditions that could result in the bispectral index indicating an incorrect hypnotic state. Anesth Analg 2005;101:765–73. 10.1213/01.ane.0000167269.62966.af. [DOI] [PubMed] [Google Scholar]
  • [41].England MR. The changes in bispectral index during a hypovolemic cardiac arrest. Anesthesiology 1999;91:1947–9. 10.1097/00000542-199912000-00050. [DOI] [PubMed] [Google Scholar]
  • [42].Azim N, Wang CY. The use of bispectral index during a cardiopulmonary arrest: A potential predictor of cerebral perfusion. Anaesthesia 2004;59:610–2. 10.1111/j.1365-2044.2004.03736.x. [DOI] [PubMed] [Google Scholar]
  • [43].Mérat S, Lévecque JP, Gulluche Le Y, Diraison Y, Brinquin L, Hoffmann JJ. BIS monitoring may allow the detection of severe cerebral ischemia. Can J Anesth 2001;48:1066–9. 10.1007/BF03020370. [DOI] [PubMed] [Google Scholar]
  • [44].Sekhon MS, Gooderham P, Menon DK, Brasher PMA, Foster D, Cardim D, et al. The Burden of Brain Hypoxia and Optimal Mean Arterial Pressure in Patients With Hypoxic Ischemic Brain Injury After Cardiac Arrest. Crit Care Med 2019;47:960–9. 10.1097/CCM.0000000000003745. [DOI] [PubMed] [Google Scholar]
  • [45].Giattino CM, Gardner JE, Sbahi FM, Roberts KC, Cooter M, Moretti E, et al. Intraoperative frontal alpha-band power correlates with preoperative neurocognitive function in older adults. Front Syst Neurosci 2017;11:24. 10.3389/fnsys.2017.00024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [46].Erdogan MA, Demirbilek S, Erdil F, Aydogan MS, Ozturk E, Togal T, et al. The effects of cognitive impairment on anaesthetic requirement in the elderly. Eur J Anaesthesiol 2012;29:326–31. 10.1097/EJA.0b013e32835475c6. [DOI] [PubMed] [Google Scholar]
  • [47].Shao YR, Kahali P, Houle TT, Deng H, Colvin C, Dickerson BC, et al. Low frontal alpha power is associated with the propensity for burst suppression: An electroencephalogram phenotype for a “vulnerable brain”. Anesth Analg 2020;131:1529–39. 10.1213/ANE.0000000000004781. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [48].Zaal IJ, Devlin JW, Peelen LM, Slooter AJC. A systematic review of risk factors for delirium in the ICU. Crit Care Med 2015;43:40–7. 10.1097/CCM.0000000000000625. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1713325-supplement-1.doc (84.5KB, doc)

RESOURCES