Journal of Clinical Anesthesia (2016) 35, 456464

Original Contribution

Cerebral oxygenation in the beach chair position for

shoulder surgery in regional anesthesia: impact on
cerebral blood ow and neurobehavioral outcome
Jos A. Aguirre MD, MSc a,, Olivia Mrzendorfer MMed a ,
Muriel Brada MMed a , Andrea Saporito MD, MHA b ,
Alain Borgeat MD (Professor for Anesthesiology)a , Philipp Bhler MD a
a
Division of Anesthesiology, Balgrist University Hospital, Zurich, Switzerland
b
Anaesthesiology Department, Bellinzona Regional Hospital, Bellinzona, Switzerland

Received 10 March 2016; revised 9 August 2016; accepted 14 August 2016

Keywords:
Abstract
Beach chair position,
Study objective: Beach chair position is considered a potential risk factor for central neurological events
neurocognitive outcome;
particularly if combined with low blood pressure. The aim of this study was to assess the impact of regional
Cerebral blood ow;
anesthesia on cerebral blood ow and neurobehavioral outcome.
Cerebral oxygenation;
Design: This is a prospective, assessor-blinded observational study evaluating patients in the beach chair po-
Intraoperative monitoring;
sition undergoing shoulder surgery under regional anesthesia.
Near-infrared spectroscopy
Setting: University hospital operating room.
Patients: Forty patients with American Society of Anesthesiologists classes I-II physical status scheduled
for elective shoulder surgery.
Interventions: Cerebral saturation and blood ow of the middle cerebral artery were measured prior to an-
esthesia and continued after beach chair positioning until discharge to the postanesthesia care unit. The an-
esthesiologist was blinded for these values. Controlled hypotension with systolic blood pressure 100 mm
Hg was maintained during surgery.
Measurements: Neurobehavioral tests and values of regional cerebral saturation, bispectral index, the mean
maximal blood ow of the middle cerebral artery, and invasive blood pressure were measured prior to re-
gional anesthesia, and measurements were repeated after placement of the patient on the beach chair position
and every 20 minutes thereafter until discharge to postanesthesia care unit. The neurobehavioral tests were
repeated the day after surgery.
Main results: The incidence of cerebral desaturation events was 5%. All patients had a signicant blood
pressure drop 5 minutes after beach chair positioning, measured at the heart as well as the acoustic meatus
levels, when compared with baseline values (P b .05). There was no decrease in either the regional cerebral
saturation (P = .136) or the maximal blood ow of the middle cerebral artery (P = .212) at the same time
points. Some neurocognitive tests showed an impairment 24 hours after surgery (P b .001 for 2 of 3 tests).

Disclosures: This work was supported by a grant of the European Society of Regional Anesthesia and Pain Medicine.
Corresponding author at: Consultant Anesthetist, Division of Anesthesiology, Balgrist University Hospital, Forchstrasse 340, 8008 Zurich; Switzerland. Tel.: +41 44
386 31 03; fax: +41 44 386 11 09.
E-mail addresses: jose.aguirre@balgrist.ch (J.A. Aguirre), oliviam@access.uzh.ch (O. Mrzendorfer), Muriel.brada@balgrist.ch (M. Brada),
Andrea.Saporito@eoc.ch (A. Saporito), alain.borgeat@balgrist.ch (A. Borgeat), karlphilippbuehler@gmail.com (P. Bhler).

http://dx.doi.org/10.1016/j.jclinane.2016.08.035
0952-8180/ 2016 Elsevier Inc. All rights reserved.
Anesthesia, cerebral blood flow and beach chair position
Conclusions: Beach chair position in patients undergoing regional anesthesia for shoulder surgery had no
major impact on cerebral blood ow and cerebral oxygenation. However, some impact on neurobehavioral
outcome 24 hours after surgery was observed.
2016 Elsevier Inc. All rights reserved.
1. Introduction
Beach chair position (BCP) is commonly used for shoulder
surgery compared with the lateral decubitus position (LDP) to
facilitate shoulder joint access and optimal visualization [1].
However, these advantages are contrasted by challenging he-
modynamic changes, and several case reports are blaming
the BCP for devastating central neurologic complications [2]
such as cranial nerve injury [3], visual loss [4], and cerebral in-
farction [5,6]. However, the real impact of BCP on these se-
vere complications remains unclear [7,8].
Combining BCP and head xation can lead to cerebral tis-
sue hypoperfusion during surgery caused by mechanical ob-
struction of cerebral blood ow (CBF) due to the extreme
rotation of the head in combination with the traction in the op-
erated arm [9]. Additional risk factors for reduced CBF are the
frequently used combination of general anesthesia and con-
trolled hypotension for shoulder arthroscopy as well as open
shoulder surgery [10]. Furthermore, congenital variations of
the circle of Willis anatomy with incomplete expression have
been described in 59% to 79% of adults [11,12]. These con-
genital variations present an additional risk for patients in the
BCP undergoing shoulder surgery [13]. Near-infrared spec-
troscopy (NIRS) constitutes a useful tool to improve current
monitoring by providing additional information regarding the
regional cerebral tissue oxygenation [7]. Recent studies have
shown that when NIRS is used for a cerebral oximetry moni-
toring, it reliably detects cerebral hypoperfusion during shoul-
der surgery in the BCP [14,15]. In 1 case, a cause and effect
relation between regional cerebral (capillary) oxygen satura-
tion (rScO2), mean arterial pressure (MAP), and end-tidal
CO2 was reported. [15] Murphy et al. compared the incidence
of cerebral desaturation during shoulder surgery under general
anesthesia in the BCP vs the LDP, demonstrating a signicant
reduction of cerebral oxygen saturation in the BCP and
highlighting the importance of this position in cerebral desa-
turation [14]. Recently, 2 studies have reported the importance
of maintaining an adequately high systemic blood pressure in
the BCP to allow for cerebral autoregulation and adequate
CBF [16,17]. However, no study has described the correlation
between CBF [18] and cerebral oxygenation in a controlled
hypotension protocol for shoulder surgery in the BCP.
The aim of this prospective, observational, assessor-
blinded study was to evaluate the prevalence of regional cere-
bral oxygen desaturation in American Society of Anesthesiol-
ogists (ASA) I-II patients undergoing shoulder surgery in the
BCP under regional anesthesia with conscious sedation and a
controlled hypotension protocol (systolic blood pressure 80-
100 mm Hg at heart level) [19]. Our hypothesis was that
457
NIRS reliably detects changes in cerebral oxygenation during
regional anesthesia in the BCP and that the incidence of cere-
bral desaturation events (CDEs) would correlate with CBF im-
pairment and neurobehavioral decline. Our primary outcome
was the incidence of CDEs (expressed as a drop of absolute
rScO2 to a value b55% for N15 seconds of baseline and in rel-
ative terms as a decrease in rScO2 20% compared with the
baseline value) [5] during regional anesthesia with conscious
sedation under controlled hypotension protocol.
Secondary outcomes were the effects of blood pressure
(measured at heart level and at the level of the acoustic meatus)
on rScO2, the effects of the CDEs on the neurological and neu-
robehavioral outcome, as well as the effects of BCP on CBF
and the correlations between CDEs, neurobehavioral, and
CBF decline.
2. Materials and methods
The study was approved by the local Ethics Committee
(Kantonale Ethikkommission Zrich, KEK-Zh-Nr: 2012-
0112). Written informed consent was obtained from each pa-
tient. Forty ASA I-II adult patients scheduled for elective, uni-
lateral shoulder surgery were included in this prospective,
assessor-blinded, single-center cohort study. Exclusion criteria
were a history of central neurological diagnosis (transient is-
chemic attack, stroke, bleeding, syncope, chronic headache,
cervical disk herniation, spinal cord injury, recent vision im-
pairment/loss, cerebral tumor or metastasis, orthostatic hypo-
tension), recent myocardial infarction (b6 months), known
relevant carotid stenosis (N40%) or known ow disturbance
of vertebral arteries, pregnancy, allergies to any drug used
for anesthesia, and known neurobehavioral disorders or base-
line Minimental State Examination test result b24.
The evening prior to surgery, a standardized neurologic ex-
amination (pupil size and reaction, lateralization tests of both
extremities, Glasgow Coma Scale, Minimental State Examina-
tion test) and neurobehavioral tests (Trail Making Test [TMT]
A, TMT B, Grooved Pegboard) were conducted as baseline
measurements by an anesthesiologist not involved in the fur-
ther anesthetic management of the patient. TMT A and TMT
B are neuropsychological tests focusing on visual attention
and task switching. They consist of 2 parts in which the patient
is instructed to connect a set of 25 dots (numbers in A, alternat-
ing numbers and letters in B) as fast as possible. They provide
information about visual search speed, speed of processing,
scanning, mental exibility, as well as executive functioning.
The Grooved Pegboard test assesses ne motor control and
speed. Patients have to t notched pegs into matching holes
458
as quickly as possible. The time taken to ll all holes is record-
ed. All 3 tests were suggested by Murkin et al. in a consensus
to assess neurobehavioral outcomes after cardiac surgery and
have been used in other studies focusing on cerebral oxygena-
tion [20,21]. The patients were evaluated by those tests both
preoperatively and postoperatively (24 hours after surgery)
by the same anesthesiologist using only the nonoperated
arm. Prior to surgery, the carotid arteries of the patients were
assessed by ultrasonography to exclude signicant occlusions.
Cerebral oximetry values are known to be affected by several
factors, such as type of anesthetic used, depth of anesthesia,
FIO2, PaCO2, as well as blood pressure management [22].
Therefore, anesthetic management was standardized in all pa-
tients, and the anesthesiologist and the anesthesia nurse in
charge of each patient were blinded for the rScO2 value during
the whole procedure by covering the NIRS monitor and
switching off the CDE alarms. All patients were premedicated
with 0.1 mg/kg oral midazolam (maximum 7.5 mg) 1 hour pri-
or to arrival in the induction room. There, standard monitoring
was installed (oxygen saturation, electrocardiogram, noninva-
sive blood pressure cuff). An intravenous access was placed on
the nonoperated arm). In addition, all patients got a bispectral
index (BIS) electrode on the forehead for comparison for
rScO2 value changes, and also a radial arterial line for invasive
blood pressure measurement was placed on the nonoperated
arm. Furthermore, 2 NIRS sensors were applied on both sides
of the forehead just above the BIS electrode (INVOS near-
infrared spectroscopy monitoring; Somametics, Covidien,
MA) [23]. All patients had a middle cerebral artery ow
velocity (Vmax MCA: maximal blood ow) measurement at
the nonsurgical site (EZ-Dop, 2-MHz probe for transtemporal
measurement; Neurolite, Switzerland). Baseline oxygenation
parameters and Vmax MCA measurements were set/performed
5 minutes after initial monitoring installation. Arterial oxygen
saturation and rScO2 values were measured in patients while
breathing a 40% air/oxygen mixture. Blood pressure was
recorded at heart and acoustic meatus levels by placing the
transducer at the corresponding height at the different mea-
surement time points. All measurements were recorded at
baseline, 5 minutes after induction of regional anesthesia, 5
minutes after beach chair positioning, at the surgical start,
and then every 20 minutes thereafter until discharge to posta-
nesthesia care unit (PACU) or ward. Measurements were also
performed in the case of any adverse events (systemic deoxy-
genation, hypoventilation states). BIS and rScO2 values were
continuously recorded by the monitoring system for evalua-
tion after surgery. An interscalene catheter using neurostimu-
lation was placed as described elsewhere [24] in all patients
to avoid a negative inuence of opioids on neurobehavioral
tests. The interscalene block was performed with 40 mL of
ropivacaine 0.5%. To obtain a mild or moderate sedation level
and analgesia for comfort during beach chair positioning of the
spontaneous breathing patient, a combined target controlled
infusion (target-controlled infusion pump, Graseby; SIMS
Graseby Limited Watford, Herts, UK) was applied using
propofol (0.5-0.8 g/mL effect site concentration) and
J.A. Aguirre et al.
remifentanil (0.5-0.8 ng/mL effect site concentration). The tar-
get sedation level during surgery was an Observer's Assess-
ment of Alertness/Sedation Scale score of 4 to 3 to ensure a
mild to moderate sedated and cooperative patient according
to the modied Observer's Assessment of Alertness/Sedation
Scale [25]. The inspired oxygen fraction was kept at 40%.
All interventions were performed in BCP with an angle of
65 from horizontal. Prior to surgery, cefuroxim 1.5 g intrave-
nous (IV) antibiotic prophylaxis was given. A controlled hy-
potension protocol for a systolic blood pressure at heart level
(SAPheart) 100 mm Hg was used according to our clinical
practice. An intervention protocol for a SAPheart N100 mm
Hg for more than 3 minutes was included as follows: (1) deep-
ening of sedation with propofol/remifentanil according to clin-
ical ndings, (2) clonidine 75-150 g IV, and (3) labetalol
10 mg bolus IV. The intervention protocol for a SAPheart
b80 mm Hg for more than 3 minutes consisted of (1) crystal-
loid bolus of 250 mL, (2) ephedrine IV 2.5-5 mg, and (3) nor-
epinephrine IV 5-10 g. Using a Bair Hugger Temperature
Management System for air warming, body temperature was
kept N36.5C and measured every 10 minutes at tympanic
level. After surgery, patients were eligible for PACU bypass
once they had reached an Aldrette score 8 and a VAS
b 30 which were also the criteria for discharge from PACU
to the ward [26]. Postoperative analgesia was achieved using
a patient-controlled regional analgesia regimen with a contin-
uous ow rate of 5 mL/h of ropivacaine 0.3% and additional
possible bolus of 5 mL every 20 minutes for the rst 48 hours
postoperatively. Additional standard analgesia consisted of
acetaminophen and metamizol to avoid centrally acting
agents. PaCO2 was measured from arterial blood samples at a
standard temperature of 37C (Radiometer ABL 700; Radiom-
eter, Copenhagen, Denmarl), and SaO2 was determined by di-
rect oximetry (multiwavelength hemoximetry, ABL 800;
Radiometer Medical A/S, Akandevej 21 DK-2700 Bronshoj,
Denmark).
2.1. Statistics
The sample size was determined based on the primary out-
come variable, the incidence of CDEs. Cerebral regional satu-
ration values (rScO2) below a predetermined critical threshold
(20% decrease from baseline or absolute value b55% for
N15 seconds) were used to dene these events. In published
studies of patients undergoing surgery in the BCP under gen-
eral anesthesia, CDEs were observed in 20%-80% of the sub-
jects [10,14,27], and this is considered to be of clinical
relevance [21]. A mean of 71% and a standard deviation
(SD) of 6% were introduced in our calculation according to
Kim et al. [28]. According to published studies using regional
anesthesia in the BCP, we hypothesized that there would
be 50% fewer CDEs under a regional anesthesia regimen
[29-31]. To detect a clinically signicant drop in rScO2
with an risk of .05 and a power of 80%, a sample size of
30 patients was needed. To compensate for possible dropouts,
assuming a dropout rate of 6%, and to further increase the
Anesthesia, cerebral blood flow and beach chair position
power, we included 40 patients. Discrete data were compared
using 2 or Fisher exact test as necessary. Ordinal data and
continuous data that were not normally distributed are present-
ed as median and range. These data were compared between
groups using the Mann-Whitney U test and within groups
using Wilcoxon signed rank test. Normally distributed contin-
uous data are presented as mean SD. These data were com-
pared using the unpaired t test. The rScO2 values at different
time points were compared using an analysis of variance for
repeated measurements. Possible relationships between rScO2
and physiological variables were analyzed using linear re-
gression analysis and quantied using the Pearson correla-
tion test. A P b .05 was considered to be signicant.
Statistical analyses were performed using SigmaStat Version
16 (SPSS Science, Chicago, IL).
3. Results
No patient had to be excluded from the study. The patient
and surgical characteristics are summarized in Tables 1 and 2.
Hemodynamic alterations were seen 5 minutes after beach
chair positioning with a drop in MAP measured at heart
(P = .004) and acoustic meatus (P b .001) level (Fig. 1).
There was no concomitant drop in the Vmax MCA (P =
.212) at the same time point (Fig. 2). No signicant change
in heart rate was registered at any time point. At baseline, bilat-
eral rScO2 values were equivalent between both sides (left and
right) with differences b5%. There was no difference 5 mi-
nutes after BCP with a slight raise of rScO2 values compared
with baseline from 65.68% 4.07% to 67% 4.8% (P =
.136) in either side.
There was a signicant increase in rScO2 values 5 minutes
after regional anesthesia compared with the respective base-
lines (P b .05). The rScO2 values compared with baseline from
the time point of beach chair positioning until discharge to the
PACU remained stable but were lower compared with base-
line without reaching statistical signicance (P = .08). The
Vmax MCA was slightly increased after the BCP without
reaching statistical signicance (P = .212) and returned to
values near baseline values at PACU discharge (P = .20).
There was a strong positive correlation between Vmax MCA
and rScO2 values at all time points (5 minutes after BCP:
r = 0.739, P N .001), but there was no correlation between
Vmax MCA or rScO2 and blood pressure at acoustic meatus
or heart level or its ratio at any time point.
The results of neurocognitive tests were impaired 24 hours
after regional anesthesia only for the TMT A and TMT B
(P b .001) but not for the Grooved Pegboard Test (P = .177)
(Table 2, Fig. 3) without any impairment in the neurological
tests. There was no difference in PaCO2 compared with base-
line at any time point (P in all N .09). No differences compared
with the baselines were noted in arterial oxygen saturation
throughout surgery. The incidence of CDEs was 5%, showing
in all events an rScO2 value decrease of 20% from baseline
459
Table 1 Patient characteristics
Variables Patients (n = 40)
Sex (M/F) 21 (52.5%)/19 (47.5%)
Age (y) 45.88 9.706 (42.77/48.98)
Weight (kg) 77.63 8.705 (74.84/80.41)
Height (cm) 172.73 6.349 (170.69/174.76)
BMI (kg/m2) 25.93 2.39 (25.16/26.69)
ASA physical status I 15 (37.5%)
II 25 (62.5%)
Preoperative hemoglobin (g/dL) 12.28 1.30 (11.87/12.69)
Blood transfusion, n (%) 1 (2.5%)
Coronary artery disease, n (%) 11 (27.5%)
Arterial hypertension, n (%) 14 (35.0%)
Diabetes mellitus, n (%) 3 (7.5%)
Smoker, n (%) 16 (40.0%)
COPD, n (%) 4 (10.0%)
Data expressed as number (%) or mean standard deviation. 95% con-
dence interval (CI) calculated where appropriate. COPD = chronic ob-
structive pulmonary disease.
and with no event showing an absolute rScO2 value b55%
for more than 15 seconds. The duration of the episodes ranged
from 60 seconds to 2.5 minutes. There was no neurobehavioral
impairment compared with baseline in the CDE subgroup
(P N .05 for all tests) (Table 3). Patients with CDEs showed
no difference in the blood pressure drop 5 minutes after beach
chair positioning measured at either the heart (P = .615) or the
acoustical meatus (P = .369) level compared with patients
without CDEs. At the same time point, the changes in the
rScO2 values (P = .310) were not different between the sub-
groups, but the drop in the Vmax MCA (P = .005) was signif-
icant in the CDE subgroup.
There was no difference between CDE and non-CDE pa-
tients concerning the neurobehavioral tests (Table 3), as there
was no alteration of the neurological tests (apart from the
blocked arm), the Glasgow Coma Scale, and the Minimental
State Examination test in both subgroups at baseline and after
24 hours. The BIS values did not correlate with the rScO2
values, and also the subgroup with CDEs did not show any
BIS alterations during these episodes compared with other pa-
tients in the same group. All patients fullled PACU bypass
criteria, and there was no incidence of postoperative nausea
and vomiting.
4. Discussion
This study showed a low incidence (5%) of CDEs in ASA
I-II patients after regional anesthesia in BCP for shoulder sur-
gery, and only 5% of patients had to be treated for hypotension
(SAPheart b 80 mm Hg). A partial impact on neurocognitive
outcome 24 hours after surgery without any neurological con-
sequences artery was identied (Table 2).
The incidence of stroke during shoulder surgery in the BCP
is rare (0.003%-0.07%) [30,32]. However, the 2 reviews by
460 J.A. Aguirre et al.

Table 2 Surgical characteristics and perioperative measurements

Variables Patients (N = 40) P value
Surgery Rotator cuff repair (25)
Shoulder arthroplasty (10)
Latarjet (1)
Osteosynthesis (4)
Surgery time (min) 76.38 29.13 (67.06/85.69)
Crystalloid infusion (mL) 525.00 219.27 (454.88/595.12)
Colloid infusion (mL) 525.00 110.36 (489.70/560.30)
PONV 0
Treatment for SAPhe b 80 mm Hg 2 (5.0%)
Treatment for SAPhe N 100 mm Hg 1 (2.5%)
CDEs with rScO2 drops 20% of BL 2 (5.0%)
CDEs with absolute rScO2 value b55% for N15 s 0 (0.0%)
PACU bypass 40 (100.0%)
MAPhe BL/ 94.35 14.67 (89.66/99.04)/ .004
5 min after BCP (mm Hg) 84.60 12.19 (80.70/88.50)
MAPam BL/ 94.28 14.18 (89.74/98.81)/
5 min after BCP (mm Hg) 63.90 15.08 (59.08/68.72) b.001

rScO2 right BL/ 65.68 4.07 (64.37/66.98)/ .136

5 min after BCP (%) 67.00 4.75 (65.48/68.52)
rScO2 left BL/ 65.10 4.27 (64.00/67.01)/ .150
5 min after BCP (%) 67.10 4.80 (66.12/68.85)
Vmax MCA BL/ 58.43 3.84 (57.20/59.65)/ .212
5 min after BC (cm/s) 59.73 4.48 (58.29/61.16)
TMT A BL/ 31.30 1.27 (30.90/31.70)/
24 h after surgery (s) 32.68 1.18 (32.30/33.05) b.001
TMT B BL/ 78.28 3.35 (77.20/79.35)/ b.001
24 h after surgery (s) 81.00 3.412 (79.91/82.09)
Grooved Pegboard BL/ 60.25 5.17 (58.60/61.90)/ .177
24 h after surgery (s) 61.78 5.40 (60.05/63.50)
paCO2 BL/ 4.59 0.24 (4.52/4.67)/ .090
5 min after BCP (kPa) 4.69 0.21 (4.62/4.76)
Pulse rate BL/ 67.20 14.62 (62.52/71.88)/ .228
5 min after BCP (b/min) 63.80 13.07 (59.62/67.98)
Data expressed as number (%) or mean standard deviation. 95% CI calculated where appropriate.
Am = acoustical meatus; BL = baseline; he = heart level; PONV = postoperative nausea and vomiting; SAP = systolic arterial pressure.
Statistically signicant compared with baseline.

YaDeau et al. [30] and Rohrbaugh et al. [31] include N95% of
shoulder surgery performed only under regional anesthesia
which has been shown to offer stable hemodynamic conditions
[21,27]. Moreover, the survey by Friedman et al. [32] which
calculated an estimated incidence of intraoperative cerebrovas-
cular events of 0.00382%0.00461% had only a feedback of
32%, suggesting an unknown number as demonstrated by
many case reports [13]. Therefore, noninvasive monitoring
and surrogate parameters like delirium or postoperative cogni-
tive dysfunction (POCD) for intracerebral events have been in-
troduced in clinical practice [21,33]. Different studies have
prospectively assessed the impact of regional anesthesia on
the rScO2 in the BCP. In the YaDeau et al. [29] study, 1 risk
factor for cerebrovascular disease was present in 17% and 2
risk factors in 12% of the patients. Hypotension occurred in
76% of observations (99% of patients) with a mean duration
of 4.26 seconds, but cerebral desaturation was seen in only
0.77% of cases (10% of patients), with a mean duration of
426 seconds. The combination of both hypotension and cere-
bral desaturation occurred in 10% of patients. By unadjusted
modeling, hypotension was associated with cerebral desatura-
tion (OR = 3.21; P = .02). If time-trend adjusted, cerebral
desaturation was associated with time from baseline but not
with hypotension (P = .14). However, when adjusted for base-
line factors, the statistical analysis showed a still positive but
nonsignicant association with hypotension (P = .34) and a
signicant association with the presence of risk factors for ce-
rebrovascular disease (P = .01). These results are in accor-
dance with our ndings: 5% of patients needed treatment for
hypotension according to our criteria with a higher incidence
of cardiovascular risk factors (Table 3) and a maximum dura-
tion of CDE of 2.5 minutes. YaDeau et al. [29] concluded that
Anesthesia, cerebral blood flow and beach chair position 461

*
*
*

Fig. 1 Perioperative blood pressure course. Blood pressure values at different time points at heart level (blue line) and acoustic meatus level (red
line). Data shown as mean values of all patients. *Statistically signicant compared with baseline. BC = beach chair position; cut = surgery start;
RA = general anesthesia.

hypotension was frequent but cerebral desaturation was rare
among ASA I-II patients. In our work, only 5% had CDEs,
which is in accordance with the results of Koh et al. [27]. How-
ever, in the YaDeau et al. study, there were no standardized
controlled hypotension protocol and no measured blood pres-
sure at the brain level.
Two studies compared a regional anesthesia (interscalene
block) to a general anesthesia regimen, assessing their effect
on rScO2. Koh et al. [27] compared general anesthesia with
sevourane to regional anesthesia for shoulder surgery in
BCP. They found a higher incidence of CDEs in the general
anesthesia group (56.7%) compared with regional anesthesia
(0%) and a greater hemodynamic instability in the general an-
esthesia group with a decrease in MAP N20% in 73.3% and
10% of cases after general and regional anesthesia, respective-
ly. Aguirre et al. [21] compared in a prospective, assessor-
blinded cohort study with a standardized controlled hypoten-
sion protocol total intravenous general anesthesia to a regional
anesthesia regimen for shoulder surgery in BCP. They found a
higher incidence of CDEs in the general anesthesia group
(71.1%) compared with regional anesthesia (2.2%) and a
greater hemodynamic instability in the general anesthesia
group. Risk factors and CDEs could be correlated for coronary
artery disease (0.367, P = .013) and arterial hypertension
(0.414, P = .006) with odds ratio of 2.0 and 0.12, respectively.
However, Koh et al. had no controlled hypotension protocol
and used phenylephrine for hypotension and CDE treatment,
a drug which has been shown to interfere with CBF [18,34].
Aguirre et al. [21] found no correlation between POCD and
CDE in the regional anesthesia group, which is in accordance
with our ndings. Three studies have assessed patient risk
factors to CDEs. Salazar et al. described a body mass in-
dex (BMI) 34 kg/m2 as a risk factor for CDE (OR = 12.4)
after general anesthesia with sevourane for shoulder surgery
in BCP [33]. In a second study, they again described
BMI 34 kg/m2 as a risk factor for CDE (OR = 12.4) but
showed that duration of surgery was not a risk factor [35].
Animal studies have shown that neurologic impairment
is related to the severity and duration of desaturation events
causing cerebral ischemia, which remains undened when
anesthesiologists are not blinded for cerebral oxygenation
data [36,37].
Our results emphasize the importance of a proper neuro-
cognitive assessment. Gross neurologic tests, Minimental

Fig. 2 Perioperative regional cerebral saturation and CBF course. CBF expressed as maximal blood ow (mean) of the middle cerebral artery
(red line) and regional cerebral oxygen saturation (blue line) at different time points during regional anesthesia. Data shown as mean values of all
patients.
462
Fig. 3 Preoperative and postoperative neurobehavioral tests.
Neurobehavioral test results expressed as seconds (mean values of
all patients) needed to perform the tests at baseline (blue bars) and
24 hours after surgery (red bars) and the difference (green bars).
*Statistically signicant compared with baseline.
State Examination test, and Glasgow Coma Scale, which are
often used as surrogate parameters, are not sensitive enough
as shown in our collective to assess POCD [38,39]. However,
it is interesting that, in the CDE group, no POCD was regis-
tered despite a reduction in the maximal blood ow (mean)
of the middle cerebral artery. This could be due to the short du-
ration of the events or simply due to the fact that 2 events are
not enough to create POCD. However, despite a low incidence
of hypotension, a stable CBF of the middle cerebral artery, and
no impairment of rScO2 values during BCP (Table 2), we
found an impairment in the TMT A and B 24 hours after sur-
gery (P b .001). However, no differences in the Grooved Peg-
board Test (P = .177) were observed. There could be different
explanations for these ndings. As shown by Murphy et al.
[14], comparing patients in the BCP with patients in the LDP
for shoulder surgery, the MAP decrease in both groups was
signicant compared with the respective baselines but not be-
tween the groups. However, the impairment of rScO2 in both
groups was signicant compared not only with the respective
baselines but also between the groups. Obviously, the drop
in blood pressure or its magnitude is not the only factor
inuencing rScO2 impairment. This is in accordance with the
data published by YaDeau et al. [29] who observed the occur-
rence of hypotension in 99% of patients but registered a CDE
only in 10%. Recently, Laam et al. [40] made a similar nd-
ing comparing BCP with LDP for shoulder surgery under gen-
eral anesthesia: patients in BCP had more often a loss of
cerebral autoregulation and lower rScO2 values. However, no
differences in cognitive outcome and in serum biomarker
levels (S100, neuron-specic enolase, and glial brillary
acidic protein) were found between the 2 groups [40]. We
found no correlation between Vmax MCA or rScO2 and blood
J.A. Aguirre et al.
pressure at the acoustic meatus and heart level or its ratio, a ra-
tio describing the dependency of systemic blood pressure on
the degree of elevation in the BCP recently suggested by Trip-
let et al. [41], at any time point, further underlying the fact that
systemic blood pressure is only 1 variable inuencing CBF
and oxygenation.
We found a strong positive correlation between Vmax
MCA and rScO2 values at all time points (5 minutes after
BCP: r = 0.739, P N .001). The studies of McCulloch et al.
[16] and Soeding et al. [17] demonstrated by Doppler
measurements of CBF the importance of an adequate MAP
to maintain CBF autoregulation. However, many blood
pressureindependent intracerebral mechanisms leading to
shunts of blood ow have been described, which additionally
could lead to areas of relative hypoperfusion [42]. These
complex, systemic blood pressure, regional regulatory mecha-
nisms could have negatively inuenced some of the neurobe-
havioral test results in our study.
Aguirre et al. associated cardiovascular risk factors as coro-
nary artery disease and arterial hypertension with an odds ratio
of 2.0 and 0.12, respectively, to CDEs [21]. This nding was
recently highlighted by Mori et al. in a study on 91 patients
in BCP for shoulder surgery under general anesthesia includ-
ing 63 patients with cardiac risk factors [43]. Maintaining the
MAP above 60 mm Hg at heart level led to stable rScO2 values
even in the cardiac risk patients, suggesting that cerebral oxy-
genation monitoring in this vulnerable risk population could
prevent CDEs. Our patients with CDEs were anemic compared
with the non-CDE population, and the only blood transfusion was
used for one of those patients (Table 3). The impact of anemia
on cerebral oxygenation remains unclear even in signicant
anemia as shown by Cem et al. on coronary arterial bypass pa-
tients during cardiopulmonary bypass where, even for anemia
of 7 g/dL, the changes in rScO2 were within acceptable limits
[44]. They concluded that cerebral oxygenation monitoring
can only assist in decision making related to blood transfusion.
Limitations of our study include the lack of a control group, as
we rst wanted to assess the changes in CBF and their possible
correlations with cerebral saturation in the regional anesthesia
population. Moreover, the neurocognitive tests were performed
only once 24 hours after surgery. Perhaps earlier and later tests
would have shown possible early impairment or late recovery
of neurocognitive function. In addition, the CBF was only
assessed by the middle cerebral artery blood ow. The vertebral
artery might also be of clinical importance considering the
xed head position and the movements of the torso during
intraoperative traction. However, these measurements are
technically difcult to achieve during surgery and beyond
the possibilities of intraoperative anesthesia monitoring.
To conclude, we can state that ASA I-II patients undergo-
ing shoulder surgery in regional anesthesia maintain stable
CBF and cerebral oxygenation. Moreover, the incidence of hy-
potension requiring treatment and CDEs is rare in this group.
However, a partial impairment on neurobehavioral outcome
without neurologic symptoms was identied. Further studies
analyzing the effect of regional anesthesia, comparing general
Anesthesia, cerebral blood flow and beach chair position 463

Table 3 CDEs subgroup vs non-CDEs subgroup

Variables CDE (n = 2) 95% CI Non-CDE (n = 38) 95% CI P (P)
Sex (M/F), n (%) M: 1 (50.0%) M: 20 (52.6%) 1.000
F: 1 (50.0%) F: 18 (47.4%)
Age (y) 55.50 0.71 49.15/61.85 45.37 9.70 42.18/48.56 .108
BMI (kg/m2) 28.76 2.35 7.61/49.92 25.78 2.33 25.01/26.54 .092
ASA physical status, n (%) I: 0 (0%) I: 15 (39.5%) .635
II: 2 (100%) II: 23 (60.5%)
Preoperative hemoglobin (g/dL) 10.75 0.92 2.49/19.01 12.36 1.27 11.94/12.78 .126
Blood transfusion, n (%) 1 (50.0%) 0 (0.0%) .073
Coronary artery disease, n (%) 2 (100%) 9 (23.7%) .199
CDEs with rScO2 drops 20% of BL 2 (100%) 0 (0%) b.001
Arterial hypertension, n (%) 2 (100%) 12 (31.6%) .274
Diabetes mellitus, n (%) 0 (0.0%) 3 (7.9%) 1.000
Smoker, n (%) 0 (0.0%) 16 (42.1%) 1.000
COPD, n (%) 0 (0.0%) 4 (10.5%) 1.000
MAPhe BL/ 92.50 10.61/ 2.80/187.80/ 94.45 14.95/ 89.53/99.36 .615/
5 min after BCP (mm Hg) 88.00 9.90 0.94/176.94 84.42 12.39 80.35/88.49 (.667)
MAPam BL/ 94.00 11.31/ 7.65/195.65/ 94.29 14.43/ 89.55/99.03 .369/
5 min after BCP (mm Hg) 57.00 18.39 108.18/222.18 64.26 15.09 59.30/69.22 (.333)
rScO2 right BL/ 64.50 0.71/ 58.15/70.85/ 65.74 4.16/ 64.37/67.11 .338/
5 min after BCP (%) 64.52 0.72 58.12/70.87 67.11 4.85 65.51/68.70 (.667)
rScO2 left BL/5 min after BCP (%) 64.40 0.70/ 58.10/70.77/ 65.20 4.10/ 64.00/66.95 .310/
64.20 0.69 58.11/70.20 66.10 4.65 65.00/68.01 (.710)
Vmax MCA BL/5 min after BCP (cm/s) 56.50 0.71/ 50.15/62.85/ 58.55 3.90/ 57.27/59.84 .005 /
54.50 0.71 48.15/60.85 60.03 4.39 58.58/61.47 (.333)
TMT A BL/ 32.00 1.41/ 19.29/44.71/ 31.26 1.27/ 30.85/31.68 .785/
24 h after surgery (s) 33.00 1.41 20.29/45.71 32.66 1.19 32.27/33.05 (.667)
TMT B BL/ 76.50 0.71/ 70.15/82.85/ 78.37 3.41/ 77.25/79.49 .400/
24 h after surgery (s) 77.50 3.54 45.73/109.27 81.18 3.35 80.08/82.29 (1.0)
Grooved Pegboard BL/24 h after surgery (s) 63.00 2.83/ 37.59/88.41/ 60.11 5.24/ 58.38/61.83 .615/
65.00 4.24 26.88/103.12 61.61 5.45 59.82/63.40 (.667)
Comparison between desaturation event and nondesaturation event patients. Data expressed as number (%) or mean standard deviation. 95% CI calculated
where appropriate. P = compared with non-CDE group; P = compared with the respective baseline.
Statistically signicant.

with regional anesthesia in an ASA I-IV population, and ana-
lyzing the impact on neurobehavioral changes are warranted.
Acknowledgments
The authors would like to acknowledge the help provided
by Gina Votta-Velis, MD, PhD, (University of Illinois at
Chicago) for her precious help in manuscript preparation and
language/grammar corrections.
References
[1] Li X, Eichinger JK, Hartshorn T, Zhou H, Matzkin EG, Warner JPA.
Comparison of the lateral decubitus and beach-chair positions for shoul-
der surgery: advantages and complications. J Am Acad Orthop Surg
2015;23:18-28.
[2] Dippmann C, Winge S, Nielsen HB. Severe cerebral desaturation during
shoulder arthroscopy in the beach-chair position. Arthroscopy 2010;26(9
Suppl):S148-50.
[3] Boisseau N, Rabarijaona H, Grimaud D, Raucoules-Aime M. Tapia's
syndrome following shoulder surgery. Br J Anaesth 2002;88:869-70.
[4] Bhatti MT, Enneking FK. Visual loss and ophthalmoplegia after shoulder
surgery. Anesth Analg 2003;96:899-902.
[5] Pohl A, Cullen DJ. Cerebral ischemia during shoulder surgery in the up-
right position: a case series. J Clin Anesth 2005;17:463-9.
[6] Verbrugge SJ, Klimek M, Klein J. A cerebral watershed infarction after
general anaesthesia in a patient with increased anti-cardiolipin antibody
level. Anaesthesist 2004;53:341-6.
[7] Pant S, Bokor DJ, Low AK. Cerebral oxygenation using near-infrared
spectroscopy in the beach-chair position during shoulder arthroscopy un-
der general anesthesia. Arthroscopy 2014;30:1520-7.
[8] Nielsen HB. Systematic review of near-infrared spectroscopy determined
cerebral oxygenation during non-cardiac surgery. Front Physiol 2014;5:
93.
[9] Toole JF. Effects of change of head, limb and body position on cephalic
circulation. N Engl J Med 1968;279:307-11.
[10] Moerman AT, De Hert SG, Jacobs TF, De Wilde LF, Wouters PF. Cere-
bral oxygen desaturation during beach chair position. Eur J Anaesthesiol
2012;29:82-7.
464
[11] Manninen H, Makinen K, Vanninen R, Ronkainen A, Tulla H. How of-
ten does an incomplete circle of Willis predispose to cerebral ischemia
during closure of carotid artery? Postmortem and clinical imaging stud-
ies. Acta Neurochir 2009;151:1099-105.
[12] van Raamt AF, Mali WP, van Laar PJ, van der Graaf Y. The fetal variant
of the circle of Willis and its inuence on the cerebral collateral circula-
tion. Cerebrovasc Dis 2006;22:217-24.
[13] Drummond JC, Lee RR, JP H Jr. Focal cerebral ischemia after surgery in
the "beach chair" position: the role of a congenital variation of circle of
Willis anatomy. Anesth Analg 2012;114:1301-3.
[14] Murphy GS, Szokol JW, Marymont JH, Greenberg SB, Avram MJ,
Vender JS, et al. Cerebral oxygen desaturation events assessed by near-
infrared spectroscopy during shoulder arthroscopy in the beach chair
and lateral decubitus positions. Anesth Analg 2010;111:496-505.
[15] Fischer GW, Torrillo TM, Weiner MM, Rosenblatt MA. The use of ce-
rebral oximetry as a monitor of the adequacy of cerebral perfusion in a
patient undergoing shoulder surgery in the beach chair position. Pain
Pract 2009;9:304-7.
[16] McCulloch TJ, Liyanagama K, Petchell J. Relative hypotension in the
beach-chair position: effects on middle cerebral artery blood velocity.
Anaesth Intensive Care 2010;38:486-91.
[17] Soeding PF, Wang J, Hoy G, Jarman P, Phillips H, Marks P, et al. The
effect of the sitting upright or 'beach chair' position on cerebral blood
ow during anaesthesia for shoulder surgery. Anaesth Intensive Care
2011;39:440-8.
[18] Soeding PF, Hoy S, Hoy G, Evans M, Royse CF. Effect of phenylephrine
on the haemodynamic state and cerebral oxygen saturation during anaes-
thesia in the upright position. Br J Anaesth 2013;111:229-34.
[19] Tobias JD. Strategies for minimizing blood loss in orthopedic surgery.
Semin Hematol 2004;41(Suppl 1):145-56.
[20] Murkin JM, Newman SP, Stump DA, Blumenthal JA. Statement of con-
sensus on assessment of neurobehavioral outcomes after cardiac surgery.
Ann Thorac Surg 1995;59:1289-95.
[21] Aguirre J, Borgeat A, Trachsel T, Cobo Del Prado I, De Andres J, Buhler
P. Cerebral oxygenation in patients undergoing shoulder surgery in beach
chair position: comparing general to regional anesthesia and the impact
on neurobehavioral outcome. Rev Esp Anestesiol Reanim 2014;61:64-72.
[22] Fassoulaki A, Kaliontzi H, Petropoulos G, Tsaroucha A. The effect of
desurane and sevourane on cerebral oximetry under steady-state con-
ditions. Anesth Analg 2006;102:1830-5.
[23] Casati A, Fanelli G, Pietropaoli P, Proietti R, Tufano R, Danelli G, et al.
Continuous monitoring of cerebral oxygen saturation in elderly patients
undergoing major abdominal surgery minimizes brain exposure to poten-
tial hypoxia. Anesth Analg 2005;101:740-7.
[24] Borgeat A, Dullenkopf A, Ekatodramis G, Nagy L. Evaluation of the lat-
eral modied approach for continuous interscalene block after shoulder
surgery. Anesthesiology 2003;99:436-42.
[25] Ghisi D, Fanelli A, Tosi M, Nuzzi M, Fanelli G. Monitored anesthesia
care. Minerva Anestesiol 2005;71:533-8.
[26] Aldrete JA. The post-anesthesia recovery score revisited. J Clin Anesth
1995;7:89-91.
[27] Koh JL, Levin SD, Chehab EL, Murphy GS. Cerebral oxygenation in the
beach chair position: a prospective study on the effect of general anesthe-
sia compared with regional anesthesia and sedation. J Shoulder Elbow
Surg 2013;22:1325-31.
[28] Kim MB, Ward DS, Cartwright CR, Kolano J, Chlebowski S, Henson
LC. Estimation of jugular venous O2 saturation from cerebral oximetry
J.A. Aguirre et al.
or arterial O2 saturation during isocapnic hypoxia. J Clin Monit Comput
2000;16:191-9.
[29] Yadeau JT, Liu SS, Bang H, Shaw PM, Wilfred SE, Shetty T, et al. Ce-
rebral oximetry desaturation during shoulder surgery performed in a sit-
ting position under regional anesthesia. Can J Anaesth 2011;58:986-92.
[30] Yadeau JT, Casciano M, Liu SS, Edmonds CR, Gordon M, Stanton J,
et al. Stroke, regional anesthesia in the sitting position, and hypotension:
a review of 4169 ambulatory surgery patients. Reg Anesth Pain Med
2011;36:430-5.
[31] Rohrbaugh M, Kentor ML, Orebaugh SL, Williams B. Outcomes of
shoulder surgery in the sitting position with interscalene nerve block: a
single-center series. Reg Anesth Pain Med 2013;38:28-33.
[32] Friedman DJ, Parnes NZ, Zimmer Z, Higgins LD, Warner JJ. Prevalence
of cerebrovascular events during shoulder surgery and association with
patient position. Orthopedics 2009.
[33] Salazar D, Sears BW, Aghdasi B, Only A, Francois A, Tonino P, et al.
Cerebral desaturation events during shoulder arthroscopy in the beach
chair position: patient risk factors and neurocognitive effects. J Shoulder
Elbow Surg 2013;22:1228-35.
[34] Meng L, Cannesson M, Alexander BS, Yu Z, Kain ZN, Cerussi AE, et al.
Effect of phenylephrine and ephedrine bolus treatment on cerebral oxy-
genation in anaesthetized patients. Br J Anaesth 2011;107:209-17.
[35] Salazar D, Sears BW, Andre J, Tonino P, Marra G. Cerebral desaturation
during shoulder arthroscopy: a prospective observational study. Clin
Orthop Relat Res 2013;471:4027-34.
[36] Kurth CD, McCann JC, Wu J, Miles L, Loepke AW. Cerebral oxygen
saturation-time threshold for hypoxicischemic injury in piglets. Anesth
Analg 2009;108:1268-77.
[37] Hou X, Ding H, Teng Y, Zhou C, Tang X, Li S, et al. Research on the
relationship between brain anoxia at different regional oxygen satura-
tions and brain damage using near-infrared spectroscopy. Physiol Meas
2007;28:1251-65.
[38] Highton D, Elwell C, Smith M. Noninvasive cerebral oximetry: is there
light at the end of the tunnel? Curr Opin Anaesthesiol 2010;23:576-81.
[39] Kocaoglu B, Ozgen SU, Toraman F, Karahan M, Guven O. Foreseeing
the danger in the beach chair position: are standard measurement
methods reliable? Knee Surg Sports Traumatol Arthrosc 2015;23:
2639-44.
[40] Laam A, Joshi B, Brady K, Yenokyan G, Brown C, Everett A, et al.
Shoulder surgery in the beach chair position is associated with dimin-
ished cerebral autoregulation but no differences in postoperative cogni-
tion or brain injury biomarker levels compared with supine positioning:
the anesthesia patient safety foundation beach chair study. Anesth Analg
2015;120:176-85.
[41] Triplet JJ, Lonetta CM, Everding NG, Moor MA, Levy JC. Association
between temporal mean arterial pressure and brachial noninvasive blood
pressure during shoulder surgery in the beach chair position during gen-
eral anesthesia. J Shoulder Elbow Surg 2015;24:127-32.
[42] Willie CK, Tzeng YC, Fisher JA, Ainslie PN. Integrative regulation of
human brain blood ow. J Physiol 2014;592:841-59.
[43] Mori Y, Yamada M, Akahori T, Hatakeyama N, Yamazaki M, Fujiwara
Y, et al. Cerebral oxygenation in the beach chair position before and dur-
ing general anesthesia in patients with and without cardiovascular risk
factors. J Clin Anesth 2015;27:457-62.
[44] Cem A, Serpil UO, Fevzi T, Murat O, Umit G, Esin E, et al. Efcacy of
near-infrared spectrometry for monitoring the cerebral effects of severe
dilutional anemia. Heart Surg Forum 2014;17:E154-9.