British Journal of Anaesthesia, 118 (4): 551–62 (2017)

doi: 10.1093/bja/aex008
Clinical Practice

Association of intraoperative cerebral and muscular

tissue oxygen saturation with postoperative
complications and length of hospital stay after major
spine surgery: an observational study
L. Meng1,2,*, J. Xiao3, K. Gudelunas2,6, Z. Yu4, Z. Zhong5 and X. Hu6
1
Department of Anesthesiology, Yale University School of Medicine, New Haven, CT 06520, USA, 2Department
of Anesthesia and Perioperative Care, University of California San Francisco, San Francisco, CA 94143, USA,
3
Department of Bioengineering, University of California Berkeley, Berkeley, CA 94709, USA, 4Department of
Statistics, University of California Irvine, Irvine, CA 92697, USA, 5Department of Spine Surgery, Nanfang
Hospital, Southern Medical University, Guangzhou, Guangdong 510515, China and 6Physiological Nursing,
University of California San Francisco, San Francisco, CA 94143, USA

*Corresponding author. E-mail: lingzhong.meng@yale.edu

Abstract
Background. Compromised tissue oxygenation is one of the root causes of dysfunction of various organs and postoperative
complications. Oxygenation of different tissue beds may follow different patterns of change during physiological
derangement.
Methods. Patients undergoing elective major posterior spine surgery participated in this prospective observational study.
Cerebral tissue oxygen saturation (SctO2) was monitored on the upper forehead and muscular tissue oxygen saturation
(SmtO2) on the lower leg. The associations of various oxygenation indices with postoperative composite complications and
length of hospital stay (LOH) were investigated.
Results. The number of composite complications per patient was 3 (2) while the LOH was 6 (3) days (n ¼ 102). Multiple SmtO2
indices (maximum, minimum, mean, median, and area under curve (AUC)) were associated with composite complications
(univariate analysis, P < 0.05). No SctO2 indices were associated with complications. Multiple SmtO2 indices (maximum,
mean, median, and AUC) showed differences (P < 0.05) between patients with composite complications 3 and >3, respect-
ively. SmtO2 standard deviation, AUC, and AUC weighted, and SctO2 standard deviation, were associated with LOH (univari-
ate analysis, P < 0.05). Two SmtO2 indices (AUC and AUC weighted), showed differences (P < 0.05) between the patients with
an LOH 6 and >6 days, respectively. SmtO2, but not SctO2, indices improved the adjusted R2 for composite complications
(þ54.0%, P ¼ 0.0001) and LOH (þ19.0%, P ¼ 0.02) based on multiple linear models.
Conclusions. Muscular tissue oxygenation has a stronger association with postoperative complications and length of hos-
pital stay than cerebral tissue oxygenation after major spine surgery.

Key words: postoperative complications

Editorial decision: December 2, 2016; Accepted: Month 0, 0000

C The Author 2017. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved.
V
For Permissions, please email: journals.permissions@oup.com

551
 552 | Meng et al.
Editors key points
• Inadequate tissue oxygenation is a potential cause of
postoperative complications.
• Near-infrared spectroscopy can be used to assess the bal-
ance at tissue level between oxygen supply and demand.
• The authors studied associations between muscle and
brain tissue oxygenation, and outcome after major spi-
nal surgery.
• Muscle tissue oxygenation indices were weakly associ-
ated with postoperative complications and hospital
length of stay.
The constant utilization of oxygen by a live tissue bed requires
continuous and sufficient delivery of oxygen to meet the meta-
bolic demand. Tissue oxygenation describes the balance be-
tween the rates of oxygen consumption and supply in a specific
tissue bed. Compromised oxygenation is the result of mis-
matching between tissue oxygen consumption and supply,
which can cause organ dysfunction and perioperative morbidity
and mortality. Real-time tissue oxygenation monitoring during
high-risk surgery facilitates timely detection and rectification of
a compromised tissue oxygenation, before the emergence of an
adverse consequence. However, the first question is the oxygen-
ation of which tissue bed, as there are many, is worthy of being
monitoring during major surgeries.
The advent of tissue oximetry based on near-infrared spectros-
copy enables non-invasive, continuous and bedside monitoring of
the haemoglobin oxygen saturation in mixed arteriolar, capillary
and venular blood in the tissue bed being probed.1 2 The modern
tissue oximeter can monitor distinctive tissue beds depending on
the location of the probe being applied. Cerebral tissue oxygen sat-
uration (SctO2) is measured when the probe is placed on the upper
forehead; while muscular tissue oxygen saturation (SmtO2) is
measured when the probe is positioned peripherally with a
muscle mass underneath.3 4 Different tissue beds have different
schemes of oxygen consumption and supply, and distinctive com-
pensating ability for ischaemia and hypoxia. It is possible that the
same insult exerts distinctive impacts on oxygenation of different
tissue beds. The tissue bed that provides the most valuable infor-
mation in relation to patient outcome should be prioritized for
oxygenation monitoring. At present, how the oxygenation of dif-
ferent tissue beds (e.g. SctO2 vs SmtO2), is associated with patient
outcome after major surgeries has not been reported.
The aim of this prospective observational study is to com-
pare the associations of SctO2 vs SmtO2 with postoperative com-
plications and length of hospital stay (LOH) in patients
undergoing major spine surgery. We chose spine surgery owing
to the fact that major spine surgery involving osteotomy and
multi-segment fusion can incur large blood loss and require
intraoperative transfusion, which may adversely affect tissue
oxygenation and necessitate a prolonged hospitalization. The
hypothesis was that SmtO2 has a better association with patient
outcome because, when SmtO2 is deranged by various intrao-
perative insults, SctO2 may remain stable owing to the robust
autoregulation of cerebral blood flow.
Methods
This is a prospective observational single-cohort study. The
study was approved by the Institutional Review Board for
Clinical Investigations at the University of California San
Francisco, San Francisco, California, USA. Both verbal and writ-
ten informed consents were obtained from the patient before
surgery.
Patients
Patients who were undergoing elective posterior lumbar or
thoracolumbar spine surgery in a prone position and with the
potential for osteotomy and multi-segment fusion were re-
cruited. Our goal was to recruit 100 patients in this study, be-
cause the previous studies investigating the association
between SctO2 and postoperative outcome had about 100 pa-
tients in whom either monitoring only or monitor-guided care
was implemented.5–8 The exclusion criteria were: 1) patient re-
fusal; 2) emergent surgery; 3) age <18 yr old; 4) ASA physical sta-
tus score > III; 5) skin condition precluding the application of the
oximetry probe.
Anaesthesia
The administration of premedications such as midazolam, fen-
tanyl, gabapentin, and oxycodone was at the discretion of the
anaesthetist. In the operating room, following the establish-
ment of standard monitoring including ECG, pulse oximetry,
and non-invasive bp, anaesthesia was induced using i.v. lido-
caine, fentanyl and propofol and a tracheal tube was placed
with the facilitation of either succinylcholine or rocuronium. A
radial arterial catheter was placed for continuous bp monitoring
after tracheal intubation. All patients received neurophysio-
logical monitoring including motor and somatosensory-evoked
potentials and electromusculography. Anaesthesia was typic-
ally maintained using i.v. infusion of propofol, fentanyl, lido-
caine and ketamine with or without a volatile anaesthetic agent
at a low minimum alveolar concentration. Bp was typically
maintained using phenylephrine infusion. Tranexamic acid was
used in some patients when large amount of blood loss was
anticipated. Cell salvaging was available in all patients.
Tissue oxygenation monitoring
Tissue oxygenation was monitored using a tissue oximeter
based on near-infrared spectroscopy (FORE-SIGHT Elite,
CASMED, Inc., Branford, Connecticut, USA). The oximeter had
four cables with each cable connected to an adhesive probe.
Two probes were placed on the upper forehead to monitor the
left and right frontal lobe SctO2, respectively. Another two
probes were placed on the lower legs, four fingers below the tib-
ial tuberosity and two fingers lateral to the anterior edge of the
tibial shaft, to monitor the SmtO2 of the left and right lower leg,
respectively. All four probes were placed on the patient after tra-
cheal intubation. The monitoring and data recording were
started after the patient was turned to prone position and
stopped at the time when the procedure was done. The monitor
was positioned at the back of anaesthesia machine and blinded
to the anaesthesia team.
The tissue oximeter generated a new data point every two s
that was continuously recorded by a research computer. The
median data of SctO2 and SmtO2 within each min were used for
tissue oxygenation indices derivation. Seven indices of tissue
oxygenation were derived that include the maximum, min-
imum, mean, median, standard deviation, area under curve
(AUC), and AUC weighted. AUC was the sum of the differences

between the mean and the data point of every min that was less
than the mean. AUC weighted was the AUC divided by the time
of recording. For each index, SctO2 and SmtO2 were treated
independently.
Clinical covariates
Patient characteristics including age, sex, weight, height, and
ASA physical status score were collected. Past medical history
including diagnosis of hypertension, diabetes mellitus, periph-
eral vascular disease, cerebrovascular disease, coronary artery
disease, chronic lung disease, and chronic kidney disease
were recorded. Surgical information that were recorded were
history of previous spine surgery, fusion or not, number of seg-
ments, and surgical time. The anaesthetic agents, non-
anaesthetic drugs, the input and output during surgery were
also recorded.
Postoperative outcomes
The primary outcome measures were postoperative composite
complications and LOH. All complications were identified based
on chart review of the electronic medical records at the
University of California San Francisco. Information of intensive
care unit (ICU) admission and length of ICU stay were also
recorded.
Statistical analysis
As the outcome variables (composite complications and LOH)
are count data and skewed, we used log transformation to re-
duce the skewness. It is well known that linear regression with
log-transformed dependent variables is an alternative to
Poisson regression and they often give similar results. We chose
linear regression with log-transformation as it can provide R2,
which provides the percentage of variance in the dependent
variable that can be explained by one or a set of independent
variables. Constant 1 was added before taking log for the vari-
able of composite complications because its minimum is zero.
The association between each outcome variable and each clin-
ical covariate or oxygenation index was first tested using linear
regression with Student’s t-test for continuous variables and
two-sample Student’s t-test for binary variables.
The next analysis examined the contribution of oxygenation
indices, in addition to clinical covariates, to each outcome
measure using multiple linear regression. The selection of clin-
ical covariates was based on both clinical significance (i.e. the
results of univariate analysis and clinical experience) and the
avoidance of collinearity. To avoid collinearity (i.e. the presence
of highly correlated independent variables), for each cluster of
highly correlated set of variables, we only used one variable to
represent the set. For example, as the amount of packed red
blood cells, the amount of fresh frozen plasma, the amount of
cell saver blood, and the estimated blood loss were highly corre-
lated, we only included the packed red blood cells in the mul-
tiple regressions. The clinical covariates that were included in
analysis were age, ASA score, packed red blood cells, usage of
phenylephrine, vasopressin, and epinephrine, surgical time,
number of spinal segments being fused, and history of chronic
kidney disease. The contribution of oxygenation indices was
quantified by comparing the multiple R2 and the adjusted R2 be-
tween two multiple regression models. The first model was ob-
tained by including the selected clinical covariates only. In this
Tissue oxygenation and postoperative outcome | 553
step, oxygenation indices were excluded. To obtain the second
model, we started from the first model and examined the add-
itional contribution of SctO2 and SmtO2 to the outcome. If more
than two indices were selected in this step, only the top two
most significant oxygenation indices were included. The mul-
tiple R2 and the adjusted R2, with oxygenation indices excluded
and included, respectively, were compared to evaluate the im-
provement of the model fitting by oxygenation indices. SctO2
and SmtO2 were treated independently in multiple linear
models.
Based on the averages of postoperative composite complica-
tions (3) and LOH (6 days), the patients were stratified into
two groups using the criteria of having a total of  3 or > 3 post-
operative complications or being hospitalized  6 or > 6 days, re-
spectively. We tested if the patient characteristics, clinical
covariates and oxygenation indices were significantly different
between the binary strata. The binary strata based on composite
complications or LOH were examined separately. The means of
the continuous variables were compared using the Student’s t-
test. The proportions of the binary variables were compared
using the v2 test.
P value < 0.05 was regarded as statistically significant.
Results
A total of 108 patients participated in the study. Six patients
were excluded from the analysis because of the arm instead of
the leg location of SmtO2 monitoring (n ¼ 5) and cervical, rather
than lumbar or thoracolumbar, spinal fusion (n ¼ 1). The patient
characteristics, past medical history, surgical information, tis-
sue oxygenation indices, anaesthetic agents, non-anaesthetic
drugs, intraoperative input and output, and outcomes of the re-
maining 102 patients are summarized in Table 1.
None of SctO2 indices were significantly associated with
postoperative composite complications based on univariate
analysis (Table 2) and none of the SctO2 indices showed signifi-
cant difference between the binary strata with composite com-
plications  3 and > 3, respectively (Table 3). In contrast,
multiple SmtO2 indices (maximum, minimum, mean, median,
and AUC) were significantly associated with postoperative com-
posite complications based on univariate analysis (Table 2) and
multiple SmtO2 indices (maximum, mean, median, and AUC)
showed significant difference between the binary strata with
composite complications  3 and > 3, respectively (Table 3). Of
interest, the values of all SmtO2 indices, especially the max-
imum, mean and median, were higher in patients with more
postoperative complications than less (composite complica-
tions > 3 vs.  3, Table 3 and Fig. 1).
Among all SctO2 indices, only the standard deviation was
significantly associated with LOH based on univariate analysis
(Table 2) and none of the SctO2 indices showed significant dif-
ference between the binary strata with an LOH  6 and > 6 days,
respectively (Table 4). In contrast, multiple SmtO2 indices
(standard deviation, AUC and AUC weighted) were significantly
associated with LOH based on univariate analysis (Table 2) and
two SmtO2 indices (AUC and AUC weighted) showed significant
difference between the binary strata with an LOH  6 and
> 6 days, respectively (Table 4). The mean of SmtO2 was not sig-
nificantly different between patients with LOH  6 and > 6 days,
respectively (Table 4 and Fig. 2).
With oxygenation indices excluded, the selected clinical
covariates accounted for composite complications with a mul-
tiple R2 ¼ 0.32 and adjusted R2 ¼ 0.25. With oxygenation indices
554 | Meng et al.

Table 1 Continued
Table 1 Patient characteristics, clinical information and tissue
oxygenation of the study population (n ¼ 102). *Data are in Variables* Value & count
mean (SD) for continuous variables and in count (n) and per-
centage (%) for binary variables. ASA ¼ ASA physical status; PRBC (ml) 602 (780)
AUC ¼ area under curve; PRBC ¼ packed red blood cell;
Fresh frozen plasma (ml) 457 (821)
ICU ¼ intensive care unit; ALI ¼ acute lung injury; ARDS ¼ adult
Platelet (ml) 51 (141)
respiratory distress syndrome; PONV ¼ postoperative nausea
and vomiting; INR ¼ international normalized ratio Cellsaver (ml) 383 (550)
Urine output (ml) 731 (537)
Estimated blood loss (ml) 1488 (1618)
Variables* Value & count
Hospital course
Patient characteristics Length of hospital stay (days) 6 (2.8)
Age (yr) 63 (9) Number of ICU admission (n (%)) 58 (57%)
Sex ¼ male (n (%)) 43 (42%) ICU (days) 1.5 (1.8)
Weight (kg) 79 (20) Postoperative complications
Height (cm) 168 (12) Composite (total count, n) 2.6 (2.1)
ASA ¼ III (n (%)) 59 (58%) Hypotension (n (%)) 27 (27%)
Past medical history Arrhythmia (n (%)) 17 (17%)
Hypertension (n (%)) 52 (51%) Heart failure (n (%)) 2 (2%)
Diabetes mellitus (n (%)) 14 (14%) Myocardial ischaemia or infarction (n (%)) 3 (3%)
Peripheral vascular disease (n (%)) 2 (2%) Pulmonary oedema (n (%)) 1 (1%)
Cerebrovascular disease (n (%)) 10 (10%) Intubation > 24 h (n (%)) 9 (9%)
Coronary artery disease (n (%)) 17 (17%) ALI or ARDS (n (%)) 5 (5%)
Chronic lung disease (n (%)) 17 (17%) Pulmonary infection (n (%)) 3 (3%)
Chronic kidney disease (n (%)) 30 (29%) Pleural effusion (n (%)) 1 (1%)
Surgical information Shortness of breath (n (%)) 1 (1%)
Fusion ¼ yes (n (%)) 89 (87%) Hypoxia (n (%)) 3 (3%)
Number of segments (n) 7 (5.6) Stroke (n (%)) 0
Previous spine surgery (n (%)) 65 (64%) Neurocognitive change (n (%)) 16 (16%)
Surgical time (h) 5 (1.8) Spinal nerve injury (n (%)) 2 (2%)
Cerebral tissue oxygen saturation (SctO2) Seizure (n (%)) 1 (1%)
SctO2 maximum (%) 78 (6) Constipation (n (%)) 24 (24%)
SctO2 minimum (%) 61 (9) PONV (n (%)) 11 (11%)
SctO2 mean (%) 70 (5) Oliguria (n (%)) 4 (4%)
SctO2 median (%) 69 (5) Urinary infection (n (%)) 5 (5%)
SctO2 standard deviation (%) 3 (1) Creatinine elevation (n (%)) 19 (19%)
SctO2 AUC (%*min) 500 (412) Thrombocytopenia (Platelet < 100,000) (n (%)) 31 (30%)
SctO2 AUC weighted (%*min*hr1) 1.58 (0.94) Coagulopathy (INR > 1.5) (n (%)) 20 (20%)
Muscular tissue oxygen saturation (SmtO2) Requirement of PRBC transfusion (n (%)) 39 (38%)
SmtO2 maximum (%) 83 (8) Wound infection (n (%)) 6 (6%)
SmtO2 minimum (%) 63 (15) Wound dehiscence (n (%)) 11 (11%)
SmtO2 mean (%) 75 (11) Spinal haematoma (n (%)) 1 (1%)
SmtO2 median (%) 75 (11) Skin ulcer (n (%)) 1 (1%)
SmtO2 standard deviation (%) 4 (2) Severe weight loss (n (%)) 1 (1%)
SmtO2 AUC (%*min) 458 (580) Leucocytosis, unknown aetiology (n (%)) 1 (1%)
SmtO2 AUC weighted (%*min*hr1) 1.59 (2.35) Hyponatremia (n (%)) 2 (2%)
Anaesthetic agents & drugs Meralgia paresthetica (n (%)) 1 (1%)
Sevoflurane (n (%)) 46 (45%)
Desflurane (n (%)) 72 (71%)
Isoflurane (n (%)) 2 (2%)
Propofol (n (%)) 99 (97%)
included, the model selected none of the SctO2 indices; in con-
Fentanyl (n (%)) 86 (84%)
trast, it additionally selected SmtO2 minimum and AUC
Ketamine (n (%)) 85 (83%)
Lidocaine (n (%)) 57 (56%) weighted that led to a significant improvement of the multiple
Magnesium (n (%)) 1 (1%) R2 from 0.32 to 0.46 (þ41.7%) and the adjusted R2 from 0.25 to
Sufentanil (n (%)) 2 (2%) 0.38 (þ54.0%) (P ¼ 0.0001).
Tranexamic acid (n (%)) 59 (57.8%) With oxygenation indices excluded, the selected clinical
Phenylephrine (n (%)) 96 (94%) covariates accounted for LOH with a multiple R2 ¼ 0.34 and ad-
Phenylephrine (mg) 9.6 (7.9) justed R2 ¼ 0.26. With oxygenation indices included, the model
Norepinephrine (n (%)) 6 (6%) selected none of the SctO2 indices; in contrast, it additionally se-
Vasopressin (n (%)) 7 (7%) lected SmtO2 minimum and AUC weighted that led to a signifi-
Epinephrine (n (%)) 6 (6%) cant improvement of the multiple R2 from 0.34 to 0.40 (þ17.9%)
Intraoperative input & output and the adjusted R2 from 0.26 to 0.31 (þ19.0%) (P ¼ 0.02).
Crystalloid (ml) 2434 (1042) The adjusted R2 with oxygenation indices excluded and
Colloid (ml) 1052 (834) included, respectively, and the improvement of R2 attributable
to the inclusion of muscular tissue oxygenation in the fitting
Continued
model (multiple linear models) were illustrated (Fig. 3).
 Tissue oxygenation and postoperative outcome | 555

Table 2 Association of patient characteristics, clinical covariates and oxygenation indices with postoperative composite complications
and length of hospital stay based on univariate analysis. *Binary covariates with  4 times of occurrence were excluded in analysis.
†
P < 0.05; ‡P < 0.01. ASA ¼ ASA physical status; AUC ¼ area under curve; PRBC ¼ packed red blood cell; ALI ¼ acute lung injury; ARDS ¼ adult
respiratory distress syndrome; PONV ¼ postoperative nausea and vomiting

Variables* Composite complications Length of hospital stay

2
R P value R2 P value

Patient characteristics
Age 0.04 0.03† 0.02 0.19
Sex 0.05 0.04† 0.02 0.18
Weight 0.03 0.1 0.03 0.09
Height 0.06 0.01† 0.04 0.05
ASA 0.09 0.003‡ 0.03 0.07
Past medical history
Hypertension 0.02 0.12 0.001 0.73
Diabetes mellitus 0.008 0.38 0.002 0.65
Cerebrovascular disease 0.003 0.46 0.008 0.27
Coronary artery disease 0.002 0.63 0.009 0.27
Chronic lung disease 0.0006 0.83 0.002 0.70
Chronic kidney disease 0.06 0.01† 0.01 0.18
Surgical information
Spinal fusion 0.002 0.59 0.01 0.20
Number of segments 0.09 0.003‡ 0.15 <0.001‡
Previous spine surgery 0.003 0.59 0.01 0.25
Surgical time 0.12 <0.001‡ 0.11 <0.001‡
Cerebral tissue oxygen saturation (SctO2)
SctO2 maximum 0.004 0.55 0.0002 0.89
SctO2 minimum 0.0008 0.78 0.04 0.05
SctO2 mean 0.005 0.47 0.002 0.69
SctO2 median 0.004 0.51 0.001 0.72
SctO2 standard deviation 0.04 0.05 0.05 0.03†
SctO2 AUC 0.002 0.70 0.008 0.38
SctO2 AUC weighted 0.003 0.58 0.0005 0.83
Muscular tissue oxygen saturation (SmtO2)
SmtO2 maximum 0.10 0.001‡ 0.04 0.06
SmtO2 minimum 0.07 0.007‡ 0.003 0.60
SmtO2 mean 0.09 0.002‡ 0.01 0.27
SmtO2 median 0.08 0.004‡ 0.01 0.30
SmtO2 standard deviation 0.001 0.70 0.04 0.04†
SmtO2 AUC 0.04 0.04† 0.11 <0.001‡
SmtO2 AUC weighted 0.02 0.22 0.07 0.009‡
Anaesthetic agents & drugs
Sevoflurane 0.01 0.30 0.03 0.08
Desflurane 0.007 0.42 0.07 0.004‡
Fentanyl 0.06 0.02† 0.03 0.11
Ketamine 0.05 0.02† 0.08 0.02†
Lidocaine 0.01 0.25 0.006 0.46
Tranexamic acid 0.02 0.12 0.14 <0.001‡
Phenylephrine 0.12 <0.001‡ 0.16 <0.001‡
Norepinephrine 0.01 0.26 0.02 0.22
Vasopressin 0.04 0.04† 0.04 0.04†
Epinephrine 0.05 0.02† 0.06 0.01†
Intraoperative input & output
Crystalloid 0.10 0.002‡ 0.06 0.02†
Colloid 0.12 <0.001‡ 0.1 0.001‡
PRBC 0.22 <0.001‡ 0.22 <0.001‡
Fresh frozen plasma 0.18 <0.001‡ 0.21 <0.001‡
Platelet 0.15 <0.001‡ 0.10 0.002‡
Cellsaver 0.13 <0.001‡ 0.12 <0.001‡
Urine output 0.004 0.51 0.02 0.17
Estimated blood loss 0.19 <0.001‡ 0.17 <0.001‡
Postoperative complications
Composite (total count) 0.34 <0.001‡
Hypotension 0.03 0.06

Continued
 556 | Meng et al.
Table 2 Continued
Variables* Composite complications
R2
Arrhythmia
Intubation > 24 h
ALI & ARDS
Neurocognitive change
Constipation
PONV
Urinary infection
Creatinine elevation
Thrombocytopenia
Coagulopathy
Requirement of PRBC transfusion
Wound infection
Wound dehiscence
Discussion
This prospective observational study showed that muscular tis-
sue oxygenation has a stronger association with postoperative
complications and length of hospital stay than cerebral tissue
oxygenation, in patients undergoing major posterior spine sur-
gery. Cerebral tissue oxygenation has weak, if any, association
with these outcome measures.
Maintenance of tissue oxygenation is one of the prerequis-
ites of the proper function of any organ. Tissue hypoxia, de-
pending on the severity and duration, can cause organ
dysfunction and increase morbidity and mortality.9 Tissue oxy-
genation is essentially determined by the rates of tissue oxygen
consumption and supply, with the supply primarily determined
by organ perfusion and arterial blood oxygen content. In anaes-
thetized patients undergoing major surgeries, tissue oxygen-
ation of different organs may have distinctive patterns of
change as a result of different responses to variations in, but not
limited to, anaesthetic depth, cardiac output, arterial bp,
haemoglobin, and arterial blood carbon dioxide tension. It is
illustrated by the fact that anaesthetic agents such as isoflurane
and propofol have a profound dose-dependent suppressive ef-
fect on cerebral metabolic rate of oxygen.10 11 In contrast, the ef-
fect of anaesthetic agents on muscular metabolic rate of oxygen
is likely minimal, even though direct evidence is lacking. The
complexity of tissue oxygenation can be further demonstrated
by the fact that different organs have different tolerance for
ischaemia and hypoxia. The muscle can tolerate one to two h of
ischaemia without clinically overt injury,12 while the brain can
be permanently injured even if ischaemia only lasts for a few
min. As a vital organ, the brain is prioritized during haemo-
dynamic derangement. For instance, the acute decrease of car-
diac output leads to a parallel decrease of muscular perfusion in
a comparable magnitude; in comparison, the decrease of cere-
bral perfusion is only about one third of that of cardiac output.13
Taken together, tissue oxygenation of different organs may fol-
low different patterns of change during major surgeries because
of the distinctive schemes of oxygen consumption, supply, and
regulation in the face of various insults.
The goal of monitoring is to facilitate clinical decision-
making and improve patient outcome. It is important to under-
stand how the monitoring correlates with the outcome that
matters to patients. It was shown in cardiac surgical patients
Length of hospital stay
P value R2 P value
0.11 <0.001‡
0.15 0.001‡
0.0006 0.80
0.04 0.06
0.07 0.001‡
0.005 0.51
0.001 0.67
0.03 0.08
0.06 0.004‡
0.08 <0.001‡
0.16 <0.001‡
0.002 0.66
0.08 0.01†
that SctO2 is associated with outcomes including stroke,14 delir-
ium,15 neurocognition,5 7 8 major organ dysfunction,6 length of
hospital stay,7 and mortality in both adult16 and pediatric17 pa-
tients. Reduced SctO2 is associated with cognitive dysfunction
in patients undergoing thoracic surgery with one-lung ventila-
tion.18 Improved SctO2 is associated with a better cognitive
function and shorter post-anaesthesia care unit and hospital
stay in patients undergoing non-cardiothoracic surgeries.19 20 In
contrast, research on the association between SmtO2 and clin-
ical outcome is relatively new. It was shown that SmtO2 moni-
tored on the thenar eminence is associated with a composite of
major complications and mortality after major non-cardiac sur-
gery,21 and probability of successfully weaning from mechanical
ventilation.22 23 SmtO2 monitored on brachioradialis muscle is
associated with the mortality in patients with community-
acquired pneumonia.24 It has also been shown that the thenar
SmtO2 is able to significantly improve the predictive value of
Simplified Acute Physiology Score II and Sequential Organ
Failure Assessment scores in septic shock patients.25 In patients
undergoing open urological surgery, significant associations be-
tween thenar SmtO2 and haemoglobin concentration, central
venous oxygen saturation and high risk level were demon-
strated.26 However, the care guided by thenar SmtO2 monitoring
failed in improving patient outcome in a pilot small-scale
randomized controlled trial (n ¼ 25 in both intervention and
control groups).27 Taken together, the available evidence shows
that tissue oxygenation, monitored at either a central or a per-
ipheral location, is associated with various clinical variables de-
pending on the patient population being studied. Still, the
fundamental question is whether the care guided by tissue oxy-
genation monitoring, SctO2 or SmtO2, can provide the patient
with a beneficial outcome.
Spine surgery with osteotomy and multi-segment fusion is a
major complication-prone procedure.28 Massive haemorrhage
and transfusion of red blood cell, fresh frozen plasma, platelet,
cryoprecipitate, and cell saver blood pose a serious threat to
haemodynamic stability, adequacy of organ perfusion, and tis-
sue oxygen delivery in complicated major spine surgeries.28 29
Therefore, tissue oxygenation monitoring during major spine
surgery may offer valuable information, facilitate clinical
decision-making, and improve patient outcome. The only study
that investigated the impact of tissue oxygenation monitoring
 Tissue oxygenation and postoperative outcome | 557

Table 3 Comparison between patients with postoperative composite complications  3 or > 3 (counts). *Data are in mean (SD) for continu-
ous variables and in percentage (%) for binary variables. Binary covariates with  4 times of occurrence were excluded in analysis.
†
P < 0.05; ‡P < 0.01. ASA ¼ ASA physical status; AUC ¼ area under curve; PRBC ¼ packed red blood cell; ICU ¼ intensive care unit; ALI ¼ acute
lung injury; ARDS ¼ adult respiratory distress syndrome; PONV ¼ postoperative nausea and vomiting

Variables* Composite complications P value

 3 (n ¼ 70) > 3 (n ¼ 32)

Patient characteristics
Age (yr) 62.7 (7.9) 66.2 (10.5) 0.10
Sex ¼ male (%) 47.1 34.4 0.32
Weight (kg) 81.1 (20.0) 77.2 (20.5) 0.36
Height (cm) 169.2 (11.6) 166.2 (11.8) 0.24
ASA (I-III) 2.5 (0.6) 2.8 (0.4) 0.006‡
Past medical history
Hypertension (%) 47.8 59.4 0.39
Diabetes mellitus (%) 11.6 18.8 0.51
Cerebrovascular disease (%) 10.1 9.7 1.00
Coronary artery disease (%) 15.9 18.8 0.95
Chronic lung disease (%) 18.8 12.9 0.66
Chronic kidney disease (%) 26.1 37.5 0.35
Surgical information
Spinal fusion (%) 87.0 90.6 0.84
Number of segments (n) 6.0 (5.0) 8.9 (6.2) 0.02†
Previous spine surgery (%) 63.8 65.6 1.00
Surgical time (h) 4.6 (1.6) 5.8 (2.0) 0.004‡
Cerebral tissue oxygen saturation (SctO2)
SctO2 maximum (%) 78.1 (5.4) 77.6 (6.2) 0.70
SctO2 minimum (%) 61.4 (9.3) 61.1 (7.0) 0.86
SctO2 mean (%) 69.8 (5.1) 69.1 (5.5) 0.57
SctO2 median (%) 69.5 (5.0) 69.0 (5.5) 0.68
SctO2 standard deviation (%) 3.1 (1.5) 3.5 (1.4) 0.19
SctO2 AUC (%*min) 460.7 (682.7) 452.5 (245.7) 0.93
SctO2 AUC weighted (%*min*hr1) 1.7 (2.8) 1.3 (0.6) 0.27
Muscular tissue oxygen saturation (SmtO2)
SmtO2 maximum (%) 81.8 (8.3) 86.4 (4.8) <0.001‡
SmtO2 minimum (%) 62.0 (15.6) 66.2 (12.2) 0.15
SmtO2 mean (%) 73.2 (11.8) 77.6 (7.0) 0.02†
SmtO2 median (%) 73.4 (12.0) 77.8 (7.4) 0.02†
SmtO2 standard deviation (%) 3.7 (1.8) 4.3 (2.8) 0.25
SmtO2 AUC (%*min) 431.7 (368.3) 649.1 (466.3) 0.02†
SmtO2 AUC weighted (%*min*hr1) 1.4 (0.8) 1.9 (1.2) 0.07
Anaesthetic agents & drugs
Sevoflurane (%) 44.3 46.9 0.98
Desflurane (%) 70.0 71.9 1.00
Fentanyl (%) 80.0 93.8 0.14
Ketamine (%) 77.1 96.9 0.03†
Lidocaine (%) 52.9 62.5 0.49
Tranexamic acid (%) 57.1 59.4 1.00
Phenylephrine (%) 98.5 100.0 1.00
Norepinephrine (%) 1.4 15.6 0.02†
Vasopressin (%) 2.9 15.6 0.52
Epinephrine (%) 0 18.8 0.001‡
Intraoperative input & output
Crystalloid (ml) 2234 (906) 2875 (1193) 0.009‡
Colloid (ml) 925 (824) 1333 (799) 0.02†
PRBC (ml) 351 (479) 1154 (1008) <0.001‡
Fresh frozen plasma (ml) 207 (406) 1004 (1175) <0.001‡
Platelet (ml) 6 (48) 151 (212) <0.001‡
Cellsaver (ml) 233 (256) 710 (825) 0.003‡
Urine output (ml) 682 (500) 837 (605) 0.21
Estimated blood loss (ml) 988 (773) 2620 (2334) <0.001‡
Postoperative complications
ICU admission (%) 44.3 84.4 <0.001‡
Length of hospital stay (days) 5.3 (2.8) 7.7 (2.3) <0.001‡

Continued
 558 | Meng et al.

Table 3 Continued

Variables* Composite complications P value

 3 (n ¼ 70) > 3 (n ¼ 32)

Hypotension (%) 17.1 46.9 0.004‡

Arrhythmia (%) 5.7 40.6 <0.001‡
Intubation > 24 h (%) 2.9 21.9 0.006‡
ALI & ARDS (%) 1.4 12.5 0.06
Neurocognitive change (%) 10.0 28.1 0.04†
Constipation (%) 15.7 40.6 0.01†
PONV (%) 11.4 9.4 1.00
Urinary infection (%) 1.4 12.5 0.06
Creatinine elevation (%) 8.6 40.6 <0.001‡
Thrombocytopenia (%) 15.7 62.5 <0.001‡
Coagulopathy (%) 5.7 50.0 <0.001‡
Requirement of PRBC transfusion (%) 24.3 68.8 <0.001‡
Wound infection (%) 5.7 9.4 0.80
Wound dehiscence (%) 4.3 25.0 0.005‡

Complications<=3 (n=70) Complications<=3 (n=32)

90
Muscular tissue oxygen saturation (%)

85

80

75

70

65

60

55

50
1
17
33
49
65
81
97
113
129
145
161
177
193
209
225
241
257
273
289
305
321
337
353
369
385
401
417
433
449
465
481

Time (mins)

Fig 1 Muscular tissue oxygen saturation (%) per min. The mean and standard deviation of patients with complications  3 are illustrated by the dark blue line and
the light blue area, respectively. The mean and standard deviation of patients with complications > 3 are illustrated by the dark orange line and the light orange
area, respectively. The plotting stops at the time point where  4 patients remaining in the pool.

on outcome in spine surgery, concluded that SctO2-guided care
reduces the probability of cognitive deficiency after lumbar
spine surgery in prone position.30 However, this study had only
13 patients and did not monitor SmtO2. In contrast, our study
had a larger scale, incorporated both SctO2 and SmtO2 monitor-
ing, and used different outcome measures.
This study has limitations. It is a single-cohort observational
study. Therefore, the causal effect of tissue oxygenation on pa-
tient outcome cannot be determined. Whether there are effect-
ive interventions to restore the deranged SmtO2 and whether
the care guided by SmtO2 monitoring, in comparison to the care
guided by SctO2 monitoring, can improve patient outcome
cannot be determined by this study. The outcomes of postoper-
ative composite complications and length of hospital stay have
multiple determinants. It is likely that some outcome determin-
ants were not included in this study. Therefore, there is a
chance that the multiple linear models may have overestimated
the associative value of tissue oxygenation with the outcomes
being investigated in this study. There is also a chance that the
association between SctO2 and patient outcome may be under-
estimated because of the small sample size of this study. The
pre-induction value of tissue oxygenation is frequently used as
a reference for diagnostic and therapeutic purposes; however, it
was not obtained in this study. Therefore, the intraoperative
 Tissue oxygenation and postoperative outcome | 559

Table 4 Comparison between patients with length of hospital stay  6 or > 6 days. *Data are in mean (SD) for continuous variables and in
percentage (%) for binary variables. Binary covariates with  4 times of occurrence were excluded in analysis. †P < 0.05; ‡P < 0.01.
ASA ¼ ASA physical status; AUC ¼ area under curve; PRBC ¼ packed red blood cell; ICU ¼ intensive care unit; ALI ¼ acute lung injury;
ARDS ¼ adult respiratory distress syndrome; PONV ¼ postoperative nausea and vomiting

Variables* Length of hospital stay P value

 6 days (n ¼ 61) > 6 days (n ¼ 41)

Patient characteristics
Age (yr) 63.4 (7.5) 64.4 (10.6) 0.58
Sex ¼ male (%) 47.5 36.6 0.37
Weight (kg) 82.7 (20.7) 75.7 (18.8) 0.08
Height (cm) 170.2 (11.6) 165.2 (11.4) 0.04†
ASA (I-III) 2.5 (0.6) 2.6 (0.5) 0.58
Past medical history
Hypertension (%) 46.7 58.5 0.33
Diabetes mellitus (%) 11.7 17.1 0.63
Cerebrovascular disease (%) 8.5 12.2 0.79
Coronary artery disease (%) 21.7 9.8 0.19
Chronic lung disease (%) 18.6 14.6 0.80
Chronic kidney disease (%) 23.3 39.0 0.14
Surgical information
Spinal fusion (%) 91.7 82.9 0.31
Number of segments (n) 5.7 (4.6) 8.8 (6.4) 0.009‡
Previous spine surgery (%) 60.0 70.7 0.37
Surgical time (h) 4.6 (1.6) 5.5 (1.9) 0.01†
Cerebral tissue oxygen saturation (SctO2)
SctO2 maximum (%) 78.2 (5.4) 77.5 (6.1) 0.55
SctO2 minimum (%) 62.6 (8.2) 59.4 (9.0) 0.07
SctO2 mean (%) 69.9 (4.8) 69.0 (5.7) 0.44
SctO2 median (%) 69.7 (4.7) 68.7 (5.7) 0.38
SctO2 standard deviation (%) 3.0 (1.4) 3.5 (1.5) 0.13
SctO2 AUC (%*min) 453.7 (712.9) 464.8 (297.4) 0.91
SctO2 AUC weighted (%*min*hr1) 1.7 (3.0) 1.4 (0.6) 0.34
Muscular tissue oxygen saturation (SmtO2)
SmtO2 maximum (%) 82.2 (8.2) 84.9 (6.5) 0.07
SmtO2 minimum (%) 63.1 (15.6) 63.6 (13.4) 0.87
SmtO2 mean (%) 73.8 (11.7) 75.8 (8.8) 0.34
SmtO2 median (%) 74.0 (11.9) 75.9 (9.2) 0.38
SmtO2 standard deviation (%) 3.6 (1.7) 4.4 (2.7) 0.08
SmtO2 AUC (%*min) 398.6 (288.6) 650.6 (514.4) 0.006‡
SmtO2 AUC weighted (%*min*hr1) 1.4 (0.7) 1.9 (1.2) 0.02†
Anaesthetic agents & drugs
Sevoflurane (%) 52.5 34.1 0.11
Desflurane (%) 60.7 85.4 0.01†
Fentanyl (%) 83.6 85.4 1.00
Ketamine (%) 78.7 90.2 0.21
Lidocaine (%) 54.1 58.5 0.81
Tranexamic acid (%) 41.0 82.9 <0.001‡
Phenylephrine (mg) 98.2 100.0 1.00
Norepinephrine (mg) 3.3 9.8 0.35
Vasopressin (units) 3.3 12.2 0.18
Epinephrine (mg) 3.3 9.8 0.35
Intraoperative input & output
Crystalloid (ml) 2230 (793) 2740 (1282) 0.03†
Colloid (ml) 906 (849) 1272 (772) 0.03†
PRBC (ml) 384 (592) 929 (911) 0.001‡
Fresh frozen plasma (ml) 217 (478) 814 (1069) 0.002‡
Platelet (ml) 29 (103) 84 (180) 0.08
Cellsaver (ml) 247 (295) 586 (751) 0.008‡
Urine output (ml) 716 (533) 754 (550) 0.73
Estimated blood loss (ml) 1082 (1080) 2108 (2066) 0.006‡
Postoperative complications
ICU admission (%) 41.0 80.5 <0.001‡

Continued
 560 | Meng et al.

Table 4 Continued

Variables* Length of hospital stay P value

 6 days (n ¼ 61) > 6 days (n ¼ 41)

Composite (total count, n) 1.7 (1.6) 4.0 (2.2) <0.001‡

Hypotension (%) 19.7 36.6 0.10
Arrhythmia (%) 6.6 31.7 0.002‡
Intubation > 24 h (%) 3.3 17.1 0.04†
ALI & ARDS (%) 4.9 4.9 1.00
Neurocognitive change (%) 8.2 26.8 0.02†
Constipation (%) 16.4 34.1 0.07
PONV (%) 9.8 12.2 0.96
Urinary infection (%) 3.3 7.3 0.65
Creatinine elevation (%) 13.1 26.8 0.14
Thrombocytopenia (%) 24.6 39.0 0.18
Coagulopathy (%) 11.5 31.7 0.02†
Requirement of PRBC transfusion (%) 27.9 53.7 0.02†
Wound infection (%) 6.6 7.3 1.00
Wound dehiscence (%) 6.6 17.1 0.18

LOH<=6 days (n=61) LOH>6 days (n=41)

90
Muscular tissue oxygen saturation (%)

85

80

75

70

65

60

55

50
1
16
31
46
61
76
91
106
121
136
151
166
181
196
211
226
241
256
271
286
301
316
331
346
361
376
391
406
421
436
451
466
481

Time (mins)

Fig 2 Muscular tissue oxygen saturation (%) per min. The mean and standard deviation of patients with length of hospital stay (LOH)  6 days are illustrated by
the dark blue line and the light blue area, respectively. The mean and standard deviation of patients with LOH > 6 days are illustrated by the dark orange line and
the light orange area, respectively. The plotting stops at the time point where  4 patients remaining in the pool.

tissue oxygenation change was not referenced to this baseline
value in this study. In addition, the observation that the patients
with more postoperative complications seem having a higher
SmtO2 value than those with less postoperative complications
is interesting. However, whether this is a true pattern and what
the clinical implications are remain to be determined. In this
study, the SmtO2 was measured in the lower leg, which is differ-
ent to the previous studies which used the arm, especially the
thenar muscle, for SmtO2 measurement. We chose the lower leg
owing to the fact that the thenar muscle and the area adjacent
to it were used for evoked potential monitoring in our patients.
The constant stimulation and muscle twitches would have
interfered with the oximetry monitoring. In addition, we
hypothesized that there may be a likelihood for the tissue bed
that is far from the heart to be more affected by compromised
haemodynamics than the tissue bed that is close to the heart.
Intuitively, a tissue bed that has the least priority, or the first
one to be sacrificed during haemodynamic instability may have
 Tissue oxygenation and postoperative outcome | 561

0.5

0.45 Oxygenation excluded Muscular oxygenation included

0.4
*
0.35
**
0.3
Adjusted R2

0.25

0.2

0.15

0.1

0.05

0
Composite complications Length of hospital stay

Fig 3 The adjusted R2 based on multiple linear models with oxygenation indices excluded and muscular oxygenation indices included, respectively. *P ¼ 0.0001;
**P ¼ 0.02.

a better association with the overall outcome (i.e. more sensi-
tive). All patients in our study were in a prone position, which
may lead to venous pooling and is often accompanied by oe-
dema formation in the dependent parts of the body including
the forehead, which are two factors that may influence SctO2
measurements. This limitation needs to be recognized even
though the modern cerebral oximetry has incorporated meas-
ures to reduce the contamination by the extracranial tissue
layers and the compression of the forehead by the foam sup-
porting head frame may actually reduce the extent of the
position-related blood pooling and oedema formation. Other
limitations are that the sample size was not pre-determined be-
fore the study and the technology we used in this study (FORE-
SIGHT Elite, CASMED) is different to that in the previous studies
(INVOS, Covidien/Medtronic). The technological differences,
including the severity of the extracranial contamination,31 could
have contributed to the different findings between this study
and the previous studies.5–8 14–20
In summary, intraoperative muscular tissue oxygen satur-
ation has a stronger association with postoperative complica-
tions and length of hospital stay than cerebral tissue oxygen
saturation after major posterior spine surgery. Cerebral tissue
oxygenation has weak, if any, association with these outcome
measures in this patient population. Whether the care guided
by SmtO2 monitoring can improve patient outcome needs to be
addressed by randomized controlled trials.
Authors’ contributions
planning:/planning: L.M., X.H.
Study conduct: L.M., K.G.
Data analysis: L.M., Z.Y.
Writing paper: L.M., J.X.
Revising paper: all authors
Acknowledgements
We thank CAS Medical Systems, Inc., Branford, Connecticut,
for providing the FORE-SIGHT ELITE Tissue Oximeter at no
cost.
Declaration of interest
L.M. is a consultant to CAS Medical Systems, Inc.
Funding
This work was supported by the Inaugural Anesthesia
Department Awards for Seed Funding for Clinically-Oriented
Research Projects from the Department of Anesthesia and
Perioperative Care, University of California San Francisco,
San Francisco, California (to L.M.).
References
1. Jöbsis FF. Noninvasive, infrared monitoring of cerebral and
myocardial oxygen sufficiency and circulatory parameters.
Science 1977; 198: 1264–7
2. Mancini DM, Bolinger L, Li H, Kendrick K, Chance B, Wilson
JR. Validation of near-infrared spectroscopy in humans.
J Appl Physiol (1985) 1994; 77: 2740–7
3. De Blasi RA, Fantini S, Franceschini MA, Ferrari M, Gratton E.
Cerebral and muscle oxygen saturation measurement by
frequency-domain near-infra-red spectrometer. Med Biol Eng
Comput 1995; 33: 228–30
4. Scheeren TW, Schober P, Schwarte LA. Monitoring tissue
oxygenation by near infrared spectroscopy (NIRS): back-
ground and current applications. J Clin Monit Comput 2012; 26:
279–87
 562 | Meng et al.
5. Yao FS, Tseng CC, Ho CY, Levin SK, Illner P. Cerebral oxygen
desaturation is associated with early postoperative neuro-
psychological dysfunction in patients undergoing cardiac
surgery. J Cardiothorac Vasc Anesth 2004; 18: 552–8
6. Murkin JM, Adams SJ, Novick RJ, et al. Monitoring brain oxy-
gen saturation during coronary bypass surgery: a random-
ized, prospective study. Anesth Analg 2007; 104: 51–8
7. Slater JP, Guarino T, Stack J, et al. Cerebral oxygen desatur-
ation predicts cognitive decline and longer hospital stay
after cardiac surgery. Ann Thorac Surg 2009; 87: 36–44
8. Colak Z, Borojevic M, Bogovic A, Ivancan V, Biocina B,
Majeric-Kogler V. Influence of intraoperative cerebral oxim-
etry monitoring on neurocognitive function after coronary
artery bypass surgery: a randomized, prospective study. Eur J
Cardiothorac Surg 2015; 47: 447–54
9. Eastwood G, Bellomo R, Bailey M, et al. Arterial oxygen ten-
sion and mortality in mechanically ventilated patients.
Intensive Care Med 2012; 38: 91–8
10. Madsen JB, Cold GE, Hansen ES, Bardrum B. The effect of iso-
flurane on cerebral blood flow and metabolism in humans
during craniotomy for small supratentorial cerebral tumors.
Anesthesiology 1987; 66: 332–6
11. Lagerkranser M, Stånge K, Sollevi A. Effects of propofol on
cerebral blood flow, metabolism, and cerebral autoregula-
tion in the anesthetized pig. J Neurosurg Anesthesiol 1997; 9:
188–93
12. Drolet BC, Okhah Z, Phillips BZ, et al. Evidence for safe tourni-
quet use in 500 consecutive upper extremity procedures.
Hand (N Y) 2014; 9: 494–8
13. Meng L, Hou W, Chui J, Han R, Gelb AW. Cardiac output and
cerebral blood flow: the integrated regulation of brain perfu-
sion in adult humans. Anesthesiology 2015; 123: 1198–208
14. Goldman S, Sutter F, Ferdinand F, Trace C. Optimizing intrao-
perative cerebral oxygen delivery using noninvasive cerebral
oximetry decreases the incidence of stroke for cardiac surgi-
cal patients. Heart Surg Forum 2004; 7: E376–81
15. Schoen J, Meyerrose J, Paarmann H, Heringlake M, Hueppe M,
Berger KU. Preoperative regional cerebral oxygen saturation
is a predictor of postoperative delirium in on-pump cardiac
surgery patients: a prospective observational trial. Crit Care
2011; 15: R218
16. Heringlake M, Garbers C, Ka € bler JH, et al. Preoperative cere-
bral oxygen saturation and clinical outcomes in cardiac sur-
gery. Anesthesiology 2011; 114: 58–69
17. Suemori T, Skowno J, Horton S, Bottrell S, Butt W, Davidson
AJ. Cerebral oxygen saturation and tissue hemoglobin con-
centration as predictive markers of early postoperative out-
comes after pediatric cardiac surgery. Paediatr Anaesth 2016;
26: 182–9
18. Tang L, Kazan R, Taddei R, Zaouter C, Cyr S, Hemmerling TM.
Reduced cerebral oxygen saturation during thoracic surgery
predicts early postoperative cognitive dysfunction. Br J
Anaesth 2012; 108: 623–9
19. Casati A, Fanelli G, Pietropaoli P, et al. Continuous monitor-
ing of cerebral oxygen saturation in elderly patients
undergoing major abdominal surgery minimizes brain ex-
posure to potential hypoxia. Anesth Analg 2005; 101: 740–7
20. Ballard C, Jones E, Gauge N, et al. Optimised anaesthesia to
reduce post operative cognitive decline (POCD) in older pa-
tients undergoing elective surgery, a randomised controlled
trial. PLoS One 2012; 7: e37410
21. Abdelmalak BB, Cata JP, Bonilla A, et al. Intraoperative tissue
oxygenation and postoperative outcomes after major non-
cardiac surgery: an observational study. Br J Anaesth 2013;
110: 241–9
22. Poriazi M, Kontogiorgi M, Angelopoulos E, et al. Changes in
thenar muscle tissue oxygen saturation assessed by near-
infrared spectroscopy during weaning from mechanical ven-
tilation. Minerva Anestesiol 2014; 80: 666–75
23. Gruartmoner G, Mesquida J, Masip J, et al. Thenar oxygen sat-
uration during weaning from mechanical ventilation: an ob-
servational study. Eur Respir J 2014; 43: 213–20
24. Claverias L, Marı M, Marın-Corral J, et al. The prognostic value
of muscle regional oxygen saturation index in severe
community-acquired pneumonia: a prospective observa-
tional study. J Intensive Care 2016; 4: 7
25. Payen D, Luengo C, Heyer L, et al. Is thenar tissue hemoglobin
oxygen saturation in septic shock related to macrohemody-
namic variables and outcome? Crit Care 2009; 13: S6
26. Spruit RJ, Schwarte LA, Hakenberg OW, Scheeren TW.
Association of intraoperative tissue oxygenation with sus-
pected risk factors for tissue hypoxia. J Clin Monit Comput
2013; 27: 541–50
27. van Beest PA, Vos JJ, Poterman M, Kalmar AF, Scheeren TW.
Tissue oxygenation as a target for goal-directed therapy in
high-risk surgery: a pilot study. BMC Anesthesiol 2014; 14: 122
28. Bianco K, Norton R, Schwab F, et al; International Spine Study
Group. Complications and intercenter variability of three-
column osteotomies for spinal deformity surgery: a retro-
spective review of 423 patients. Neurosurg Focus 2014; 36: E18
29. Elgafy H, Bransford RJ, McGuire RA, Dettori JR, Fischer D.
Blood loss in major spine surgery: are there effective meas-
ures to decrease massive hemorrhage in major spine fusion
surgery? Spine (Phila Pa 1976). 2010; 35: S47–56
30. Trafidło T, Gaszyn  ski T, Gaszyn  ski W, Nowakowska-
Domagała K. Intraoperative monitoring of cerebral NIRS ox-
imetry leads to better postoperative cognitive performance:
a pilot study. Int J Surg 2015; 16: 23–30
31. Davie SN, Grocott HP. Impact of extracranial contamination
on regional cerebral oxygen saturation: a comparison of
three cerebral oximetry technologies. Anesthesiology 2012;
116: 834–40
Handling editor: A. R. Absalom