Download A noninvasive estimation of mixed venous oxygen saturation using

Pediatric Anesthesia 2005
15: 495–503
doi:10.1111/j.1460-9592.2005.01488.x
A noninvasive estimation of mixed venous oxygen
saturation using near-infrared spectroscopy by
cerebral oximetry in pediatric cardiac surgery
patients
T I A A . T O R T O R I E L L O M D F A A P * , S T E PH E N A . S T A Y E R M D †,
A N T O N I O R. M O T T M D F A A P * , E . D E A N M cK E N Z I E M D ‡,
C H A R L E S D . F R A S ER M D ‡, D E A N B . A N D R O P O U L O S M D †
A N D A N TH O N Y C . C H A N G M D F A A P *
*The Lillie Frank Abercrombie Section of Paediatric Cardiology, †Paediatric Cardiovascular
Anaesthesia and ‡Congenital Heart Surgery, The Heart Center, Texas Children’s Hospital,
Baylor College of Medicine, Houston, TX, USA
Summary
Background: Near-infrared spectroscopy (NIRS) is a noninvasive
optical monitor of regional cerebral oxygen saturation (rSO2). The aim
of this study was to validate the use of NIRS by cerebral oximetry in
estimating invasively measured mixed venous oxygen saturation
(SvO2) in pediatric postoperative cardiac surgery patients.
Methods: Twenty patients were enrolled following cardiac surgery
with intraoperative placement of a pulmonary artery (PA) or superior
vena cava (SVC) catheter. Five patients underwent complete biventricular repair – complete atrioventricular canal (n ¼ 3) and other
(n ¼ 2). Fifteen patients with functional single ventricle underwent
palliative procedures – bidirectional Glenn (n ¼ 11) and Fontan
(n ¼ 4). Cerebral rSO2 was monitored via NIRS (INVOS 5100) during
cardiac surgery and 6 h postoperatively. SvO2 was measured from
blood samples obtained via an indwelling PA or SVC catheter and
simultaneously correlated with rSO2 by NIRS at five time periods: in
the operating room after weaning from cardiopulmonary bypass, after
sternal closure, and in the CICU at 2, 4, and 6 h after admission.
Results: Each patient had five measurements (total ¼ 100 comparisons). SvO2 obtained via an indwelling PA or SVC catheter for all
patients correlated with rSO2 obtained via NIRS: Pearson’s correlation
coefficient of 0.67 (P < 0.0001) and linear regression of r2 ¼ 0.45
(P < 0.0001). Separate linear regression of the complete biventricular
repairs demonstrated an r ¼ 0.71, r2 ¼ 0.50 (P < 0.0001). Bland–
Altman analysis showed a bias of +3.3% with a precision of 16.6% for
rSO2 as a predictor of SvO2 for all patients. Cerebral rSO2 was a more
accurate predictor of SvO2 in the biventricular repair patients (bias
)0.3, precision 11.8%), compared with the bidirectional Glenn and
Fontan patients.
Correspondence to: Tia A. Tortoriello MD, FAAP, Division of Cardiology, Children’s Medical Center, The University of Texas Southwestern,
1935 Motor Street, Dallas, TX 75235, USA (email: [email protected]).
2005 Blackwell Publishing Ltd
495
4 96
T .A . T O R TO RI E LL O E T A L .
Conclusions: Regional cerebral oximetry via NIRS correlates with SvO2
obtained via invasive monitoring. However, the wide limits of
agreement suggest that it may not be possible to predict absolute
values of SvO2 for any given patient based solely on the noninvasive
measurement of rSO2. Near-infrared spectroscopy, using the INVOS
5100 cerebral oximeter, could potentially be used to indicate trends in
SVO2, but more studies needs to be performed under varying clinical
conditions.
Keywords: mixed venous oxygen saturation; near-infrared spectroscopy; cardiothoracic surgery; cerebral oximetry; perioperative care
Introduction
Near-infrared spectroscopy (NIRS) is a noninvasive
optical technique to assess microcirculatory oxygenation (1,2). NIRS relies on the relative transparency
of biological tissues to near-infrared light (700–
900 nm) where oxygenated and deoxygenated
hemoglobin have distinct absorption spectra. By
measuring the attenuation of light at several wavelengths and distances between emitter and detector,
it is possible to determine a value for cerebral
oxygen saturation (rSO2) (3).
Cerebral oximetry differs from pulse oximetry in
several respects. Although both use near-infrared
light signals, pulse oximetry monitors the pulsatile
signal component reflecting arterial circulation,
whereas cerebral oximetry monitors the nonpulsatile
signal component reflecting tissue circulation (arterioles, capillaries, and venules). Because the cerebral
microcirculation contains arterial, venous, and capillary components, cerebral saturation represents a
‘weighted average’ of the tissue circulation, with
approximately 75–85% of the signal originating
from venules (4,5). Because NIRS monitoring is
noninvasive and portable, it can provide real-time
measurements of these changes at the bedside (5–7).
In postoperative cardiac surgery patients, maintaining an adequate cardiac output is a critical
determinant of outcome. When the cardiac output
begins to decline there is a decrease in the mixed
venous oxygen saturation (less than the normal 65–
70%) because of reduced tissue blood flow and
greater oxygen extraction (increased A ) V oxygen
difference) (8). Thus, measurement of the mixed
venous oxygen saturation (SvO2) is an indicator of
cardiac output. Catheters placed in the pulmonary
artery in these postoperative patients allow the
direct measurement of SvO2. In patients with intracardiac shunting or cavopulmonary anastamosis,
such as those undergoing bidirectional Glenn (BDG)
or Fontan completion, oxygen saturation in the
superior vena cava (SVC) can be used as a substitute
for SvO2. Complications from an indwelling intracardiac catheter, however, include bleeding at the
time of removal, catheter breakage and entrapment
at the time of removal, infection, thrombus formation and embolization (9–11). Invasively measured
SvO2 from a central venous catheter does not
provide continuous monitoring, and acute deteriorations of SvO2 may be missed. In addition, repeated
blood sampling leads to increased blood loss, central
venous catheter colonization, contamination, and
increased risk of infection. It is prudent, therefore, to
develop noninvasive measures to monitor SvO2.
The purpose of this study was to compare cerebral
rSO2 with oximetrically measured SvO2 from an
indwelling pulmonary artery or SVC catheter in
infants and children after cardiac surgery. We
hypothesize that the two methods would correlate
closely, and that the noninvasive cerebral rSO2
would be adequate for monitoring SvO2.
Methods
Patient selection and data collection
After approval from our Institutional Review Board
for human subject research and informed parental
consent, patients were enrolled in the study protocol. Inclusion criteria included consecutive patients
2005 Blackwell Publishing Ltd, Pediatric Anesthesia, 15, 495–503
E S T I M A T I O N O F M I X E D V E N O U S O X Y G EN S A T U R A T I O N B Y N I R S
who underwent a reparative or palliative cardiac
surgical procedure in which an intracardiac
pulmonary artery catheter or percutaneous SVC
catheter in patients with cavopulmonary anastomosis was used.
The INVOS 5100 Cerebral Oximeter (Somanetics
Corp., Troy, MI, USA), is a two-channel (R + L)
cerebral oximeter, with an adult (>40 kg) and a
pediatric (£40 kg) Somasensor. The single-use sensors differ by shape and dimension, and are placed
on the right, left, or bilateral forehead by a selfadhesive layer. The INVOS 5100 sensor has two
detectors to measure the ratio of oxyhemoglobin to
total hemoglobin, with the resulting percentage
equal to the value for regional cerebral oxygen
saturation (rSO2). The proximal detector receives a
signal from the peripheral tissue and the distal
detector receives a signal from the extra- and
intracranial tissues; by subtracting the proximal
from the distal value, rSO2 is obtained. Depending
on which sensor is connected, the INVOS 5100 uses
sensor-dependent algorithms to calculate rSO2, with
the pediatric algorithm being adjusted for the
stronger signal reflected because of the thinner skull
of a child allowing more ambient light to enter the
head.
Currently all patients who undergo cardiac
surgery at our institution have continuous cerebral
oximetry monitoring in the operating room and
those patients entered into the study had cerebral
oximetry monitoring continued in the cardiac
intensive care unit (CICU) for 6 h after their
arrival. Pediatric Somasensors were placed on the
right or left forehead for all patients by a single
investigator (TAT). Care was taken to ensure
proper adhesion of the sensor to the forehead
and securing of cables from the sensor to the
monitor. Oxymetric measurements of SvO2 (Radiometer ABL 700; Diamond Diagnostics, Holliston,
MA, USA) were made at five time periods and
simultaneous rSO2 was correlated. The time periods were: in the operating room after weaning
from cardiopulmonary bypass, in the operating
room after sternal closure, and at 2, 4, and 6 h
after arrival to the CICU. Simultaneous arterial
blood gases were also measured at these time
periods. Six hours after arrival to the CICU, the
cerebral oximeter probe was removed and the
study protocol was terminated.
2005 Blackwell Publishing Ltd, Pediatric Anesthesia, 15, 495–503
4 97
Statistical analysis
Results are presented as mean ± SD or as mean ± SEM
when indicated. Measured SvO2 was correlated with
rSO2 measurements using Pearson’s correlation
coefficient. Analysis of agreement between the two
measurements was assessed using the method of
Bland and Altman (12,13). Statistics within subjects
was performed in accordance with a theory previously suggested by the same authors (14). Bias was
calculated as the mean difference between rSO2 and
the SvO2 for each patient; a positive bias (mean
difference) indicates that the rSO2 measure was
higher on average. Precision of bias estimate was
defined as 2 SD of the mean difference. Analysis of
covariance (ANOVA) was used to determine intrasubject variation. The relationship between rSO2
and SvO2 was also determined by linear regression.
Data were analyzed using SPSS for Windows
(version 11.0; SPSS, Inc., Chicago, IL, USA). Statistical significance was accepted at P < 0.05.
Results
Twenty patients were enrolled, 13 females and seven
males, with a mean age of 1.7 years (range 5 months
to 8 years) from July 2002 through January 2003. The
mean weight was 8.9 kg (range 5.1–24.4). Their
diagnoses were diverse and included acyanotic
and cyanotic forms of congenital heart disease
(Table 1). There were five patients who underwent
complete biventricular repairs (repair of complete
atrioventricular canal, n ¼ 3; other, n ¼ 2), and 15
patients with functional single ventricle who underwent palliative procedures (BDG, n ¼ 11; Fontan,
n ¼ 4). No functional single ventricle patients were
noted to have venovenous collaterals and no biventricular repair patients had postoperative intracardiac shunting by transesophageal echocardiogram.
No complications occurred during cerebral oximetry
monitoring in the operating room or in the CICU.
Measurements were made for the 20 patients at
five time periods after weaning from cardiopulmonary bypass, giving a total number of 100 observations. Table 2 delineates the physiological variables
at each of the five time periods after weaning from
cardiopulmonary bypass. The median rSO2 measured by NIRS was 67.5 ± 9.8%, with a range of
43–90%. Corresponding measurements of SvO2
4 98
T .A . T O R TO RI E LL O E T A L .
Table 1
Patient characteristics of the study population
Patient
Age (year)
Congenital
heart lesion
1
2
3
0.5
0.6
1.2
HLHS s/p Norwood
HLHS s/p Norwood
HLHS s/p Norwood
4
0.7
5
0.8
6
7
8
9
0.6
2.0
0.8
0.6
10
11
0.8
0.8
12
3.0
13
14
15
0.8
4.5
5.0
16
1.1
Heterotaxy,
RV dominant AVC,
PA s/p BTS
Dextrocardia,
VI, VSD, PA s/p BTS
HLHS s/p Norwood
PA/IVS s/p BTS
HLHS s/p Norwood
DILV, l-TGA, VSD,
PS s/p BTS
Ebstein’s s/p BTS
Dextrocardia, TA,
PA s/p BTS
Heterotaxy,
RV dominant AVC,
PA s/p BTS, BDG
DORV s/p PAB, BDG
TA, VSD s/p BDG
Heterotaxy,
DILV, d-TGA,
PA s/p BTS, BDG
PAPVR, VSD, PHTN
17
18
19
0.4
0.4
0.9
20
0.5
Mean ±
SD
CAVC, PHTN
CAVC
VSD, supraMV ring,
PHTN
CAVC
Cardiac surgery
Preop SaO2
CPB (min)
AXC (min)
HCA (min)
BDG
BDG
BDG,
aortic arch repair
BDG
72%
84%
50%
55
42
103
–
–
23
–
–
18
73%
45
–
–
BDG
72%
39
–
–
BDG
BDG
BDG
BDG
75%
77%
70%
70%
30
78
31
166
–
16
–
98
–
–
–
–
BDG
BDG
85%
82%
98
64
–
–
–
–
Fontan
76%
132
88
–
Fontan
Fontan
Fontan
77%
84%
86%
223
131
172
156
90
121
–
–
4
Warden repair
RPV, VSD cl
CAVC repair
CAVC repair
VSD cl,
resection MV ring
CAVC repair
94%
175
107
–
84%
86%
89%
194
167
159
141
113
100
–
–
–
92%
181
132
–
79% ± 10.0
114.3 ± 62.9
1.7 ± 2.0
98.8 ± 42.3
11.0 ± 9.9
AVC, atrioventricular canal; AXC, cross clamp time; BDG, bidirectional Glen; BTS, Blalock-Taussig shunt; CAVC, complete atrioventricular
canal; cl, closure; CPB, cardiopulmonary bypass time; DILV, double inlet left ventricle; DORV, double outlet right ventricle; HCA,
hypothermic circulatory arrest time; HLHS, hypoplastic left heart syndrome; IVS, intact ventricular septum; MV, mitral valve; PA,
pulmonary atresia; PAB, pulmonary artery band; PAPVR, partial anomalous pulmonary venous return; PHTN, pulmonary hypertension;
PS, pulmonary stenosis; RV, right ventricle; SaO2, arterial oxygen saturation; TA, tricuspid atresia; TGA, transposition of the great arteries;
s/p, status post; VI, ventricular inversion; VSD, ventricular septal defect.
were 63.0 ± 10.0%, with a range of 32–83%. The
mean rSO2 for the entire patient population was
slightly higher than the mean SvO2 at each of the five
time periods, with the trend in mean rSO2 following
the trend in the mean SvO2 at each of the five time
periods (Figure 1). Results from individual subjects
are presented in Figure 2; with two representative
subjects from the biventricular repair group, BDG
group, and Fontan group respectively. Correlation
and agreement between NIRS measurement of rSO2
and cooximetry measurement of SVO2 is presented
in Table 3. Figure 3 demonstrates the linear regres-
sion plot and Figure 4 the Bland–Altman plot of
rSO2 and SvO2 for all patients.
When comparing the differences between mean
rSO2 values measured by NIRS and mean SvO2
values from the pulmonary artery or SVC catheter,
considerable intersubject variation was found. However, the mean intrasubject standard deviation of the
difference between rSO2 by NIRS and SvO2 was low
(4.0 ± 2.1%). This indicated a consistent bias within
subjects. Therefore, all rSO2 values by NIRS were
corrected for bias by the intrasubject mean difference
between rSO2 by NIRS and SvO2 values measured
2005 Blackwell Publishing Ltd, Pediatric Anesthesia, 15, 495–503
E S T I M A T I O N O F M I X E D V E N O U S O X Y G EN S A T U R A T I O N B Y N I R S
4 99
Table 2
Physiological variables at each of the five time periods after cardiopulmonary bypass for the study population
Time 1 (n ¼ 20)
Mean time from CPB (min)
rSO2 (%)
SvO2 (%)
HR (bÆmin)1)
MAP (mmHg)
LAP mean (mmHg)
CVP mean (mmHg)
PAP mean (mmHg)
SaO2 (%)
PECO2 (mmHg)
pHa
pHv
PaO2 (mmHg)
PvO2 (mmHg)
PaCO2 (mmHg)
PvCO2 (mmHg)
Lactate
Rectal temperature (C)
23.4
64.0
61.1
132.0
58.3
5.2
7.6
16.9
94.3
29.4
7.39
7.35
79.2
30.8
40.2
45.0
1.6
36.1
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
17.2
10.6
9.8
22.6
8.2
1.5
3.2
7.2
5.9
4.3
0.06
0.07
56.3
5.3
9.4
6.2
0.8
0.5
Time 2 (n ¼ 20)
52.8
65.6
61.6
131.6
58.7
6.0
7.9
16.8
94.4
32.0
7.41
7.33
88.3
31.3
38.9
46.2
1.5
36.4
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
20.5
10.6
8.3
23.3
8.8
1.5
3.5
7.2
5.3
5.4
0.07
0.10
63.3
4.9
6.7
7.0
0.7
0.6
Time 3 (n ¼ 20)
209.7
68.1
64.2
134.1
65.5
6.4
6.5
16.2
92.5
37.1
7.38
7.34
110.2
32.9
39.7
46.4
1.5
37.2
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
33.4
12.1
8.8
22.3
11.7
3.0
3.2
8.3
6.4
6.2
0.05
0.04
82.4
4.6
5.4
6.5
0.8
0.8
Time 4 (n ¼ 20)
320.0
67.5
63.8
130.8
65.1
5.8
6.4
16.3
91.1
38.8
7.39
7.36
103.3
33.1
39.5
47.1
1.7
37.6
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
76.8
13.4
9.9
22.5
8.5
2.2
2.8
9.4
7.3
17.0
0.04
0.05
76.2
5.4
4.6
9.2
0.9
0.8
Time 5 (n ¼ 20)
410.1
66.6
64.4
132.5
66.6
5.8
6.8
17.4
92.7
37.5
7.39
7.35
96.1
33.5
40.1
47.3
2.0
37.1
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
96.3
12.0
8.6
18.7
13.7
2.4
3.6
11.1
6.2
6.9
0.05
0.05
69.2
5.2
5.2
5.9
1.0
0.9
Data are shown as mean ± SD.
CPB, cardiopulmonary bypass time; CVP, central venous pressure; PECO2, end tidal carbon dioxide; HR, heart rate; LAP, left atrial
pressure; MAP, mean arterial blood pressure; PaCO2, arterial carbon dioxide tension; PaO2, arterial oxygen tension; PAP, pulmonary
arterial pressure; pHa, arterial pH; pHv, systemic venous pH; PvCO2, systemic venous oxygen tension; PvO2, systemic venous oxygen
tension; rSO2, cerebral oxygen saturation; SaO2, arterial oxygen saturation; SvO2, mixed venous oxygen saturation.
from the pulmonary artery or SVC catheter. The
correlation coefficient found by analysis of covariance between intrasubject bias-adjusted rSO2 values
and SvO2 was 0.63 (P < 0.0001). A Bland–Altman
plot of the bias-adjusted NIRS rSO2 values and SvO2
values revealed limits of agreement of )10.1 and
13.4, still within a wide range of variability.
Discussion
rSO2
80
SvO2
Percent oxygen saturation
75
70
65
60
55
50
1
2
3
4
5
Time period
Figure 1
Mean cerebral oxygen saturation (rSO2) and mixed venous oxygen
saturation (SvO2) ± SD for the entire patient population plotted at
each of the five time periods. 1 ¼ after weaning from cardiopulmonary bypass, 2 ¼ after sternal closure, 3 ¼ 2 h after arrival
to the cardiac ICU, 4 ¼ 4 h after arrival to the cardiac ICU, and
5 ¼ 6 h after arrival in the cardiac ICU. There were 20 measurements (one per patient) at each time-point.
2005 Blackwell Publishing Ltd, Pediatric Anesthesia, 15, 495–503
Mixed venous oxygen saturation is commonly used
as an indicator of the adequacy of whole-body
oxygenation, as it reflects the balance between tissue
oxygen delivery and oxygen consumption. The
utility of serial measurements of SvO2 in the care
of critically ill patients has been documented by
many investigators (15–19), and is facilitated by the
availability of pulmonary artery catheters that provide continuous measurement of SvO2 (20,21). In
adult cardiac surgery patients, Jamieson et al. (22)
demonstrated a more than 10% fall in SvO2 before
mean arterial blood pressure, heart rate or pulmonary capillary wedge pressure were noted to change,
aiding in the identification of early manifestations of
inadequate tissue perfusion.
Measurement of SvO2 is a reliable and valuable
indicator of cardiopulmonary function in the immediate postoperative period, even in infants with
complicated repair of cardiac malformations (23–26).
5 00
T .A . T O R TO RI E LL O E T A L .
Figure 2
Cerebral oxygen saturation (rSO2 -s-), corresponding mixed venous oxygen saturation (SvO2 - -) and arterial oxygen saturation (SaO2 - -)
at the five time periods: in the operating room after weaning from cardiopulmonary bypass, in the operating room after chest closure, and
at 2, 4, and 6 h after arrival in the CICU. The top panel represents two bidirectional Glenn patients (patients 3 and 10), the middle panel two
Fontan patients (patients 12 and 15), and the bottom panel two biventricular repair patients (patients 18 and 19).
Monitoring of SvO2 improves survival following the
Norwood procedure for stage I palliation of hypoplastic left heart syndrome by providing a more
precise estimation of the pulmonary to systemic flow
ratio (Qp/Qs) and earlier identification of decreased
cardiac output (24).
2005 Blackwell Publishing Ltd, Pediatric Anesthesia, 15, 495–503
E S T I M A T I O N O F M I X E D V E N O U S O X Y G EN S A T U R A T I O N B Y N I R S
Table 3
Correlation and agreement between NIRS measurement of rSO2
and co-oximetry measurement of SVO2
All patients
Biventricular repair patients
BDG patients
Fontan patients
r-value
P-value
Bias
Precision
0.67
0.71
0.60
0.45
<0.0001
<0.0001
<0.0001
0.04
+3.3%
)0.3%
+3.6
+7.1%
16.6%
11.8%
18.4%
14.0%
5 01
Bias = 3.3%
Precision = 16.6%
BDG, bidirectional Glenn; NIRS, near-infrared spectroscopy; r,
correlation coefficient; rSO2, regional cerebral oxygen saturation;
SVO2, mixed venous oxygen saturation.
In this preliminary study with a small number of
patients we have shown a correlation between rSO2
and SvO2 (r ¼ 0.67). The strength of this relationship
was similar among the three surgical groups (biventricular repair, BDG, and Fontan): strongest in the
biventricular repair patients (r ¼ 0.71) and weakest
in the Fontan patients (r ¼ 0.45). This difference may
be related to cerebral venous congestion that
patients with single ventricle physiology develop
after cavopulmonary anastamosis. Like other tissues,
the cerebral vasculature has a greater volume of
venous blood than arterial blood. Various sources
estimate between 70 (4) and 90% (27) venous blood
by volume. Although these volumes cannot be
measured in vivo, INVOS cerebral oximeter values
correlate well when a 75% venous volume is
assumed during volume changes which occur as a
result of changes in PaCO2 (28). The compartment
ratio of 75% venous and 25% arterial is likely not
100
90
r = 0.67, P < 0.001
n = 100
rSO2
80
70
60
50
40
30
20
35 40 45 50 55 60 65 70 75 80 85 90
SvO2
Figure 3
Linear regression plot of correlation between cerebral oxygen
saturation (rSO2) and mixed venous oxygen saturation (SvO2) for
all patients.
2005 Blackwell Publishing Ltd, Pediatric Anesthesia, 15, 495–503
Figure 4
Bland–Altman plot of the differences plotted against the averages
of cerebral oxygen saturation (rSO2) and mixed venous oxygen
saturation (SvO2) for all data points.
constant, but changes continuously in response to
changes in cerebral vascular resistance, pulmonary
vascular resistance, and tissue oxygen demands.
When the compartment ratios change without a
change in blood volume, accuracy of the cerebral
oximeter does not change. However, comparison
between the rSO2 and SvO2 does change. We would,
thus, anticipate a greater discrepancy of rSO2 and
SvO2 among the BDG and Fontan patients.
The differences in venous congestion and changes
in the ratio between the arterial and venous compartments may have contributed to the intra- and
intersubject differences between rSO2 and SvO2 that
we observed. In addition to the obvious reasons for
intersubject variability differences between rSO2 and
SvO2, which include type of cardiac surgery, type of
congenital heart disease, and compartment ratio,
there are a number of other possibilities which could
play a role. Among these is the position of the sensor
around the curved head of children, ambient light,
differences in transfusion of blood components, and
soft tissue differences that include swelling with
impact on absorption. The cerebral rSO2 values
measured in our study revealed a large range of
cerebral oxygenation values similar to those found
in previous studies which were performed in children with structurally normal hearts (3,29–31). Prior
studies have attributed the large range of cerebral
oxygenation values to differences in positioning the
sensors on the forehead, as well as individual and
5 02
T .A . T O R TO RI E LL O E T A L .
age-dependent anatomical and physiological differences (3). In addition, because the venous proportion
strongly determines near-infrared spectroscopic
measurement of cerebral oxygenation state, the
value is influenced by cerebral blood flow and
cerebral arteriovenous oxygen extraction, which
may differ among subjects and was not measured
in our study. Variations in extracranial blood flow
also affect rSO2 (32–34), although at least 85% of
rSO2 is exclusively from the brain.
The scatter of the Bland–Altman plot demonstrates that the difference between rSO2 and SvO2
increases as the average decreases. Thus, the agreement between the two measurements falls outside
)2 SD when the average of rSO2 and SvO2 is
approximately <50%. This occurred at all three
measurements in the same patient, when rSO2 was
32, 35, and 41%. The cerebral oximeter probe was
inadvertently removed from this patient in the
operating room, and a new probe had to be placed
upon arrival to the CICU. Incorrect placement of the
probe on the forehead or not acquiring a firm seal
with the skin may have created false readings for
rSO2 in this patient. Again, stressing the importance
of securing the sensor position and sensor-skin
coupling by firm taping. Precision in the Bland–
Altman analysis was 16.6% for rSO2 as a predictor of
SvO2 for all patients, which in a clinical setting is a
wide range of variability. The wide limits of agreement between the two methods suggest that it is not
possible to accurately predict absolute SvO2 for any
given patient based solely on rSO2 readings. NIRS,
using the INVOS 5100 cerebral oximeter, appears to
be most useful for indicating trends in SvO2 rather
than absolute numbers.
The limitations to our study include the small
patient number, limited patient population, and
general stability of the majority of the patients prior
to and following their surgical procedure. We chose
this patient population based on the availability of a
pulmonary artery catheter or SVC catheter to obtain
the blood sample for SvO2 measurement. At our
institution, all BDG and Fontan patients have a
central venous catheter placed into the SVC via the
internal jugular vein in the operating room to
monitor the pulmonary artery pressure and calculate
the transpulmonary gradient postoperatively. The
biventricular repair patients included complete
atrioventricular canal patients and those with
preoperative concerns for pulmonary hypertension,
who also have pulmonary artery catheters placed in
the operating room at the time of their intracardiac
repair. In future studies it would be ideal to include
infants with hypoplastic left heart syndrome, and
compare NIRS by cerebral oximetry to continuous
superior vena cava oximetry (SvO2). The large range
of arterial oxygen saturation may also present a
further limitation, as an important component of
rSO2 was not constant. Thus, it would be interesting
to measure rSO2 in patients with normal SaO2
(>97%) while SvO2 is the changing variable during
major surgery. Another limitation may be the length
of time that we chose to leave the cerebral oximeter
probe in place, which was 6 h after arrival to the
CICU. We chose this time frame because the vast
majority of the patients would be functional single
ventricle patients who would undergo weaning of
analgesia and sedation and likely undergo tracheal
extubation within 6–8 h postoperatively.
In conclusion, this study suggests that it is not
possible to predict absolute values of SvO2 for any
given patient based solely on the noninvasive
measurement of cerebral rSO2. Near-infrared spectroscopy, using the INVOS 5100 cerebral oximeter, is
most useful for indicating trends in SvO2 rather than
absolute numbers. Further studies are required to
determine whether this noninvasive method can
provide a reliable trend monitor of SvO2 in a more
diverse group of patients under a wider range of
hemodynamic conditions.
Acknowledgements
The Somanetics Corporation (Troy, MI, USA) provided the INVOS 5100 cerebral oximeter monitoring
system and the pediatric Somasensors that were
used for data collection in this study. The authors
wish to thank Chris Knigge and Dana Capocaccia
from the Somanetics Corporation for their help in
conducting the study.
References
1 Jobsis FF. Non-invasive infrared monitoring of cerebral and
myocardial oxygen sufficiency and circulatory parameters.
Science 1977; 198: 1264–1267.
2 McCormick PW, Stewart M, Goetting MG et al. Noninvasive
cerebral optical spectroscopy for monitoring cerebral oxygen
delivery and hemodynamics. Crit Care Med 1991; 19: 89–97.
2005 Blackwell Publishing Ltd, Pediatric Anesthesia, 15, 495–503
E S T I M A T I O N O F M I X E D V E N O U S O X Y G EN S A T U R A T I O N B Y N I R S
3 Dullenkopf A, Frey B, Baenziger O et al. Measurement of cerebral oxygenation state in anaesthetized children using the
INVOS 5100 cerebral oximeter. Paediatr Anaesth 2003; 13: 384–
391.
4 Mchedlishvili GI. Arterial Behavior and Blood Circulation in the
Brain. New York: Plenum, 1986: 55–60.
5 Watzman HM, Kurth CD, Montenegro LM et al. Arterial and
venous contributions to near infrared cerebral oximetry. Anesthesiology 2000; 93: 947–953.
6 Kurth CD, Steven JM, Nicolson SC. Cerebral oxygenation
during pediatric cardiac surgery using deep hypothermic circulatory arrest. Anesthesiology 1995; 82: 74–82.
7 Kurth CD, Steven JM, Nicolson SC et al. Cerebral oxygenation
during cardiopulmonary bypass in children. J Thorac Cardiovasc Surg 1997; 113: 71–79.
8 Roth SJ. Postoperative care. In: Chang AC, Hanley FL, Wernovsky G, Wessel DL, eds. Pediatric Cardiac Intensive Care, 1st
edn. Baltimore: Williams & Wilkins, 1998: 163–188.
9 Smith-Wright D, Green T, Lock J et al. Complications of vascular catheterization in critically ill children. Crit Care Med
1984; 12: 1015–1017.
10 Petaja J, Lundstrom U, Sairenen H et al. Central venous
thrombosis after cardiac operations in children. Thorac Cardiovasc Surg 1996; 112: 883–889.
11 Damen J, Van der Tweel I. Positive tip cultures and related risk
factors associated with intravascular catheterization in pediatric cardiac patients. Crit Care Med 1988; 16: 221–228.
12 Altman DG, Bland JM. Measurement in medicine: the analysis
of method comparison studies. Statistician 1983; 32: 307–317.
13 Bland JM, Altman DG. Statistical methods for assessing
agreement between two methods of clinical measurement.
Lancet 1986; 1: 307–310.
14 Bland JM, Altman DG. Calculating correlation coefficients with
repeated observations: part I correlation within subjects. BMJ
1995; 310: 446.
15 Martin WE, Cheung PW, Johnson CC et al. Continuous monitoring of mixed venous oxygen saturation in man. Anesth
Analg (Cleve) 1973; 52: 784–793.
16 De La Rocha AG, Edmonds JF, Williams WG et al. Importance
of mixed venous oxygen saturation in the care of critically ill
patients. Can J Surg 1978; 21: 227–229.
17 Krauss XH, Verdouw PD, Hughenholtz PG et al. On-line
monitoring of mixed venous oxygen saturation after cardiothoracic surgery. Thorax 1975; 30: 636–643.
18 Boyd AD, Tremblay RE, Spencer FC et al. Estimation of cardiac
output soon after intracardiac surgery with cardiopulmonary
bypass. Ann Surg 1959; 150: 613–626.
19 Scheinman MM, Brown MA, Rapaport E. Critical assessment
of use of central venous oxygen saturation as a mirror of mixed
venous oxygen in severely ill cardiac patients. Circulation 1969;
40: 165–172.
20 Svedjeholm R, Hakanson E, Szabo Z. Routine SvO2 measurement after CABG surgery with a surgically introduced
2005 Blackwell Publishing Ltd, Pediatric Anesthesia, 15, 495–503
21
22
23
24
25
26
27
28
29
30
31
32
33
34
5 03
pulmonary artery catheter. Eur J Cardiothorac Surg 1999: 16:
450–457.
Vedrinne C, Bastien O, De Varax R et al. Predictive factors for
usefulness of fiberoptic pulmonary artery catheter for continuous oxygen saturation in mixed venous blood monitoring
in cardiac surgery. Anesth Analg 1997; 85: 2–10.
Jamieson WRE, Turnbull KW, Larrieu AJ et al. Continuous
monitoring of mixed venous oxygen saturation in cardiac
surgery. Can J Surg 1982; 25: 538–543.
Schranz D, Schmitt S, Oelert H et al. Continuous monitoring of
mixed venous oxygen saturation in infants after cardiac surgery. Intensive Care Med 1989; 15: 228–232.
Tweddell JS, Hoffman GM, Mussatto KA et al. Improved survival of patients undergoing palliation of hypoplastic left heart
syndrome: lessons learned from 115 consecutive patients.
Circulation 2002; 106 (Suppl. 1): I82-I89.
Rossi AF, Sommer RJ, Lotvin A et al. Usefulness of intermittent
monitoring of mixed venous oxygen saturation after stage I
palliation for hypoplastic left heart syndrome. Am J Cardiol
1994; 73: 1118–1123.
Hoffman GM, Ghanayem NS, Kampine JM et al. Venous saturation and the anaerobic threshold in neonates after the
Norwood procedure for hypoplastic left heart syndrome. Ann
Thorac Surg 2000; 70: 1515–1520.
Todd MM, Weeks J. Comparative effects of propofol, pentobarbital and isoflurane on cerebral carbon dioxide in man. Circ
Res 1978; 43: 324–331.
Henson LC, Cartwright C, Chlebowski SM et al. Estimation of
jugular venous O2 saturation from cerebral oximetry or arterial
O2 saturation during isocapnic hypoxia. Anesthesiology 1997; 87
(3A): A402.
Hirsch W, Sitka U, Scheibe B et al. Regional cerebral oxygen
measurement and near infrared spectroscopy in newborns.
Paediatr Grenzgeb 1997; 36: 155–165.
Misra M, Stark J, Dujovny M et al. Transcranial cerebral
oximetry in random normal subjects. Neurol Res 1998; 20:
137–141.
Rodriguez-Nunez A, Couceiro J, Alonso C et al. Cerebral
oxygenation in children with syncope during head-upright tilt
test. Peadiatr Cardiol 1997; 18: 406–409.
Samra SK, Stanley JC, Zelenock GB et al. An assessment of
contributions made by extracranial tissues during cerebral
oximetry. J Neurosurg Anesth 1999; 11: 1–5.
Germon TJ, Evans PD, Barnett NJ et al. Cerebral near infrared
spectroscopy: emitter-detector separation must be increased.
Br J Anesth 1999; 82: 831–837.
Edwards AD, Richardson C, van der Zee P et al. Measurements of hemoglobin flow and blood flow by near-infrared
spectroscopy. J Appl Physiol 1993; 74: 1884–1890.
Accepted 19 June 2004