British Journal of Anaesthesia 105 (2): 179–84 (2010)
Advance Access publication 10 June 2010 . doi:10.1093/bja/aeq123
PAEDIATRICS
Size of internal jugular vs subclavian vein in small infants: an
observational, anatomical evaluation with ultrasound
C. Breschan 1*, M. Platzer 1, R. Jost 2, H. Stettner 3 and R. Likar 1
1
Department of Anesthesia, LKH Klagenfurt, St Veiterstrasse 47, 9020 Klagenfurt, Austria
Department of Anesthesia, KH Spittal/Drau, Spittal/Drau, Austria
3
Department of Statistics, University of Klagenfurt, Klagenfurt, Austria
2
* Corresponding author. E-mail: [email protected]
Key points
† Ultrasound observation
reveals that the IJV is
significantly larger than
SCV.
† A positive correlation
exists between the size of
the infant and that of
both IJV and SCV.
† Other risk factors, e.g.
distortion of the IJV
during cannulation
attempts and higher
catheter-related infection
rate for IJV, should also
be considered.
Background. The primary goal of this study was to compare the size and depth of the
internal jugular vein (IJV) and the subclavian vein (SCV) in infants under general
anaesthesia. A secondary goal was to determine the correlation of weight, height, head
circumference, and age to the size and depth of these veins.
Methods. Sixty small infants weighing from 1.4 to 4.5 kg were included. Using ultrasound,
the diameters via short-axis (SAX) and long-axis (LAX) views, cross-sectional area (CSA),
and depth of the left and right IJV and SCV were measured.
Results. The diameter of the IJV was 7.9% larger on average than that of the SCV as
measured via the SAX and LAX views (mean: 3.1 vs 2.9 mm; Wilcoxon’s signed-rank test:
P,0.01). The CSA of the IJV was 27% larger on average than that of the SCV (mean:
10.2 vs 8.0 mm2; Wilcoxon’s signed-rank test: P,0.01). Seventy-five per cent of the
neonates showed a larger CSA of the IJV. The SCV was 8.4% deeper on average from the
skin surface than the IJV (mean: 6.4 vs 5.9 mm; Wilcoxon’s signed-rank test: P,0.01).
There was a significant positive correlation between weight, height, head circumference,
and age to the size and depth of the veins (Spearman’s rank correlation: P,0.01).
Conclusions. Because of its most likely larger size, the IJV can be recommended as the
better choice for cannulation in comparison with the SCV. However, other factors should
also be considered.
Keywords: infant; internal jugular vein; subclavian vein; ultrasound
Accepted for publication: 12 April 2010
Central venous access is often mandatory when caring for
critically ill newborn and premature infants. The internal
jugular vein (IJV) and the subclavian vein (SCV) are the
most frequently used vessels for central venous cannulation.
However, percutaneous catheter insertion in neonates is a
challenge even for the experienced anaesthetist. The most
significant factor leading to cannulation failure is the small
size of the veins in neonates and preterm infants.1 2
The IJV seems to be the preferred choice for central
venous catheter (CVC) insertion in small infants. One of the
reasons could be the assumed larger size compared with
the SCV as often quoted. To our knowledge, there are no
data confirming this assumption in living neonates. A study
of 21 consecutive autopsy specimens of infants ,1 yr of
age and weighing ,6 kg revealed no significant difference
in diameter between the internal jugular and subclavian
venous system, on either the right or left side.3
The primary objective of this study was to compare the
diameter, cross-sectional area (CSA), and depth from the
skin surface between IJV and SCV in small infants, neonates,
and premature babies on the right and left sides while being
positioned as for central venous cannulation under general
anaesthesia before undergoing surgery.
A secondary objective was to determine the correlation
between age, weight, height, and head circumference with
the diameter, CSA, and depth from the skin of these veins.
Methods
Study population
This study was conducted at the Klagenfurt General Hospital
in Austria. After approval by the ethics committee of Land
Kaernten (Ref: A13/08) and parental informed consent, the
veins of 61 small infants were investigated. This study
included premature babies, neonates, and small infants
who received general anaesthesia while undergoing
general paediatric surgery. Study inclusion criteria only comprised babies younger than 42 days of life or if born
& The Author [2010]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved.
For Permissions, please email: [email protected]
BJA
prematurely, until 42 days had elapsed from the calculated
birth date. Furthermore, only babies weighing ,4.6 kg at
the time of the operation were included. Only patients
whose IJV and SCV could clearly be identified via ultrasound
in terms of an optimally placed ultrasound probe were eventually evaluated.
Breschan et al.
C
Baby 3.2kg
SCM
CA
C
IJV
Protocol for ultrasound investigation
All measurements were performed in the operating theatre
during general anaesthesia before surgery. The patients
received a standardized anaesthetic including muscle relaxation, tracheal intubation, a solution of 2/3 Ringer’s lactate
in 1/3 glucose 5% at a rate of 10 ml kg h21, and a sevoflurane
maintenance concentration aimed at achieving age-specific
MAC values according to the end-tidal gas concentration
monitoring. Before surgical incision, a standard-sized towel
roll measuring 4 cm in diameter in babies ,3 kg and 4.5
cm in those more than 3 kg was placed under the shoulders.
The occiput of the head was stabilized in a standard ring
head holder of 1.5 cm in height. Using a protractor, the
patient’s head was turned 308 to the side opposite to
where the investigation was being carried out. No Trendelenburg positioning was applied.
The same ultrasound device with 10 –13 MHz and smallest
available linear probe of 2.5×1 cm was used (Micromaxx,
Sonosite, Inc., Bothell, WA, USA). The US unit was set on its
highest resolution with a depth of 1.9 cm. No colour flow
was applied.
Three consecutive measurements were performed at the
end of expiration under controlled ventilation for each anatomical location on the right and left sides. No PEEP was
applied during the mechanical ventilation, and peak airway
pressures were kept within the range of 10 –15 cm H2O.
Ultrasound gel was applied to the US probe. The probe was
then placed gently over the skin without compressing the
vein while freezing the image at the end of expiration in
order to obtain the largest surface of the vein for the
measurements.
By placing the ultrasound probe perpendicular to the skin
at the level of the cricoid cartilage, the cross-sectional view
of the right IJV (¼circular) was obtained (Fig. 1). The image
was frozen. The diameter, CSA, and depth from the skin
surface were measured via this short-axis (SAX) view. The
maximum anterior –posterior SAX diameter was determined
by placing the dotted line provided by the ultrasound image
between the most distant anterior –posterior points of the
wall of the vein (Fig. 1). The depth from the skin surface
was obtained by placing the dotted line between the wall
of the vein closest to the skin surface and the skin surface
(Fig. 1). The CSA was determined by adjusting the dotted
ellipse provided by the ultrasound image along the wall of
the vein (Fig. 1). The size within this area was then automatically calculated and displayed by the ultrasound device
(Fig. 1).4 By placing the ultrasound probe along the long
axis (LAX) of the right IJV, the longitudinal view (¼tubular)
was obtained. After freezing the image, the diameter was
180
Ant
CL
Lat
A-A: 4.2mm
B-B:5.1mm; A: 16.5mm2; C:14.6mm
1,9
C-C: 6.3mm
Fig 1 Cross-sectional view of the right IJV in a 3.2 kg infant.
Ultrasound probe placed transversely to the neck. Inset: CL, clavicle. Ultrasound image: IVJ, internal jugular vein; CA, carotid
artery; SCM, sternocleidomastoid muscle. Dotted line: A– A,
anterior – posterior diameter of IJV via SAX view. Ellipse: B –B,
medial –lateral distance of the ellipse; A, area (CSA) within the
ellipse; C, circumference of the ellipse. Dotted line: C – C, depth
of IJV from the skin surface.
Baby 4kg
PM
SCV
B
Ant
SCA
Caud
PL
A-A:4.2mm; A:10mm2; C:10.6mm
CL
B:3.2mm
Fig 2 Cross-sectional view of the right SCV in a 4 kg infant. Sagittal ultrasound probe placed below the clavicle. Inset: CL, clavicle.
Ultrasound image: SCV, subclavian vein; SCA, subclavian artery;
PL, pleura; PM, pectoral muscles. Ellipse: A –A, cranial – caudal distance of the ellipse; A, area within the ellipse; C, circumference of
the ellipse. Dotted line: B –B, anterior – posterior diameter of SCV
via SAX view.
measured by placing the dotted line between the most
distant points of the wall of the vein.
In order to obtain the longitudinal view of the right SCV,
the ultrasound probe was placed at the supraclavicular
area as described by Pirotte and Veyckemans.5 After having
obtained an optimal longitudinal view of the SCV, the
image was frozen and the maximum anterior –posterior
BJA
Size of internal jugular vs subclavian vein in neonates
characteristic data of the included babies are presented in
Table 1. Weight, height, head circumference, and PCA were
not normally distributed (Kolmogorov–Smirnov test:
P.0.95).
diameter and depth from the skin surface were determined
via this LAX view. The cross-sectional view of the right SCV
was obtained by the sagittal placement of the ultrasound
probe below the clavicle (Fig. 2).6 After having obtained an
optimal cross-sectional view of the SCV, the image was
frozen and the maximum anterior –posterior diameter and
CSA were noted via this SAX approach.
The measurements were then repeated on the left IJV and
SCV in the same way.
All investigations were performed by the same physician
experienced in ultrasound technique without randomization
or blinded evaluation.
Characteristics of veins
The parameters of the veins are presented in Table 2. The
diameter, CSA, and depth of all four veins were normally distributed (Kolmogorov –Smirnov test: P,0.95).
The SAX and LAX approach resulted in almost identical
mean values of 3.1 and 3.0 mm of the left IJV and 3.1 mm
of the right IJV diameter, respectively (Table 2). The SAX
and LAX views resulted in nearly identical mean values of
2.8 and 2.9 mm of the left SCV diameter and 2.9 mm of
the right SCV, respectively (Table 2).
Data collection
The following data were recorded by a computerized and
written protocol (Windows Access; Microsoft Corporation,
Redmond, WA, USA) at the time of the measurements:
date; patient’s weight, height, head circumference, and postconceptual age (PCA), that is, weeks¼gestational+postnatal age; diameter via SAX and LAX views, CSA, and depth
from the skin of the left and right IJV and SCV.
Comparison of differences between the veins
The diameter via the SAX and LAX views, CSA, and depth from
the skin surface did not differ significantly between the left
and right IJV and between the left and right SCV (Wilcoxon’s
signed-rank test: P.0.05). Comparing the difference in the
diameter between the IJV and the SCV, the IJV was found to
be significantly larger than the SCV on both the left and
right sides by 7.9% on average as measured via the SAX and
LAX approach (mean: 3.1 vs 2.9 mm; Wilcoxon’s signed-rank
test: P,0.01). The CSA of the IJV was also significantly
larger than that of the SCV on both sides by 27% on average
(mean: 10.2 vs 8.0 mm2; Wilcoxon’s signed-rank test:
P,0.01). Sixty-seven per cent of the patients showed a
larger diameter of the IJV via the SAX view and 63.5% via
the LAX view. The CSA of the IJV was larger than the SCV in
75.5% of the babies. The SCV was found to be significantly
deeper from the skin surface by 8.4% on average when
compared with the IJV on both sides (mean: 6.4 vs 5.9 mm;
Wilcoxon’s signed-rank test: P,0.01).
Statistical methods
The significance of the results was measured by P-values.
P-values ,0.05 were deemed as statistically significant.
The Kolmogorov–Sminov test was used to test the normality
of the distribution of the data within the groups. The correlations between the variables (body weight, height, head circumference, and post-conceptual age and diameter, CSA,
and depth of veins) were characterized by the Spearman
rank correlation coefficient, rS. Two-sided significance
testing was used for correlations.
To compare the difference in paired values between the
IJV and the SCV, the Wilcoxon signed-rank test was used.
For means, 95% confidence intervals (CIs) were reported.
All calculations were made in R (Version 2.7.0, 2008; The R
Foundation for Statistical Computing, ISBN 3-900051-07-0,
http://cran.r-project.org) and HP-RPL (Ver. 2.08, 2006;
Hewlett-Packard Company, San Diego, CA, USA).
Correlation of veins
The correlation between body weight, height, and head circumference was significantly positive (Spearman’s rank correlation coefficient, rS, between 0.62 and 0.87; P,0.01).
There was a significant positive correlation between height,
weight, head circumference, and PCA with the diameter,
CSA, and depth of all four veins including the SAX and LAX
views (Spearman’s rank correlation coefficient, rS, between
0.205 and 0.69; P,0.01) (Figs 3 and 4).
Results
Study population
Sixty-one patients were investigated. One infant was
excluded because the left SCV could not be clearly identified
as such possibly due to an anatomical variation. The
Table 1 Characteristic data of patients. Non-normal distribution of weight, height, head circumference, and PCA (Kolmogorov– Smirnov test:
P.0.95)
Min
Body weight (kg)
1.4
Med
3.2
Mean
3.2
Max
4.5
SD
SEM
95% CI
0.87
0.11
2.99 – 3.44
Height (cm)
35
52
52.1
63
6.07
0.78
50.50– 53.63
Head circumference (cm)
28
35
35.8
43
2.99
0.39
34.90– 36.53
PCA (weeks)
33
42
41.9
46
3.91
0.51
40.80– 42.88
181
BJA
Breschan et al.
Table 2 Parameters of veins. Normal distribution of diameter, CSA, depth of all four veins (Kolmogorov–Smirnov test: P≤0.95). Depth and
diameter via SAX and LAX views: number of patients, n¼60; CSA: number of patients, n¼51
Min
Med
Mean
Max
SD
SEM
95% CI
Left IJV diameter SAX (mm)
1.7
3.0
3.1
5.0
0.73
0.09
2.90 – 3.28
Left IJV diameter LAX (mm)
1.5
3.0
3.0
4.8
0.73
0.09
2.85 – 3.23
Left IJV CSA (mm2)
3.0
8.7
10.0
18.3
4.68
0.66
8.71 – 11.30
Left IJV depth (mm)
3.8
5.8
5.8
9.0
1.09
0.14
5.54 – 6.11
Right IJV diameter SAX (mm)
1.7
3.1
3.1
5.1
0.73
0.09
2.95 – 3.33
Right IJV diameter LAX (mm)
1.7
3.1
3.1
5.3
0.72
0.09
2.92 – 3.29
Right IJV CSA (mm2)
3.3
10.3
10.3
19.0
4.53
0.63
9.04 – 11.59
Right IJV depth (mm)
3.8
5.7
6.0
9.0
1.16
0.15
5.66 – 6.26
Left SCV diameter SAX (mm)
1.8
2.8
2.8
4.5
0.60
0.08
2.70 – 3.01
Left SCV diameter LAX (mm)
1.5
2.9
2.9
4.6
0.65
0.08
2.71 – 3.05
Left SCV CSA (mm2)
2.3
7.3
8.1
18.0
3.72
0.52
7.05 – 9.14
Left SCV depth (mm)
4.6
6.6
6.5
8.1
1.02
0.13
6.20 – 6.73
Right SCV SAX (mm)
1.6
2.9
2.9
4.4
0.61
0.08
2.71 – 3.03
Right SCV LAX (mm)
1.8
2.9
2.9
4.5
0.63
0.08
2.70 – 3.03
2.7
8.0
7.9
16.0
3.25
0.46
6.99 – 8.82
Right SCV depth (mm)
4.0
6.4
6.4
9.3
1.05
0.13
6.13 – 6.67
20
20
15
15
LSCV–CSA (mm2)
RIJV–CSA (mm2)
Right SCV CSA (mm2)
10
5
5
0
0
1
2
3
4
5
6
Weight (kg)
Fig 3 Significant positive correlation between weight and CSA of
right IJV (Spearman’s rank correlation coefficient; rS ¼0.73).
Equation of regression line: CSA¼22.37+3.87×weight. Number
of patients: n¼51.
Discussion
In this study, we clearly showed that the IJV has a significantly larger size compared with the SCV. The size did not
differ between the left and right IJV nor between left and
right SCV.
The assumption of the IJV being larger than the SCV in
small children has been supported by the fact that the
head is comparatively large, whereas the upper extremities
182
10
1
2
3
4
Weight (kg)
5
6
Fig 4 Significant positive correlation between weight and CSA of
left SCV (Spearman’s rank correlation coefficient; rS ¼0.56).
Equation of regression line: CSA¼0.16+2.42×weight. Number
of patients: n¼51.
are rather small in this age group. However, the in vivo evidence for this was lacking. Opposite results found in the
autopsy study by Cobb and colleagues3 can be explained
by the smaller patients included in this study. Previous
studies have investigated the IJV diameter in children via
ultrasound or venography.1 7 – 9
We also found a significant positive correlation between
weight, height, head circumference, and the size and depth
of both the IJV and the SCV. It can be stated that the smaller
BJA
Size of internal jugular vs subclavian vein in neonates
the neonate, the smaller his/her IJV and SCV. In contrast to our
findings, Murat and colleagues7 showed a weak correlation
between the size of the patients and that of the veins.
However, they mostly included older children in their study.
The significantly greater depth of the SCV in comparison
with the IJV as measured in this study is clinically not relevant because the site of the determination of the depth
of the SCV is not the puncture site and does not show the
path of the exploratory needle.
In order to minimize errors of our measurements caused
by the US technique, we determined the diameter via the
SAX and LAX views and the CSA. Each measurement
was performed three times. The CSA and SAX diameter
were measured exactly on the same picture (Figs 1 and 2).
However, calculating the CSA by applying the formula,
diameter2 ×M/4, using the maximum anterior –posterior
SAX diameter as a single parameter would usually result in
a smaller CSA than the actually measured one by the ultrasound device. This discrepancy can be explained by the fact
that veins usually appear ellipsoid on a cross-sectional
view. There is no doubt that the determination of the CSA
via the ultrasound image was the most accurate measurement, because the actual size of an area enclosed by an
adjusted ellipse was measured.
In addition to finding the vein, another difficult part of CVC
catheterization in neonates is often the introduction of the
guide-wire into the vein due to the small size.7 This is an
additional reason to choose the largest vein for central
venous line placement. According to the results provided by
this study, this would be the IJV over the SCV in 75% of the
neonates. However, the question arises as to whether
the difference in the diameter in the range of 0.25 mm for
the mean values is really clinically relevant? In addition,
during actual cannulation attempts, there will be significant
distortion and compression of the vessels by the exploratory
needle, and in our experience, this seems to be disproportionately more important for the IJV than for the SCV.
The advantage when entering the SCV is the fact that it is
fixed to the clavicle thus preventing major anterior wall
collapse. This phenomenon (i.e. anterior wall collapse),
frequently seen when US guidance is used for IJV puncture
in infants, can lead to transfixion of the vein, venous
haematoma, or failed puncture.4 Nevertheless, even a
small difference in terms of the size can mean the difference
between failed and successful catheter placement in these
small veins.
Some clinicians favour the femoral vein for central venous
lines. Potentially, life-threatening complications upon cannulation virtually do not occur.10 However, as with the IJV, distortion and compression of the femoral vein during the
needle approach represent a significant problem mainly in
neonates. Furthermore, femoral venous catheters can
obstruct easily by kinking in mobile infants and the catheterrelated infection rate might also be higher compared with
the IJV and SCV.11
One minor drawback of our investigation was that the CSA
of the SCV was not measured exactly at the same place
where the diameter of the longitudinal view and the depth
of the SCV were determined (Fig. 2). In fact, it could have
been the proximal end of the axillary vein because it was
not possible to obtain a good cross-sectional view of the
SCV by placing the ultrasound probe at the supraclavicular
area. On the other hand, the two positions where our
measurements were carried out are very closely located to
each other and there are no other large veins draining into
the SCV which could have had an impact on the size of the
vein. It must also be mentioned that the CSA of the IJV
and the SCV could not be measured in nine patients due to
technical reasons.
Further studies comparing the success rates of actual subclavian and internal jugular line placement could clarify
which approach to central venous line placement is more
effective.
In conclusion, our study shows a statistically significant
larger IJV than SCV in neonates. There was a positive correlation between the size of the infant and that of the veins,
meaning that the bigger the baby, the larger both the IJV
and the SCV. These were the results of our ultrasound investigation. Although the results seem to support the recommendation of using the IJV as the first choice for the
placement of a CVC, other factors such as interindividual
variability in terms of the size of the veins, significant distortion of the IJV during cannulation attempts, and the higher
catheter-related infection rate of internal jugular venous
lines compared with subclavian venous lines in surgical neonates must be considered as well.12
Conflict of interest
None declared.
References
1 Nakayama S, Yamashita M, Osaka Y et al. Right internal jugular
vein venography in infants and children. Anesth Analg 2001; 93:
331– 4
2 Tripathi M, Pandey M. Modified anchoring maneuver using pilot
puncture needle to facilitate internal jugular vein puncture for
small children. Paediatr Anaesth 2008; 18: 1050–4
3 Cobb LM, Vincour CD, Wagner LW et al. The central venous
anatomy in infants. Surg Gynecol Obstet 1987; 165: 230–4
4 Pirotte T. Ultrasound-guided vascular access in adults and children: beyond the Internal Jugular Vein puncture. Acta Anaesth
Belg 2008; 59: 157– 66
5 Pirotte T, Veyckemans F. Ultrasound-guided subclavian vein cannulation in infants and children: a novel approach. Br J Anaesth
2007; 98: 509– 14
6 Chan VWS, Abbas S, Brull R et al. The infraclavicular region. In:
Chan VWS, ed. Ultrasound Imaging for Regional Anaesthesia,
2nd Edn. Toronto Printing Company Inc. 2009; 70–82
7 Murat MS, Arzu M, Ozge K et al. Internal jugular vein diameter in
pediatric patients: are the J-shaped guidewire diameters bigger
than internal jugular vein? An evaluation with ultrasound. Paediatr Anaesth 2008; 18: 745–51
8 Botero M, White SE, Younginer JG et al. Effects of Trendelenburg
position and positive intrathoracic pressure on internal jugular
183
BJA
vein cross-sectional area in anaesthetized children. J Clin Anesth
2001; 13: 90– 3
9 Karazincir S, Akoglu E, Balci A et al. Dimensions of internal jugular
veins in Turkish children aged between 0 and 6 years in resting
state and Valsalva maneuver. Int J Pediatr Otorhinolaryngol
2007; 71: 1247– 50
10 Karapinar B, Cura A. Complications of central venous catheterization in critically ill children. Pediatr Int 2007; 5: 593–9
184
Breschan et al.
11 Alumneef MA, Memish ZA, Balkhy HH et al. Rate, risk factors
and outcomes of catheter-related bloodstream infection in a
paediatric care unit in Saudi Arabia. J Hosp Infect 2006; 2:
207– 13
12 Breschan C, Platzer M, Jost R et al. Comparison of catheter-related
infection and tip colonization between internal jugular and subclavian venous catheters in surgical neonates. Anesthesiology
2007; 107: 946– 53