Point of Care Ultrasound For Basic Haemodynamic

Anaesthesia, 2006, 61, pages 849855
doi:10.1111/j.1365-2044.2006.04746.x
.....................................................................................................................................................................................................................
Point of care ultrasound for basic haemodynamic

assessment: novice compared with an expert operator
C. F. Royse,1 J. L. Seah,2 L. Donelan3 and A. G. Royse1
1 Associate Professors, 2 BMedSci, 3 DMU AMS, Cardiovascular Therapeutics Unit, Department of Pharmacology,
University of Melbourne, Carlton, Victoria, Australia 3010
Summary
Miniaturisation of ultrasound equipment has led to the development of hand-held echocardiography devices suitable for bedside evaluation of cardiac function. Basic assessment of the haemodynamic state can be performed using a limited transthoracic echocardiography examination. This
study evaluated a third generation device (SonoSite TitanTM) used by novice and expert operators.
Limited transthoracic examination was performed on 30 healthy volunteers by an expert and a
novice operator. The novice had performed 10 studies prior to data accrual. Agreement analysis
was performed using weighted least products regression and Bland-Altman analysis. Acceptable
results for the novice were achieved following 20 studies (including practice sessions) for basic
haemodynamic assessment and following 40 studies for all measured parameters. The SonoSite
Titan is acceptable for basic transthoracic measurements to determine the basic haemodynamic state
and cardiac output measurements. We recommend a minimum of 20 training studies for novice
operators prior to clinical use.
. ......................................................................................................
Correspondence to: A Prof. Colin Royse

E-mail: colin.royse@mh.org.au
Accepted: 18 May 2006
Point of Care ultrasound imaging (POC) is the use of

ultrasound at the patients bedside to facilitate procedures
or cardiovascular diagnosis. The use of this modality is
increasing rapidly in peri-operative medicine for haemodynamic assessment, vascular access, and ultrasound
guided nerve blockade. Although any type of ultrasound
machine can be used, hand-held cardiac ultrasound
devices have contributed to the increased use of ultrasound in peri-operative medicine because of their portable nature and relatively low-cost compared with
conventional trolley based ultrasound machines [1].
The technological development of hand-held ultrasound
devices is commensurate with the rapid growth in
miniaturisation and computer processing power. Firstgeneration devices were limited to two-dimensional
imaging; and second-generation devices incorporated
features such as Doppler imaging including colour flow.
Third generation devices added a wider array of probes
allowing abdominal, vascular, and transthoracic imaging
through the use of tightly curved linear array technology.
These devices also include the facility for on-line
measurement of data. Recently, fourth generation devices
have been produced by several manufacturers which now
2006 The Authors
Journal compilation 2006 The Association of Anaesthetists of Great Britain and Ireland
incorporate phased array transducer technology, thereby

allowing more comprehensive transthoracic imaging as
well as transoesophageal echocardiography capability.
This latest generation of hand-held devices is essentially
a fully capable echocardiography machine reduced to the
size of a large laptop computer. With earlier generation
hand-held cardiac ultrasound, the extent of cardiac
evaluation was limited by technological inadequacies,
whereas with third or fourth generation devices, the type
of cardiac evaluation (whether complete or focused) is
determined by the needs of the operator and the clinical
situation rather than by the device.
Ultrasound guided haemodynamic assessment is a
rapidly emerging use of echocardiography in peri-operative practice. Conventional pressure-based or flow-based
haemodynamic assessment (e.g. blood pressure, pulmonary artery pressure, cardiac output) is well suited to the
assessment of trends in haemodynamic change, but poorly
suited to diagnosing the cause of haemodynamic abnormality. No trial has shown any outcome advantage
associated with the use of pulmonary artery catheters in
the care of critically ill and it has been suggested that the
outcome may be worsened by their use [2].
849
C. F. Royse et al.
Point of care ultrasound
. ....................................................................................................................................................................................................................
Table 1 Ultrasound guided haemodynamic state assessment.

LV Volume
LV Systolic function
LV Filling pressure
Basic state
N
N
N
NORMAL
EMPTY
N
N
PRIMARY
DIASTOLIC
FAILURE
N
SYSTOLIC
FAILURE
SYSTOLIC and
DIASTOLIC
FAILURE
N
VASODILATION
RV
RV
RIGHT
VENTRICULAR
FAILURE
LV is left ventricle, RV is right ventricle, arrows to know to be increased or decreased.
Basic haemodynamic state can be defined as the

interaction between myocardial and vascular function at
any point in time. Patterns such as left ventricular systolic
failure, hypovolaemia, primary diastolic failure or vasodilation are examples of haemodynamic states. Echocardiography can be used in isolation or in conjunction with
pressure-based information to identify the haemodynamic state, with the triad of ventricular volume, systolic
function and left ventricular filling pressure assessments. A
summary of haemodynamic states and ultrasound-guided
assessment is shown in Table 1.
Our aims were to evaluate the third generation handheld cardiac ultrasound, Sonosite Titan (Sonosite,
Bothell, WA), and to investigate the training requirements for a novice operator, for basic haemodynamic state
evaluation.
Methods
The study was approved by the University of Melbourne

Human Ethics Committee and written informed consent
was obtained from 30 normal volunteers. All subjects
received transthoracic echocardiography examination by
an expert using the SonoSite Titan and were then
subsequently imaged by the novice operator. The study
was conducted in the Human Laboratory of the Cardiovascular Therapeutics Unit, Department of Pharmacology, University of Melbourne.
Volunteers were recruited from a normal healthy
population in a university setting. A study sample size
of 30 was determined based on previous studies from our
research group [3, 4], aiming to detect a 20% difference
between measurements such as cardiac output with a
power of 0.8 and at p < 0.05.
There was a training period of 4 weeks for the novice
operator by the expert echocardiographer before the start
of the study. In that period, there were five supervised
examination sessions and five unsupervised practice
sessions. Supervised training sessions involved the expert
echocardiographer demonstrating the use of the equipment and the skill involved in obtaining standard
transthoracic views relevant to measurements obtained
in the study. The expert operator was an echocardio850
graphy technologist with > 20 years of clinical experience. The novice was a third-year medical student
engaged in an Advanced Medical Science year of
research, and had no prior exposure to echocardiography.
Study conduct
Measurements were obtained with the subject lying down
in the left lateral position. An electrocardiogram (ECG)
and automated non-invasive blood pressure were studied.
The novice performed the study, followed by the expert.
All measurements were performed on-line as there was no
capacity to perform measurements off-line. To ensure
blinding, the novice would complete the study and then
call the expert in to conduct the second examination.
Examination sequence
1 Subject demographics included sex, age, baseline blood
pressure, height, weight, and calculated body surface
area [5].
2 Parasternal long axis view (PLAX). M-Mode measurements included left atrial size, interventricular septal
wall posterior wall thickness (at end-diastole), left
ventricular end-diastolic dimension and left ventricular
end-systolic dimension. Caliper measurements from
two-dimensional (2D) images were performed to obtain
the left ventricular outflow tract (LVOT) dimension.
3 Parasternal short axis view (PSAX). 2D measurement
of right ventricular outflow tract diameter (RVOT),
and pulsed wave Doppler velocity-time integral (VTI),
measured just proximal to the pulmonary valve.
4 Apical 4-chamber view (Ap4C). The 2D shape and
movement of the interatrial septum was used to
determine left atrial pressure state. Royse et al. [6]
identified three patterns of interatrial septal movement
which correlated with either high, low, or normal
pulmonary capillary wedge pressure (PCWP). In brief,
if the interatrial septum did not move and the curvature was fixed, bowing towards the right atrium
(fixed curvature), this was associated with PCWP
> 15 mmHg, if the direction of the interatrial septum
reversed during mid-systole with each cardiac cycle so
that the curvature was now bowing towards the left
atrium (mid-systolic reversal) this was associated with
2006 The Authors

C. F. Royse et al.
. ....................................................................................................................................................................................................................
PCWP 1015 mmHg, and if the motion of the

interatrial septum was marked and the appearance of
the septum seemed to be folded or concertinaed (systolic
buckling) this was associated with PCWP < 5 mmHg.
5 Apical 5-chamber view (Ap5C). The VTI for the left
ventricular outflow tract was measured by positioning
the pulsed wave (PW) Doppler sample volume within
0.5 cm of the aortic valve on the ventricular side [7].
6 Apical long axis view (ApLAX). The PW Doppler
sample volume was placed within 0.5 cm on the
ventricular side of aortic valve and VTI recorded.
7 Cardiac output was calculated as cross-sectional area
VTI heart rate from data obtained from each of the
PSAX, Ap5C and ApLAX views, and LVOT diameter
from the PLAX view.
Statistical analysis
Demographic data are presented as mean (SEM). Differences across study sequences were evaluated using twoway ANOVA. Agreement analysis was performed in two
stages. First, weighted least products regression analysis
was performed and the 95% confidence intervals (CI) for
the y-intercept (a) and the slope (b) were determined. If
the 95% CI for a included the value 0, there is no fixed
bias, and for b, if the 95% CI included the value 1, then
there is no proportional bias [8]. Agreement is defined by
the absence of bias. Second, the paired differences were
plotted against the paired means using the method of

Bland and Altman [9]. The mean bias and the limits of
agreement (LAgr) were calculated. This allows the reader
to better judge how close the agreement was. The LAgr
percentage refers to 2 standard deviations (SD) of the
difference as a percentage of the average value of the
mean of the differences. Weighted least products regression analysis was performed using Statistical Package for
Social Sciences 13.0 (SPSS, Chicago, IL) and the BlandAltman analysis with GRAPHPAD PRISM 4.0 (GraphPad
Software, San Diego, CA).
Results
Thirty normal healthy volunteers (17 male) participated

in this study. The ages ranged from 18 to 45 years (22.8
(6.1 years)). The body surface area was 1.9 (0.3) m2.
Heart rate and mean arterial blood pressure (MAP) did
not change significantly over the study sequences (heart
rate: 64.5 (9.8), 65.1 (11.5) beats.min)1, p > 0.05; MAP:
73.9 (8.9), 75.3 (9.7), p > 0.05). All left atrial pressure
estimations were normal (mid-systolic reversal) with
agreement between novice and expert on all occasions.
Comparison between novice and expert operator
Agreement analysis data are shown in Table 2, and the
example of left ventricular end-diastolic dimension is
Table 2 Summary of agreement: expert echocardiographer vs novice using the Sonosite Titan.
Weighted Least Products Regression
Variable
95% CI
for a
Correlation
Bland-Altman analysis
95% CI
for b
Fixed
bias
Prop
bias
Agreement
Mean
bias
LAgr
LAgr%
Fixed dimensions; cm
Aortic root diameter
Left atrial size
IVS diameter
PW thickness
LVOT diameter
RVOT diameter
)0.18
)0.30
)0.06
)0.22
0.05
0.21
1.02
0.25
1.06
1.41
0.93
0.85
)0.520.16
)0.800.20
)0.310.19
)0.580.14
)0.280.38
)0.260.68
0.911.13
0.921.23
0.751.37
0.911.90
0.801.06
0.611.09
None
None
None
None
None
None
None
None
None
None
None
None
Yes
Yes
Yes
Yes
Yes
Yes
0.95
0.84
0.65
0.52
0.89
0.70
<0.001
<0.001
<0.001
0.003
<0.001
<0.001
)0.11
)0.02
0.002
0.12
)0.11
)0.11
0.11
0.73
0.28
0.46
0.16
0.47
7.61
20.61
30.77
39.17
12.10
26.55
Dynamic dimensions; cm
LVESD
LVEDD
0.30
)1.34
0.95
1.27
)0.180.77
)2.31 to )0.36
0.811.10
1.071.46
None
Yes
None
Yes
Yes
No
0.86
0.90
<0.001
<0.001
0.15
0.004
0.62
0.55
15.33
10.99
Haemodynamic measurements; cm
RVOT VTI
)3.93
LVOT VTI from Ap5Ch
)11.2
LVOT VTI from ApLAX
)8.90
1.22
1.67
1.65
)9.892.03
)17.6 to )4.85
)13.75 to )4.05
0.861.57
1.262.08
1.321.99
None
Yes
Yes
None
Yes
Yes
Yes
No
No
0.61
0.56
0.79
<0.001
0.001
<0.001
)0.51
0.32
0.18
4.02
6.16
6.46
29.42
33.93
35.50
Calculated haemodynamic measurements; l.min)1

CO for RVOT
)0.01 0.88
)0.990.98
CO for LVOT from Ap5Ch
)2.34 1.51
)4.15 to )0.53
CO for LVOT from ApLAX
)1.25 1.31
)2.960.46
0.591.17
1.061.97
0.851.77
None
Yes
None
None
Yes
None
Yes
No
Yes
0.62
0.75
0.77
<0.001
<0.001
<0.001
)0.55
)0.11
0.18
2.25
1.72
2.11
73.20
43.05
40.95
a, intercept of weighted least products regression line; b, slope of weighted least products regression line; 95% CI; 95% confidence interval; Prop
bias; proportional bias; r, Pearsons correlation coefficient; p, significance for Pearsons correlation coefficient; LAgr, limits of agreement; LAgr%,
limits of agreement in percentage; IVS diameter, interventricular septal diameter; PW thickness, posterior wall thickness; LVOT, left ventricular
outflow tract; RVOT, right ventricular outflow tract; LVESD, left ventricular end systolic dimension; LVEDD, left ventricular end diastolic dimension;
VTI, velocity time integral; Ap5Ch, apical 5 Chamber view; ApLAX, apical long axis view; CO, cardiac output.
2006 The Authors
851
C. F. Royse et al.
. ....................................................................................................................................................................................................................
Table 3 Summary comparing first, second and third groups of
ten studies for all variables using weighted least products

regression.
Variable
First 10
Second 10
Third 10
studies
studies
studies
(Agreement) (Agreement) (Agreement)
Fixed dimensions; cm
Aortic root diameter
Left atrial size
IVS diameter
PW thickness
LVOT diameter
RVOT diameter
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Dynamic dimensions; cm
LVESD
LVEDD
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Haemodynamic measurements; cm
RVOT VTI
No
LVOT VTI from Ap5Ch
Yes
LVOT VTI from ApLAX
*
Calculated haemodynamic
CO for RVOT
CO for LVOT from Ap5Ch
CO for LVOT from ApLAX
Figure 1 Comparison between novice and expert operators
using Sonosite Titan weighted least products regression and

Bland-Altman plot for the measurement: left ventricular enddiastolic dimension. Dotted lines indicate the bias and limits of
agreement (2 SD of difference).
measurements; l.min)1
Yes
No
Yes
Yes
*
Yes
Yes
Yes
Yes
a, intercept of weighted least products regression line; b, slope of

weighted least products regression line; 95% CI; 95% confidence
interval; Prop bias; proportional bias; r, Pearsons correlation coefficient; p, significance for Pearsons correlation coefficient; LAgr, limits
of agreement; LAgr%, limits of agreement in percentage; IVS diameter, interventricular septal diameter; PW thickness, posterior wall
thickness; LVOT, left ventricular outflow tract; RVOT, right ventricular
outflow tract; LVESD, left ventricular end systolic dimension; LVEDD,
left ventricular end diastolic dimension; VTI, velocity time integral;
Ap5Ch, apical 5 Chamber view; ApLAX, apical long axis view; CO,
cardiac output.
*Measurement of VTI from ApLAX was only added after the first 10
studies.
Discussion
shown in Fig. 1. Fixed and proportional bias was

identified for left ventricular end-diastolic dimension,
and VTI measurements from Ap5C and ApLAX, as well
as cardiac output calculation from Ap5C, though the
magnitude of the mean bias and limits of agreement for
this comparison were similar to that for the expert operator analysis described above.
Progression of novice compared with expert
operator
A summary of agreement analysis using weighted least
products regression between novice and expert operators
is shown for the first, second, and third series of 10 studies
in Table 3. The example of left ventricular end-diastolic
dimension is shown in Fig. 2.
852
In this study, we found that a novice operator could reach

agreement with an expert operator in as few as 20 studies
(including practice) for basic haemodynamic state measurements and 40 studies for all measurements performed.
This study is one of the first to validate a third-generation
portable machine for basic haemodynamic state assessment and is supported by previous research validating
second generation portable echocardiography machines
[1012].
Errors in image analysis were more likely with small
structures (such as the interventricular septum) than
large structures. Alexander et al. [13] assessed the
interobserver variability between expert echocardiographers interpreting the same images and found moderate
to good agreement (Kappa: 0.530.68). This could
mainly be due to the qualitative and subjective nature
2006 The Authors

C. F. Royse et al.
. ....................................................................................................................................................................................................................
Figure 2 Comparison between novice and expert operators using Sonosite Titan for the first, second and third groups of 10 studies
weighted least products regression and Bland-Altman plot for the measurement: left ventricular end-diastolic dimension. Dotted lines
indicate the bias and limits of agreement (2 SD of difference).
of the variables they assessed, whereas in our study

quantitative measurements were performed and close
agreement was found. Galasko et al. [14] and Rugolotto
et al. [10] found an interobserver variability of only 2%
and 10.7%, respectively, between expert echocardiographers when they assessed semiquantitative variables.
In interobserver studies comparing novice against expert
operator, Wong et al. [15] found a variability of less
than 10% in measuring cardiac output, which supports
the findings of this study. Blinded intra-observer
variability could not be performed on the SonoSite
Titan because it did not have the capacity for offline
analysis as configured for this study. The primary design
features of robust and small equipment limit the image
storage available on a compact flash card system. After
this study was performed, software updates allow images
to be downloaded to a DICOM server and offline
measurements performed.
As the study progressed, we found a general trend of
increasing agreement between the novice and expert,
indicating a learning curve for the novice. There were a
few variables that did not reach agreement in the first and
second 10 studies, however, by the third 10 studies, all
variables were in agreement. As the novice had performed
10 training studies prior to data accrual, we found that an
2006 The Authors
acceptable result for the basic haemodynamic assessment

was obtained after 20 studies (including practice sessions),
and for all measurements after 40 studies. The implication
for the use of this technology is that the novice should
perform at least 20 studies prior to applying it in clinical
situations. Previous studies have attempted to train novices; however, this was done mainly for assessment of
qualitative or semiquantitative measurements. For example, novices were trained to assess four diagnoses of LV
dysfunction, mitral regurgitation, aortic valve thickening
and pericardial effusion [13], assess global LV function
[16] and determine abnormal wall motion presence [17].
In these assessments, investigators found the learning
curve to be short probably because the measurements
made were qualitative; whereas in our study, a longer
learning curve was encountered as the technical skill
required to acquire images and make quantitative measurements is more demanding. As echocardiography is a
very operator-dependent modality; the aptitude of novices may play an important role in skill acquisition. For
example, Lemola et al. [18] cited the investigators high
motivation as a factor contributing to the positive results
obtained.
With the introduction of third-generation hand-held
cardiac ultrasound, the equipment is no longer the major
853
C. F. Royse et al.
. ....................................................................................................................................................................................................................
determinant of whether an echocardiographic study is

acceptable or not. Although we have found minor
limitations in the two devices, both machines are
capable of acquiring acceptable transthoracic echocardiographic images. Technological developments with
hand-held cardiac ultrasound devices have been rapid,
with fourth-generation devices (such as the Sonosite
MicroMaxxTM; Sonosite) now available. These incorporate phased array technology allowing the same
imaging modalities as conventional trolley-based echocardiography machines, including incorporation of
transesophageal echocardiography probes. Limitations
on echocardiography studies, are now predominantly
dependent on patient factors, the study conditions and
operator expertise.
The study has several important limitations. Our study
population comprised young healthy volunteers and is not
representative of sick patients in the peri-operative or
critical care environments. Measurements were taken at
rest and there were no changes in the haemodynamic state
over the study sequence. Further studies could potentially
investigate the use of this machine in clinical practice in
wards or operating theatres as a haemodynamic monitoring tool. We also only tested one novice rather than a
number of novice operators. Our recommendation for
training could alter depending on the aptitude of the
individual operator. It was not possible to obtain data
simultaneously for both operators and there could be
variation in measurements (especially VTI) at different
time periods. We attempted to minimise this effect by
performing the studies under the same conditions and
posture. Measurements were performed on-line because
the echocardiography machine was not capable of off-line
analysis. The subsequent generation (Sonosite MicroMAXX) has the capability of exporting images in
DICOM format suitable for off-line analysis. We were
therefore unable to perform an interobserver analysis of
common images.
Conclusion. The SonoSite Titan is acceptable for basic
haemodynamic measurements to determine the haemodynamic state. We recommend a minimum of 20 training
studies for novice operators prior to they use the device in
clinical practice.
References
The Sonosite Titan was generously loaned by Sonosite

Australia. An unrestricted educational grant was provided
by Sonosite to assist the conduct of the study. We thank
Dr John Ludbrook, MD, DSc, AStat (Biomedical Statistical Consulting Pty. Ltd, Melbourne, Victoria, Australia)
for assisting with statistical analysis and for manuscript
review.
1 Vourvouri EC, Koroleva LY, Ten Cate FJ, et al. Clinical

utility and cost effectiveness of a personal ultrasound
imager for cardiac evaluation during consultation rounds
in patients with suspected cardiac disease. Heart 2003; 89:
72730.
2 Connors AF Jr, Speroff T, Dawson NV, et al. The effectiveness of right heart catheterization in the initial care of
critically ill patients. SUPPORT Investigators. JAMA 1996;
276: 88997.
3 Royse CF, Royse AG, Wong CT. Assessment of left ventricular function during off-pump coronary artery bypass
surgery. Annals of Thoracic and Cardiovascular Surgery 2003; 9:
3717.
4 Royse CF, Royse AG, Wong CT, Soeding PF. The effect
of pericardial restraint, atrial pacing, and increased heart rate
on left ventricular systolic and diastolic function in patients
undergoing cardiac surgery. Anesthesia and Analgesia 2003;
96: 12749.
5 Green B, Duffull SB. What is the best size descriptor to use
for pharmacokinetic studies in the obese? British Journal of
Clinical Pharmacology 2004; 58: 11933.
6 Royse CF, Royse AG, Soeding PF, Blake DW. Shape and
movement of the interatrial septum predicts change in pulmonary capillary wedge pressure. Annals of Thoracic and
Cardiovascular Surgery 2001; 7: 7983.
7 Royse AG, Royse CF, Shah P, et al. Radial artery harvest
technique, use and functional outcome. European Journal of
Cardio-Thoracic Surgery 1999; 15: 18693.
8 Ludbrook J. Comparing methods of measurements. Clinical
and Experimental Pharmacology and Physiology 1997; 24: 193
203.
9 Bland JM, Altman DG. Measuring agreement in method
comparison studies. Statistical Methods in Medical Research
1999; 8: 13560.
10 Rugolotto M, Hu BS, Liang DH, Schnittger I. Rapid
assessment of cardiac anatomy and function with a new
hand-carried ultrasound device (OptiGo): a comparison
with standard echocardiography. European Journal of Echocardiography 2001; 2: 2629.
11 Kawai J, Tanabe K, Yagi T, et al. Assessment of the clinical feasibility of OptiGo for hand-held echocardiography.
Journal of Cardiology 2003; 41: 819.
12 Vourvouri EC, Poldermans D, De Sutter J et al. Experience
with an ultrasound stethoscope. Journal of the American Society
of Echocardiography 2002; 15: 805.
13 Alexander J, Peterson E, Chen A, et al. Feasibility of pointof-care echocardiography by internal medicine house staff.
American Heart Journal 2004; 147: 47681.
14 Galasko GI, Lahiri A, Senior R. Portable echocardiography:
an innovative tool in screening for cardiac abnormalities in
the community. European Journal of Echocardiography 2003; 4:
11927.
15 Wong DH, Onishi R, Tremper KK, et al. Thoracic bioimpedance and Doppler cardiac output measurement:
learning curve and interobserver reproducibility. Critical
Care Medicine 1989; 17: 11948.
854
2006 The Authors

Acknowledgements

C. F. Royse et al.
. ....................................................................................................................................................................................................................
16 Hu BS, Saltiel F, Popp RL. Effectiveness of limited training

in echocardiography for cardiovascular diagnosis. Circulation
1996; 94 (Suppl.): I253.
17 Kimura BJ, Amundson SA, Willis CL, et al. Usefulness of a
hand-held ultrasound device for bedside examination of left
ventricular function. American Journal of Cardiology 2002; 90:
10389.
2006 The Authors

18 Lemola K, Yamada E, Jagasia D, Kerber RE. A hand-carried

personal ultrasound device for rapid evaluation of left ventricular function: use after limited echo training. Echocardiography 2003; 20: 30912.
19 Bom N, van der Steen AF, de Jong N, Roelandt JR. Early,
recent and future applications of echocardiography. Clinical
Physiology and Functional Imaging 2004; 24: 1416.
855

Point of Care Ultrasound For Basic Haemodynamic

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Point of Care Ultrasound For Basic Haemodynamic

Caricato da

Copyright:

Formati disponibili

Anaesthesia, 2006, 61, pages 849855