Vous êtes sur la page 1sur 8

British Journal of Anaesthesia 108 (1): 1007 (2012)

Advance Access publication 20 October 2011 . doi:10.1093/bja/aer336


Intraoperative use of transoesophageal Doppler to predict

response to volume expansion in infants and neonates
O. Raux *, A. Spencer, R. Fesseau, G. Mercier, A. Rochette, S. Bringuier, K. Lakhal, X. Capdevila and
C. Dadure
Department of Anaesthesia and Critical Care Medicine, Montpellier University Hospital, Montpellier, France
* Corresponding author. E-mail: o-raux@chu-montpellier.fr

Editors key points

TED derived stroke

volume accurately
predicts responsiveness
to VE.
VE based on standard
monitoring data is
inappropriate in more
than 50% of patients.
TED monitoring can avoid
undue fluid loading.

Methods. Neonates and infants under general anaesthesia without myocardial dysfunction
were prospectively included when the attending anaesthetist, blinded to TED
measurements, decided to provide VE based on clinical appreciation and standard
monitoring data. Standard and TED-derived data were recorded before and after VE. After
VE, patients were classified as responders and non-responders, if their indexed stroke
volume (iSV) increased by more than 15% or not, respectively. The attending anaesthetist
assessment of VE responsiveness was recorded at the end of VE.
Results. Fifty patients aged 42 (4) post-conceptional weeks were included, among which 26
(52%) were responders. Baseline iSV was the only parameter associated with VE
responsiveness. Baseline iSV was fairly correlated with VE-induced changes in iSV (r 20.64)
and was associated with an area under the receiver operating characteristic curve of 0.90
(0.80, 0.99). Using a cut-off of 25 ml m22, baseline iSV predicted volume responsiveness with
a sensitivity of 92% and a specificity of 83%. Attending anaesthetists assessment of VE
effectiveness agreed only moderately with TED measurements of iSV changes.
Conclusions. TED-derived iSV measurement during volatile anaesthesia is useful to predict and
follow VE responsiveness in neonates and infants without myocardial dysfunction.
Keywords: Doppler ultrasonography, transoesophageal; infants; monitoring, intraoperative;
Accepted for publication: 16 August 2011

The assessment of perioperative hypovolaemia and the

trigger for volume expansion (VE) in paediatric anaesthesia
are based on the integration of multiple variables. Clinical
endpoints such as arterial pressure, heart rate (HR), pulse
pressure, and urine output play an important role in appraising a patients volume status. The type of intervention
(e.g. bowel surgery), current or predicted surgical events
(e.g. bleeding), and laboratory findings (e.g. haematocrit,
lactate) are also considered before VE. The goal of VE is an
attempt to increase stroke volume (SV) and, consequently,
cardiac output (CO). VE assumes that the patient is on the
ascending portion of the Frank Starling curve and has recruitable CO. Once the left ventricle (LV) is functioning near the
flat part of the FrankStarling curve, fluid loading has little
effect on CO.1 In a paediatric context, VE may in fact be

harmful as the immaturity of neonatal diastolic cardiac function increases the risk of congestive pulmonary oedema in
cases of undue fluid loading.2 As clinical appreciation of
the volaemic status is known to be unreliable in ventilated
children,3 minimally invasive tools that could predict
patient responsiveness to VE would be extremely useful. A
non-invasive, user-friendly monitor allowing for the
measurement of CO and effective management of haemodynamic instability in neonates and infants would be a
major advance in practice. Transoesophageal Doppler (TED)
has been shown to effectively measure CO in newborns
and children4 and, therefore, could allow for the assessment
of the efficacy of VE. Moreover, among many velocity TED
waveform-derived indices, corrected flow time (cFT) has
been shown to reflect LV filling.5 The aim of the study was

& The Author [2011]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved.
For Permissions, please email: journals.permissions@oup.com

Downloaded from http://bja.oxfordjournals.org/ at Semmelweis University on August 5, 2012

Doppler (TED) is useful to
guide intraoperative
volume expansion (VE) in
neonates and young

Background. Volume expansion (VE) in neonates or infants during volatile anaesthesia may
lead to fluid overload if inappropriate. Transoesophageal Doppler (TED), a non-invasive
cardiac output monitoring technique, can provide a comprehensive estimation of the
volaemic status. We evaluated whether intraoperative TED-derived parameters can
predict volume responsiveness.


Intraoperative use of TED to predict response to VE

to determine the utility of TED to guide VE in neonates and

young infants without myocardial dysfunction undergoing
general tracheal anaesthesia with mechanical ventilation.
The primary objective was to prospectively evaluate the performance of standard clinical monitoring parameters and
TED-derived parameters including CO, indexed SV (iSV),
peak velocity (PV), and mean acceleration (MA) to predict
volume responsiveness. The secondary objective was to
compare the paediatric anaesthetists opinion of the responsiveness to VE based on standard clinical monitoring endpoints with those changes measured by TED parameters.


Statistical analysis
Patients were initially classified as responders (R) if their iSV
increased by 15% after VE. Patients with ,15% change
in iSV were classified as non-responders (NR). Results are
expressed as mean (SD) or median (10th, 90th percentiles),
where appropriate. R/NR groups and baseline/post-VE
measurements were compared using Students, Wilcoxons,
x 2, or Fischers exact test, as appropriate. Covariables significantly associated with VE responsiveness in the univariate
analysis (P,0.05) were entered in a logistic regression analysis to explore the relative importance of each covariable on
VE responsiveness [step-by-step procedure based on the
a-to-exit0.20)]. Interactions between covariables were
assessed if relevant. Adjusted odds ratios and 95% confidence intervals were calculated. Receiver operating characteristic (ROC) curves were fitted for variables significantly
associated with VE responsiveness in the multivariate analysis. The area under the ROC curve was calculated using the
trapezoid rule. For each variable, a cut-off value was determined to maximize both sensitivity and specificity. Linear
correlations were searched between variables associated
with VE responsiveness and VE-induced iSV changes, using
the Spearman rank method. k coefficient was calculated to
assess the concordance between R/NR classification and
anaesthetist perceived responsiveness to VE. Statistical significance threshold was set at 5%. Statistical analysis was
performed using SASTM version 9.0 (SAS InstituteTM , Cary,

Results are presented in Tables 1 and 2 for both R and NR at
baseline and post-VE and are expressed as mean (SD) or
median (10th, 90th percentiles), where appropriate. Fifty-six
patients were felt to be hypovolaemic and required VE. Six
patients were excluded from the study due to inability to


Downloaded from http://bja.oxfordjournals.org/ at Semmelweis University on August 5, 2012

This prospective study was conducted at the Centre Hospitalier Universitaire de Montpellier after Research Ethics Board
approval (Comite de Protection des Personnes SudMediterranee III, Ref: 2009.12.01 bis) and the study was
registered with EudraCT (Ref: 2009-A01185-52). Informed
consent from patients caregivers was obtained before
inclusion in the study. Neonates and infants weighing
between 2.5 and 5.0 kg, aged ,6 months, undergoing a surgical procedure under volatile anaesthesia requiring tracheal
intubation and controlled ventilation were consented for the
possible entry into the study. Exclusion criteria were preoperative haemodynamic instability or catecholamine infusion, known congenital heart disease with haemodynamic
consequences, hypothermia (temperature ,358C), and oesophageal malformation. Patients were optimized before operation and deemed haemodynamically stable and clinically
euvolaemic before the induction of anaesthesia. Anaesthesia
was maintained using a volatile anaesthetic, but otherwise
was not strictly standardized. Ventilation was set to maintain
end-tidal CO2 between 25 and 35 mm Hg. Arterial pressure
was measured using a standard non-invasive cuff applied
to the upper limb. Patients were formally entered into the
study if the attending paediatric anaesthetist felt that it
was necessary to provide a fluid bolus during the maintenance of anaesthesia. The need for VE was based on traditional endpoints such as clinical history, type of surgery,
perioperative events, and standard monitoring data excluding TED measurements. Upon selection into the study, a
second anaesthesiologist was called to perform TED
measurements. TED-derived data and standard monitoring
data were registered immediately before and after VE. The
second anaesthetist inserted a 5 mm diameter, 4 MHz, flexible TED probe (Cardio QPTM, Deltex MedicalTM , Chichester, UK)
through the mouth of the patient in an attempt to place its
tip in the mid-oesophagus. The attending anaesthetist was
blinded to TED waveform- or TED-derived data. The optimal
position of the probe was suggested by an audible,
maximal pitch and a sharply defined velocity waveform
with minimal spectral dispersion. Five consecutive TED
measurements pre- and post-VE were completed and averaged. The probe was rotated to display the best aortic
blood flow signal before each measurement. Cardiac index
(CI), iSV, cFT (length of time of systolic blood flow adjusted

for HR, that is, divided by the square root of the heart cycle
time), PV (maximal velocity during systole), and MA (PV
divided by the acceleration time during systole) of the descending aorta velocity waveform were recorded. VE consisted
of an infusion of 10 30 ml kg21 of crystalloid (normal saline
0.9%) or colloid (albumin 4% or Voluvenw hydroxyethyl starch
130/0.4 6%) over a period of 2040 min. Patients were
excluded if: (i) a correct TED signal was impossible to
obtain, (ii) an anaesthetic or a surgical event with haemodynamic consequences occurred during VE, such as epidural
bolus, surgical incision, surgical bleeding, compression of
great vessels, or pneumoperitoneum inflation, or (iii) there
was concurrent use of a vasoactive agent. Volatile anaesthetic concentration and mechanical ventilation were kept
constant during the VE period. At the end of VE, the attending anaesthesiologist was asked to classify the VE as positive
(improved haemodynamic status) or negative (unchanged or
worsened haemodynamic status). Each patient could only be
included once in the study.


Raux et al.

obtain a correct and stable TED signal (two), hypothermia

,358C (one), a surgical incision during VE (one), epidural
bolus during VE (one), and inflation of pneumoperitoneum
during VE (one). The remaining 50 patients were aged 22
(1, 101) post-natal days, 42 (4) post-conceptional weeks,
and weighed 3.6 (0.9) kg. A TED probe was inserted in each
patient at a mean depth of 14 (2) cm. Patient characteristic
data, surgical procedures, VE characteristics, and respiratory
and haemodynamic parameters are reported in Tables 1 and


R (n526)

NR (n524)


Patient characteristic data

Birth term (weeks)

37 (4)

36 (4)

Post-natal age (days)

11 (1, 63)

45 (1, 126) 0.010

Post-conceptional age

41 (3)

44 (5)


Birth weight (kg)

2.9 (1)

2.6 (0.9)


Current weight (kg)

3.3 (0.8)

3.8 (1)


11.6 (2.7)


Preoperative haemoglobin 14.6 (3.5)

(dg ml21)


Surgical procedures (n)

Chylothorax drainage

Diaphragmatic hernia

Oesogastric junction



Abdominal wall defect

Liver or bile duct surgery

Adrenal or renal surgery

Bowel or colonic surgery

Bladder surgery

Penis surgery

Sacrococcygeal teratoma

Peritoneal catheter

Central venous access


Haemodynamic data presented in Table 2 such as HR, iSV,

cFT, PV, and MA improved after VE in the group of responders.
Table 3 shows that variations in haemodynamic parameters
between pre- and post-VE were statistically different in R and
NR, except for mean arterial pressure (MAP). The anaesthetists assessment of volume responsiveness agreed with
TED responsiveness in 73% of cases (k coefficient0.46,
95% confidence interval: 0.22, 0.71). The anaesthetist correctly noted that 24 of the 26 (92%) R had benefitted from
VE. However, they correctly identified that only 10 of the
24 (42%) NR did not improve after VE.


Anaesthesia induction dosages


Propofol (mg kg

6.5 (2.5)

7.5 (4.5)


Fentanyl (mg kg21)

0.9 (0.2)

1.1 (0.5)


Sufentanil (mg kg21)


0.2 (0.2)




Characteristics of volume expansion

Crystalloids (% of
Colloids (% of patients)
Volume (ml kg21)



As shown in Table 2, before VE, baseline HR, CI, iSV, cFT, and
PV differed significantly between R and NR. In the multivariate model, iSV was found to be the only parameter significantly associated with VE responsiveness [P0.0001,
adjusted odds ratio of 1.39 (1.13, 1.70)]. As seen in
Figure 1, baseline iSV correlated negatively with VE-induced
iSV changes (r 20.64; P,0.0001). Using a cut-off of 25 ml
m22, baseline iSV predicted volume responsiveness with a
sensitivity of 92% and a specificity of 83%. Figure 2 shows
that the area under the curve of iSV was found to be 0.90
(0.80, 0.99).

Clinical judgement of VE effectiveness


Predicting volume responsiveness


16.8 (9.4)

14.4 (3.5)


Time between incision

and VE (min)

53 (48)

40 (25)


Duration of VE (min)

24 (10)

28 (9)


The results of this study demonstrate that (i) TED-derived iSV

can predict responsiveness to VE with a sensitivity of 92%
and a specificity of 83%; (ii) VE based on standard clinical
monitoring data may be inappropriate in as much as 50%
of neonatal and young infants undergoing general volatile
anaesthesia; and (iii) the anaesthetists perceived responsiveness to VE correlated only moderately with TED
measured iSV changes.
It is known that clinical appreciation of the volaemic
status in ventilated children in an intensive care unit (ICU)
setting is unreliable and there are no data to demonstrate
that volaemic status may be better assessed perioperatively.3 In our study, VE based on standard clinical

Downloaded from http://bja.oxfordjournals.org/ at Semmelweis University on August 5, 2012

Table 1 Patient characteristic data, surgical procedures,

anaesthesia induction dosages and characteristics of VE. R and
NR, responders and non-responders to VE; Crystalloids, Ringers
lactate or physiologic serum; Colloids, hydroxyethyl starch or
albumin; VE, volume expansion; results are mean (SD) or median
(10th, 90th percentiles); P, R vs NR

2, according to responsiveness to VE. Of these 50 patients, 26

(52%) were classified as R with an iSV increase of 15% after
VE, whereas 24 (48%) were classified as NR. Anaesthesia
induction consisted of sevoflurane in 50% and 79%
(P0.032) and propofol in 92% and 96% (P0.609) of R
and NR, respectively. Fentanyl, sufentanil, or remifentanil
were used as opioids in 36, six, and two patients, respectively. Anaesthesia was maintained with sevoflurane in 77%
and 62% of R and NR, respectively (P0.231) whereas desflurane was used for maintenance in the remaining patients.
Nitrous oxide was added in 23% and 29% of R and NR,
respectively (P0.560). Preoperative epidural anaesthesia
with 0.6 0.9 ml kg21 ropivacaine 0.2% was combined to
general anaesthesia in 42% and 62% of R and NR, respectively (P0.160).


Intraoperative use of TED to predict response to VE

Table 2 Anaesthetic, respiratory, and haemodynamic parameters. R and NR, responders and non-responders to VE; VE, volume expansion; MAC,
minimum alveolar concentration: 3.3% and 9.16% were considered as MAC of sevoflurane and desflurane, respectively. Twenty per cent
potentiation by N2O was considered when associated; HR, heart rate; MAP, mean arterial pressure; CI, cardiac index; iSV, indexed stroke volume;
cFT, corrected flow time; PV, peak velocity; MA, mean acceleration; results are mean (SD); *P,0.05 baseline vs after VE; P, R vs NR at baseline

R (n526)

NR (n524)


After VE


After VE

Anaesthetic and respiratory parameters

Volatile agent (% MAC)

0.8 (0.3)

0.7 (0.3)*

0.7 (0.2)

0.7 (0.2)


Tidal volume (ml kg21)

9.4 (2.2)

9.3 (2.6)

9.3 (1.6)

8.7 (2.2)


Respiratory frequency (mn21)

28 (4)

28 (4)

27 (3)

27 (3)


3 (1)

3 (1)

2 (1)

2 (1)


Inspiratory peak pressure (cm H2O)

19 (3)

19 (3)

20 (4)

22 (5)


SpO2 (%)

98 (2)

98 (2)

99 (2)

99 (2)


PEEP (cm H2O)

Central temperature (8C)

36 (0.7)*

35.8 (0.8)

35.8 (0.8)


Haemodynamic parameters
HR (min21)

144 (19)

135 (16)*

129 (18)

127 (15)

MAP (mm Hg)

32 (9)

37 (8)*

32 (7)

36 (7)*

CI (litre min21 m22)

2.9 (0.6)

3.8 (0.8)*

3.7 (0.8)

3.6 (0.8)


iSV (ml m22)

20.7 (4.6)

28.4 (7)*

29.2 (6.5)

28.8 (6.3)


cFT (ms)

320 (28)

353 (29)*

354 (32)

368 (25)*


100 (23)

129 (38)*

127 (28)

120 (25)*


MA (cm s22)

12.6 (2.4)

14.7 (3.6)*

13.7 (2.7)

13.1 (2.9)










VE-induced iSV changes (%)

PV (cm s21)








Baseline iSV (ml m2)



Fig 1 Baseline iSV plotted against VE-induced changes in iSV. iSV,

indexed stroke volume; VE, volume expansion. Vertical line,
cut-off values of iSV for optimal sensitivity and specificity25
ml m22; horizontal line, cut-off values of iSV change (15%) distinguishing R and NR patients. r 20.64; P,0.0001.

monitoring data was shown to be inappropriate in close to

50% of the patients classified as NR. Before VE, MAP and HR
were not found to be different between R and NR. Complex
paediatric patients undergoing a balanced anaesthetic and
subjected to various surgical interventions may explain
some of these challenges in the assessment of volume
status. Similar to previous studies, MAP and HR did not
reliably reflect CO in our population of anaesthetized paediatric patients.6 7 However complex, the assessment of
need and responsiveness to VE is important, especially in
neonates where immature diastolic cardiac function could

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Fig 2 ROC curves for baseline iSV to predict responsiveness to VE.

iSV, indexed stroke volume.

rapidly lead to pulmonary oedema.2 Furthermore, anaesthetists had difficulty distinguishing whether a patient had
responded positively or negatively to VE. In 28% of our
patients, the clinicians felt that NR had effectively
responded to VE. Inaccurate appraisal of volume status
after inaugural fluid challenge may expose patients to
further undue fluid loading.
Little research has focused on the assessment of preload
dependence and the paediatric hearts ability to convert an
augmentation of preload into an increase in CO. Static
indices such as central venous access (CVP) and pulmonary
arterial occlusive pressure (PAOP) are considered to be the


Downloaded from http://bja.oxfordjournals.org/ at Semmelweis University on August 5, 2012

36.1 (0.7)


Raux et al.

Table 3 Variations of haemodynamic parameters during VE. R

and NR, responders and non-responders to volume expansion; HR,
heart rate; MAP, mean arterial pressure; CI, cardiac index; iSV,
indexed stroke volume; cFT, corrected flow time; PV, peak velocity;
MA, mean acceleration; variations are calculated as percentage of
baseline value; results are mean (SD); P, R vs NR
HR (%)
MAP (%)
CI (%)

R (n526)
25.5 (6)
+18.9 (28.4)
+31 (23.7)

NR (n524)
21.5 (7)
+14.3 (19.5)


21.7 (10)


iSV (%)

+37.4 (21.8)

20.9 (11.4)


cFT (%)

+10.6 (8.2)

21.5 (7)


PV (%)

+29.2 (20.7)

24.4 (10.7)


MA (%)

+18.4 (19.2)

21.9 (19.6)



Downloaded from http://bja.oxfordjournals.org/ at Semmelweis University on August 5, 2012

poor predictors of fluid responsiveness in adults.8 10

Dynamic indices which account for cardiopulmonary interactions in mechanically ventilated patients offer improved
accuracy in predicting VE responsiveness.1 11 Examples of
dynamic indices include transthoracic echocardiography
(TTE) or TED measurements of aortic blood flow velocity or
SV, arterial pulse pressure, and pulse oximetry plethysmographic waveform breath-related changes.12 15 Unfortunately, the performance of these dynamic indices in the
paediatric literature has been disappointing due to low
tidal volumes, higher arterial elastic properties, and
increased chest wall compliance which may dampen the
transmission of transpulmonary pressure changes.16 Two
paediatric ICU studies including ventilated children
between the ages of 8 months and 4 yr demonstrated that
variations in aortic blood flow velocity, inferior vena cava
diameter measured by TTE, or both, but not variations in
pulse pressure, predicted response to VE.16 17 However, it
would be challenging to complete intraoperative TTE in neonates undergoing a surgical intervention. Respiratory variations in aortic blood flow velocity as measured by TED has
been found to be reliable in adult studies, but this requires
specialized software that is able to perform sophisticated
analysis of the Doppler signal in real time.13 To our knowledge, there are no studies investigating dynamic indices in
neonates or young infants.
Preload-dependent patients with a diminished SV are
functioning on the ascending part of the FrankStarling
curve. It is therefore not surprising that in our study, iSV
was found to effectively predict the responsiveness to VE in
our paediatric population with preserved LV function. Maturational changes in neonatal cardiac function and altered LV
filling dynamics may be more dependent on age-related
changes in diastolic rather than systolic properties.18 Patients
presenting with preoperative haemodynamic instability and
possible associated cardiac dysfunction were excluded from
our study. The low concentration of sevoflurane (,1
minimum alveolar concentration) used in our study
(Table 2) has been shown to produce minor decreases in contractibility.19 Desflurane and nitrous oxide possess direct

negative inotropic actions independent of changes in autonomic nervous system tone based on studies in denervated
animals.20 21 Desfluranes sympathoexcitation properties
can mask its direct negative inotropic effects.22 Patients
receiving desflurane, nitrous oxide, or both were equally distributed among the R and NR groups, suggesting a minimal
effect of these agents on cardiac contractility. Nevertheless,
this is a factor that may have led to a decreased iSV
threshold predicting VE responsiveness in our study. A
recent meta-analysis which included 29 studies examined
the performance of dynamic indices in the assessment of
volume responsiveness.1 In five of eight studies examining
SV measures before VE, SV was reduced in patients who
were found to be responders when compared with nonresponders.23 27 Among the three other studies, two
focused on patients treated with b-blockers15 28 and consequently may modify the relationship between preload and
CO. The last study used CO data derived from combined
TED aortic blood velocity and aortic diameter measurements,29 which has been found to increase the measurement error.30 Other studies included in this meta-analysis
did not stipulate SV values before VE, but did report
reduced CO before VE in patient responders compared with
non-responders.13 31 34 In a study which included mechanically ventilated postoperative cardiac surgery patients
without cardiac dysfunction, the area under the ROC curve
describing the value of CI in predicting responsiveness to
VE was 0.74 (0.07) vs 0.63 (0.07) and 0.58 (0.08) for PAOP
and CVP, respectively, and 0.98 (0.01) for respiratory-induced
changes in arterial pulse pressure.34 Among the data derived
from TED-based blood flow velocity waveform, cFT has been
proposed as a value of preload index.5 8 29 However, this parameter is also influenced by afterload and, therefore, may
lack specificity for predicting responsiveness to VE, which
seems to be the case in our study.8 35 PV and MA, two
indices derived from the shape of aortic blood velocity waveform and related to myocardial contractility, increased
significantly in the R group but not in the NR group
(Table 3). However, these indices are also known to be influenced by changes in preload and afterload changes.36 Alternatively, this may reflect known complex interactions
between cardiac loading conditions and contractility.
In daily practice, the decision of VE is based on multiple
factors including quantitative clinical endpoints, anaesthetic
depth, and the dynamic clinical and surgical context. There
were no predefined arterial pressure values used to prompt
VE in our study. However, MAP in both the R and NR groups
were low before VE (mean32 mm Hg), knowing that acceptable MAP can usually be grossly defined as equivalent to the
number of weeks of gestational age (i.e. 37 and 36 weeks in R
and NR, respectively).37 This observation supports the preoccupation with MAP values and its significant role in anaesthetic management.38 Systolic arterial hypotension
associated with a normal CI and iSV in the NR group [3.7
(0.8) litre min21 m22 and 29.2 (6.5) ml m22, respectively]
suggests a state of excessive arterial vasodilatation.39 The
hypotensive effect of sevoflurane is well described and is


Intraoperative use of TED to predict response to VE

paediatric patients which was used in our study (Cardio

QPTM, Deltex MedicalTM ).
The accuracy of aortic blood flow velocity requires that the
angle of insonation (angle between ultrasound beam and
aortic blood flow) is ,208, otherwise this produces large
errors due to its cosine relationship with flow. Estimated
aortic diameter, angle of insonation, and the descending
aorta to total-blood-flow ratio used as a constant are
reasons why it is reported that TED seems most useful to
track changes in CO rather than producing absolute
values.42 Despite the potential for imprecision in TEDmeasured CO, our results show that iSV can effectively
predict response to VE when compared with standard clinical
monitoring data. Coupling aortic cross-sectional area to
Doppler CO measurements was identified as a major
source of error;42 a technique that was not utilized in our
study. Accuracy is best when TED measurements are performed by an experienced operator43 and when the probe
is repositioned before each measurement to obtain the
best blood flow signal;44 two requirements that were met
in our study.
In adults, many studies have shown that intraoperative
optimization of TED-derived CO can reduce postoperative
complications and duration of hospital stay post-abdominal
and orthopaedic surgery, in particular in the elderly
patient.36 45 48

Monitoring CO by TED in neonates and young infants without
cardiac dysfunction undergoing volatile anaesthesia can be
used to predict responsiveness to VE with a sensitivity of
92% and a specificity of 83%. In this study, TED showed
that VE based on standard clinical monitoring data may be
inappropriate in as much as 50% of patients in this age
group. Furthermore, the lack of responsiveness to VE was
not detected in over 25% of patients, possibly leading to
undue fluid loading. TED is a minimally invasive and quickly
learned monitoring technique that is useful to guide intraoperative VE and could be shown in the future to improve
postoperative outcomes in neonates and young infants.

Declaration of interest
None declared.

Departmental sources. Department of Anaesthesia and Critical Care Medicine, Montpellier University Hospital, France.

1 Marik PE, Cavallazzi R, Vasu T, Hirani A. Dynamic changes in arterial waveform derived variables and fluid responsiveness in
mechanically ventilated patients: a systematic review of the literature. Crit Care Med 2009; 37: 2642 7
2 Schmitz L, Xanthopoulos A, Koch H, Lange PE. Doppler flow parameters of left ventricular filling in infants: how long does it


Downloaded from http://bja.oxfordjournals.org/ at Semmelweis University on August 5, 2012

especially significant in the youngest paediatric patients.40

This effect, mostly attributable to a decrease in systemic vascular resistance, may explain the arterial hypotension among
the NR patients, despite the low concentration of volatiles.22
While less well defined in paediatric patients, desflurane has
been shown to produce similar arterial pressure effects in
adults receiving sevoflurane.41 Although not examined in
our study, vasopressor therapy or a switch to an alternative
anaesthetic drug may have been more effective than VE in
the NR group. The increase in MAP observed in NR patients
post-VE, without an improvement in CI, may be explained
by other factors such as the subsequent reduction in
opioids given to the NR during the period of arterial
There are some limitations to our study. Results from our
study showed that R were younger, lighter, and less anaemic
than NR, factors that are associated with each other. This
may highlight that the youngest patients are those who generally undergo emergent major surgery, such as abdominal
surgery, often associated with volaemic disorders. A strict
standardized anaesthetic protocol was not used. VE tended
to be of greater volume, completed more rapidly, and more
colloid based in the R group (Table 1). Even though not statistically different, these differences between groups may
have introduced bias in our study. All patients received sevoflurane or desflurane, two volatile agents with similar
haemodynamic effects.41 This pragmatic methodology
attempted to avoid a protocol with strict inclusion criteria
in a patient population including neonates and young
infants, often operated on an urgent-to-emergent setting.
Furthermore, we feel that this allowed an evaluation of TED
in realistic conditions.
Although TED, when compared with transpulmonary thermodilution, has been shown to provide a clinically accurate
estimate of CO in non-cardiac paediatric critical care
patients, this is the first study to evaluate the use of TED
measurements in neonates and young infants in the intraoperative setting. TED is user-friendly with a good reproducibility including an intraobserver variability of about 3%,42 but
contains a few limitations. The product of the blood velocitytime integral (VTI; or stroke distance, the distance of
a column of blood within the aorta which travels past the
probe with each ejection from the LV) and the aortic crosssectional area (aCSA) provides a measure of SV (stroke
volumeVTIaCSA). Assuming a linear proportionality
between aortic diameter and the patients height, Tibby
and colleagues4 established nomograms on the entire paediatric population, from neonates to older children, showing
satisfactory correlation with thermodilution measures
(r 2 0.88, standard error of the estimate0.266). These
authors attempted to introduce a correction factor accounting for the variation in aortic cross-sectional area during
haemodynamic changes, but this did not improve the predictive value of the model. They concluded that TED provides a
clinically accurate estimate of CO across the entire paediatric
age range and is able to follow changes in CO. These results
permitted the commercialization of a monitor adapted to





























preload recruitable stroke work relationship in chronically instrumented dogs. Anesthesiology 1990; 73: 1148 57
Pagel PS, Kampine JP, Schmeling WT, Warltier DC. Influence of
volatile anesthetics on myocardial contractility in vivo: desflurane
versus isoflurane. Anesthesiology 1991; 74: 9007
Ebert TJ, Harkin CP, Muzi M. Cardiovascular responses to sevoflurane: a review. Anesth Analg 1995; 81(Suppl.): S11 22
Berkenstadt H, Margalit N, Hadani M, et al. Stroke volume variation as a predictor of fluid responsiveness in patients undergoing brain surgery. Anesth Analg 2001; 92: 984 9
Reuter DA, Kirchner A, Felbinger TW, et al. Usefulness of left ventricular stroke volume variation to assess fluid responsiveness in
patients with reduced cardiac function. Crit Care Med 2003; 31:
1399 404
Solus-Biguenet H, Fleyfel M, Tavernier B, et al. Non-invasive prediction of fluid responsiveness during major hepatic surgery. Br
J Anaesth 2006; 97: 80816
Charron C, Fessenmeyer C, Cosson C, et al. The influence of tidal
volume on the dynamic variables of fluid responsiveness in critically ill patients. Anesth Analg 2006; 102: 1511 7
Belloni L, Pisano A, Natale A, et al. Assessment of
fluid-responsiveness parameters for off-pump coronary artery
bypass surgery: a comparison among LiDCO, transesophageal
J Cardiothorac Vasc Anesth 2008; 22: 2438
Cannesson M, Desebbe O, Rosamel P, et al. Pleth variability index
to monitor the respiratory variations in the pulse oximeter
plethysmographic waveform amplitude and predict fluid responsiveness in the operating theatre. Br J Anaesth 2008; 101: 200 6
Lee JH, Kim JT, Yoon SZ, et al. Evaluation of corrected flow time in
oesophageal Doppler as a predictor of fluid responsiveness. Br J
Anaesth 2007; 99: 3438
Wodey E, Gai V, Carre F, Ecoffey C. Accuracy and limitations of
continuous oesophageal aortic blood flow measurement during
general anesthesia for children: comparison with transcutaneous
echography-Doppler. Paediatr Anaesth 2001; 11: 309 17
Monnet X, Rienzo M, Osman D, et al. Passive leg raising predicts
fluid responsiveness in the critically ill. Crit Care Med 2006; 34:
1402 7
Natalini G, Rosano A, Taranto M, Faggian B, Vitorielli E,
Bernardini A. Arterial versus plethysmographic dynamic indices
to test responsiveness for testing fluid administration in hypotensive patients: a clinical trial. Anesth Analg 2006; 103: 1478 84
Biais M, Nouette-Gaulain K, Cottenceau V, Revel P, Sztark F. Uncalibrated pulse contour-derived stroke volume variation predicts
fluid responsiveness in mechanically ventilated patients undergoing liver transplantation. Br J Anaesth 2008; 101: 7618
Auler JO, Galas F, Hajjar L, Santos L, Carvalho T, Michard F. Online
monitoring of pulse pressure variation to guide fluid therapy after
cardiac surgery. Anesth Analg 2008; 106: 12016
Singer M, Allen MJ, Webb AR, Bennett ED. Effects of alterations in
left ventricular filling, contractility, and systemic vascular resistance on the ascending aortic blood velocity waveform of
normal subjects. Crit Care Med 1991; 19: 1138 45
Singer M. Oesophageal Doppler. Curr Opin Crit Care 2009; 15:
244 8
Nuntnarumit P, Badda-Ellzey HS. Blood pressure measurements
in the newborn. Clin Perinatol 1999; 26: 98196
Michel F, Lando A, Aubry C, Arnaud S, Merrot T, Martin C. Experience with remifentanil sevoflurane balanced anesthesia for
abdominal surgery in neonates and children less than 2 years.
Paediatr Anaesth 2008; 18: 5328

Downloaded from http://bja.oxfordjournals.org/ at Semmelweis University on August 5, 2012

take for the maturation of the diastolic function in a normal left

ventricle to occur? Pediatr Cardiol 2004; 25: 482 91
Tibby SM, Hatherill M, Marsh MJ, Murdoch IA. Clinicians abilities to
estimate cardiac index in ventilated children and infants. Arch Dis
Child 1997; 77: 5168
Tibby SM, Hatherill M, Murdoch IA. Use of transesophageal
Doppler ultrasonography in ventilated pediatric patients: derivation of cardiac output. Crit Care Med 2000; 28: 204550
Singer M, Bennett ED. Noninvasive optimization of left ventricular
filling using esophageal Doppler. Crit Care Med 1991; 19:
1132 7
Osthaus WA, Huber D, Carstzn B, et al. Correlation of oxygen delivery with central venous oxygen saturation, mean arterial pressure
and heart rate in piglets. Paediatr Anaesth 2006; 16: 944 7
Kluckow M, Evans N. Relationship between blood pressure and
cardiac output in preterm infants requiring mechanical ventilation. J Pediatr 1996; 129: 506 12
Tibby SM, Hatherill M, Durward A, Murdoch IA. Are transoesophageal Doppler parameters a reliable guide to paediatric haemodynamic status and fluid management? Intensive Care Med
2001; 27: 201 5
Kumar A, Anel R, Bunnell E, et al. Pulmonary artery occlusion
pressure and central venous pressure fail to predict ventricular
filling volume, cardiac performance, or the response to volume
infusion in normal subjects. Crit Care Med 2004; 32: 6919
Marik PE, Baram M, Vahid B. Does central venous pressure predict
fluid responsiveness? A systematic review of the literature and
the tale of seven mares. Chest 2008; 134: 1728
Michard F, Boussat S, Chemla D, et al. Relation between respiratory changes in arterial pulse pressure and fluid responsiveness in
septic patients with acute circulatory failure. Am J Respir Crit Care
Med 2000; 162: 1348
Feissel M, Michard F, Mangin I, Ruyer O, Faller JP, Teboul JL. Respiratory changes in aortic blood velocity as an indicator of fluid
responsiveness in ventilated patients with septic shock. Chest
2001; 119: 867 73
Monnet X, Rienzo M, Osman D, et al. Esophageal Doppler monitoring predicts fluid responsiveness in critically ill ventilated
patients. Intensive Care Med 2005; 31: 1195201
Feissel M, Teboul JL, Merlani P, Badie J, Faller JP, Bendjelid K.
Plethysmographic dynamic indices predict responsiveness in
septic ventilated patients. Intensive Care Med 2007; 33: 993 9
Cannesson M, Attof Y, Rosamel P, et al. Respiratory variations in
pulse oximetry plethysmographic waveform amplitude to
predict fluid responsiveness in the operating room. Anesthesiology 2007; 106: 110511
Durand P, Chevret L, Essouri S, Haas V, Devictor D. Respiratory
variations in aortic blood flow predict fluid responsiveness in ventilated children. Intensive Care Med 2008; 34: 888 94
Choi DY, Kwak HJ, Park HY, Kim YB, Choi CH, Lee JY. Respiratory
variation in aortic blood flow velocity as a predictor of fluid
responsiveness in children after repair of ventricular septal
defect. Pediatr Cardiol 2010; 31: 116670
Harada K, Takahashi Y, Tamura M, Orino T, Takada G. Serial echocardiographic and Doppler evaluation of left ventricular systolic
performance and diastolic filling in premature infants. Early
Hum Dev 1999; 54: 16980
Wodey E, Pladys P, Copin C, et al. Comparative hemodynamic
depression of sevoflurane versus halothane in infants: an echocardiographic study. Anesthesiology 1997; 87: 795 800
Pagel PS, Kampine JP, Schmeling WT, Warltier DC. Effects of
nitrous oxide on myocardial contractility as evaluated by the

Raux et al.


Intraoperative use of TED to predict response to VE

39 Heymann M. Care of the high risk neonate. In: Klaus MH,

Fanaroff AA, eds. The Heart, 4th Edn. Philadelphia: WB Saunders
Co., 1993; 34573
40 Lerman J, Sikich N, Kleinman S, Yentis S. The pharmacology of sevoflurane in infants and children. Anesthesiology 1994; 80: 81424
41 Ebert TJ, Muzi M, Lopatka CW. Neurocirculatory responses to sevoflurane in humans. A comparison to desflurane. Anesthesiology
1995; 83: 8895
42 Chew MS, Poelaert J. Accuracy and repeatability of pediatric
cardiac output measurement using Doppler: 20-year review of
the literature. Intensive Care Med 2003; 29: 188994
43 Funk DJ, Moretti EW, Gan TJ. Minimally invasive cardiac output
monitoring in the perioperative setting. Anesth Analg 2009;
108: 88797
44 Krishnamurthy B, McMurray TJ, McClean E. The peri-operative use
of oesophageal Doppler monitor in patients undergoing coronary





artery revascularisation. A comparison with the continuous

cardiac output monitor. Anaesthesia 1997; 52: 6249
Sinclair S, James S, Singer M. Intraoperative intravascular volume
optimisation and length of hospital stay after repair of proximal
femoral fracture: randomised controlled trial. Br Med J 1997;
315: 90912
Ven R, Steele A, Richardson P, Poloniecki J, Grounds M, Newman P.
Randomized controlled trial to investigate influence of the fluid challenge on duration of hospital stay and perioperative morbidity in
patients with hip fractures. Br J Anaesth 2002; 88: 6571
Gan TJ, Soppitt A, Maroof M, et al. Goal-directed intraoperative
fluid administration reduces length of hospital stay after major
surgery. Anesthesiology 2002; 97: 8206
Wakeling HG, McFall MR, Jenkins CS, et al. Intraoperative oesophageal
Doppler guided fluid management shortens postoperative hospital
stay after major bowel surgery. Br J Anaesth 2005; 95: 63442

Downloaded from http://bja.oxfordjournals.org/ at Semmelweis University on August 5, 2012