Content uploaded by Gilles A Orliaguet
Author content
All content in this area was uploaded by Gilles A Orliaguet on Nov 19, 2014
Content may be subject to copyright.
Stroke volume variation and indexed stroke volume
measured using bioreactance predict fluid responsiveness
in postoperative children
†
E. Vergnaud, C. Vidal, J. Verche
`re, J. Miatello, P. Meyer, P. Carli and G. Orliaguet*
Service d’Anesthe
´sie Re
´animation, Ho
ˆpital Universitaire Necker-Enfants Malades, Universite
´Paris Descartes,
Assistance Publique Ho
ˆpitaux de Paris, 149 rue de Sevres, 75743 Paris Cedex 15, France
* Corresponding author. E-mail: gilles.orliaguet@nck.aphp.fr
Editor’s key points
†This study assessed the
ability of dynamic
haemodynamic variables
measured using a
bioreactance device to
predict fluid
responsiveness.
†Stroke volume (SV) and
stroke volume variation
(SVV) measured by
bioreactance were
predictive for fluid
responsiveness.
†Optimal threshold values
for SV and SVV were 29 ml
m
22
and 10%,
respectively.
Background. Postoperative fluid management can be challenging in children afterhae morrhagic
surgery. The goal of this study was to assess the ability of dynamic cardiovascular variables
measured using bioreactance (NICOM
w
, Cheetah Medical, Tel Aviv, Israel) to predict fluid
responsiveness in postoperative children.
Methods. Children sedated and mechanically ventilated, who require volume expansion (VE)
during the immediate postoperative period, were included. Indexed stroke volume (SVi),
cardiac index, and stroke volume variation (SVV) were measured using the NICOM
w
device.
Responders (Rs) to VE were patients showing an increase in SV measured using
transthoracic echocardiography of at least 15% after VE. Data are median [95% confidence
interval (CI)].
Results. Thirty-one patients wereincluded, but one patient was excluded because of the lack of
calibration of the NICOM
w
device. Before VE, SVi [33 (95% CI 31 –36) vs 24 (95% CI 21 –28) ml
m
22
;P¼0.006] and SVV [8 (95% CI 4 – 11) vs 13 (95% CI 11– 15)%; P¼0.004] were significantly
different between non-responders and Rs. The areas under the receiver operating
characteristic curves of SVi and SVV for predicting fluid responsiveness were 0.88 (95% CI
0.71– 0.97) and 0.81 (95% CI 0.66–0.96), for a cut-off value of 29 ml m
22
(grey zone 27–29 ml
m
22
) and 10% (grey zone 9–15%), respectively.
Conclusions. The results of this study show that SVi and SVV non-invasively measured by
bioreactance are predictive of fluid responsiveness in sedated and mechanically ventilated
children after surgery.
Keywords: children; equipment, monitors; measurement techniques, transthoracic electrical
impedance; monitoring, cardiopulmonary
Accepted for publication: 23 August 2014
Postoperative fluid management after haemorrhagic surgery
can be challenging. On the one hand, undiagnosed and uncor-
rected hypovolaemia may lead to organ dysfunction;
1
on the
other hand, fluid overload may be associated with impaired
oxygenation.
2
Therefore, to avoid hypovolaemia and overload,
it is mandatory to accurately assess the preload state of the
patients.
Static haemodynamic variables, corresponding to static
measurement of cardiac filling pressures [mainly central
venous pressure (CVP)], are unable to reliably assess fluid
responsiveness in adults
34
and in children.
5–10
A more
dynamic approach using variables based on the heart– lung
interaction induced by mechanical ventilation has been pro-
posed.
10
When the two cardiac ventricles are working on the
steep portion of the Frank– Starling curve, the respiratoryvaria-
tions are high; the heart is preload-dependent; and the patient
is more likely to be a fluid responder (R). In contrast, if at least
one of the two ventricles works on the plateau of this relation-
ship, then the respiratory variations are low and the heart is
preload independent; the patient is more likely to be a fluid
non-responder (NR). The use of respiratory variation of stroke
volume or surrogates to predict fluid responsiveness has
some limitations that are now identified:
11
(i) spontaneous
breathing activity; (ii) cardiac arrhythmias; (iii) increased ab-
dominal pressure; (iv) open-chest surgery; (v) high-frequency
ventilation, with respiratory rates over 40 bpm; (vi) insufficient
variations in intra-thoracicpressure [tidal volume (V
t
),7mlkg
21
or decreased compliance of the respiratory system].
†
This study was presented, in part, during the 2013 annual meeting of the French Society of Anaesthesia and Intensive Care.
&The Author 2014. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
British Journal of Anaesthesia Page 1 of 7
doi:10.1093/bja/aeu361
BJA Advance Access published October 14, 2014
at JEPU ProfCORIAT on November 19, 2014http://bja.oxfordjournals.org/Downloaded from
In adult patients, dynamic variables, such as pulse pressure
variation (PPV) and stroke volume variation (SVV), have been
shown to reliably predict fluid responsiveness.
12 –14
In contrast,
conflicting results have been obtained regarding the predictive
values of dynamic variables in children.
5681015–17
Continuous efforts have been made to developnon-invasive
methods for cardiac output (CO) and other haemodynamic
variables measurement. One of the most widely studied non-
invasive techniques relies upon bioimpedance, which involves
the analysis of intra-beat variations in transthoracic voltage in
response to applied high-frequency transthoracic currents.
18
However, the analysis of published studies found discrepancies
between CO measured using bioimpedance and thermodilu-
tion, and also numerous artifacts that interfered with meas-
urement.
19 20
Bioreactance is an innovative evolution of the
bioimpedance device, processing, and analysing the electrical
signal using a new technique. The bioreactance technique
sends a high-frequency current, with known low amplitude,
through the thorax and measures the frequency and phase
modulations resulting from the changes in thoracic blood
volume. With bioreactance, a much higher signal quality is
obtained, because it is not the changes in the amplitude of
the signal that are measured but the changes in frequency,
allowing to significantly reduce background noise. In addition,
bioreactance does not depend upon the distance between the
electrodes for the calculations of CO, which significantly
reduces the potential for error in the results. Controversial
results have been published regarding the reliability of bioreac-
tance for predicting fluid responsiveness in adult patients.
21 –23
Only one study is available on the use of bioreactance for pre-
dicting fluid responsiveness in children.
16
In this study, SVV
measured by bioreactance reliably predicted fluid responsive-
ness after cardiac surgery in children.
16
The aim of this study was to assess the ability of dynamic
haemodynamic variables measured using a bioreactance
device to predict fluid responsiveness in children aftercraniosy-
nostosis repair.
Methods
This prospective single-centre observational study was
approved by the Institutional Review Board (IRB) (i.e. Comite
´
de Protection des Personnes Ile-de-France VI) and was con-
ducted in the paediatric neurosurgical intensive care unit of
the Necker University Hospital (Paris, France). The patients
and their parents (or legal guardians) were informed about
the study, but the requirement for signed informed consent
was waived by the IRB because of the design of the study, a pro-
spective observational study without any change in the stand-
ard management of the children.
Patient population
Postoperative children, aged 0–16 yr, sedated and mechanic-
ally ventilated, and who require volume expansion (VE) in the
early postoperative period (before postoperative hour 2) after
craniosynostosis repair, could be included in the study.
Exclusion criteria were: patient or legal guardian refusal,
cardiac dysrhythmia, severe systolic cardiac dysfunction,
significant valvular heart disease, and intra-cardiac shunt.
Patient management and monitoring
Intraoperative anaesthesia management and fluid mainten-
ance were left at the discretion of the anaesthetist in charge
of the patient. In our institution, invasive monitoring is stand-
ard practice for paediatric patients undergoing craniosynosto-
sis repair. Briefly, routine monitoring included ECG, invasive
arterial and CVP, core temperature, pulse oximetry, end-tidal
carbon dioxide (E′
CO2), and urine output (bladder catheter). Iso-
volaemic compensation of blood loss was observed, with fluid
replacement based upon haemodynamic variables, to main-
tain mean arterial pressure in the range of 45– 55 mm Hg
and CVP.2 mm Hg, using artificial colloids and blood transfu-
sion. Transfusion of packed red blood cells (PRBC) was used
to maintain intraoperative haemoglobin level in the range
of 7–10 g dl
21
and fresh-frozen plasma (FFP) as required
(FFP:PRBC transfusion ratio of 1.5–2). During the immediate
postoperative period, sedation was maintained using a con-
tinuous i.v. infusion of midazolam, to maintain a Richmond Agi-
tation Sedation Scale between 23 and 21, and analgesia was
ensured with i.v. morphine and i.v. paracetamol. Mechanical
ventilation was provided using: a V
t
of 7–8 ml kg
21
body
weight, a PEEP of 3 – 4 cm H
2
O, an I/Eratio of 1/1.5 to 1/1.7
(Servo-I Universel, System version 6.1 or 7.0, Maquet Critical
Care, Sweden), while the respiratory rate was set to maintain
an E′
CO2between 35 and 40 mm Hg. Children were connected
to an IntelliVue MP70TM patient monitor (Philips Medical
Systems, Suresnes, France), which continuously recorded
heart rate (HR, beats min
21
), systolic/diastolic/mean invasive
arterial pressure (SAP/DAP/MAP, mm Hg), and CVP (mm Hg).
Bioreactance measurements
The bioreactance-based non-invasive CO and stroke volume
measurement system relies upon an analysis of relative
phase shifts of an oscillating current that occur when the
current passes through the thorax.
24
The NICOM
w
device
(NICOM
w
, Cheetah Medical, Tel Aviv, Israel) monitors CO
using four electrodes. Two upper electrodes are placed on the
right and left mid-clavicular lines of the thorax, respectively,
and two lower electrodes are placed on the midpoints of the
right and left 12th ribs, respectively. After placing the electro-
des and entering patients’ characteristics, the NICOM
w
device automatically calibrates and then provides continuous
values of: stroke volume (SV), indexed stroke volume (SVi),
CO, cardiac index, and SVV.
Echocardiographic measurements
Transthoracic echocardiography (TTE) was performed using
a Siemens Acuson CV70TM (Siemens Medical Solutions,
Issaquah, WA, USA). All measurements were performed by
the same investigators (E.V. and C.V.). The aortic diameter
(D, mm) was measured at the aortic annulus level in a
two-dimensional view from the parasternal long-axis view.
BJA Vergnaud et al.
Page 2 of 7
at JEPU ProfCORIAT on November 19, 2014http://bja.oxfordjournals.org/Downloaded from
Pulsed Doppler waves aortic flow was recorded at the exact
level of the aortic annulus in the apical five-chamber view.
Maximal and minimal aortic velocity–time integral (VTI) was
measured over a single respiratory cycle on two sets of mea-
surements; then VTI
mean
was calculated. Stroke volume mea-
sured using TTE (SV
TTE
) was calculated as follows:
SVTTE (ml)=VTImean ×
p
(D)2
4
Data recording
In addition to the usual patient’s characteristics, respiratory
settings and haemodynamic variables were recorded, and
also data measured by thoracic bioreactance (SVi and cardiac
index) immediately before and after VE. The PPV index, based
on Aboy and colleagues,
25 26
was continuously displayed by
the Philips
w
monitor (Intellivue MP70, Philips Medical System).
In accordance with our unit policy, children who were clinically
judged to require VE (tachycardia, hypotension, oliguria, or
delayed capillary refilling) received an i.v. bolus of 20 ml kg
21
of artificial colloids (Plasmion
w
, Fresenius Kabi, France) over
15–30 min.
Statistical analysis
Data were analysed with the Shapiro– Wilk and Kolmogorov–
Smirnov tests. Data are expressed as mean (SD)ormedian
(95% confidence interval), as appropriate, and as numbers of
patients (%). Rs to VE were defined as patients showing an
SV
TTE
increment of at least 15% after VE and non-responders
(NRs) as those showing an SV incrementof ,15%. Comparisons
between R and NR were assessed using a two-sample Student
t-test or a Mann– Whitney U-test, as appropriate. The predictive
ability of a variable for fluid responsiveness was assessed using
receiver operating characteristics (ROC) curve analysis. For each
variable, a threshold value was determined to maximize both
sensitivity and specificity. The grey zone corresponds to a
range of values for which the variable of interest does not pro-
vide conclusive information. Inconclusive responses are defined
for pre-challenge values of the variable of interest with a sensitiv-
ity ,90% or specificity ,90% (diagnosis tolerance of 10%).
27 28
Grey zone limits are expressed as (low limit–high limit). A power
analysis showed that 25 patients were necessary to detect a dif-
ference of 0.3 between SVV and CVP areas under the ROC curves
(5% type I error rate, 80% power, two-tailed t-test). Allowing for
possible dropouts, 31 patients were included in the current study.
Data were also analysed according to two age strata: 0–3 and
.3–16 yr. A P-value of 0.05 was considered statistically signifi-
cant. The statistical analysis was performed with the BiostaTGV
software (INSERM and Pierre et Marie Curie University, Paris,
France, http://marne.u707.jussieu.fr/biostatgv/).
Results
Between August 2012 and July 2013, 31 patients were included
after craniosynostosis surgical repair (Fig. 1). One patient was
excluded from the study because of the lack of calibration of
the NICOM
w
device. Therefore, the data from 30 patients
were included in the analysis.
158 patients operated for craniosynostosis repair
for August 2012 to July 2013
56 patients sedated and mechanically
ventilated requiring volume expansion
before postoperative hour 2
31 patients included
30 patients with data available
included in the analysis
15 patients responders 15 patients non-responders
I.V. bolus of 20 ml kg–1 of
artificial colloid over 15–30 min
One excluded because of the lack of
calibration of the NICOM‚ device
13 legal guardian refusal
12 no NICOM‚ device available
Fig 1 Flow chart.
Fluid responsiveness prediction using NICOM in children BJA
Page 3 of 7
at JEPU ProfCORIAT on November 19, 2014http://bja.oxfordjournals.org/Downloaded from
Baseline patient characteristics were similar for the Rs and
the NRs (Table 1).
Table 2shows the haemodynamic variables in the Rs and
NRs, before and after VE. Before VE, HR [106 (95% CI 86– 116)
vs 130 (95% CI 106–144); P¼0.016], PPV [7.0% (95% CI 5.9 –
10.0) vs 10.0% (95% CI 9.0–12.9), P¼0.0149], SVi [33 (95% CI
31–36) vs 24 (95% CI 21 –28) ml m
22
;P¼0.006], and SVV [8
(95% CI 4–11) vs 13 (95% CI 11 –15)%; P¼0.004] were signifi-
cantly different between NR and R (Table 2). The areas under
the ROC curves of SVi (cut-off value: 29 ml m
22
) and SVV
(cut-off value: 10%) for predicting fluid responsiveness were,
respectively, 0.88 (95% CI 0.71– 0.97) and 0.81 (95% CI 0.66–
0.96) (Table 3, Fig. 2). Corresponding grey zone limits were
(26–29 ml m
22
) and (9 –15%), respectively. Data from the
two age strata (0– 3 and 3– 16 yr) are presented in Table 4.
Discussion
In this study, SVi and SVV measured by bioreactance were
found to be predictive of fluid responsiveness, with optimal
threshold values of 29 ml m
22
and 10%, respectively, in
sedated and mechanically ventilated children after craniosy-
nostosis surgical repair.
In children, there is an urgent need for a tool enabling to
predict fluid responsivenessto avoid undue fluid loading, carry-
ing a risk of impaired oxygenation.
2
In our study, we have noted
that VE based on standard clinical monitoring was inappropri-
ate in 50% of children. This result is in accordance with the
adult literature
12 29
and has been also recently confirmed in
paediatric patients.
5–810
Before VE, MAP and CVP were not
different between R and NR in our study. In contrast, HR was
Table2 Differences in haemodynamic variables between R and NR before and after VE. Data are expressed as median (95% CI). HR, heart rate; MAP,
mean arterial pressure; CVP, central venous pressure; PPV, pulse pressure variation; SVi, indexed stroke volume; SVV, stroke volume variation.
*P,0.05 vs before VE (baseline) within a group.
†
P,0.05 vs NRs
Before VE After VE
NRs Rs NRs Rs
HR (beats min
21
) 106 (86 –116) 130 (106–144)
†
108 (89–18) 135 (104–145)
†
MAP (mm Hg) 73 (61–91) 67 (63–85) 83 (67–91) 80 (74–93)
CVP (mm Hg) 5.5 (4.0–9.6) 8.0 (3.9–9.0) 10.0 (6.3–16.4) 13.5 (7.0–16.5)*
PPV (%) 7.0 (5.9–10.0) 10.0 (9.0–12.9)
†
5.0 (4.2–8.0) 6.0 (5.2–8.2)
SVi (ml m
22
) 33 (31–36) 24 (21–28)
†
34 (30–38) 35 (26–36)*
SVV (%) 8 (4–11) 13 (11–15)
†
7 (4–11) 10 (7– 12)
Cardiac index (litre min
21
m
22
) 3.3 (2.8–3.8) 3.1 (2.4–3.6) 3.3 (3.0–4.1) 3.8 (3.3–5.0)*
Table 1 Baseline characteristics of the Rs and the NRs to VE. Data are median (95% CI). BSA, body surface area. *VE was performed using artificial
colloid (Plasmion
w
)
NRs (n515) Rs (n515) P-value
Age (months) 60.0 (32.8–96.8), range (4 –137) 27.0 (11.5–32.7), range (7–139) 0.077
Body weight (kg) 18.0 (11.5–24.5) 13.0 (9.1–15.0) 0.191
BSA (m
2
) 0.72 (0.52 –0.94) 0.57 (0.42– 0.64) 0.158
Aortic diameter (cm) 13.7 (10.7– 15.1) 11.6 (9.7–13.5) 0.329
VE (ml kg
21
)* 20.0 (20.0 –20.0) 20.0 (19.6– 20.0) 0.838
PEEP (cm H
2
O) 3 (3–3) 3 (3–5) 0.287
Mean airway pressure (cm H
2
O) 8.0 (8.0 –9.0) 8.0 (7.2–8.0) 0.135
V
t
(ml kg
21
) 8.3 (7.1–9.2) 8.0 (6.9– 11.1) 0.85
Haematocrit (%) 32.2 (29.7 –37.9) 32.3 (28.7– 35.6) 0.62
Table3 Prediction of the fluid responsiveness by the ROC curvesof CVP, PPV, SVi, and SVV measured using the NICOM
w
device. AUC, area under the
ROC curve; Grey zone, range of values with a sensitivity ,90% or specificity ,90%
Cut-off
value
AUC (95% CI) Specificity
(%)
Sensitivity
(%)
Positive predictive
value (%)
Negative predictive
value (%)
Grey zone
CVP 8 mm Hg 0.59 (0.35–0.80) 40 80 57 66 3– 9 mm Hg
PPV 8% 0.77 (0.57–0.91) 78 69 72 76 6– 10%
SVi 29 ml m
22
0.88 (0.71–0.97) 93 80 71 100 27 –29 ml m
22
SVV 10% 0.81 (0.66–0.96) 93 80 74 90 9– 15%
BJA Vergnaud et al.
Page 4 of 7
at JEPU ProfCORIAT on November 19, 2014http://bja.oxfordjournals.org/Downloaded from
significantly different between R and NR. However, while
increased HR may indicate some degree of hypovolaemia,
the interpretation of tachycardia in the perioperative period
is not univocal and may be related to different causes, includ-
ing, for example, pain or other cause of discomfort. Lastly,
anaesthetists may be confronted with difficulty distinguishing
whether a patient had responded positively or negatively to
VE.
7
Therefore, a tool enabling them to know when to start
and when to stop fluid loading in children could be useful.
Thoracic bioimpedance was among the firstand most widely
used non-invasive method of measuring CO.
30
However, the
analysis of published studies found discrepancies between CO
measured using bioimpedance and thermodilution and also
numerous artifacts that interfered with measurement.
19 20
Bioreactance has the advantage over bioimpedance of getting
a much higher signal quality, by reducing the background
noise. Unlike bioimpedance, bioreactance does not use static
impedance and does not depend upon the distance between
the electrodes for the calculations of CO, which significantly
reduces the potential for error in the results. Paediatric studies
assessing the performance of bioimpedance and bioreactance
devices for the measurement of CO are limited and have pro-
vided contrasting results.
31–34
Up to now, only one study has
assessed the ability of haemodynamic variables provided by
the NICOM
w
device to predict fluid responsiveness in mechanic-
ally ventilated children.
16
This study has shown that SVV mea-
sured by NICOM
w
reliably predicted fluid responsiveness in
children during mechanical ventilation after cardiac surgery.
16
Thus, the results from this study are in agreement with our
own results, with an optimal cut-off value of SVVof 10% which
is similar to the cut-off value found in our study.
16
Dynamic variables such as SVV, pulse pressure, and systolic
pressure variation are clinically reliable variables for predicting
fluid responsiveness in adults.
3 4 12 29
However, they are under-
explored in childrenand the few results available are controver-
sial. In paediatric patients, the respiratory variation in aortic
blood flow velocity (DV
peak
, derived by echocardiography) is
now considered as reliable for predicting fluid responsive-
ness.
10
More recently, measurements of SVV
16
and SVi
735
were also found to be predictive of fluid responsiveness in chil-
dren. In our study, PPV was significantly different between Rs
and NRs before VE (Table 2). However, the area under the
ROC curve for PPV was ,0.80 (Table 3), confirming that periph-
eral dynamic index does not predict fluid responsiveness in
children,
10
in contrast to central index, such as SVi and SVV in
children more than 3 yr of age (Table 4).
Among many advantages of SVV and SVi measured using
bioreactance, one may underline: a totally non-invasive aspect
(as opposed to oesophageal Doppler which may be considered
100
80
60
40
20
0
0 20406080100
Sensitivity
100-specificity
CVP
PPV
SVi
SVV
Fig 2 Prediction of the fluid responsiveness by the ROC curves of
indexed stroke volume (SVi, cut-off value: 29 ml m
22
) and stroke
volume variation (SVV, cut-off value: 10%) measured using the
NICOM
w
device. The straight line is a reference line. The areas
under the ROC curves of SVi and SVV are 0.88 (95% CI 0.71 –0.97)
and 0.81 (95% CI 0.66–0.96), respectively.
Table4 Prediction of the fluid responsiveness by the ROC curves of CVP, PPV,SVi, and SVV measured using the NICOM
w
device in children under and
more than 3 yrof age. AUC, area under the ROC curve; Grey zone, range of values with a sensitivity ,90% or specificity ,90%; R, responder to VE; NR,
non-responder to VE
Cut-off
value
AUC (95% CI) Sensitivity
(%)
Specificity
(%)
Positive predictive
value (%)
Negative predictive
value (%)
Grey zone
Children ,3yr(n¼17, R¼12, NR¼5)
CVP 6 mm Hg 0.74 (0.38–0.95) 100 43 34 100 3–5 mm Hg
PPV 12% 0.60 (0.33–0.83) 28 100 37 100 9 –11%
SVi 26 ml m
22
0.92 (0.68–0.99) 75 100 100 63 27–28 ml m
22
SVV 9% 0.57 (0.31 –0.81) 91 40 80 100 7 –17%
Children .3yr(n¼13, R¼3, NR¼10)
CVP 5 mm Hg 0.76 (0.41–0.96) 100 43 44 100 8–9 mm Hg
PPV 8% 0.81 (0.48 –0.97) 67 100 100 87 6–6%
SVi 33 ml m
22
0.81 (0.49–0.96) 100 60 52 100 28–32 ml m
22
SVV 10% 0.97 (0.70– 1.00) 100 90 81 100 10 –11%
Fluid responsiveness prediction using NICOM in children BJA
Page 5 of 7
at JEPU ProfCORIAT on November 19, 2014http://bja.oxfordjournals.org/Downloaded from
as semi-invasive), the possibility of continuous monitoring (as
opposed to echocardiography), and the fact that it is easy to
use (‘plug and play’).
In our study, Rs to VE were defined as patients showing an
increase in SV
TTE
of at least 15% after VE. We have decided to
choose this cut-off value of 15% based on the results of previ-
ous studies in children, which suggested that this difference is
clinically significant.
10
One possible limitation of our study is related to the relative-
ly wide age range of patients included in our study. This could
potentially result in a heterogeneous population, in particular
from a cardiovascular physiology point of view. However, no
patient was aged more than 12 yr and ,6 months, while the
major differences in cardiovascular physiology are usually
observed between children under and more than 6 months of
age.
36
In the NR and R groups, the age ranged from 6 to 137
and 7 to 139 months, respectively, without patients younger
than 6 months. Moreover, other authors studying fluid respon-
siveness in children have included patients with wider age
range. For example, in a very recent study assessing the
utility of SVV as a predictor of fluid responsiveness in children,
the authors have included 13 children aged 2 months to 14
yr.
37
On the other hand, we also performed an analysis after
stratification for age under and above 3 yr. Such a subgroup
analysis was able to provide further insights into the effect of
age on fluid responsiveness assessment (Table 4). However,
we recognize that these results should be interpreted with
caution because stratifying by age may generate a problem
of statistical power. Indeed, power and sample size calcula-
tions were based on the total number of patients and not on
the number of patients included in each subgroup.
Bioreactance has its own shortcomings. The mathematical
model of the bioreactance is based on several anatomical and
physiological assumptions, including the fact that blood resist-
ivity is supposed to remain constant. However, blood resistivity
is proportional to haematocrit; therefore, significant haemodi-
lution might skew bioreactance CO measurement. Neverthe-
less, previous studies have shown that this limitation was
not significant for haematocrit values ranging from 25% to
45%.
38
All the patients included in our study had haematocrit
values included in this range. In addition, the bias and precision
of the NICOM
w
device measurements are comparable with
what is published in the literature for other methods of CO
measurements used in the paediatric population.
39 40
In conclusion, the results of this study show that SViand SVV
non-invasively measured by bioreactance are predictive of
fluid responsiveness in sedated and mechanically ventilated
postoperative children.
Authors’ contributions
E.V.:study design, patient recruitment, datacollection, and data
analysis; C.V.: patient recruitment, data collection, and data
analysis; J.V., J.M., and P.M.: patient recruitmentand data collec-
tion; P.C.: data analysis and writing up of the paper. G.O.: study
design, data analysis, writing up of the paper, and archiving of
the study files.
Acknowledgement
The authors thank Frances O’Donovan, MD, FFARCSI (Staff
Anaesthesiologist, Department of Anaesthesia, Children’s
University Hospital, Dublin, Ireland), for kindly reviewing the
manuscript.
Declaration of interest
None declared.
References
1 Maitland K, Pamba A, Newton CR, Levin M. Response to volume re-
suscitation in children with severe malaria. Pediatr Crit Care Med
2003; 4: 426– 31
2 Arikan AA, Zappitelli M, Goldstein SL, Naipaul A, Jefferson LS,
Loftis LL. Fluid overload is associated with impaired oxygenation
and morbidity in critically ill children. Pediatr Crit Care Med 2012;
13: 253–8
3 Marik PE, Cavallazzi R. Does the central venous pressure predict fluid
responsiveness? An updated meta-analysis and a plea for some
common sense. Crit Care Med 2013; 41: 1774– 81
4 Osman D, Ridel C, Ray P, et al. Cardiac filling pressuresare not appro-
priate to predict hemodynamic response to volume challenge. Crit
Care Med 2007; 35:64–8
5 Durand P, ChevretL, Essouri S, Haas V, Devictor D. Respiratory varia-
tions in aortic blood flow predict fluid responsiveness in ventilated
children. Intensive Care Med 2008; 34: 888–94
6 Renner J, Broch O, Gruenewald M, et al. Non-invasive prediction of
fluid responsiveness in infants using pleth variability index. Anaes-
thesia 2011; 66: 582–9
7 Raux O, Spencer A, Fesseau R, et al. Intraoperative use of transoeso-
phageal Doppler to predict response to volumeexpansion in infants
and neonates. Br J Anaesth 2012; 108: 100– 7
8 Renner J, Broch O, Duetschke P, et al. Prediction of fluid responsive-
ness in infants and neonates undergoing congenital heart surgery.
Br J Anaesth 2012; 108: 108– 15
9 Stricker PA, Lin EE, Fiadjoe JE, et al. Evaluation of central venous
pressure monitoring in children undergoing craniofacial recon-
struction surgery. Anesth Analg 2013; 116: 411– 9
10 Gan H, Cannesson M, Chandler JR, Ansermino JM. Predicting fluid re-
sponsiveness in children: a systematic review. Anesth Analg 2013;
117: 1380– 92
11 Monnet X, Teboul JL. Assessment of volume responsiveness during
mechanical ventilation: recent advances. Crit Care 2013; 17: 217
12 Michard F, Teboul JL. Predicting fluid responsiveness in ICU patients:
a critical analysis of the evidence. Chest 2002; 121: 2000– 8
13 Sandroni C, Cavallaro F, Marano C, Falcone C, De Santis P,
Antonelli M. Accuracy of plethysmographic indices as predictors
of fluid responsiveness in mechanicallyventilated adults: a system-
atic review and meta-analysis. Intensive Care Med 2012; 38:
1429– 37
14 Guerin L, Monnet X, Teboul JL. Monitoring volume and fluid respon-
siveness: from static to dynamic indicators. Best Pract Res Clin
Anaesthesiol 2013; 27: 177– 85
15 Broch O, Bein B, Gruenewald M, et al. Accuracyof the pleth variability
index to predict fluid responsiveness depends on the perfusion
index. Acta Anaesthesiol Scand 2011; 55: 686– 93
16 Lee JY, Kim JY, Choi CH, Kim HS, LeeKC, Kwak HJ. The ability of stroke
volume variation measured by a noninvasive cardiac output
monitor to predict fluid responsiveness in mechanically ventilated
children. Pediatr Cardiol 2014; 35: 289–94
BJA Vergnaud et al.
Page 6 of 7
at JEPU ProfCORIAT on November 19, 2014http://bja.oxfordjournals.org/Downloaded from
17 Pereira de Souza Neto E, Grousson S, Duflo F, et al. Predicting fluid
responsiveness in mechanically ventilated children under general
anaesthesia using dynamic parameters and transthoracic echo-
cardiography. Br J Anaesth 2011; 106: 856– 64
18 Kubicek WG, Karnegis JN, Patterson RP, Witsoe DA, Mattson RH. De-
velopment and evaluation of an impedance cardiac output system.
Aerosp Med 1966; 37: 1208– 12
19 de Waal EE, Konings MK, Kalkman CJ, Buhre WF. Assessment of
stroke volumeindex with three different bioimpedance algorithms:
lack of agreement compared to thermodilution. Intensive CareMed
2008; 34: 735– 9
20 Fellahi JL, Caille V, Charron C, Deschamps-Berger PH, Vieillard-
Baron A. Noninvasive assessment of cardiac index in healthy
volunteers: a comparison between thoracic impedance cardiog-
raphy and Doppler echocardiography. Anesth Analg 2009; 108:
1553– 9
21 Benomar B, Ouattara A, Estagnasie P, Brusset A, Squara P. Fluid re-
sponsiveness predicted by noninvasive bioreactance-based
passive leg raise test. Intensive Care Med 2010; 36: 1875–81
22 Marik PE, Levitov A, Young A, Andrews L. The use of bioreactance
and carotid Doppler to determine volume responsiveness and
blood flow redistribution following passive leg raising in hemo-
dynamically unstable patients. Chest 2013; 143: 364– 70
23 Kupersztych-Hagege E, Teboul JL, Artigas A, et al. Bioreactance is
not reliable for estimating cardiac output and the effects of
passive leg raising in critically ill patients. Br J Anaesth 2013; 111:
961– 66
24 Keren H, Burkhoff D, Squara P. Evaluation of a noninvasive continu-
ous cardiac output monitoring system based on thoracic bioreac-
tance. Am J Physiol Heart Circ Physiol 2007; 293: H583–9
25 Aboy M, McNames J, Thong T, Phillips CR, Ellenby MS, GoldsteinB. A
novel algorithm to estimate the pulse pressure variation index
deltaPP. IEEE Trans Biomed Eng 2004; 51: 2198–203
26 Cannesson M, Slieker J, Desebbe O, et al. The ability of a novel algo-
rithm for automatic estimation of the respiratory variations in ar-
terial pulse pressure to monitor fluid responsiveness in the
operating room. Anesth Analg 2008; 106: 1195– 200
27 Cannesson M, Le Manach Y,Hofer CK, et al. Assessing the diagnostic
accuracy of pulse pressure variations for the prediction of fluid re-
sponsiveness: a ‘gray zone’ approach. Anesthesiology 2011; 115:
231– 41
28 Coste J, Pouchot J. A grey zone for quantitative diagnostic and
screening tests. Int J Epidemiol 2003; 32: 304– 13
29 Monnet X, Teboul JL. Volume responsiveness. Curr Opin Crit Care
2007; 13: 549– 53
30 Shoemaker WC, Belzberg H, Wo CC, et al. Multicenter study of
noninvasive monitoring systems as alternatives to invasive
monitoring of acutely ill emergency patients. Chest 1998; 114:
1643– 52
31 Ballestero Y, Lopez-HerceJ, Urbano J, et al. Measurement of cardiac
output in children by bioreactance. Pediatr Cardiol 2011; 32:469–72
32 Noori S, Drabu B, Soleymani S, Seri I. Continuous non-invasive
cardiac output measurements in the neonate byelectrical velocim-
etry: a comparison with echocardiography. Arch Dis Child Fetal
Neonatal Ed 2012; 97: F340– 3
33 Weisz DE, Jain A, McNamara PJ, EL-Khuffash A. Non-invasive
cardiac output monitoring in neonates using bioreactance: a com-
parison with echocardiography. Neonatology 2012; 102:61–7
34 Vergnaud E, Vidal C, Montmayeur Verchere J, et al. Noninvasive
cardiac output measurement using bioreactance in postoperative
pediatric patients. Paediatr Anaesth Advance Access published on
May 12, 2014, doi:10.1111/pan.12412
35 Julien F, Hilly J, Sallah TB, et al. Plethysmographic variability index
(PVI) accuracy in predicting fluid responsiveness in anesthetized
children. Paediatr Anaesth 2013; 23: 536–46
36 Miller-Hance W, Wilmot I, Andropoulos DB. Developmental physi-
ology of the cardiovascular system. In: Gregory GA, Andropoulos DB,
eds. Gregory’s Pediatric Anesthesia. Oxford: Wiley-Blackwell, 2012;
60–94
37 McLean JR, Inwald DP. The utility of stroke volume variability as a
predictor of fluid responsiveness in critically ill children: a pilot
study. Intensive Care Med 2014; 40: 288–9
38 Squara P, Denjean D, Estagnasie P,Brusset A, Dib JC, Dubois C. Non-
invasive cardiac output monitoring (NICOM): a clinical validation.
Intensive Care Med 2007; 33: 1191– 4
39 Chew MS, Poelaert J. Accuracy and repeatability of pediatric cardiac
output measurement using Doppler: 20-year review of the litera-
ture. Intensive Care Med 2003; 29: 1889–94
40 Wippermann CF, Huth RG, Schmidt FX, Thul J, Betancor M,
Schranz D. Continuous measurement of cardiac output by the
Fick principle in infants and children: comparison with the thermo-
dilution method. Intensive Care Med 1996; 22: 467–71
Handling editor: M. M. R. F. Struys
Fluid responsiveness prediction using NICOM in children BJA
Page 7 of 7
at JEPU ProfCORIAT on November 19, 2014http://bja.oxfordjournals.org/Downloaded from