Comparison of normal saline, hypertonic saline and hypertonic saline colloid resuscitation fluids in an infant animal model of hypovolemic shock

Page 1

E

Cr

JJa

b

c

a

ARRA

KHHSHCP

1

oDtctIci

i

eT

0d

Resuscitation 83 (2012) 1159– 1165

Contents lists available at SciVerse ScienceDirect

Resuscitation

jo u rn al hom epage : www.elsev ier .com/ locate / resusc i ta t ion

xperimental paper

omparison of normal saline, hypertonic saline and hypertonic saline colloidesuscitation fluids in an infant animal model of hypovolemic shock�

avier Urbanoa,b, Jesús López-Hercea,∗, María J. Solanaa,b, Jimena del Castilloa, Marta Botrána,ose M. Bellónc

Paediatric Intensive Care Department, Hospital General Universitario Gregorio Maranón, Instituto de Investigación Sanitaria del Hospital Gregorio Maranón, Madrid, SpainResearch Contract, Health Research Found, Health Institute “Carlos III”, Madrid, SpainPreventive and Quality Control Service, Hospital General Universitario Gregorio Maranón, Madrid, Spain

r t i c l e i n f o

rticle history:eceived 14 September 2011eceived in revised form 14 January 2012ccepted 5 February 2012

eywords:ypovolemiaemorrhagehockypertonic saline solutionolloidsediatrics

a b s t r a c t

Purpose: Incorrect resuscitation after hypovolemic shock is a major contributor to preventable pediatricdeath. Several studies have demonstrated that small volumes of hypertonic or hypertonic–hyperoncoticsaline can be an effective initial resuscitation solution. However, there are no pediatric studies to recom-mend their use. The aim of this study is to determine if in an infant animal model of hemorrhagic shock,the use of hypertonic fluids, as opposed to isotonic crystalloids, would improve global hemodynamic andperfusion parameters.Methods: Experimental, randomized animal study including thirty-four 2-to-3-month-old piglets. 30 minafter controlled 30 mL kg−1 bleed, pigs were randomized to receive either normal saline (NS) 30 mL kg−1

(n = 11), 3% hypertonic saline (HS) 15 mL kg−1 (n = 12), or 5% albumin plus 3% hypertonic saline (AHS)15 mL kg−1 (n = 11).Results: High baseline heart rate (HR) and low mean arterial pressure (MAP), cardiac index (CI), braintissue oxygenation index (bTOI), and lactate were recorded 30 min after volume withdrawal, with no sig-nificant differences between groups. Thirty minutes after volume replacement there were no significant

differences between groups for HR (NS, 188 ± 14; HS, 184 ± 14; AHS, 151 ± 14 bpm); MAP (NS, 80 ± 7;HS, 86 ± 7; AHS, 87 ± 7 mmHg); CI (NS, 4.1 ± 0.4; HS, 3.9 ± 0.4; AHS, 5.1 ± 0.4 mL min−1 m−2); lactate (NS,2.8 ± 0.7; HS, 2.3 ± 0.6; AHS, 2.4 ± 0.6 mmol L−1); bTOI (NS, 43.9 ± 2.2; HS, 40.1 ± 2.5; AHS, 46.1 ± 2.3%).Conclusions: In this model of hypovolemic shock, hypertonic fluids achieved similar end-points as twicethe volume of NS. Animals treated with albumin plus hypertonic saline presented prolonged increase inblood volume parameters and recovery of the oxygen debt.

. Introduction

Hypovolemic shock, often due to diarrhea or hemorrhage sec-ndary to injury, is one of the main causes of childhood death.1

espite the fact that incorrect resuscitation being a major con-ributor in the preventable pediatric death, few studies have beenonducted in pediatric population.2 The aim of resuscitation iso restore organ perfusion, not only hemodynamic parameters.3

nternational guidelines on the treatment of hypovolemic shock inhildren recommend administering a fast bolus of 20 mL kg−1 ofsotonic crystalloid when peripheral perfusion is inadequate, even

� A Spanish translated version of the summary of this article appears as Appendixn the final online version at doi:10.1016/j.resuscitation.2012.02.003.∗ Corresponding author at: Paediatric Intensive Care Department, Hospital Gen-

ral Universitario Gregorio Maranón, C/Doctor Castelo 47, 28009 Madrid, Spain.el.: +34 915290308; fax: +34 915868018.

E-mail address: [email protected] (J. López-Herce).

300-9572/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.resuscitation.2012.02.003

© 2012 Elsevier Ireland Ltd. All rights reserved.

if the blood pressure is normal and repeating with a second bolusof 20 mL kg−1 if heart rate, conscious level, and capillary filling donot improve.4 Prehospital fluid resuscitation is limited by difficultyin delivering large volumes of fluid in the field, increasing the totaltime of prehospital care.5,6 Several studies have demonstrated thatsmall volumes of hypertonic saline or hypertonic–hyperoncoticsaline can be an effective initial resuscitation solution.7–9 Hyper-tonic fluids can produce a greater increase in cardiac output andin peripheral perfusion than isotonic crystalloid solutions, a favor-able modulation of the immune system, and an improvement insurvival.10–12 However, there have been conflicting results fromclinical studies.13–15 There are very few studies that have analyzedthe effect in children and in infant animal models, and insufficientdata to make recommendations for or against the use of hyper-

tonic or hyperoncotic fluids in the treatment of shock associatedwith hypovolemia.16

The objective of this study was to compare the efficacy of normalsaline (NS), 3% hypertonic saline (HS), and the combination of 5%

Page 2

1 itation 83 (2012) 1159– 1165

amiw

2

twSIe

2

p(sftltearat

2

(asti(wia(cttgTw(bwBgacat

2

CfpO

160 J. Urbano et al. / Resusc

lbumin and 3% hypertonic saline (AHS) in an experimental animalodel of hypovolemic shock. Our hypothesis is that hypertonic flu-

ds would achieve similar hemodynamic and perfusion endpointsith less volume infusion than normal saline.

. Materials and methods

The experimental protocol was approved by the local Insti-utional Ethics Committee for Animal Research. The experimentsere performed in the Department of Experimental Medicine and

urgery, Gregorio Maranón University Hospital, Madrid, Spain.nternational guidelines for ethical conduct in the care and use ofxperimental animals were applied throughout the study.

.1. Anesthesia and instrumentation

Thirty-four healthy 2-to-3-month-old (9.8 ± 2.0 kg) Marylandigs were used. After premedication with intramuscular ketamine15 mg kg−1) and atropine (0.02 mg kg−1) and monitoring, anesthe-ia was induced by intravenous boluses of propofol (5 mg kg−1),entanyl (5 �g kg−1) and atracurium (0.5 mg kg−1). Ventilation washen maintained using a mechanical ventilator (Dräger SA2, Baby-og N, Lubeck, Germany) with a respiratory rate of 20 breaths min−1,idal volume of 10 mL kg−1, FiO2 of 40%, and positive end-xpiratory pressure of 3 cm H2O. Ventilation was adjusted tochieve PaCO2 between 35 and 45 mmHg. Sedation and muscleelaxation (propofol 10 mg kg−1 h−1, fentanyl 10 �g kg−1 h−1, andtracurium 2 mg kg−1 h−1, by continuous infusion) were main-ained throughout the procedure.

.2. Hemodynamic and perfusion monitoring

Monitoring included ECG, peripheral oxygen saturationVisconnet® monitor, RGB Madrid, Spain), respiratory volumesnd pressures, and FiO2 and EtCO2 measured by means of apirometer connected to the endotracheal tube and an S5® moni-or (Datex Ohmeda, Madison, WI, USA). Brain tissue oxygenationndex (bTOI) was monitored by near-infrared spectroscopy (NIRS)INVOS® Cerebral Oximeter monitor, Somanetics, Troy, MI, USA)ith sensors positioned on the skin of forehead. A 4F catheter was

nserted into the femoral artery to measure the blood pressurend cardiac output using a femoral arterial thermodilution systemPiCCO®, Pulsion Medical Systems, Munich, Germany). A 7.5Fatheter was inserted into the external jugular vein to measurehe central venous pressure (CVP) and a 4F catheter was insertedo measure jugular venous oxygen saturation (SjvO2). To measureastric intramucosal pH (pHi), a 7F tonometric catheter (TRIP,onometrics Division, Instrumentarium Corp., Helsinki, Finland)as passed into the stomach and connected to an S5® Monitor

Datex-Ohmeda, Madison, USA). No histamine type 2 (H2) receptorlockers were administered. The tonometer balloon was filledith air, and automatic sampling was performed every 10 min.lood gases were analyzed using the GEM Premier 3000® bloodas analyzer (Instrumentation Laboratory, Lexington, KY, USA),nd standard complete blood counts, coagulation studies, and bio-hemistry (including electrolytes, total protein, albumin, aspartateminotransferase, alanine aminotransferase, urea, creatinine, androponin) were performed.

.3. Experimental design

The experiment was divided into five different periods (Fig. 1):

annulation, controlled hemorrhage, stabilization, infusion, and

ollow up for 30 and 60 min after infusion. Following surgicalreparation, the animals were allowed to stabilize for 30 min.nce a steady state was achieved and baseline (baseline) data

Fig. 1. Experimental timeline.

were gathered, hypovolemic shock was induced by the withdrawalof 30 mL kg−1 of blood over 30 min (shock30′). After a 30-minstabilization period (Res0′), animals were randomized to receivean intravenous bolus of 30 mL kg−1 of NS, 15 mL kg−1 of HS, or15 mL kg−1 of AHS. The fluid was administered over a period of30 min (Res30′). During the next hour, no resuscitation effort wasmade, and parameters were recorded each 30 min (Res60′ andRes90′, respectively). HS was prepared by the addition of 12.5 mL of20% NaCl to each 100 mL of NS. AHS was prepared by the additionof 22 mL of 20% NaCl plus 50 mL of 20% albumin to each 128 mL ofNS.

The following parameters were recorded at baseline and every30 min during the experiment: blood temperature, inspiratory tidalvolume, EtCO2, cardiac rhythm, heart rate (HR), systolic and dia-stolic blood pressures, mean arterial pressure (MAP), CVP, cardiacindex (CI), intrathoracic blood volume index (ITBI), global enddiastolic volume index (GEDVI), systolic volume index (SVI), leftventricular contractility (Dt/Dpmax), systemic vascular resistanceindex (SVRI), extravascular lung water index (ELWI), stroke volumevariation (SVV), peripheral hemoglobin saturation, gastric pHi, andbTOI by NIRS. Cardiac output (CO) was taken as the average of 2consecutive measurements using 5 mL boluses of 0.9% NS at a tem-perature below 8 ◦C administered via the central venous catheter.Arterial, venous, and jugular bulb blood gases and lactate concen-tration were also determined every 30 min. On completion of theexperiment, all successfully resuscitated animals were sacrificedby the administration of sedative overdose and the intravenousinjection of potassium chloride.

In order to achieve balance and reduce bias in the allocation ofparticipants to treatment arms, three blocks of different random-ized sequences of 12 assignments to the three treatments wereperformed. The randomization scheme was generated by a com-puter program in the web site http://www.randomization.com.Investigators were not blind to the treatment.

2.4. Statistics

The statistical analysis was performed using the SPSS statisti-cal package, version 18.0 (SPSS Inc., Chicago, USA). The Pearson’sChi-squared test and the Fisher’s exact test were used for qualita-tive variables analysis. Multiple analysis of variance (MANOVA) forrepeated measures was used to study the changes in the parametersover the course of the experiment and between-group comparison.The Bonferroni test was used to adjust multiple comparisons. Dataare shown as means and standard deviations, otherwise specified.p Values less than 0.05 were considered significant.

3. Results

A total of 34 animals were included in the study. Eleven pigletsreceived NS, 12 HS, and 11 AHS. The study groups did not differ withrespect to weight (NS group, 9.3 ± 1.3 kg; HS group, 10.2 ± 2.0 kg;AHS group, 10.1 ± 2.6 kg; p = 0.525) or baseline hemodynamic,

Page 3

J. Urbano et al. / Resuscitation 83 (2012) 1159– 1165 1161

Table 1Differences among baseline, hemorrhage, infusion and 30 min evolution.

Parameter Group Baseline–Res0′ Res0′–Res30′ Res30′–Res60′ Res60′–baseline

Mean p Mean p Mean p Mean p

HR(bpm)

NS −74 0.000 22 0.070 −7 1.000 −59 0.001HS −70 0.000 21 0.127 −10 1.000 −59 0.001AHS −50 0.000 18 0.296 5 1.000 −26 1.000

MAP(mmHg)

NS 16 0.237 −19 0.080 12 0.327 −9 1.000HS 10 1.000 −19 0.019 7 1.000 2 1.000AHS 19 0.068 −19 0.008 2.5 1.000 −2 1.000

CI(L min−1 m−2)

NS 1.4 0.005 −2.0 0.002 1.0 0.004 −0.4 1.000HS 1.5 0.003 −1.9 0.005 0.7 0.109 −0.3 1.000AHS 1.8 0.000 −2.9 0.000 −0.6 0.187 0.5 1.000

GEDVI(mL m−2)

NS 126 0.001 −119 1.000 74 0.226 −81 1.000HS 131 0.001 −67 1.000 22 1.000 −86 1.000AHS 169 0.000 −301 0.071 45 1.000 87 1.000

SVI(mL m−2)

NS 17.8 0.000 −13.9 0.003 8.4 0.037 −12.2 0.000HS 20.8 0.000 −11.8 0.021 4.7 0.000 −13.7 0.000AHS 20.4 0.000 −20.7 0.000 3.6 1.000 −3.3 1.000

SVRI(dyn s cm−5 m−2)

NS 296 1.000 333 1.000 41 1.000 152 1.000HS −526 0.345 629 0.189 −129 1.000 27 1.000AHS −486 0.722 798 0.040 −175 1.000 −136 1.000

pHi NS 0.22 0.001 −0.36 1.000 −0.04 1.000 −0.15 0.080HS 0.10 0.800 0.00 1.000 −0.06 0.060 −0.05 1.000AHS 0.12 0.102 0.01 1.000 −0.03 1.000 −0.11 0.360

SvO2

(%)NS 38.1 0.000 −21.2 0.004 7.9 0.707 −24.8 0.001HS 16.0 0.049 −12.2 0.378 6.3 1.000 −10.2 0.937AHS 23.9 0.001 −19.6 0.100 5.2 1.000 −9.4 1.000

bTOI(%)

NS 8.7 0.124 −3.9 1.000 3.2 1.000 −8.0 0.207HS 9.0 0.143 −3.3 1.000 3.3 1.000 −9.0 0.128AHS 6.5 0.636 0.7 1.000 1.6 1.000 −8.8 0.670

Lactate(mmol L−1)

NS −2.6 0.000 0.9 0.001 0.2 1.000 1.5 0.060HS −1.7 0.237 0.4 1.000 0.1 1.000 1.3 0.163AHS −1.8 1.000 0.3 1.000 0.4 0.620 1.0 0.750

Baseexcess(mmol L−1)

NS 7.6 0.068 −0.3 1.000 −0.8 0.096 −6.4 0.000HS −0.3 0.005 0.7 1.000 −0.7 0.315 −4.2 0.012AHS 4.1 0.003 1.1 0.101 −1.8 0.000 −3.5 0.078

Na(mmol L−1)

NS 0.8 0.000 −1.8 0.013 0.9 0.134 0.8 1.000HS −0.1 0.001 −10.8 0.000 2.1 0.000 8.8 0.000AHS 0.5 0.001 −9.3 0.000 1.1 0.134 7.7 0.000

H −1 m−

( ramuco lactat

pT(

3

hpvsebFiOrfsrbT

R, heart rate (bpm); MAP, mean arterial pressure (mmHg); CI, cardiac index (L minmL m−2); SVRI, systemic vascular resistance index (dyn s cm−5 m−2); and pHi, intxygenation index by near infrared spectroscopy (NIRS); and lactate, arterial blood

erfusion, and laboratory parameters. For supplementary data inables S1–S3 the reader is referred to the web version of the articleAppendix A published online).

.1. Response to acute hypovolemia

Following volume withdrawal, the animals presented deepemodynamic changes regarding HR, MAP, CI, and blood volumearameters (GEDVI and ITBI), although no statistically significantariations were detected in CVP or SVRI values. Systemic perfu-ion parameters such as central venous saturation (SvO2), basexcess (BE), lactate and gastric pHi were affected equally. SjvO2,ut not bTOI, showed significant changes (Table 1 and Table S2,ig. 4). Albumin concentration decreased and troponin levelsncreased significantly, with no differences between the groups.ther parameters, including core temperature, PaO2, PaCO2, arte-

ial oxygen saturation, EtCO2, electrolytes, and kidney and liverunction parameters, remained stable (Table S3). There were no

ignificant differences between groups regarding hemodynamic,espiratory, perfusion parameters and blood-gas profiles at theeginning of the infusion except for gastric pHi (Tables S1–S3).he NS group presented significantly lower gastric pHi and higher

2); GEDVI, global end diastolic volume index (mL m−2); SVI, systolic volume indexosal gastric pH. SvO2, mixed venous blood oxygen saturation; bTOI, brain tissue

e (mmol L−1).

gastric intramucosal PCO2 values than the other groups beforestarting the infusion.

3.2. Hemodynamic response to volume expansion (Table 1 andTable S1, Figs. 2 and 3)

Volume expansion produced a significant increase in the CI,MAP, ITBI, GEDVI, and SVI, which reached values similar to baselineat the end of the volume expansion phase. Fluid infusion producedan initial fall in the HR. No significant variations were observedin SVRI. During the post infusion observation period the CI wasmaintained in all the groups, whereas blood volume parameters(ITBI, GEDVI) and SVI progressively decreased, and HR increased inanimals treated with crystalloids (NS and HS) in comparison withbaseline and immediate postinfusion values. Similarly, MAP val-ues progressively decreased in the group treated with NS; whereasthese changes were not observed in the animals that received

hypertonic colloid (AHS). Moreover, AHS group showed SVI valueshigher than NS group 30 min after the end of infusion, and SVI andGEDVI values significantly higher than in HS group 60 min after theend of the infusion. No significant differences were found between

Page 4

1162 J. Urbano et al. / Resuscitation 83 (2012) 1159– 1165

0

20

40

60

80

100

120

Baseline Shock30' Res0' Res30' Res60' Res90'

NS HS AHS

§

†

##

† †

#

††

0

1

2

3

4

5

6

7

Baseline Shock30' Res0' Res30' Res60' Res90'

NS HS AHS

##

#

††

†

†

† §

§

§

#

†

##

0

50

100

150

Heart

Rate

(b

pm

)C

ard

iac In

dex (

L/m

in/m

2)

Mean

Art

eri

al P

ressu

re (

mm

Hg

)

200

250 A

B

C

Baseline Shock30' Res0' Res30' Res60' Res90'

NS HS AHS

##

##

###

†

###

##

Fig. 2. Heart rate (A), cardiac index (B) and mean arterial pressure (C) at baselineand during the evolution of the experiment: end of controlled bleeding (Shock 30′);beginning of infusion, 30 min after the end of controlled bleeding (Res 0′); end ofinfusion (Res 30′); follow up 30 min after the end of the infusion (Res 60′); follow up6bg

td

3e

litsgguau

A

B

0

500

1.000

1.500

2.000

2.500

3.000

SV

RI (d

yn

e-s

ec

/cm

5/m

2)

Glo

bal E

nd

Dia

sto

lic V

olu

me In

dex

(ml/m

2)

Baseline Shock30' Res0' Res30' Res60' Res90'

NS HS AHS

†

‡

0

100

200

300

400

500

600

Baseline Shock30' Res0' Res30' Res60' Res90'

NS HS AHS

†

‡

# #

##

#

†

†

#

#

#

§

Fig. 3. Systemic vascular resistance index (SVRI) (A), and global end diastolic index(GEDVI) (B) at baseline and during the evolution of the experiment: end of controlledbleeding (Shock 30′); beginning of infusion, 30 min after the end of controlled bleed-ing (Res 0′); end of infusion (Res 30′); follow up 30 min after the end of infusion (Res60′); follow up 60 min after the end of infusion (Res 90′). (#) Significant difference(p < 0.05) from baseline, same group. (†) p < 0.05 from hemorrhage, same group. (*)

0 min after the end of infusion (Res 90′). (#) Significant difference (p < 0.05) fromaseline, same group. (†) p < 0.05 from hemorrhage, same group. (*) p < 0.05 fromroup NS, (‡) p < 0.05 from group HS. (§) p < 0.05 from infusion, same group.

he treatment groups except in SVI and GEDVI as has already beenescribed above.

.3. Response of cerebral and systemic perfusion to volumexpansion (Table 1 and Table S2; Figs. 4 and 5)

On completion of the volume expansion, the SvO2 rose to base-ine values in all groups. After this, a progressive fall was observedn the animals that received NS. Sixty minutes after completion ofhe infusion, only the groups treated with hypertonic fluids pre-ented values comparable to baseline values; the values in the AHSroup at this time point were significantly higher than those of the

roup treated with NS. As observed with the SvO2, the SjvO2 val-es increased after the infusion to levels similar to baseline values,nd subsequently fell. BE levels increased in all groups after vol-me expansion, with no differences between the groups; baseline

p < 0.05 from group NS, (‡) p < 0.05 from group HS. (§) p < 0.05 from end of infusion,same group.

values were not reached. The median lactate clearance 60 min aftercompleting volume expansion was greater in the groups treatedwith HS or AHS although the differences were not significant [NS:15.6% (IQR, −4.2%; 34.6%); HS: 35.3% (IQR, −25.9%; 43.4%); AHS:39.2% (IQR, 35.8%; 47.7%), p = 0.295]. Intramucosal gastric pHi in thegroup treated with NS remained lower than in the groups treatedwith hypertonic solutions. The bTOI values showed no significantchanges (Table 1 and Table S2; Fig. 5). There was a fall in the venous-arterial PCO2 gradient after volume expansion to baseline values inall groups (Table S2). There were no other significant differencesbetween the groups in the perfusion parameters.

3.4. Changes in the blood tests in response to volume expansion(Table 1 and Table S3)

As expected, serum sodium concentrations increased after theinfusion of hypertonic fluids; AHS group had higher serum albuminlevels; and serum sodium and chloride concentrations remainedsignificantly lower in the NS group until the end of the experiment.The maximum serum sodium concentrations varied between 141and 144 mmol L−1 in the NS group, between 150 and 152 mmol L−1

in the HS group, and between 153 and 154 mmol L−1 in the AHSgroup. Serum chloride concentrations increased immediately afterinfusion and remained high in the three groups 60 min later. Nochanges were observed in arterial pH, base excess, or bicarbonate

in the postinfusion period. Creatinine levels showed no varia-tions, whereas statistically significant changes were observed inthe urea levels but stayed within the normal limits. Maximumlevels of urea were identified at R90′ NS: 38.0 (2.9) mg dL−1; HS:

Page 5

J. Urbano et al. / Resuscitation 83 (2012) 1159– 1165 1163

0

10

20

30

40

50

60

70

80 A

B

C

Sv

O2

(%)

Lacta

te (

mm

ol/

L)

pH

i

Baseline Shock30' Res0' Res30' Res60' Res90'

NS HS AHS

*

#

#

#

#

†

##

†

†

# §

0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

4,0

4,5

5,0

Baseline Shock30' Res0' Res30' Res60' Res90'

NS HS AHS

#

#

#

#†

6,70

6,80

6,90

7,00

7,10

7,20

7,30

7,40

Baseline Shock30' Res0' Res30' Res60' Res90'

NS HS AHS

* *

***

† ##

#

Fig. 4. Venous blood oxygen saturation (SvO2) (A), arterial lactate (B), gastric intra-mucosal pH (pHi) (C) at baseline and during the evolution of the experiment: end ofcontrolled bleeding (Shock 30′); beginning of infusion, 30 min after the end of con-trolled bleeding (Res 0′); end of infusion (Res 30′); 30 min follow up after the end ofthe infusion (Res 60′); 60 min follow up after the end of infusion (Res 90′). (#) Signif-isi

4or

4

Nap

tpr

0

10

20

30

40

50

60

70

Bra

in T

iss

ue

ox

yg

en

ati

on

In

de

x (

%)

Baseline Shock30' Res0' Res30' Res60' Res90'

NS HS AHS

Fig. 5. Brain tissue oxygenation index by near infrared spectroscopy (bTOI) at base-line and during the evolution of the experiment: end of controlled bleeding (Shock30′); beginning of infusion, 30 min after the end of controlled bleeding (Res 0′); endof infusion (Res 30′); 30 min follow up after the end of the infusion (Res 60′); 60 min

′

cant difference (p < 0.05) from baseline, same group. (†) p < 0.05 from hemorrhage,ame group. (*) p < 0.05 from group NS, (‡) p < 0.05 from group HS. (§) p < 0.05 fromnfusion, same group.

1.1 (3.5) mg dL−1; AHS 36.8 (3.1) mg dL−1, with no differencesbserved among groups. Coagulation and liver function parametersemained stable (data not shown).

. Discussion

This is the first experimental study to compare the effects ofS, HS, and AHS in a pediatric animal model of hypovolemic shock,nd to analyze their effect on hemodynamic and tissue perfusionarameters.

In this model of hypovolemic shock, it was observed that con-rolled hemorrhage produces marked changes in hemodynamicarameters as well as in cerebral and tissue perfusion, as has beeneported previously.17 It was also observed that the infusion of

follow up after the end of infusion (Res 90 ). (#) Significant difference (p < 0.05) frombaseline, same group. (†) p < 0.05 from hemorrhage, same group. (*) p < 0.05 fromgroup NS, (‡) p < 0.05 from group HS. (§) p < 0.05 from infusion, same group.

hypertonic fluids, with or without colloids, produced a similarresponse to the infusion of twice the volume of isotonic fluid withno significant alteration of acid–base balance.

4.1. Effects of hypertonic fluids

The use of hypertonic fluids produces an immediate mobi-lization of intracellular water into the intravascular extracellularspace, increasing hydrostatic pressure and intracellular osmolal-ity. This shift normalizes the volume of the vascular endothelialcells, previously increased due in part to a dysfunction of inter-change across the cell membrane during shock, and improvesthe microcirculation.18,19 Positive hemodynamic effects have beenobserved; including a greater increase in preload, diuresis and anearly fall in afterload.20,21 However, studies of the effects on tis-sue and cerebral perfusion have been contradictory and there hasbeen no demonstrated improvement in survival.14,22,23 In order toprolong the expansor effect, hypertonic fluids have been combinedwith colloid solutions, aiming to increase the oncotic pressure of theplasma and retain mobilized fluid within the intravascular space fora longer period.23 Some authors have also found that the adminis-tration of colloids after controlled hemorrhage can improve oxygentransport to the tissues.22,24–26 However, there are no previousstudies that have analyzed resuscitation with a combination of 5%albumin plus 3% hypertonic saline.

4.2. Hemodynamic response

Despite the fact that 3 types of infusion fluid achieving similarglobal hemodynamic parameter values after expansion, these datasuggest that there was a longer increase in CI in animals treatedwith HS, and even longer with AHS. The prolonged increase ofintravascular volume plus the absence of a fall in the blood volumeparameters in the group receiving AHS leads us to consider that thealbumin remains within the intravascular compartment.26

4.3. Metabolic response. Systemic and cerebral perfusion

None of the 3 types of infusion fluid completely normalized

the parameters that estimate systemic and cerebral perfusion. Thiscould be related to a reduction in oxygen transport due to the loss ofred blood cells and hemodilution secondary to treatment, despitethe improvement in the hemodynamic situation.27

Page 6

1 itation

espit

cIlt

iibea

maonett

arddcage

4

oSpoehibdpsiossbcfao

5

ctN

1

1

164 J. Urbano et al. / Resusc

The use of SvO2 has been recommended as an objective param-ter of hemodynamic stabilization in hypovolemic shock.28 In ourtudy, only the group that received AHS achieved SvO2 values com-arable with baseline values and these values remained higher than

n the other groups for a longer period; this could suggest betterissue perfusion.

Arterial lactate concentration and venous base deficit (BD) areonsidered as parameters that reflect the oxygen debt closely.29

n our study, the 3 treatments reduced but did not normalize theactate concentration and BD, suggesting persistence of a degree ofissue oxygen debt.

Intramucosal PCO2 may be increased both by a reductionn its clearance secondary to decreased blood flow and byncreased production secondary to tissue hypoxia and anaero-ic metabolism.30–32 In our study pHi increased with volumexpansion, without reaching baseline values. This would suggest

persistence of mesenteric hypoperfusion.Near infrared spectroscopy has recently been introduced as a

ethod to measure brain tissue oxygenation in patients33 and innimal models of hypovolemic shock.34 In our study, brain tissuexygenation index (bTOI) declined moderately although not sig-ificantly after the hemorrhage, and it did not increase with thexpansion. However, the sensor position could favor signal con-amination by the sagittal venous sinus and interference by otherissues.35

Jugular venous blood oxygen saturation evaluates the bal-nce between oxygen delivery to the brain and brain oxygenequirements.36 However, only 1 study has analyzed this aspecturing hypovolemic shock and one other in septic shock.37,38 Ourata coincide with those of the study by Chen, in which a signifi-ant fall in the SjvO2 values was observed after hemorrhage, with

rise after volume expansion, suggesting that the reduced oxy-en delivery to the brain was compensated by increased oxygenxtraction.

.4. Limitations

Our study has certain limitations. These results would applynly to hypovolemic shock secondary to controlled hemorrhage.pecific studies would be necessary in other conditions as theathophysiology is not the same in different types of shock. Sec-ndary effects such as bradycardia and hypotension are well knownffects of the sedative drugs used to maintain anesthesia, and couldave attenuated the physiologic responses during shock. Mechan-

cal ventilation could have interfered with measurement of thelood volume; however, these treatments are also used in chil-ren with severe hypovolemic shock. Despite randomization ofosthemorrhage treatment, the group treated with NS presentedignificantly lower gastric pHi values than the other 2 groups afternducing hypovolemia. This could indicate a greater redistributionf mesenteric flow and greater cell damage, which could repre-ent a disadvantage compared with the other groups. The PiCCOystem is not the gold standard for evaluating the cardiac output,ut it has been shown to have an acceptable correlation and con-ordance with the method that uses a Swan Ganz catheter, withewer adverse effects and less complexity of placement in childrennd experimental animal models.39,40 Furthermore, the objectivef this study was to compare the different treatments.

. Conclusions

In conclusion, in this infant animal model of hypovolemic shock,ontrolled acute hemorrhage produced major alterations of sys-emic, cerebral, and intestinal perfusion, lactic acidosis, cerebralIRS, and gastric tonometry. Animals treated with albumin plus

1

83 (2012) 1159– 1165

hypertonic saline presented a prolonged increase in CI and bloodvolume parameters, with a more prolonged recovery of the oxy-gen debt. Both hypertonic crystalloid (HS) and hypertonic colloid(AHS) fluids achieved similar hemodynamic and end-organ perfu-sion endpoints with half the volume compared to normal saline.Acceptable levels of serum sodium were achieved without signifi-cant alterations of acid–base balance.

Although this experimental study cannot be directly extendedto clinical practice, our results suggest that the albumin hypertonicsaline (AHS) could be a better suitable liquid for the initial resusci-tation rather than normal saline or the hypertonic saline. However,before doing so, randomized clinical trials in children with hemor-rhagic shock would be necessary to test AHS as out of hospital orin-hospital initial resuscitation treatment.

Conflicts of interest statement

None of the authors have declared a conflict of interest.

Sources of funding

This study was supported by a research grant MM 0535-2007from the Mutua Madrilena and from the Maternal, Child Healthand Development Network (REDSAMID) within the framework ofthe VI National I + D + i Research Program (2008–2011).

Acknowledgments

The authors gratefully acknowledge the generous help ofYolanda Ballestero, Diego Vinciguerra, Ana García-Figueruelo andSantiago Mencía from the Pediatric Intensive Care Unit; andMercedes Adrados and Natalia Sánchez from the Department ofExperimental Medicine and Surgery of the Gregorio Maranón Gen-eral University Hospital for their collaboration in the experiments.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.resuscitation.2012.02.003.

References

1. Patton GC, Coffey C, Sawyer SM, et al. Global patterns of mortality in young peo-ple: a systematic analysis of population health data. Lancet 2009;374:881–92.

2. Dykes EH, Spence LJ, Young JG, Bohn DJ, Filler RM, Wesson DE. Preventable pae-diatric trauma deaths in a metropolitan region. J Pediatr Surg 1989;24:107–10.

3. Fiddian-Green RG, Haglund U, Gutierrez G, Shoemaker WC. Goals for the resus-citation of shock. Crit Care Med 1993;21:S25–31.

4. Rossaint R, Bouillon B, Cerny V, et al. Management of bleeding following majortrauma: an updated European guideline. Crit Care 2010;14:R52.

5. Smith JP, Bodai BI, Hill AS, Frey CF. Prehospital stabilization of critically injuredpatients: a failed concept. J Trauma 1985;25:65–70.

6. Kaweski SM, Sise MJ, Virgilio RW. The effect of prehospital fluids on survival intrauma patients. J Trauma 1990;30:1215–8.

7. Bulger EM, Jurkovich GJ, Nathens AB, et al. Hypertonic resuscitation of hypo-volemic shock after blunt trauma: a randomized controlled trial. Arch Surg2008;143:139–48.

8. Simma B, Burger R, Falk M, Sacher P, Fanconi S. A prospective, random-ized, and controlled study of fluid management in children with severe headinjury: lactated Ringer’s solution versus hypertonic saline. Crit Care Med1998;26:1265–70.

9. Wade CE, Grady JJ, Kramer GC. Efficacy of hypertonic saline dextran fluidresuscitation for patients with hypotension from penetrating trauma. J Trauma2003;54:S144–8.

0. Velasco IT, Pontieri V, Rocha e Silva Jr M, Lopes OU. Hyperosmotic NaCl andsevere haemorrhagic shock. Am J Physiol 1980;239:H664–73.

1. Pascual JL, Ferri LE, Seely AJ, et al. Hypertonic saline resuscitation of haemor-

rhagic shock diminishes neutrophil rolling and adherence to endothelium andreduces in vivo vascular leakage. Ann Surg 2002;236:634–42.

2. Younes RN, Aun F, Ching CT, et al. Prognostic factors to predict outcome fol-lowing the administration of hypertonic/hyperoncotic solution in hypovolemicpatients. Shock 1997;7:79–83.

Page 7

itation

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

J. Urbano et al. / Resusc

3. Mattox KL, Maningas PA, Moore EE, et al. Prehospital hypertonic saline/dextraninfusion for post-traumatic hypotension. The U.S.A. Multicenter Trial. Ann Surg1991;213:482–91.

4. Wade CE, Kramer GC, Grady JJ, Fabian TC, Younes RN. Efficacy of hypertonic7.5% saline and 6% dextran-70 in treating trauma: a meta-analysis of controlledclinical studies. Surgery 1997;3:609–16.

5. Shackford SR, Bourguignon PR, Wald SL, Rogers FB, Osler TM, Clark DE.Hypertonic saline resuscitation of patients with head injury: a prospective, ran-domized clinical trial. J Trauma 1998;44:50–8.

6. Kwan I, Bunn F, Roberts I. WHO Pre-Hospital Trauma Care Steering Committee.Timing and volume of fluid administration for patients with bleeding. CochraneDatabase Syst Rev 2003:CD002245.

7. López-Herce J, Rupérez M, Sánchez C, García C, García E. Haemody-namic response to acute hypovolaemia, rapid blood volume expansion andadrenaline administration in an infant animal model. Resuscitation 2006;68:259–65.

8. Guyton AC. The body fluid compartments: extracellular and intracellular flu-ids; interstitial fluid and oedema. In: Guyton AC, editor. Textbook of MedicalPhysiology. 1st ed. Philadelphia: W.B Saunders; 1991. p. 274–85.

9. Pascual JL, Khwaja KA, Chaudhury P, Christou NV. Hypertonic saline and themicrocirculation. J Trauma 2003;54:S133–40.

0. Velasco IT, Rocha e Silva M, Oliveira MA, Oliveira MA, Silva RI. Hypertonic andhyperoncotic resuscitation from severe haemorrhagic shock in dogs: a compar-ative study. Crit Care Med 1989;17:261–4.

1. Kien ND, Kramer GC, White DA. Acute hypotension caused by rapid hypertonicsaline infusion in anesthetized dogs. Anesth Analg 1991;73:597–602.

2. Braz JR, do Nascimento Jr P, Paiva Filho O, et al. The early systemic and gas-trointestinal oxygenation effects of haemorrhagic shock resuscitation withhypertonic saline and hypertonic saline 6% dextran-70: a comparative studyin dogs. Anesth Analg 2004;99:536–46.

3. Chiara O, Pelosi P, Brazzi L, et al. Resuscitation from haemorrhagic shock:experimental model comparing normal saline, dextran, and hypertonic salinesolutions. Crit Care Med 2003;31:1915–22.

4. Knotzer H, Pajk W, Maier S, et al. Comparison of lactated Ringer’s, gelatineand blood resuscitation on intestinal oxygen supply and mucosal tissue oxygentension in haemorrhagic shock. Br J Anaesth 2006;97:509–16.

5. Layon AJ, Gallagher TJ. Five percent human albumin in lactated Ringer’s solu-

tion for resuscitation from haemorrhagic shock: efficacy and cardiopulmonaryconsequences. Crit Care Med 1990;18:410–3.

6. Horstick G, Lauterbach M, Kempf T, et al. Early albumin infusion improves globaland local haemodynamics and reduces inflammatory response in haemorrhagicshock. Crit Care Med 2002;30:851–5.

4

83 (2012) 1159– 1165 1165

7. Carcillo JA, Fields Al. American College of Critical Care Medicine Task ForceCommittee Members Clinical practice parameters for haemodynamic sup-port of paediatric and neonatal patients in septic shock. Crit Care Med2002;30:1365–78.

8. Kremzar B, Spec-Marn A, Kompan L, Cerovic O. Normal values of SvO2

as therapeutic goal in patients with multiple injuries. Intensive Care Med1997;23:65–70.

9. Schmelzer TM, Perron AD, Thomason MH, Sing RF. A comparison of centralvenous and arterial base deficit as a predictor of survival in acute trauma. Am JEmerg Med 2008;26:119–23.

0. Mathias DW, Clifford PS, Klopfenstein HS. Mixed venous blood gases are superiorto arterial blood gases in assessing acid–base status and oxygenation duringacute cardiac tamponade in dogs. J Clin Invest 1988;82:833–8.

1. Gutierrez G. Cellular energy metabolism during hypoxia. Crit Care Med1991;19:619–26.

2. Miami Trauma Clinical Trials Group. Splanchnic hypoperfusion-directed thera-pies in trauma: a prospective, randomized trial. Am Surg 2005;71:252–60.

3. Soul JS, du Plessis AJ. New technologies in paediatric neurology. Near-infraredspectroscopy. Semin Pediatr Neurol 1999;6:101–10.

4. Meybohm P, Renner J, Boening A, et al. Impact of norepinephrine and fluidon cerebral oxygenation in experimental haemorrhagic shock. Pediatr Res2007;62:440–4.

5. Nemoto EM, Erdmann W, Strong E, Rao GR, Moossy J. Regional brain PO2 afterglobal ischemia in monkeys: evidence for regional differences in critical perfu-sion pressures. Stroke 1979;10:44–52.

6. Pérez A, Minces PG, Schnitzler EJ, Agosta GE, Medina SA, Ciraolo CA. Jugularvenous oxygen saturation or arteriovenous difference of lactate content andoutcome in children with severe traumatic brain injury. Pediatr Crit Care Med2003;4:33–8.

7. Chen S, Zhu X, Wang Q, et al. The early effect of Voluven, a novel hydroxyethylstarch (130/0.4), on cerebral oxygen supply and consumption in resuscitationof rabbit with acute haemorrhagic shock. J Trauma 2009;66:676–82.

8. Wohl JS, Baggs A, Lin JL, Fink MP, Dhupa N. Use of jugular venous blood, comparedwith mixed venous blood, for measurement of venous oxygenation indices in aporcine model of endotoxic shock. Am J Vet Res 1997;58:910–4.

9. López-Herce J, Rupérez M, Sánchez C, García C, García E. Correlation betweencardiac output measured by the femoral arterial thermodilution technique pul-

monary arterial and that measured by contour pulse analysis in a paediatricanimal model. J Clin Monit Comput 2006;20:19–23.

0. Gil J, Cecchetti C, Menéndez S, Cambra FJ, López-Herce J, Rodríguez-Núnez A.Preliminary clinical experience with PiCCO system in children with shock. AnPediatr (Barc) 2009;71:135–40.