Julien JabotJean-Louis TeboulChristian RichardXavier Monnet

Passive leg raising for predicting fluidresponsiveness: importance of the posturalchange

Received: 5 March 2008Accepted: 27 August 2008Published online: 16 September 2008� Springer-Verlag 2008

Electronic supplementary materialThe online version of this article(doi:10.1007/s00134-008-1293-3) containssupplementary material, which is availableto authorized users.

J. Jabot � J.-L. Teboul � C. Richard �X. Monnet ())Service de Reanimation Medicale, AP-HP,Hopital de Bicetre, Centre HospitalierUniversitaire de Bicetre, 78, rue du GeneralLeclerc, 94270 Le Kremlin-Bicetre, Francee-mail: xavier.monnet@bct.aphp.frTel.: ?33-1-45213544Fax: ?33-1-45213551

J. Jabot � J.-L. Teboul � C. Richard �X. MonnetFaculte de MedecineParis-Sud, Univ Paris-Sud, EA 4046,94270 Le Kremlin-Bicetre, France

Abstract Objective: For predict-ing fluid responsiveness by passiveleg raising (PLR), the lower limbs canbe elevated at 45� either from the 45�semi-recumbent position (PLRSEMI-

REC) or from the supine position(PLRSUPINE). PLRSUPINE could havea lower hemodynamic impact thanPLRSEMIREC since it should notrecruit the splanchnic venous reser-voir. Design: Prospective studySetting: A 24-bed medical intensivecare unit. Patients and partici-pants: A total of 35 patients withcirculatory failure who responded toan initial PLRSEMIREC by an increasein cardiac index C 10%. Interven-tions: PLRSEMIREC, a transfer fromthe semi-recumbent to the supineposition and PLRSUPINE were per-formed in all patients in a randomorder before fluid expansion (500 mLsaline). Measurements andresults: PLRSEMIREC, supine trans-fer and PLRSUPINE significantlyincreased the pulse-contour derivedcardiac index (PiCCOplus) by 22(17–28)%, 9 (5–15)% and 10(7–14)% (P \ 0.05 vs. PLRSEMIREC

for the latter two), respectively. Thesemaneuvers significantly increased theright ventricular end-diastolic area(echocardiography) by 20 (14–29)%,9 (5–16)% and 10 (5–16)% (P \ 0.05vs. PLRSEMIREC for the latter two)and the central venous pressure by 33(22–50)%, 15 (10–20)% and 20 (15–29)% (P \ 0.05 vs. PLRSEMIREC for

the latter two), respectively. Volumeexpansion significantly increasedcardiac index by 27 (21–38)% and allpatients were responders to volumeexpansion. If an increase in cardiacindex C 10% is considered as apositive response to PLRSUPINE, 15(43%) patients would have beenunduly predicted as non-responders tofluid administration by PLRSUPINE.Conclusions: PLRSEMIREC induceslarger increase in cardiac preload thanPLRSUPINE and may be preferred forpredicting fluid responsiveness.

Keywords Passive leg raising �Fluid responsiveness � Centralvenous pressure � Sepsis

Abbreviations

PLR Passive leg raisingPLRSEMIREC Passive leg raising

performed byelevating the patient’slegs and bysimultaneouslytransferring the trunkfrom the semi-recumbent position toa horizontal position

PLRSUPINE Passive leg raisingperformed byelevating the patient’slegs from the supineposition

Intensive Care Med (2009) 35:85–90DOI 10.1007/s00134-008-1293-3 ORIGINAL

Introduction

Transferring a patient to the passive leg raising (PLR)position (in which the lower limbs are elevated at 45�while the trunk is lying supine) transfers venous bloodfrom the legs to the intrathoracic compartment andincreases cardiac preload [1]. The effects of this‘‘endogenous fluid challenge’’ on aortic blood flow [2, 3]or cardiac output [4, 5] enable to test fluid responsivenesswith accuracy. As a result, leg raising has been recom-mended as a part of hemodynamic monitoring by recentinternational recommendations [6].

In the previous studies dealing with PLR, the posturalchange ended at a similar PLR position (lower limbselevated at 45� and trunk in supine position). However, atbaseline, the position was either semi-recumbent (trunkelevated at 45� (PLRSEMIREC) [3, 4] or at 30� [5]) orsupine (PLRSUPINE) [2, 7] (Fig. 1). Compared toPLRSEMIREC, PLRSUPINE could have a lower hemody-namic impact since it should not recruit the blood of thesplanchnic reservoir. Nonetheless, PLRSUPINE could beconsidered as an alternative to PLRSEMIREC in someparticular conditions such as increased abdominal pres-sure in which PLRSEMIREC can be contraindicated.

We compared the hemodynamic effects of PLRSEMI-

REC and of PLRSUPINE [8] and compared the ability of thetwo methods for predicting fluid responsiveness in viewof standardizing the technique.

Materials and methods

Study population

We included patients if:

1. They presented an acute circulatory failure due tosepsis [9] (see ESM),

2. A prior PLRSEMIREC maneuver had increased the pulsecontour-derived cardiac output (PiCCOplus� devicev6.0, Pulsion Medical SystemsTM, Munich, Germany)by more than 10%.

Non-inclusion criteria were deep venous thrombosis orelastic compression stocking, head trauma or an increasein the intra-abdominal pressure suspected by clinicalcontext and examination. We included 35 patients whosecharacteristics are listed in Table 1. All patients or next ofkin gave their deferred consent.

Study design

After the initial PLRSEMIREC, the patients were transferredin the semi-recumbent position. They were then submittedto the following postural maneuvers (Fig. 1):

• a PLRSEMIREC which consisted of transferring thepatients from the 45� semi-recumbent position to thePLR position;

• a ‘‘supine transfer’’ which consisted of transferring thepatients from the 45� semi-recumbent position to thesupine position followed by a PLRSUPINE whichconsisted of transferring the patients from the supineposition to the PLR position (Fig. 1).

These two sequences of postural changes were 2-minlong each [3, 4] and were performed successively in eachpatient in a randomised order. Finally, 500 mL salinewere administered intravenously over 10 min.

PLRSEMIREC

PLRSEMIREC

PLRSUPINE

PLRSUPINESUPINE TRANSFER

SUPINE TRANSFER

PRE-INCLUSIONPLR

VOLUME EXPANSION

BASE1 BASE3BASE2

BASE3BASE1BASE2

Fig. 1 Study protocol with the two sequences of postural changesat which the patients were randomized. PLR Passive leg raising,PLRSEMIREC passive leg raising performed by elevating thepatient’s legs and by simultaneously transferring the trunk fromthe semi-recumbent position to a horizontal position, PLRSUPINE

passive leg raising performed by elevating the patient’s legs fromthe supine position

Table 1 Characteristics of the patients (n = 35)

Age (years) 63 (32–82)Gender (M/F) 18/17SAPS II 40 (33–100)ARDS (n, %) 25 (71)Reason for ICU admissionCommunity acquired pneumonia 12 (34)Gastric bleeding 5 (14)Mesenteric ischemia 3 (8)Post cardiac arrest 3 (8)Diarrhea and hypovolemia 2 (6)Pyelonephritis 2 (6)Tumor lysis syndrome 2 (6)Acute alcoholic hepatitis 2 (6)Peritonitis 2 (6)Status epilepticus 2 (6)

LV ejection fraction 45 (30–50)Sepsis (n, %) 35 (100)Vasopressors (n, %) 24 (69)Norepinephrine 23Dose of norepinephrine (lg/kg per min) 0.84 (0.24–3.33)Dobutamine 1

Intubation (n, %) 30 (86)Triggering of the ventilator (n) 5

Cardiac arrhythmias (n, %) 8 (23)

Data are expressed as median (25–75 interquartile) or as n (%)

86

Measurements

After each position change and after volume expansion, werecorded the heart rate, the systemic arterial pressure, thepulse contour-derived cardiac output (PiCCOplus� devicev6.0, Pulsion Medical SystemsTM, Munich, Germany), thecentral venous pressure and the right ventricular end-dia-stolic area by an apical 4-chamber view at transthoracicechocardiography (EnVisor B.0, Philips Medical System,Andover, MA). The pulse contour-derived cardiac outputwas calibrated by transpulmonary thermodilution atbaseline.

Data analysis

Patients in whom the fluid administration induced anincreased cardiac index by more than 15% were definedas ‘‘responders’’ to fluid administration [3, 10–12]. In anattempt to estimate the endogenous volume recruited bythe PLRSEMIREC, we noted during fluid infusion the vol-ume that had been infused at the time when cardiac outputprecisely reached the value that it had reached during thePLRSEMIREC maneuver. We tested whether fluidresponders were detected by a PLRSUPINE-inducedincrease in cardiac index by more than 10%, which wasthe cut-off value of cardiac index increase in response toPLRSEMIREC previously reported to discriminate fluidresponders and non responders [3, 4].

Comparisons of hemodynamic variables betweenbefore versus during PLR, between before versus aftersupine transfer and between before versus after volumeexpansion, comparisons of the changes in hemodynamicvariables induced by PLRSUPINE, PLRSEMIREC, supinetransfer and volume expansion were assessed using apaired Student t test (Statview 5.0 software, Abacusconcepts, Berkeley, CA). Data were expressed as median(25–75% interquartile range). A P value B 0.05 wasconsidered statistically significant.

Results

Results are expressed like if in all patients the sequence ofpostural changes would have been: PLRSEMIREC, supinetransfer and then PLRSUPINE.

The PLRSEMIREC maneuver, the supine transfer andthe PLRSUPINE maneuver increased cardiac index by 22(17–28), 9 (5–15) and 10 (7–14)%, respectively (Table 2;Fig. 2).

The PLRSEMIREC maneuver increased the right ven-tricular end-diastolic area by 20 (14–29)% and the centralvenous pressure by 33 (22–50)%. The supine transferincreased the right ventricular end-diastolic area by 9 (5–16)% and the central venous pressure by 15 (10–20)%. T

ab

le2

Ev

olu

tio

no

fh

emo

dy

nam

icv

aria

ble

sd

uri

ng

the

dif

fere

nt

po

stu

ral

chan

ges

Bas

e1

PL

RS

EM

IRE

CB

ase

2S

UP

INE

PL

RS

UP

INE

Bas

e3

Po

stV

E

Hea

rtra

te(b

eats

/min

)9

6(8

4–

11

6)

96

(84

–1

13

)9

8(8

4–

11

4)

97

(83

–1

11

)9

8(8

3–

11

5)

98

(85

–1

13

)9

7(8

4–

11

3)§

Sy

sto

lic

arte

rial

pre

ssu

re(m

mH

g)

10

8(9

4–

11

6)

12

0(1

07

–1

28

)*1

09

(92

–1

17

)1

07

(98

–1

22

)1

15

(10

9–

12

7)#

10

4(9

5–

11

5)

12

1(1

10

–1

30

)§

Mea

nar

teri

alp

ress

ure

(mm

Hg

)6

8(6

4–

77

)7

5(6

9–

84

)*7

0(6

2–

76

)6

8(6

3–

79

)7

3(6

9–

85

)#6

8(6

1–

77

)7

8(6

7–

84

)§

Car

dia

cin

dex

(L/m

inp

erm

2)

2.4

(2.1

–3

.0)

3.0

(2.7

–4

)*2

.4(2

.0–

3.0

)2

.7(2

.3–

3.3

)�2

.9(2

.5–

4)#

2.4

(2.1

–3

.0)

3.2

(2.6

–4

)§

Rig

ht

ven

tric

ula

ren

d-d

iast

oli

car

ea(c

m2)

16

(15

–1

8)

19

(19

–2

1)*

17

(15

–1

8)

18

(17

–1

9.8

)�2

1(1

9–

22

)#1

6(1

5–

17

)2

2(2

0–

23

)§

Cen

tral

ven

ou

sp

ress

ure

(mm

Hg

)7

(6–

8)

9(9

–1

0)*

7(5

–8

)8

(6–

9)�

9(7

–1

0)#

6(6

–8

)1

1(9

–1

1)§

Dat

aar

eex

pre

ssed

asm

edia

n(2

5–

75

inte

rqu

arti

le),

(n=

35

)*P

\0

.05

,P

LR

SE

MIR

EC

ver

sus

Bas

e1;

�P

\0

.05

,S

UP

INE

ver

sus

Bas

e2;

#P

\0

.05

,P

LR

SU

PIN

Ev

ersu

sS

UP

INE

;§

P\

0.0

5P

ost

VE

ver

sus

Bas

e3P

LR

SE

MIR

EC

pas

siv

ele

gra

isin

gp

erfo

rmed

by

elev

atio

no

fth

ele

gs

fro

mth

ese

mi-

recu

mb

ent

po

siti

on

,P

LR

SU

PIN

Ep

assi

ve

leg

rais

ing

per

form

edb

yel

evat

ion

of

the

leg

sfr

om

the

sup

ine

po

siti

on

,P

ost

VE

po

st-v

olu

me

exp

ansi

on

87

The PLRSUPINE maneuver increased the right ventricularend-diastolic area by 10 (5–16)% and the central venouspressure by 20 (15–29)% (Table 2; Fig. 3).

Volume expansion significantly increased cardiacindex by 27 (21–38)%. All patients were responders tovolume expansion. In all patients, the PLRSEMIREC

maneuver of the protocol had increased cardiac index bymore than 10%, i.e. that maneuver reproduced the effectsof the PLRSEMIREC performed before inclusion in thestudy. The effects of the PLRSEMIREC on cardiac indexwere reached when 312 (250–350)mL of saline had beeninfused.

The PLRSUPINE increased cardiac index by more than10% in only 20 patients. Thus, if an increase in cardiacindex by more than 10% would have been considered as apositive response to PLRSUPINE as it is used forPLRSEMIREC [3], 15 (43%) patients would have beenunduly predicted as non-responders to fluid administra-tion by a PLRSUPINE test. In these 15 patients, thePLRSUPINE had increased cardiac index by 5 (2–8)%while the subsequent volume expansion increased cardiacindex in a lower extent than in the other ones [23 (19–27)vs. 33 (24–45)%, respectively, P \ 0.05].

Discussion

This study demonstrated (1) that cardiac preload andcardiac index increased more during PLRSEMIREC thanduring PLRSUPINE, (2) that the 10% increase in cardiac

index, which has been previously found as the diagnosticcut-off for predicting fluid responsiveness by thePLRSEMIREC test [3, 4], could not be used for thePLRSUPINE test.

Lifting the legs up to 45� was supposed to transfervenous blood from the legs toward the intrathoraciccompartment [13], increasing the intrathoracic bloodvolume [14] and inducing an increase in the cardiacpreload [7, 15] as confirmed in the present study by themarked increase in central venous pressure and in theright ventricular end-diastolic area during PLR. As aclinical application, PLR has been developed as an‘‘endogenous fluid challenge’’ [16, 17] that we found tobe of around 300 mL. PLR predicts fluid responsivenesswith accuracy, especially in case of arrhythmias orspontaneous breathing [3–5].

However, there is a debate regarding the baselineposture before moving the patient to the final PLR posi-tion (lower extremities at 45� and trunk supine) [1]. Thepresent study evidenced that PLRSEMIREC exerted larger‘‘auto-fluid loading’’ effect than PLRSUPINE. All patientsof the study were preload responsive and, as a conse-quence of a larger increase in preload, the increase incardiac index was significantly larger during PLRSEMIREC

than during PLRSUPINE. Interestingly, the difference incardiac preload and cardiac index changes found betweenthe two postural maneuvers corresponded to thoseinduced by transferring the patient’s trunk from the semi-recumbent to the supine position before performing the

0

25

12.5

37.5

PLRSEMIRECSUPINE

TRANSFERPLRSUPINE

VOLUMEEXPANSION

% c

hang

e in

CI f

rom

bas

elin

e **

**

Fig. 2 Changes in cardiac index (CI) induced by the differentpostural changes, expressed in % change [median (25–75IR)] fromthe value measured before the postural change, n = 35; PLRPassive leg raising, PLRSEMIREC passive leg raising performed byelevating the patient’s legs and by simultaneously transferring thetrunk from the semi-recumbent position to a horizontal position,PLRSUPINE passive leg raising performed by elevating the patient’slegs from the supine position

0

50

25

75

PLRSEMIRECSUPINE

TRANSFERPLRSUPINE

VOLUME EXPANSION

**

% c

hang

e in

CV

P fr

omba

selin

e

**

Fig. 3 Changes in central venous pressure (CVP) induced by thedifferent postural changes, expressed in % change [median(25–75IR)] from the value measured before the postural change,n = 35; PLR Passive leg raising, PLRSEMIREC passive leg raisingperformed by elevating the patient’s legs and by simultaneouslytransferring the trunk from the semi-recumbent position to ahorizontal position, PLRSUPINE passive leg raising performed byelevating the patient’s legs from the supine position

88

PLRSUPINE, suggesting that PLRSEMIREC induced theadditional recruitment of the vast splanchnic reservoir.

Importantly, if a 10% increase in cardiac index wasused for defining a ‘‘positive’’ PLRSUPINE test, it wouldhave falsely classified more than 40% of the patients asfluid non-responders while they actually were responders.In those ‘‘false negative’’ patients, volume expansionincreased cardiac output by a lower extent than in theother ones, indicating that their preload dependence wasof lower degree and suggesting that the PLRSUPINE testwas not sensitive enough to detect it.

The main limitation of our study is that for ethicalreasons, we included only responders to volumeexpansion such that it was not be possible to directlycompare the accuracy of the PLR methods and todetermine the cut-off for the PLRSUPINE test. We didnot recalibrate pulse contour cardiac output by ther-modilution at each step of the protocol to avoid theinjection of a large amount of fluid. Nevertheless, thepotential resulting error in cardiac output measurementlikely affected both postural methods, such that itshould have not altered our results. Moreover, the pulsecontour analysis, which was reported to be far more

reproducible than transpulmonary thermodilution [18]should be quite appropriate to allow a correct inter-pretation of the PLR tests [19]. Finally, weadministered the same amount of fluid in all patients,what should have been excessive in some patients andnot enough in some others. In this regard, it should beadvised to repeat PLR at the end of fluid administrationfor testing the need for additional fluid administration.

To summarize, when PLR started from the supineposition, its increasing effects on cardiac preload andcardiac index were lower than if it started from the semi-recumbent position. Moreover, the same cut-off valuecould not be used interchangeably for the two PLRmethods. This strongly suggests that the PLRSEMIREC

should be preferred for assessing fluid responsiveness inpatients with septic circulatory failure.

Acknowledgments The authors are greatly indebted to Mrs.Alexia Letierce, from the Unite de Recherche Clinique of theBicetre Hospital for helpful statistical reviewing of the manuscript.Prof. Jean-Louis Teboul and Dr. Xavier Monnet are members of theMedical Advisory Board of the Pulsion Medical Systems Company.The other authors have no financial interest to disclose.

References

1. Monnet X, Teboul JL (2008) Passiveleg raising. Intensive Care Med34:659–663

2. Lafanechere A, Pene F, Goulenok C,Delahaye A, Mallet V, Choukroun G,Chiche J, Mira J, Cariou A (2006)Changes in aortic blood flow inducedby passive leg raising predict fluidresponsiveness in critically ill patients.Crit Care 10:R132

3. Monnet X, Rienzo M, Osman D,Anguel N, Richard C, Pinsky MR,Teboul JL (2006) Passive leg raisingpredicts fluid responsiveness in thecritically ill. Crit Care Med 34:1402–1407

4. Lamia B, Ochagavia A, Monnet X,Chemla D, Richard C, Teboul JL (2007)Echocardiographic prediction ofvolume responsiveness in critically illpatients with spontaneously breathingactivity. Intensive Care Med 33:1125–1132

5. Maizel J, Airapetian N, Lorne E,Tribouilloy C, Massy Z, Slama M(2007) Diagnosis of centralhypovolemia by using passive legraising. Intensive Care Med 33:1133–1138

6. Antonelli M, Levy M, Andrews PJ,Chastre J, Hudson LD, Manthous C,Meduri GU, Moreno RP, Putensen C,Stewart T, Torres A (2007)Hemodynamic monitoring in shock andimplications for management.International Consensus Conference,Paris, France, 27–28 April 2006.Intensive Care Med 33:575–590

7. Boulain T, Achard JM, Teboul JL,Richard C, Perrotin D, Ginies G (2002)Changes in BP induced by passive legraising predict response to fluid loadingin critically ill patients. Chest121:1245–1252

8. Monnet X, Jabot J, Richard C, TeboulJL (2008) Passive leg raising forpredicting fluid responsiveness.Importance of the postural change. AmJ Respir Crit Care Med 177: A371

9. Levy MM, Fink MP, Marshall JC,Abraham E, Angus D, Cook D, CohenJ, Opal SM, Vincent JL, Ramsay G(2003) 2001 SCCM/ESICM/ACCP/ATS/SIS International SepsisDefinitions Conference. Intensive CareMed 29:530–538

10. Monnet X, Rienzo M, Osman D,Anguel N, Richard C, Pinsky MR,Teboul JL (2005) Esophageal Dopplermonitoring predicts fluidresponsiveness in critically ill ventilatedpatients. Intensive Care Med 31:1195–1201

11. Feissel M, Michard F, Mangin I, RuyerO, Faller JP, Teboul JL (2001)Respiratory changes in aortic bloodvelocity as an indicator of fluidresponsiveness in ventilated patientswith septic shock. Chest 119:867–873

12. Michard F, Boussat S, Chemla D,Anguel N, Mercat A, Lecarpentier Y,Richard C, Pinsky MR, Teboul JL(2000) Relation between respiratorychanges in arterial pulse pressure andfluid responsiveness in septic patientswith acute circulatory failure. Am JRespir Crit Care Med 162:134–138

13. Rutlen DL, Wackers FJ, Zaret BL(1981) Radionuclide assessment ofperipheral intravascular capacity: atechnique to measure intravascularvolume changes in the capacitancecirculation in man. Circulation 64:146–152

14. Caille V, Jabot J, Belliard G, CharronC, Jardin F, Vieillard-Baron A (2008)Hemodynamic effects of passive legraising: an echocardiographic study inpatients with shock. Intensive Care Med34:1239–1245

15. Thomas M, Shillingford J (1965) Thecirculatory response to a standardpostural change in ischaemic heartdisease. Br Heart J 27:17–27

89

16. Cavallaro F, Sandroni C, Antonelli M(2008) Functional hemodynamicmonitoring and dynamic indices of fluidresponsiveness. Minerva Anestesiol74:123–135

17. Durairaj L, Schmidt GA (2008) Fluidtherapy in resuscitated sepsis: less ismore. Chest 133:252–263

18. Hamzaoui O, Monnet X, Richard C,Osman D, Chemla D, Teboul JL (2008)Effects of changes in vascular tone onthe agreement between pulse contourand transpulmonary thermodilutioncardiac output measurements within anup to 6-hour calibration-free period.Crit Care Med 36:434–440

19. De Backer D (2006) Can passive legraising be used to guide fluidadministration? Crit Care 10:170

90