CURRENTOPINION Selecting the right positive end-expiratory

Carlessob, and Massimo Cressoni b

tion aiming either to oxygenation or to the full lung

wiosie tssbyigae c

methodology is preferable. Finally, the positive end-expiratory pressure adequate for full lung opening maybe comp

SummaWhen cwith higare the opositive

Keyworacute repositive

INTRODUCTIONThe physiology of po(PEEP) and its applicatdescribed by Barach ewidespread clinical uspediatric patients [2] athe treatment of acute(ARDS) [3]. The targeimprove oxygenationoutput decrease, as deet al. in 1948 [4]. Thmoreover, is a mechanism that, per se, may improveoxygenation as described by Lemaiconfirmed by Dantzker et al. [6]. Tmise tohemodauthorvariaboutput

Dipartimento di Fisiopatologia Medico-Chirurgica e dei Trapianti,Universita` degli Studi di Milano, Milan, Italy

ico.mi.it

www.c

REVIEWs found that the best oxygen transport, ale which associates oxygenation and cardiac, is reached when the PEEP provides the best

Curr Opin Crit Care 2015, 21:5057

DOI:10.1097/MCC.0000000000000166

o-criticalcare.com Volume 21 Number 1 February 2015ynamic was described by Suter et al. [7]. These Italy. Tpoliclin Copyright 2015needs with the Ospedale Maggiore Policlinico, Via Francesco Sforza 35, 20122 Milan,el: +39 02 55033232; fax: +39 02 55033230; e-mail: gattinon@reconcile the oxygenationnimazione ed Emergenza Urgenza, Fondazione IRCCS Ca` Grandare et al. [5], andhe best compro-

Correspondence to Luciano Gattinoni, Dipartimento di Anestesia, Ria-uted combining the computed tomography scan variables and the chest wall elastance.

ryompared, most of the methods give the same positive end-expiratory pressure values in patientsher and lower recruitability. The positive end-expiratory pressure/inspiratory oxygen fraction tablesnly methods providing lower positive end-expiratory pressure in lower recruiters and higherend-expiratory pressure in higher recruiters.

dsspiratory distress syndrome, computed tomography scan, esophageal pressure, oxygenation,end-expiratory pressure

sitive end-expiratory pressureion in pulmonary edema wast al. in 1938 [1]; however, itse began with Gregory et al. innd became routine practice inrespiratory distress syndromet of PEEP application was to; the concern was the cardiacscribed in detail by Cournande decrease of cardiac output,

respiratory system compliance. Suters PEEP selec-tion according to oxygenation and respiratory sys-tem compliance has been rediscovered several timesover the decades, up to the most recent papers.

In the 1990s, a new concept for PEEP use emerged,in the framework of lungprotective strategy, starting,in our opinion, from a paper of Webb and Tierney

aDipartimento di Anestesia, Rianimazione ed Emergenza Urgenza, Fon-dazione IRCCS Ca` Granda Ospedale Maggiore Policlinico and bDipar-oxygenation or respiratory mechanics change as the best marker for adequate selection. A growing interestis paid to the estimate of transpulmonary pressure, although no consensus is available on whichpressure level

Luciano Gattinonia,b, Eleonora

Purpose of reviewTo compare the positive end-expiratory pressure selecopening.

Recent findingsIncreasing positive end-expiratory pressure in patientsoutcome if the oxygenation response is greater and pperformed in a few minutes. The oxygenation responsbetter outcome in patients with acute respiratory distrerecruitment maneuver, the recruitment is not sustainedlung will unavoidably collapse. Several papers investaccording to the deflation limb of the pressurevolum Wolters Kluwer Healthth severe hypoxemia is associated with bettertive end-expiratory pressure tests may beo recruitment maneuvers was associated withsyndrome from influenza A (H1N1). If, aftersufficient positive end-expiratory pressure, theted the positive end-expiratory pressure selectionurve. It is still questionable whether to consider, Inc. All rights reserved.

[8], whPEEP aThe meprimarand clovided bmechawas firscontextarget opreven

Weing to oaccordifinally,ent meconclu

SELECPRESS

Settingmost dwork [1PEEP, sin larganalysicome wsurpriseoxygenhas be

the LOVs [13] and ExPress [16] trials and in H1N1patients [17]. These studies reported that therelation between oxygenation response following

adin(Pa. Aerntitapoch

ECSS

Settingugsoahimepr

pirrasu

voidowt, weg, th

KEY POINTS

PEEP selection methods based on PEEP/FiO2 tables arethe ohigh

To kepatielung

PEEPpressinspiwherthe lu

PEEPpressoxygenation or compliance changes as a marker ofadeq

Tailodefinappr15 c20 c

Selecting the right PEEP level Gattinoni et al.

1070-529o described, in rats, a protective effect ofgainst damage of mechanical ventilation.chanism of lung protection was attributedily to the prevention of intratidal openingsing based on a theoretical background pro-yMead et al. [9], whereas the transduction ofnical stimuli to the inflammatory reactiont described by Slutsky [10]. Therefore, in thet of open the lung and keep it open [11], the

throissueapprfor tvolunary

Resexpipres

Theindivthe lmenthe bades

uate PEEP is questionable.

ring PEEP according to the ARDS severity, ased by the Berlin definition, may be a reasonableoach: 510cmH2O PEEP in mild patients, 10mH2O PEEP in moderate patients and 15mH2O PEEP in severe patients.PEEPgestmiaPEEPhighpatierecruis imPEEP

SELPRE

nly discriminating factors between patients wither and lower recruitability.

ep the lung fully open, similar PEEP is required innts with higher and lower potential forrecruitment.

selection using the inspiratory limb of the volumeure curve is conceptually wrong becauseratory pressure reflects the opening pressure,eas PEEP relates to the closing pressure ofng.

selection along the deflation limb of the volumeure curve makes more sense, but the use of Copyright 2015 Wolters Kluwer Hea

f PEEP was nomore the oxygenation but thetionof the intratidal collapse anddecollapse.will describe, first, the PEEP selection accord-xygenation and, second, the PEEP selectionng to the lung protective strategy, and,we will attempt to describe how these differ-thods were compared and which are thesions reached so far.

TING POSITIVE END-EXPIRATORYURE TO IMPROVE OXYGENATION

PEEP based on PEEP/FiO2 tables is likely theiffused method, introduced by the ARDS Net-2] and LOVs study [13]. Higher and lowerelected by these tables, have been comparede trials recently reviewed in a Cochranes [14]. The authors concluded that the out-as unrelated to the PEEP level and (what a!) the higher the PEEP, the higher is theation. The oxygenation response to PEEPen studied in a secondary analysis [15

&

] of

respirathas been focused on the deflation limb. At the samepressurexpiratrequirethe insThe exdissipatensionto open

The inpressure curve: the recruitment maneuver

In contconsistspeciespressuratory ventire copen uexamppart of

5 Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.lth, Inc. All rights reserved.

rast with previous beliefs, it has been shownently, in humans [22] and in different animal[23,24], that recruitment is not limited to thee around the inflection point of the inspir-olumepressure curve but occurs along theurve. This indicates that the collapsed unitsp at different opening pressures. As anle, at the inflation of 30 cmH2O, a consistentthe potentially recruitable lung, which may

www.co-criticalcare.com 51e, the inspiratory volume is lower than theory one, and, conversely, the pressured to reach a given volume is greater alongpiratory limb than along the expiratory one.tra pressure required during the inflation isted in the system to overcome the surfaceand the tissue resistances and, eventually,up the collapsed lung regions.

spiratory limb of the volumejustment and decreasedmortality was stron-patients with more severe baseline hypoxe-O2/FiO2

lth

be estimated from 15% to 30%, remains closed. Toopen up these regions, opening pressures spanningfrom 30 to 4560 cmH2O are required [22,25].Therefoup diffformedLiu etmaneuoxygenassociasame ghyperingreateroleic arecruitmwith norecruitmless hereactioas recrulevels,the defThis faconceprecruitmmaintaviouslyrecruitmunavoiet al. [2ment irecruitmvidual

Thetailoredsure culeadingreviewinflatioPEEP seauthorsalthougany reastudiesassumeplete abtrue; sebe idenwas useshould

The exvolum

In the leffectsthe vol

the pressure values corresponding to the best com-pliance or before the oxygenation decrease. It mustbe noted, however, that, first, because of the sigmoid

eysormARmHi

3]tioS,repsnscatieme

surger

lura

intstrwesure lcans aila

trtiv

recisiestaugunurraldaehel preeraturbsral

Respiratory system

52 Copyright 2015 Wolters Kluwer Hea

re, the recruitment maneuver may openerent amounts of recruitable lung if per-at 30, 40 or 60 cmH2O inflation pressure.al. [17] reported that, if the recruitmentver (at 30 cmH2O for 60 s) resulted in betteration, the patients with influenza A (H1N1)-ted ARDS had a better chance of survival. Theroup [26] found, in a canine model, thatflation after recruitment maneuver wasin the surfactant model rather than in thecid model. Engel et al. [27] compared twoent maneuvers (at 45 and 15 cmH2O PEEP)recruitment. These authors concluded thatent maneuvers improve oxygenation with

modynamic impairment and inflammatoryn at lower PEEP. Actually, the authors defineditment the application of two different PEEPwhich are expiratory phenomena related tolation limb of the volumepressure curve.ct underlines the confusion originating byts such as recruitment with PEEP. Actually,ent occurs during inspiration and PEEP

ins open, if sufficient, what has been pre-recruited. If the PEEP is insufficient, theent is not sustained and the lung will

dably collapse again, as confirmed by Kheir8]. Keenan et al. [29], reviewing the recruit-ssue, wisely concluded, in our opinion, thatent maneuvers should be guided by indi-

clinician experience and patients factor.refore, although the recruitment must beon the inspiratory limb of the volumepres-rve, tailoring PEEP in the same limb is mis-. Hata et al. [3033] provided a systematicof three randomized trials that used then limb of the pressurevolume curve to tailorlected above the lower inflection point. Thesuggested a possible outcome benefit,

h the limited number of patients preventsl conclusion. In our opinion, in all of these, there is a fundamental bias. First, the authorsthat recruitment is complete or near com-ove the lower inflection point, which is notcond, when the lower inflection point cannottified, a PEEP approximately 1516cmH2Od. To be consistent with the hypothesis, PEEPhave been set equal to 0 cmH2O.

piratory limb of theepressure curve

ast few years, several papers investigated theof PEEP selection on the deflation part ofumepressure curve, usually setting PEEP at

shapalwain nin(20 cboth[22,2deflaARDlungcollashow(CT)51 psinglcruitusedmeadurinBikk

Theexpi

Thetherenemeaof thCT sstatuvent

Theposi

Thelungstudmainthrothe lpresspleumanthe m

TageabeenFesslestimpressthe apleu

www.co-criticalcare.com, Inc. All rights reserved.

anspulmonary pressure-basede end-expiratory pressure selection

ognition that the distending pressure of thethe transpulmonary pressure led to a series ofin which the PEEP level was selected toin the transpulmonary pressure positiveh the whole respiratory cycle, to maintaing always open. Because the transpulmonarye is the difference between the airway and thepressure, the estimate of this variable is

tory and the only clinical tool available isasurement of the esophageal pressure.indications and the limits of using esoph-

ressure as a surrogate of pleural pressure haveviewed by Brochard [37

&

] and by Keller and[38]. Two approaches have been proposed toe the pleural pressure from the esophageale measurement. The first one assumes thatolute values of esophageal pressure equal thepressure. To take into account the weight of

Volume 21 Number 1 February 2015of the deflation limb, the compliance ishigher in the middle part of the lung, evenal lungs. Second, the derecruitment starts,DS, at very high deflation pressures

2O), as shown by the closing pressure curve,n humans and in experimental animals, to continue at lower pressures along then limb. Actually, in supine patients withwhen PEEP is decreased, the most dependentgions along the sternumvertebral directione first and then the less dependent regions, aswith the regional computed tomographyan analysis [34], and recently confirmed inents with ARDS [35

&

]. This makes the use of aunique pressure point as a marker of dere-nt highly questionable. How the commonlyphysiological variables are different ifed at the same pressure during inflation ordeflation has been recently emphasized byet al. [36].

ng mechanics-based positive end-tory pressure selection

erest in ventilator-induced lung injury andess/strain applied to the lung structuresd the attention on the esophageal pressureement and its clinical use in the frameworkung protective strategy. Moreover, the use ofallowed a better characterization of the lungnd the individualization of the mechanicaltion settings.

ea

the mediastinum, a correction factor of 5 cmH2Omay be applied. The second method considers thevariation of the esophageal pressure equal to thevariatiomeasurpressurpressurrespiratcompapressurcludedinterchory syschange

The copositiv

The asselectioin ARDpressesCT scasufficiepressure superimposed on the most dependent lungregions and the pressure necessary to lift up thechest win sevefrom 7mild A185 cthe lunpulmonregions

COMPMODEPRESS

In thePEEP sethe meauthorthat, wmethods tested, the only one that provides appro-priatelywas thethe othprovidelower pinsteadables. Yprovidepressuret al.,stress i

and the inflection point in the inspiratory limb ofthe volumepressure curve. These authors con-cluded that stress index and oxygenation methods

EEest, Podcrtepliemlact,mes ongblp

e,ied, womrent

ting PEee einwt idiertm

oxicuataunessitmevv

y.re

the defioedorecbeasshntratcrupli

Selecting the right PEEP level Gattinoni et al.

1070-529 Copyright 2015 Wolters Kluwer H

lower PEEP in the less recruitable patientshigh PEEP arm of the ARDSNet table. All ofer systems, including the CT-derived PEEP,similar values in patients with higher or

otential for lung recruitment. Other authors,of recruitability, targeted PEEP to other vari-ang et al. found that better oxygenation wasd by applying a positive transpulmonarye than following the ARDSNet table. Huangduring a decremental PEEP trial, measuredndex, static lung compliance, oxygenation

physsiderauththe dtruearewhicIn corespiderecom

5 Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.all [35&

]. Cressoni et al. found, however, that,re ARDS, the CT scan-derived PEEP rangedto 28 cmH2O, averaging 165 cmH2O in

RDS, 165 cmH2O in moderate ARDS andmH2O in severe ARDS, and was unrelated tog recruitability, that is, to keep open 1 or 100ary units collapsed in the dependent lung, approximately the same PEEP is required.

ARISON BETWEEN DIFFERENTS OF POSITIVE END-EXPIRATORYURE SELECTION

last years, several papers compared differentlection methods. In Table 1, we summarizethods in comparison, their targets, and thes conclusions. Briefly, Chiumello et al. foundithin all of the bedside PEEP selection

to thvalu[48],PEEP

Ividepropationapprnamis usicalthe lExPrrecruthe latoryearitthe pns of the pleural pressure. Therefore, aftering the chest wall elastance, the Dpleurale may be estimated as the change in airwaye times the ratio of the chest wall to the totalory system elastance [35

&

]. Gulati et al. [39&

]red these two methods for estimating pleurale on the same group of patients. They con-that the two methods cannot be consideredangeable. Moreover, chest wall and respirat-tem elastances may vary unpredictably withs in PEEP.

mputed tomography scan-basede end-expiratory pressure selection

sumption behind the CT scan-based PEEPn is that the primary reason for lung collapseS is the excessive lung weight that com-the dependent lung regions. Therefore, then-derived PEEP is computed as the pressurent to overcome the maximal hydrostatic

set Phighturnmethin dea betcomdecralveoin fabecaseriedurivarianaryspacapplthatto candpatielth, Inc. All rights reserved.

P at higher values than indicated by thecompliance and the inflection point. In

intado et al. found that the best compliance, comparedwith the ARDSNet table, resultedeased organ dysfunction with a trend towardr outcome. In addition to respiratory systemance and transpulmonary pressure duringental PEEP trial, Rodriguez et al. found thatr dead space could add further information;it increased when transpulmonary pressurenegative and oxygenation deteriorated. In af papers, in humans [49] and in pigs [36,45],the PEEP changes, in addition to the usuales such as dynamic compliance, transpulmo-ressure, oxygenation parameters and deadthe electrical impedance tomography was. As expected, all of these studies showedhen applying PEEP, we have unavoidablypromise between regional overdistensiongional collapse. Finally, in postoperatives, Ferrando et al. [47] found advantages set-EP during a decremental PEEP trial accordingbest compliance instead of using a constantqual to 5 cmH2O PEEP, whereas Hansen et al.cardiovascular patients, found that 8 cmH2Oas substantially similar to 5 cmH2O PEEP.s evident that the different methods do pro-fferent PEEP values as they explore differenties of the system. The target of the oxygen-ethod is to provide an oxygen saturation

imately 90% without negative hemody-effects. The PEEP level to reach this targetlly lower than the one required for mechan-rgets, because the complete opening ofg is not necessary. The stress index and thestudy methods aim to sustain a completeent by setting PEEP approximately at

el of the upper inflection point of the inspir-olumepressure curve, where it loses its lin-The healthier the lung, however, the higher isssure set with these twomethods [40

&

]. Usinglation part of the volumepressure curve islogically sound, but the variable to be con-for setting PEEP is questionable. Some

s proposed as a signal of derecruitmentrease of oxygenation. This is not necessarilycause the changes in intrathoracic pressureociated with changes in hemodynamics,may influence oxygenation changes [5,6].rast, some authors consider the decrease inory system compliance as the beginning ofitment. Even in normal lung, however, theance during deflation first increases, then

www.co-criticalcare.com 53

Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.

Table

1.Po

sitiv

een

d-ex

piratory

pressure

selectionmetho

dsrepo

rted

inrece

ntliterature

Author

Population

PEEPselectionmetho

dTa

rgets

Conclusions

Chium

ello

etal.

[35&,40&]

Patientswith

ARD

SIncrea

sedrecruitmen

tstra

tegy

oftheExPressstud

yAirway

pressure

upto

2830cm

H2O

orPEEP20cm

H2O

atconstant

tidal

volume6ml/kg

IBW

PEEP/FiO

2tableistheon

lymetho

dproviding

approp

riatelylower/h

ighe

rPEEP

inlower/

higherrecruiters

Stress

inde

xPEEP

atwhich

thetim

epressure

curvelosesits

linea

rity

Esop

hagea

lpressure

PEEP

was

sete

qual

totheab

solute

valueof

esop

ha-

gea

lpressuremea

suredat

functio

nalresidua

lca

pacity

LOVstud

yPEEP

selected

accordingto

aPEEP/FiO

2table,

target-

ingSa

O2be

tween88%

and93%

CT-de

rived

PEEPmax

imal

supe

rimpo

sedpressure

Yan

get

al.[41]

Patients,with

andwithou

tIAH

Tran

spulmon

arypressure

Tran

spulmon

arypressure01

0cm

H2O

aten

dexpiratio

n,ac

cordingto

aslidingscaleba

sedon

PaO2an

dFiO

2

Tran

spulmon

arypressure

metho

dprovided

higher

PEEP

than

PEEP/FiO

2tablewith

betterox

ygen-

ationan

drespiratorymecha

nics

ARD

SNet

protocol

PEEP

selected

accordingto

aPEEP/FiO

2table

Gulatiet

al.

[39&]

Patientswith

ARD

SPes-ba

sedmetho

dE C

W-based

metho

dEn

d-expiratory

transpu

lmon

arypressure

of0cm

H2O

End-inspiratorytra

nspu

lmon

arypressure

of26cm

H2O

Absoluteesop

hagea

lpressureor

chestw

allcom

-pliancemetho

dto

settranspu

lmon

arypressure

does

noty

ield

simila

rresults

Hua

nget

al.

[42]

Pulmon

ary

patientswith

ARD

S

Oxyge

natio

nPEEP

decrem

enteduntil

PaO2/F

iO25%

diffe

rencein

PaO2/F

iO2be

tweentwoconsecu-

tivePEEP

redu

ction

PEEP

titratio

nby

stress

inde

xmight

bemore

bene

ficialfor

pulmon

arypa

tientswith

ARD

Safterarecruitmen

tman

euver

Stress

inde

xOptimal

PEEP

was

settoob

tain

astress

inde

xvalue

between0.9

and1.1

Cst

PEEP

was

redu

cedin

step

sof

2cm

H2O

startingfrom

20cm

H2O,until

thelowestP

EEPlevelp

roviding

the

max

imal

Cst

LIP2cm

H2O

Optimal

PEEPLIP2cm

H2O

Pintad

oet

al.

[43]

Patientswith

ARD

SCom

pliance-guide

dPEEP

Thehighe

ststatic

compliancewas

considered

tobe

thebe

stPEEP

during

anincrem

entaltrial.Ifat

two

diffe

rent

PEEPsthestatic

compliancewas

iden

tical,

theon

ewith

thelowestp

lateau

was

chosen

PEEP

setting

byhighestcomplianceisbe

tter

than

byPEEP/F

iO2tableto

decrea

seorgan

dysfunction

ARD

SNet

protocol

PEEP

selected

accordingto

aPEEP/FiO

2table

Rodriguezet

al.

[44]

Patientswith

ARD

SCrs

Thebe

stCrs

PEEP

was

defin

edas

thehigh

estvalue

ofPEEP

prod

ucingthehighe

rCrs

during

thede

crem

en-

taltitrationman

euver

Neg

ativevalues

oftra

nspu

lmon

arypressure

during

decrem

entalP

EEPareassociated

with

increa

sed

VD/V

Tan

dhighrisk

ofcollapse

Tran

spulmon

arypressure

ThePEEP

valuecorrespo

nded

toan

expiratory

Ptpof

0

Dea

dspac

e

Respiratory system

54 www.co-criticalcare.com Volume 21 Number 1 February 2015

Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.

Blan

kman

etal.

[45]

Postcardiac

surgery

patients

Dynam

iccompliance

PaO2/F

iO2ratio

EIT

Decremen

talP

EEPtrial.Aneven

distribu

tionof

tidal

volumeto

theno

ndep

ende

ntan

dde

pend

entlung

regions

Decremen

talP

EEPtrial.Highe

stvalue

Decremen

talP

EEPtrial.Highe

stvalue

Inpo

stca

rdiacsurgerypa

tients,theEIT-de

rived

ITVinde

xwas

compa

rablewith

dyna

mic

com-

plianceto

indicate

bestPEEP,that

is,avoids

overdisten

sion

oftheno

ndep

ende

ntregion

s

Bikker

etal.

[36]

Pigs(hea

lthyan

dafterALIindu

c-tio

n)

Crs

EELV

Tran

spulmon

arypressure

PaO2

Dea

dspac

eSh

unt

Electrica

limpe

dance

BestPEEP

atmax

imum

compliance

BestPEEP

atmax

imum

EELV

BestPEEP

was

defin

edat

transpu

lmon

arypressure

equa

ltoor

exceed

ingzero

during

endexpiratio

nBe

stPEEP

atmax

imum

PaO2

BestPEEP

atlowestd

eadspac

eBe

stPEEP

atlowestshunt

Minimal

lung

collapsean

doverdisten

sion

BestPEEP

levelsarecompa

rablewith

thediffe

r-en

tPEE

Pselectionmetho

ds.EITprovides

infor-

mationon

gasdistribu

tion

Wolfet

al.

[46]

Pigs(ARD

S)ARD

SNet

protocol

Electrica

limpe

dance

PEEP

selected

accordingto

aPEEP/FiO

2table

Regiona

lEIT-derived

compliancewas

used

tomax

imizetherecruitmen

tofde

pend

entlung

andminimizeoverdisten

sion

ofno

ndep

ende

ntlung

area

s

EIT-guide

dventila

tionresultedin

improved

respir-

atorymecha

nics,improved

gas

exchan

gean

dredu

cedhistolog

iceviden

ceof

ventila

tor-

indu

cedlung

injury

inan

animal

mod

el

Ferran

doet

al.

[47]

Patientsunde

r-go

ingthoracic

surgery

Dynam

iccompliance

5cm

H2O

PEEP

decrem

enttrial

at2cm

H2O

step

suntil

the

max

imal

dyna

mic

compliancewas

obtained

Duringon

e-lung

ventila

tion,

bestcompliance

metho

disbe

tterthan

5cm

H2O

PEEP

topreserve

oxyg

enationan

dlung

mecha

nics

Han

senet

al.

[48]

Mecha

nica

llyventila

ted

patientsafter

isolated

coron-

aryartery

bypa

ssgraft-

ingor

com-

bine

dCABG

andvalve

operations

5cm

H2O

8cm

H2O

Theuseof

8cm

H2O

PEEP

instea

dof

5cm

H2O

does

notseem

beneficial

Mau

riet

al.

[49]

Patientsrecover-

ingfrom

ARD

Saftersw

itch

topressure

supp

ort

ventila

tion

Clinical

PEEP

(72cm

H2O)

Clinical

PEEP5cm

H2O

Moreho

mog

eneo

usdistribu

tionby

EIT

Highe

rPEEP

andlower

pressure

supp

ortp

rovide

smoreho

mog

eneo

usventila

tionan

d,po

ssibly,

betterventila

tion/

perfusionmatching

Thetablesummarizes

theresults

ofrecent

stud

iescompa

ring

diffe

rent

PEEP

selectionmetho

ds.Itrepo

rtsthemetho

dscompa

red,

theirtargetsan

dtheau

thorsconclusion

s.ALI,ac

utelung

injury;ARD

S,ac

uterespiratory

distress

synd

rome;

CABG

,corona

ryartery

bypa

ssgraft;

Crs,respiratorysystem

compliance;

Cst,static

pulmon

arycompliance;

CT,

compu

tedtomog

raph

y;E C

W,chestw

alle

lastan

ce;EE

LV,en

d-expiratory

lung

volume;

EIT,

electrica

limpe

dancetomog

raph

y;FiO

2,inspiratoryox

ygen

frac

tion;

IBW

,idea

lbod

yweight;ITV,intra

tidal

gas

distribu

tion;

LIP,

lower

inflectionpo

int;PEEP,po

sitiveen

d-expiratory

pressure;Pes,esop

hagea

lpressure;Ptp,

transpu

lmon

arypressure;Sa

O2,arterial

oxyg

ensaturatio

n;VD/V

T,de

adspac

e.

Selecting the right PEEP level Gattinoni et al.

1070-5295 Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved. www.co-criticalcare.com 55

lth

stays constant and, then, decreases again, accordingto the sigmoid shape of the volumepressure curve,independent of recruitability. The use of positivetranspution aspleuralthis assageal pwith ARhave theven aesophathe chainformtendincomplideriveddo notguide ffrom mlung counits,used. Tuse higrecruitity, andis unre

CONC

The bethat we may find a PEEP level that avoids intratidalrecruitmtime thlowestand afdreamTherefoPEEP aoxygenopeninenon dability,the besPEEP inPEEP inmediatchest w[50]. Tdecadetive asproviderange o

Ackno

None.

Financial support and sponsorship

None.

fli

aut

ERDof

ighlspeout

racn to:75

regospiraEnghbaults

ournfect

hysmairveses:43ntzkof

:63terpat9.ebbsitivpo5.

eadlmo

10. SlutskRespir

ne93;ntilalume A2:1

eadelumr acuntronta

essug inv 2

olighsitivss sspirbeal trg P

ercaultsndom

X, ent maneuver with high-levelsitiv

spira13;oenpiraome? A systematic review and meta-analysis. Anesthesiology 2009;0:10981105.iel Messstre3:8

Respiratory system

56 Copyright 2015 Wolters Kluwer Hea

entderecruitment, providing in themean-e best compliance, best oxygenation anddead space, without causing hyperinflationfecting hemodynamics, reflects a wishfulthat has nothing to do with the reality.re, in our opinion, we should use a betterpproach as a reasonable compromise amongation, hemodynamics status and intratidalg and closing. Because the latter phenom-epends quantitatively on the lung recruit-which is a function of the lung severity,t compromise should be the use of highersevere ARDS (range 1520 cmH2O), lowermild ARDS (range 510 cmH2O) and inter-

ed inmoderate ARDS, paying attention to theall elastance and hemodynamic impairmenthis pragmatic approach [50], supported bys of studies and experience, is likely as effec-the more laborious PEEP trials that do not, at the end, anything else than reportedf values.

wledgements

11. Bo19

12. VevoTh34

13. Mvofoco

14. SaprlunRe

15.&

GpotreRe

It mayclassicfollowin16. M

adra

17. Liupore20

18. Phexdr11

19. Brprdi30

www.co-criticalcare.comlmonary pressure as a guide for PEEP selec-sumes that esophageal pressure equals thepressure. Unfortunately, in our opinion,umption is far from true, because the esoph-ressure is highly positive in most patientsDS, which, according to the theory, shouldeir lung completely collapsed, sometimest the end of inspiration. The changes ofgeal pressure, in contrast, better reflectnges of pleural pressure. Therefore, usefulation can be acquired to judge the real dis-g pressure of the lung once the chest wallance has been estimated. Finally, the CT-PEEP is physiologically appealing, but wehave any proof that it should be used as aor therapy. It simply tells us that in ARDS,ild to severe, if we want to keep the wholempletely open, either a few or hundreds ofapproximately the same pressure must behere is no clinical sense, in our opinion, toh pressure either in patients with higherability or in patients with lower recruitabil-, unfortunately, the CT scan-derived PEEPlated to recruitability.

LUSION

st PEEP does not exist. To pretend and claim

Con

The

REFREAPapersbeen h& of&& of

1. Batio12

2. GreN

3. Asad

4. CefJ P

5. Lecuca22

6. Daism77

7. Suin28

8. Wpoby56

9. Mpu, Inc. All rights reserved.Volume 21 Number 1 February 2015, Meade M, Mercat A, et al. Higher vs lower positive end-expiratoryure in patients with acute lung injury and acute respiratoryss syndrome: systematic review and meta-analysis. JAMA 2010;65873.e end-expiratory pressure in patients with influenza-associated acutetory distress: a single-center prospective study. Curr Ther Res Clin Exp75:8387.ix SI, Paravastu S, Columb M, et al. Does a higher positive endtory pressure decrease mortality in acute respiratory distress syn-ized controlled trial. JAMA 2008; 299:6Ma T, Qu B, et al. Efficacy of lung recruitmcts of interest

hors do not have conflicts of interest.

ENCES AND RECOMMENDEDINGparticular interest, published within the annual period of review, haveighted as:cial intereststanding interest

h AL, Martin J, Eckman M. Positive pressure respiration and its applica-the treatment of acute pulmonary edema. Ann Intern Med 1938;

4795.ry GA, Kitterman JA, Phibbs RH, et al. Treatment of the idiopathictory-distress syndrome with continuous positive airway pressure.l J Med 1971; 284:13331340.ugh DG, Bigelow DB, Petty TL, Levine BE. Acute respiratory distress in. Lancet 1967; 2:319323.and A, Motley HL, Werko L, Richards DW. Physiologic studies of theof intermittent positive pressure breathing on cardiac output in man. Amiol 1948; 126:162174.re F, Harf A, Simonneau G, et al. Gas exchange, static pressurevolumeand positive-pressure ventilation at the end of expiration. Study of 16of acute respiratory insufficiency in adults. Ann Anesthesiol Fr 1981;5441.er DR, Lynch JP, Weg JG. Depression of cardiac output is a mechan-shunt reduction in the therapy of acute respiratory failure. Chest 1980;6642.PM, Fairley B, Isenberg MD. Optimum end-expiratory airway pressureients with acute pulmonary failure. N Engl J Med 1975; 292:284

HH, Tierney DF. Experimental pulmonary edema due to intermittente pressure ventilation with high inflation pressures. Protectionsitive end-expiratory pressure. Am Rev Respir Dis 1974; 110:556

J, Takishima T, Leith D. Stress distribution in lungs: a model ofnary elasticity. J Appl Physiol 1970; 28:596608.y AS. Ventilator-induced lung injury: from barotrauma to biotrauma.Care 2005; 50:646659.

RC. The ARDS lung. New insights from computed tomography. JAMA269:21342135.tion with lower tidal volumes as compared with traditional tidal

es for acute lung injury and the acute respiratory distress syndrome.cute Respiratory Distress Syndrome Network. N Engl J Med 2000;3011308.

MO, Cook DJ, Guyatt GH, et al. Ventilation strategy using low tidales, recruitment maneuvers, and high positive end-expiratory pressurete lung injury and acute respiratory distress syndrome: a randomizedlled trial. JAMA 2008; 299:637645.CR, Rojas JI, Nervi R, et al. High versus low positive end-expiratoryre (PEEP) levels for mechanically ventilated adult patients with acutejury and acute respiratory distress syndrome. Cochrane Database Syst013; 6:CD009098.er EC, Kavanagh BP, Rubenfeld GD, et al. Oxygenation response toe end-expiratory pressure predicts mortality in acute respiratory dis-yndrome. A secondary analysis of the LOVS and ExPress trials. Am JCrit Care Med 2014; 190:7076.

of some interest to note that in patients with ARDS enrolled in theials with more severe hypoxemia, the extent of oxygenation responseEEP adjustment is associated with outcome.t A, Richard JC, Vielle B, et al. Positive end-expiratory pressure setting inwith acute lung injury and acute respiratory distress syndrome: a

46655.

ea

20. Gattinoni L, Caironi P, Cressoni M, et al. Lung recruitment in patients with theacute respiratory distress syndrome. N Engl J Med 2006; 354:17751786.

21. Chiumello D, Coppola S, Froio S, et al. Time to reach a new steady state afterchanges of positive end expiratory pressure. Intensive Care Med 2013;39:13771385.

22. Crotti S, Mascheroni D, Caironi P, et al. Recruitment and derecruitment duringacute respiratory failure: a clinical study. Am J Respir Crit Care Med 2001;164:131140.

23. Pelosi P, Goldner M, McKibben A, et al. Recruitment and derecruitment duringacute respiratory failure: an experimental study. Am J Respir Crit Care Med2001; 164:122130.

24. Quintel M, Pelosi P, Caironi P, et al. An increase of abdominal pressureincreases pulmonary edema in oleic acid-induced lung injury. Am J Respir CritCare Med 2004; 169:534541.

25. Borges JB, Okamoto VN, Matos GF, et al. Reversibility of lung collapse andhypoxemia in early acute respiratory distress syndrome. Am J Respir Crit CareMed 2006; 174:268278.

26. Yang Y, Chen Q, Liu S, et al. Effects of recruitment maneuvers with PEEP onlung volume distribution in canine models of direct and indirect lung injury. MolBiol Rep 2014; 41:13251333.

27. Engel M, Nowacki RM, Reiss LK, et al. Comparison of recruitmentmanoeuvres in ventilated sheep with acute respiratory distress syndrome.Lung 2013; 191:7786.

28. Kheir JN, Walsh BK, Smallwood CD, et al. Comparison of 2 lung recruitmentstrategies in children with acute lung injury. Respir Care 2013; 58:12801290.

29. Keenan JC, Formenti P, Marini JJ. Lung recruitment in acute respiratorydistress syndrome: what is the best strategy? Curr Opin Crit Care 2014;20:6368.

30. Hata JS, Togashi K, Kumar AB, et al. The effect of the pressurevolume curvefor positive end-expiratory pressure titration on clinical outcomes in acuterespiratory distress syndrome: a systematic review. J Intensive Care Med2014; 29:348356.

31. Amato MB, Barbas CS, Medeiros DM, et al. Effect of a protective-ventilationstrategy on mortality in the acute respiratory distress syndrome. N Engl J Med1998;

32. Ranierinflamrandom

33. Villar JexpirapersisCrit C

34. GattinpositivJAMA

35.&

Cressputedtory d

This paperhigher recrhigher and

36. Bikker IG, Blankman P, Specht P, et al. Global and regional parameters tovisualize the best PEEP during a PEEP trial in a porcine model with andwithout acute lung injury. Minerva Anestesiol 2013; 79:983992.

37.&

Brochard L. Measurement of esophageal pressure at bedside: pros and cons.Curr Opin Crit Care 2014; 20:3946.

Excellent review on the use of esophageal pressure as a surrogate of pleuralpressure.38. Keller SP, Fessler HE. Monitoring of oesophageal pressure. Curr Opin Crit

Care 2014; 20:340346.39.&

Gulati G, Novero A, Loring SH, Talmor D. Pleural pressure and optimalpositive end-expiratory pressure based on esophageal pressure versus chestwall elastance: incompatible results. Crit Care Med 2013; 41:19511957.

This paper underlines problems related to the estimate of transpulmonary pressureat the bedside.40.&

Chiumello D, Cressoni M, Carlesso E, et al. Bedside selection of positive end-expiratory pressure in mild, moderate, and severe acute respiratory distresssyndrome. Crit Care Med 2014; 42:252264.

Only the PEEP selected by the high PEEP arm of the LOV study is appropriatelylower or higher in patients with lower or higher recruitability.41. Yang Y, Li Y, Liu SQ, et al. Positive end expiratory pressure titrated by

transpulmonary pressure improved oxygenation and respiratory mechanics inacute respiratory distress syndrome patients with intra-abdominal hyper-tension. Chin Med J (Engl) 2013; 126:32343239.

42. Huang Y, Yang Y, Chen Q, et al. Pulmonary acute respiratory distresssyndrome: positive end-expiratory pressure titration needs stress index.J Surg Res 2013; 185:347352.

43. Pintado MC, de Pablo R, Trascasa M, et al. Individualized PEEP setting insubjects with ARDS: a randomized controlled pilot study. Respir Care 2013;58:14161423.

44. Rodriguez PO, Bonelli I, Setten M, et al. Transpulmonary pressure and gasexchange during decremental PEEP titration in pulmonary ARDS patients.Respir Care 2013; 58:754763.

45. Blankman P, Hasan D, Groot JE, Gommers D. Detection of best positive end-expiratory pressure derived from electrical impedance tomography parametersduring a decremental positive end-expiratory pressure trial. Crit Care 2014; 18:R95.

olf GK, Gomez-Laberge C, Rettig JS, et al. Mechanical ventilation guided byctrical impedance tomography in experimental acute lung injury. Crit Care

ed 2rranpiraal afringnse

essuntenaurintilapira:16

attinol k:10

Selecting the right PEEP level Gattinoni et al.

1070-529 Copyright 2015 Wolters Kluwer H

matory mediators in patients with acute respiratory distress syndrome: aized controlled trial. JAMA 1999; 282:5461.

, Kacmarek RM, Perez-Mendez L, Aguirre-Jaime A. A high positive end-tory pressure, low tidal volume ventilatory strategy improves outcome intent acute respiratory distress syndrome: a randomized, controlled trial.are Med 2006; 34:13111318.oni L, DAndrea L, Pelosi P, et al. Regional effects and mechanism ofe end-expiratory pressure in early adult respiratory distress syndrome.1993; 269:21222127.

oni M, Chiumello D, Carlesso E, et al. Compressive forces and com-tomography-derived positive end-expiratory pressure in acute respira-

istress syndrome. Anesthesiology 2014; 121:572581.shows that compressive forces are similar in patients with lower and

uitability. Therefore, a CT-scan derived PEEP is inappropriately similar inlower recruiters.

M47. Fe

extridu

48. HaprJ I

49. Mveex41

50. Gto80

5 Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.lth, Inc. All rights reserved.

013; 41:12961304.do C, Mugarra A, Gutierrez A, et al. Setting individualized positive end-tory pressure level with a positive end-expiratory pressure decrementter a recruitment maneuver improves oxygenation and lung mechanicsone-lung ventilation. Anesth Analg 2014; 118:657665.

n JK, Anthony DG, Li L, et al. Comparison of positive end-expiratoryre of 8 versus 5 cm H2O on outcome after cardiac operations.sive Care Med 2014. [Epub ahead of print]T, Bellani G, Confalonieri A, et al. Topographic distribution of tidaltion in acute respiratory distress syndrome: effects of positive end-tory pressure and pressure support. Crit Care Med 2013;641673.oni L, Carlesso E, Brazzi L, et al. Friday night ventilation: a safety startingit for mechanically ventilated patients. Minerva Anestesiol 2014;461057.

www.co-criticalcare.com 57338:347354.i VM, Suter PM, Tortorella C, et al. Effect of mechanical ventilation on

46. Wele