noninvasive ventilation in patients with acute cardiogenic...

Conference Proceedings

Noninvasive Ventilation in PatientsWith Acute Cardiogenic Pulmonary Edema

Sangeeta Mehta MD FRCPC, Abdul Hakeem Al-Hashim MD FRCPC,and Sean P Keenan MD FRCPC MSc

IntroductionSummary of the Randomized Controlled Trials

Continuous Positive Airway Pressure Compared to ConventionalTreatment in Patients With Acute Cardiogenic Pulmonary Edema

Noninvasive Ventilation Compared to Conventional Treatment inPatients with Acute Cardiogenic Pulmonary Edema

Continuous Positive Airway Pressure Compared to NoninvasiveVentilation in Patients With Acute Cardiogenic Pulmonary Edema

Summary

Acute cardiogenic pulmonary edema (ACPE) is a common cause of respiratory failure that neces-sitates endotracheal intubation. In some patients intubation and its attendant complications can beavoided with noninvasive ventilation (NIV). Both continuous positive airway pressure (CPAP) andNIV have been evaluated in patients with ACPE. Compared to conventional treatment, both CPAPand NIV improve vital signs and physiologic variables, and reduce intubation rate, in patients withACPE. Both CPAP and NIV appear to be well tolerated and are not associated with any seriousadverse events. Initial concern that NIV may be associated with a greater risk of myocardialinfarction than CPAP was laid to rest by later studies. Despite a physiologic rationale that NIVshould offer greater benefit than CPAP, NIV has not been found to offer any advantages regardingintubation rate or mortality compared with CPAP. We review the randomized controlled trials andsummarize the evidence on NIV and CPAP in patients with ACPE. Key words: acute cardiogenicpulmonary edema, congestive heart failure, noninvasive ventilation, mechanical ventilation, continuouspositive airway pressure, CPAP, bi-level positive airway pressure, BiPAP respiratory failure. [RespirCare 2009;54(2):186–195. © 2009 Daedalus Enterprises]

Sangeeta Mehta MD FRCPC presented a version of this paper at the 42ndRESPIRATORY CARE Journal Conference, “Noninvasive Ventilation inAcute Care: Controversies and Emerging Concepts,” held March 7-9,2008, in Cancun, Mexico.

Dr Keenan has had a relationship with Respironics.

Correspondence: Sangeeta Mehta MD FRCPC, Interdepartmental Divi-sion of Critical Care Medicine, Mount Sinai Hospital, 600 UniversityAvenue, #18-216, Toronto, Ontario, M5G 1X5, Canada. E-mail:[email protected].

Sangeeta Mehta MD FRCPC and Abdul Hakeem Al-Hashim MD FRCPCare affiliated with the Interdepartmental Division of Critical Care Med-icine, and the Division of Respirology, Department of Medicine, MountSinai Hospital, University of Toronto, Toronto, Ontario, Canada. Sean PKeenan MD FRCPC MSc is affiliated with the Department of CriticalCare Medicine, Royal Columbian Hospital, New Westminster, BritishColumbia, and with the Centre for Health Evaluation and Outcome Sci-ences, St Paul’s Hospital, and with the Division of Critical Care, Depart-ment of Medicine, University of British Columbia, Vancouver, BritishColumbia.

186 RESPIRATORY CARE • FEBRUARY 2009 VOL 54 NO 2

Introduction

Acute cardiogenic pulmonary edema (ACPE) is a com-mon cause of acute respiratory failure and often necessi-tates ventilatory support. The use of noninvasive ventila-tion (NIV) in the acute care setting in general has beenfueled by the desire to avoid the complications associatedwith intubation and invasive ventilation,1-3 includingtrauma to the larynx, pharynx, and trachea; arrhythmia;hypotension; aspiration of gastric contents; sinusitis; pneu-monia; and loss of the ability to eat and communicateverbally. By avoiding those complications NIV may, insome patient groups, reduce hospital morbidity, facilitateventilator weaning, shorten hospital stay, lower costs, andimprove patient comfort.1-3 This paper reviews the poten-tial benefit from continuous positive airway pressure(CPAP) and NIV in patients with acute respiratory failurecaused by ACPE.

We searched PubMed with the terms “noninvasive ven-tilation,” “non-invasive ventilation,” “noninvasive positivepressure ventilation,” “non-invasive positive pressure ven-tilation,” “nasal ventilation,” “bipap” (bi-level positive air-way pressure), “continuous positive airway pressure,” andterms that describe pulmonary edema and heart failure.We also scanned the citation lists of selected papers andreviewed our personal files. Two of us regularly conductliterature searches on NIV in MEDLINE, EMBASE, andthe Cochrane databases. Though there were some interest-ing cohort studies published that initially described the useof NIV and CPAP in cardiogenic pulmonary edema,4-9 inthis review we will restrict ourselves primarily to random-ized controlled trials; we use studies of other designs forbackground only. We did not include trials that have beenpublished only in abstract form. We do include informa-tion from recent systematic reviews of NIV for ACPE.

To expect benefit from any therapy in a specific patientpopulation there should be a supportive physiologic ratio-nale for its effectiveness. Although not a form of mechan-ical ventilatory assistance per se, CPAP is commonly usedto treat ACPE. Possible benefits of CPAP on ACPE in-

clude increased functional residual capacity, reduced atel-ectasis, reduced right-to-left intrapulmonary shunt, de-creased work of breathing from improved pulmonarycompliance (Table 1), and increased cardiac output fromdecreased left-ventricular preload and after-load. Presum-ably as a result of these physiologic factors, CPAP reducesmitral regurgitation in selected patients.10 NIV may pro-vide all these benefits plus inspiratory assistance to unloadthe respiratory muscles11 and alleviate respiratory distressmore quickly and effectively than CPAP alone.

Summary of the Randomized Controlled Trials

In this section we will describe the trials that have com-pared CPAP to standard therapy, NIV to standard therapy,and CPAP to NIV. The trials included 2 or 3 arms (thelatter included CPAP, NIV, and standard therapy). We willalso briefly discuss the systematic reviews of NIV andCPAP for ACPE.

Continuous Positive Airway Pressure Compared toConventional Treatment in Patients With AcuteCardiogenic Pulmonary Edema

We included 10 randomized controlled trials12-21 thatcompared CPAP to conventional treatment in patients withACPE (Table 2 and 4). The inclusion criteria were similarin most trials, and included respiratory distress, rales, andtypical chest radiographic findings. The usual exclusioncriteria were coma, inability to protect the airway, hemo-dynamic instability, or need for immediate intubation. Latertrials of both CPAP and NIV generally also excluded pa-tients with acute myocardial infarction that required revas-cularization with thrombolytics or invasive procedures. Allof the trials were conducted in only one center. A minorityof trials reported whether consecutive patients were en-rolled and whether randomization was concealed. All ofthe trials were unblinded because of the inherent difficultyof blinding the treating physician to the NIV mode. Mosttrials had either medication protocols for treatment of ACPEor similar administered medication doses in both patientgroups.

CPAP was delivered via face mask, for durations rangingfrom 3 hours12 to 9.3 hours.13 Most trials found significantimprovements in vital signs and gas exchange in the CPAPgroups. Six of the trials showed lower intubation and fail-ure rates with CPAP (at 10–12.5 cm H2O).12-15,20

Two of the trials merit extra discussion.16,17 These trialsapplied CPAP at 7.5 cm H2O via face mask to elderlypatients (� 75 y old) with ACPE. The CPAP-treated pa-tients in the trial by Kelly et al16 had more significantsymptom and physiologic improvement in the first hour,but no difference in intubation rate, compared to the con-trol group.16 However, there were no treatment failures inthe CPAP group, and only 2 treatment failures in the con-trol group, one of whom was successfully treated with

Table 1. Potential Mechanisms of Action of CPAP and NIV inPatients With Acute Cardiogenic Pulmonary Edema

CPAPIncreased functional residual capacityReduced atelectasisReduced right-to-left intrapulmonary shuntReduced work of breathing from improved pulmonary complianceIncreased cardiac output from reduced pre-load and after-loadReduced mitral regurgitation

NIVSame benefits as CPAPUnloads the respiratory muscles

CPAP � continuous positive airway pressureNIV � noninvasive ventilation

RESPIRATORY CARE • FEBRUARY 2009 VOL 54 NO 2 187

CPAP, which raises the possibility that the patients werenot that ill. Kelly et al also measured plasma epinephrine,norepinephrine, and brain natriuretic peptide, and foundno differences in neurohormone levels in the CPAP-treatedpatients versus the controls.

The trial by L’Her and colleagues17 was suspended afterthe first interim analysis because of a higher 48-h mortal-ity in the control group (7% vs 24%, P � .002), but therewas no difference in hospital mortality. However, 4 of thepatients in the control group who died had treatment with-drawn after study enrollment, which raised the mortality inthe control group. There were more “serious complica-tions” in the control group (17 vs 4, P � .002), but thatwas based on a composite outcome of coma, cardiac ar-rest, and worsening hypoxemia, and only the difference incoma was statistically significant between the groups. Fail-

ure of therapy (defined as the need for CPAP, noninvasivepressure-support ventilation [PSV], or intubation) occurredmore commonly in the control group (9% vs 30%, P � .01).Surprisingly, the most common precipitant of ACPE wasrespiratory-tract infection, which occurred more frequentlyin the control group, which raised the concern that patientsfailed to improve because they did not have ACPE.

In the trial by Lin and colleagues, all the patients had apulmonary artery catheter inserted for the study.14 Thoughpatients treated with CPAP had lower intrapulmonary shuntthan the control group, there were no differences in cardiacoutput or pulmonary artery occlusion pressures.

One small study, not included in the tables, found CPAPsafe and effective in patients with acute myocardial infarc-tion.22 Twenty-two patients with severe pulmonary edemaassociated with acute myocardial infarction were random-

Table 2. Randomized Controlled Trials That Compared CPAP to Standard Therapy in Patients With Acute Cardiogenic Pulmonary Edema

StudyYear

CPAPPressure

(cm H2O)

PatientsIntubation/Failure

RateP

(CPAP vs controlintubation/failure rate)

Other OutcomesCPAP

(n)Control

(n)CPAP

(%)Control

(%)

Rasanen12

198510 20 20 35 65 .07 CPAP group: respiratory rate, heart

rate, blood pressure, PaCO2, and

PaO2improved within 10 min.

Bersten13

199110 19 20 0 35 .005 At 30 min the CPAP group had

lower respiratory rate, heart rate,and PaCO2

; higher pH and PaO2.

CPAP group had shorter ICUstay.

Lin14

199512.5 50 50 16 36 .01 Over the 6 h of the study the

CPAP group had lowerrespiratory rate and heart rate,and higher PaO2

. All the patientshad pulmonary artery catheter.

Takeda15

19974–10 15 15 7 40 .05 Over the 12 h of the study the

CPAP group had lowerrespiratory rate and higher PaO2

/FIO2

. All the patients hadpulmonary artery catheter.

Kelly16

20027.5 27 31 0 7 NS At 1 h the CPAP group had lower

respiratory rate, heart rate,acidosis, and dyspnea. Twopatients who failed the controltherapy were successfully treatedwith CPAP.

L’Her17

20047.5 43 46 9 30 .01 Patients � 75 y old. The CPAP

group had lower respiratory rateand heart rate; higher PaO2

/FIO2;

and lower 48-h mortality (7% vs24%, P � .002). The controlgroup had more seriouscomplications.

CPAP � continuous positive airway pressureICU � intensive care unitFIO2 � fraction of inspired oxygenNS � nonsignificant

NONINVASIVE VENTILATION FOR ACUTE CARDIOGENIC PULMONARY EDEMA


ized to either CPAP or oxygen therapy. Two patients (18%)in the CPAP group and 8 patients (73%) in the oxygengroup required intubation (P � .03), and hospital mortalitywas lower in the CPAP group than in the oxygen group(9% vs 64%, P � .02).

The trials by Kelly and colleagues16 and L’Her andcolleagues17 both found early clinical and physiologic ben-efits from CPAP within the first hour, which suggests thatCPAP should be commenced as soon as possible afteradmission. No serious complications related to mask ven-tilation were reported in any of these studies. Notably,none of the trials found improvement in other outcomevariables, such as intensive-care-unit complications, inten-sive-care-unit or hospital stay, or mortality.

Several recent systematic studies explored the effective-ness of CPAP and NIV, compared to each other and tostandard treatment, in patients with ACPE.23-27 Pooling thetrials included in those reviews, which differ among them-selves and from the current review, suggested that CPAPmay be associated with lower hospital mortality (Fig. 1).Compared to standard care, CPAP improved clinical andphysiologic variables, reduced intubation rate, and mayhave significantly lowered hospital mortality.23-27

Noninvasive Ventilation Compared to ConventionalTreatment in Patients with Acute CardiogenicPulmonary Edema

In the 8 trials18-21,28-31 that compared NIV (noninvasivePSV or BiPAP) to conventional oxygen therapy in patientswith ACPE (Table 3 and 4), the inclusion and exclusioncriteria were similar to those in the trials that comparedCPAP to conventional therapy. The trial by Ferrer andco-workers enrolled a heterogeneous group of patients withhypoxemic respiratory failure, and provided some infor-mation on the subgroup with ACPE.31 Other than the trialby Ferrer and colleagues,31 all the trials enrolled patientsin the emergency department. Though most studies deliv-ered NIV predominantly via face mask, Levitt and col-leagues29 also permitted nasal mask. The NIV durationranged from 2 h to 11.4 � 3.6 h.29,30

In contrast to the trials that compared CPAP to standardtherapy, the results of the NIV trials are less consistent. In5 of the trials NIV improved physiologic variables and gasexchange,18,20,21,28,30 and in 3 trials18,20,28 NIV significantlyreduced the intubation rate. Compared to standard therapy,NIV was not associated with greater risk of adverse events;specifically, there was no greater risk of myocardial in-farction. The individual studies did not suggest a differ-ence in mortality or stay. The differences in the results ofthese trials probably relate to patient factors (eg, groupnonhomogeneity such as inclusion of non-ACPE patients,and different severity of illness between studies) and dif-

ferences in study design (eg, the individual center’s expe-rience with NIV, type of ventilator, pressures applied, de-lay in initiating NIV, differences in adjunctive treatments,and prospective definitions of NIV failure and intubationcriteria).

Masip and colleagues randomized 40 patients to nonin-vasive PSV or oxygen.28 The noninvasive PSV group hadmore rapid improvement in clinical variables (eg, respira-tory rate and oxygen saturation) and a lower intubation

Fig. 1. Meta-analysis of effect of noninvasive ventilation on mor-tality. The vertical dotted line indicates the overall mortality effectof the treatment. The boxes indicate the relative risk. The horizon-tal lines indicate the 95% confidence intervals. CPAP � continu-ous positive airway pressure. (From Reference 24, with permis-sion.)



rate (5% vs 33%, P � .04). However, the control groupmay have had a greater severity of illness at baseline. Inpatients with hypercapnia, noninvasive PSV was associ-ated with more rapid reduction in PaCO2

than was the con-trol therapy.

In the emergency department, Levitt and colleagues ran-domized 38 patients with ACPE to either BiPAP or oxy-gen therapy.29 Nonconsecutive subjects were enrolled onlywhen one of the study personnel was present. BiPAP wasinitiated at low pressures (8/3 cm H2O) and titrated, butthe final pressures were not given. BiPAP was adminis-tered for a minimum of 120 min. There were no differ-ences in vital signs, gas exchange, or intubation rate be-tween the BiPAP and oxygen groups. Four patients did nottolerate BiPAP and were crossed over to the control arm;those patients’ outcome is unclear. Possible explanationsfor the lack of benefit from BiPAP include the low pres-sures applied and baseline differences between the groups.No severity-of-illness data were provided. Other limita-

tions of the trial included nonstandardized pharmacologictreatment for ACPE, and a lack of prospectively definedcriteria for intubation.

In a rigorously conducted study with 130 patients in 5emergency departments in Italy, Nava et al defined theintubation/NIV-failure criteria, and used intention-to-treatanalysis.30 Dyspnea, respiratory rate, and gas exchangeimproved in the patients who received noninvasive PSV.Surprisingly, there was no difference in the intubation rate,except in the a-priori-defined subgroup of patients withhypercapnia, even though the inspiratory and expiratorypressures were similar to that in the study by Masip andcolleagues.28 Nava et al proposed that the limited NIVexperience of most emergency departments might explainthe failure of NIV to reduce the intubation rate.30

Compared to standard care, NIV improved clinical andphysiologic variables and reduced the intubation rate. Noneof the 3 trials found NIV to reduce mortality or stay.However, the systematic reviews23-25,27 concluded that NIV

Table 3. Randomized Controlled Trials That Compared NIV to Standard Therapy in Patients With Acute Cardiogenic Pulmonary Edema

StudyYear

Type ofVentilation

Inspiratory andExpiratoryPressure

(cm H2O)

Patients Intubation Rate P(NIV vscontrol

intubation rate)

Other OutcomesNIV(n)

Control(n)

NIV(%)

Control(%)

Masip28

2000PSV 15/5 19 18 5 33 .04 Respiratory rate decreased.

Pulse-oximetry readings increasedwithin first 15 min of NIV.

PaO2/FIO2

was higher in PSV groupfor first 2 h.

MI in 33% of control group and26% of NIV group (NS).

Levitt29

2001BiPAP 8/3 initially 21 17 24 41 NS No difference in vital signs or gas

exchange.MI in 29% of control group and

19% of BiPAP group (NS).Nava30

2003PSV 14/6 65 65 20 25 .53 PSV group had decreased dyspnea,

respiratory rate, and PaCO2, and

increased PaO2.

Less endotracheal intubation in thehypercapnic subgroup.

MI in 16% of the control groupand 16% of the NIV group.

Ferrer31

2003BiPAP 16/7 15 15 (subgroup) 26 73 ND Outcomes not reported for

subgroup with ACPE.The BiPAP group had decreased

respiratory rate and increasedPaO2

/FIO2for entire study

population.

NIV � noninvasive ventilationPSV � pressure-support ventilationFIO2 � fraction of inspired oxygenBiPAP � bi-level positive airway pressureNS � difference not significantMI � myocardial infarctionND � no data providedACPE � acute cardiogenic pulmonary edema



significantly reduced intubation and nonsignificantly re-duced mortality (Figs. 1 and 2). However, those authorscautioned that there was substantial trial heterogeneity,particularly in the pooled analysis of need for intubation.In their meta-analysis, Winck and colleagues postulatedthat (in contrast to CPAP) NIV failed to reduce mortalitybecause of the low study power caused by the small num-ber of patients in the trials.25

Continuous Positive Airway Pressure Compared toNoninvasive Ventilation in Patients With AcuteCardiogenic Pulmonary Edema

We included 11 randomized controlled trials thatcompared CPAP to NIV in patients with ACPE (Table4).18-21,32-38 The trial by Ferrari and colleagues36 was con-ducted in a high-dependence unit, but all the other trialswere conducted in emergency departments. Nine trials de-livered CPAP and NIV via face mask, one trial did notspecify how CPAP and NIV were delivered,33 and Mehtaand colleagues used nasal mask.32 Most of the trials ap-plied concealed randomization and included a suggestedmedication protocol or reported similar medication doses.Mean CPAP pressure ranged from 7.5 cm H2O to11 cm H2O. Mean NIV pressures ranged from 12/4 cm H2Oto 25/5 cm H2O.18,33 Moritz and colleagues37 differed fromthe other studies in that they used the Boussignac CPAPdevice, which is a disposable mask that generates5–10 cm H2O of CPAP with 15–25 L/min oxygen flow,without a ventilator. The duration of CPAP or NIV rangedfrom 102 min18 to 8 hours.36

In the first published trial, Mehta and colleagues32 foundthat patients treated with NIV had more rapid PaCO2

re-duction than the CPAP group, but the myocardial infarc-tion rate was higher (71% in the NIV group vs 31% in theCPAP group, P � .05). The intubation rates were similar(7% in the NIV group, 8% in the CPAP group), and werelower than the 33% intubation rate in a historical controlgroup. Morbidity and mortality were similar between the 2groups. More NIV patients than CPAP patients had chestpain on study entry (10 vs 4, P � .06), which suggestsinequalities between the groups, despite concealed ran-domization. Mehta et al concluded that most patients canbe managed successfully with CPAP alone.

Ten subsequent randomized trials compared NIV toCPAP. Overall, there does not appear to be any importantadvantage to NIV over CPAP; both modes similarly im-prove vital signs and gas exchange, and there is no sig-nificant difference in intubation rate or hospital mortality.Two trials20,32 found more rapid PaCO2

improvement withNIV than with CPAP, but that finding was not confirmedin other trials.

Four of the trials had 3 treatment arms.18-21 In their firststudy, Park and colleagues found that CPAP-treated pa-tients had a similar intubation rate to the control group,and none of the BiPAP-treated patients required intuba-tion.18 These surprising findings were explained by thesmall numbers of patients in the trial and the relatively lowCPAP pressure (7.5 cm H2O). In a subsequent larger trial,Park and colleagues found that CPAP and NIV were equallyeffective in reducing the intubation rate, compared to stan-dard care.20 In addition, 15-day mortality was significantlylower in both NIV groups, compared to standard care, but

Fig. 2. Meta-analysis of effect of noninvasive ventilation on intu-bation rate. The vertical dotted line indicates the overall mortalityeffect of the treatment. The boxes indicate the relative risk. Thehorizontal lines indicate the 95% confidence intervals. CPAP �continuous positive airway pressure. (From Reference 24, withpermission.)



Tab

le4.

Ran

dom

ized

Con

trol

led

Tri

als

Tha

tC

ompa

red

NIV

toC

PAP

inPa

tient

sW

ithA

cute

Car

diog

enic

Pulm

onar

yE

dem

a

Stud

yY

ear

Tec

hniq

ues

Com

pare

d

Pres

sure

Patie

nts

Intu

batio

nR

ate

P(C

PAP

vsN

IVin

tuba

tion

rate

)O

ther

Out

com

esC

PAP

(cm

H2O

)N

IV(c

mH

2O

)C

PAP

(n)

NIV (n)

CPA

P(%

)N

IV(%

)

Meh

ta32

1997

CPA

Pvs

BiP

AP

1015

/513

148

7N

DB

iPA

Pim

prov

edP a

CO

2,vi

tal

sign

s,an

ddy

spne

afa

ster

than

did

CPA

P.M

Iin

71%

ofN

IVgr

oup

vs31

%of

CPA

Pgr

oup.

Park

18

2001

CPA

Pvs

BiP

AP

vsO

2

7.5

12/4

97

330

.05

At

10m

inth

eB

iPA

Pgr

oup

had

low

erre

spir

ator

yra

tean

dhi

gher

P aO

2.N

odi

ffer

ence

betw

een

the

3gr

oups

at60

min

.10

patie

nts

inO

2ar

m40

%of

patie

nts

inO

2

arm

wer

ein

tuba

ted

Cro

ss33

2003

CPA

Pvs

BiP

AP

5–20

10–2

5/5

3635

113

.5N

odi

ffer

ence

sin

any

mea

sure

dva

riab

le.

Cra

ne19

2004

CPA

Pvs

BiP

AP

vsO

2

1015

/520

205

5N

SN

odi

ffer

ence

invi

tal

sign

s,ga

sex

chan

ge,

orM

Ira

tebe

twee

n3

the

grou

ps.

Hos

pita

lsu

rviv

al10

0%in

CPA

P,75

%in

BiP

AP

grou

p,an

d70

%O

2-t

hera

pygr

oup

(P�

.03)

,bu

tth

ere

was

nodi

ffer

ence

in7-

dsu

rviv

al.

20pa

tient

sin

O2

arm

No

patie

nts

inO

2ar

mw

ere

intu

bate

d

Bel

lone

34

2004

CPA

Pvs

noni

nvas

ive

PSV

1015

/522

244.

58.

3N

DB

oth

grou

psha

dsi

mila

rim

prov

emen

tin

vita

lsi

gns

and

gas

exch

ange

.N

odi

ffer

ence

inM

Ior

mor

talit

yra

te.

Park

20

2004

CPA

Pvs

BiP

AP

vsO

2

1117

/11

2727

77

.001

CPA

Pan

dB

iPA

Pgr

oups

had

sim

ilar

impr

ovem

ent

invi

tal

sign

san

dga

sex

chan

ge(b

oth

bette

rth

anO

2

grou

p).

MI

in12

%of

cont

rol

grou

pan

d4%

inB

iPA

Pan

dC

PAP

grou

ps.

26pa

tient

sin

O2

arm

42%

ofpa

tient

sin

O2

arm

wer

ein

tuba

ted

Bel

lone

35

2005

CPA

Pvs

noni

nvas

ive

PSV

1015

/518

185.

511

.1N

DE

nrol

led

only

hype

rcap

nic

patie

nts.

Bot

hgr

oups

had

sim

ilar

impr

ovem

ent

inre

spir

ator

yra

te,

pH,

and

CO

2,

at1

h.Fe

rrar

i36

2007

CPA

Pvs

noni

nvas

ive

PSV

8.8

15/7

2725

04

ND

The

grou

psha

dsi

mila

rvi

tal

sign

s,ga

sex

chan

ge,

stay

,tr

opon

inI,

MI

rate

,an

dm

orta

lity.

Mor

itz37

2007

CPA

Pvs

BiP

AP

7.7

12/4

.959

502

4N

DB

oth

grou

psha

dsi

mila

rvi

tal

sign

s,ga

sex

chan

ge,

MI

rate

,an

dm

orta

lity.

Hyp

erca

pnic

patie

nts

had

mor

eco

mpl

icat

ions

than

nonh

yper

capn

icpa

tient

s.R

uste

rhol

tz38

2008

CPA

Pvs

PAV

9.3

14/4

.219

1729

31N

DT

hegr

oups

had

sim

ilar

vita

lsi

gns,

gas

exch

ange

,M

Ira

te,

and

mor

talit

y.G

ray2

1

2008

CPA

Pvs

noni

nvas

ive

PSV

vsO

2

1014

/734

635

62.

43.

5N

SC

PAP

and

noni

nvas

ive

PSV

grou

psha

dsi

mila

rim

prov

emen

tsin

vita

lsi

gns,

gas

exch

ange

(bot

hbe

tter

than

O2

grou

p).

No

diff

eren

ces

inM

Ira

te,

stay

,or

mor

talit

ybe

twee

nth

e3

grou

ps.

367

patie

nts

inO

2ar

m2.

8%of

patie

nts

inO

2

arm

wer

ein

tuba

ted

NIV

�no

ninv

asiv

eve

ntila

tion

CPA

P�

cont

inuo

uspo

sitiv

eai

rway

pres

sure

BiP

AP

�bi

-lev

elpo

sitiv

eai

rway

pres

sure

ND

�no

data

avai

labl

eM

I�

myo

card

ial

infa

rctio

nN

S�

diff

eren

ceno

tsi

gnif

ican

tPS

V�

pres

sure

-sup

port

vent

ilatio

n



hospital mortality did not differ.20 In contrast, Crane andcolleagues found no difference in intubation rate with CPAPor NIV, compared to standard care, but the intubation rateswere extremely low (5% in the CPAP and NIV groups,and zero in the control group).19 Given that nonconsecu-tive patients were enrolled, it is possible that the sickestpatients were intubated without being considered for studyenrollment. Surprisingly, the CPAP patients were morelikely to survive to hospital discharge (100%) than werethose treated with NIV (75%) or oxygen therapy (70%,P � .03). However, it is difficult to ascribe the lowerhospital mortality rate in the CPAP group to the ventila-tion mode, given that the intubation rates were similar, andthe 7-day survival did not differ between the groups.

The largest and most recently published trial was pow-ered to detect a difference in 7-day mortality.21 Gray et alrandomized 1,069 patients in 26 Scottish emergency de-partments to CPAP, noninvasive PSV, or oxygen therapy.Patients � 16 years old with a chest radiograph consistentwith ACPE, respiratory frequency � 20 breaths/min, andpH � 7.35 were included. Though medication administra-tion was not standardized, the administered doses weresimilar in the 3 groups. Compared to standard therapy,both CPAP and noninvasive PSV were associated withsimilar improvements in dyspnea, heart rate, pH, and PaCO2

within one hour. In contrast to many other trials and tosystematic reviews, there was no difference in intubationrate or mortality with either CPAP or NIV, compared tostandard therapy; however, the intubation rates were verylow (range 2.4–3.5% in the 3 groups).

There are several potential explanations for these sur-prising results. First, several factors, such as the relativelymild respiratory rate and pH inclusion criteria, the lowintubation rates, the short durations (approximately 2 h) ofNIV application, and the lower-than-anticipated mortality(9.8% in the control group, compared to a 15% anticipatedrate), suggest that the patients were not that ill and did notrequire any ventilatory intervention. Second, crossoverswere permitted, and 19% of the patients in the oxygen armfailed therapy and were treated with CPAP or noninvasivePSV. Had those control patients all been intubated, theintubation rate in the control group might have been sig-nificantly higher than both the CPAP and noninvasive PSVgroups.

A recently published randomized trial compared pro-portional-assist ventilation to face-mask CPAP in 36 pa-tients with ACPE.38 It is not clear whether selected pa-tients were enrolled, because no data regarding patientscreening and enrollment were provided. Medical treat-ment was standardized, and there were objective intuba-tion criteria. The groups had similar improvements in clin-ical and physiologic variables, including PaCO2

, and therewas no difference in intubation rate or mortality. Thoughthey did not evaluate the work load associated with initi-

ation and titration of CPAP and proportional-assist venti-lation, the latter appears to require greater expertise. Thestudy did not find any advantages from proportional-assistventilation over CPAP for patients with ACPE.

Bellone and colleagues35 randomized 36 patients withACPE and hypercapnia to NIV (15/5 cm H2O) or CPAP(10 cm H2O). The average PaCO2

in the CPAP and BiPAPgroups was 60.5 mm Hg and 65.7 mm Hg, respectively.The study was powered to detect a 10-min difference inclinical “resolution” of ACPE. Both modes rapidly im-proved the major physiologic variables and resolved ACPE,and the intubation and mortality rates did not differ be-tween the modes. Bellone et al concluded that NIV is notsuperior to CPAP when signs of respiratory pump failureare present. Subgroup analysis of hypercapnic patients en-rolled in 2 other trials36,37 found similar PaCO2

, regardlessof assigned noninvasive strategy. However, Moritz andcolleagues observed that hypercapnic patients were morelikely to develop the combined outcome of intubation,death, or acute myocardial infarction, as well as cardio-genic shock, gastric acid aspiration, or require a switch ofthe ventilation mode, during the first 24 hours.37

Though an early trial suggested greater risk of myocar-dial infarction with NIV than with CPAP,32 none of thesubsequent trials that reported that outcome confirmed thatfinding. Two studies that specifically examined myocar-dial infarction rate found no difference in the risk of myo-cardial infarction or cardiac enzyme level between theNIV and CPAP patients.34,36 Other reported complicationsin the studies that compared CPAP to NIV were minor andinfrequent: facial erythema in 5%, vomiting in 5%,19 andgastric distention in 18% of CPAP patients and 30% ofNIV patients.20 Intolerance of CPAP and NIV was ob-served in 15% and 10% of patients, respectively.19

A study by Sharon and colleagues raised concerns aboutthe safety of BiPAP in the treatment of patients with se-vere pulmonary edema.39 Forty consecutive patients wererandomized to receive high-dose isosorbide dinitrate, orBiPAP (initially at 8/3 cm H2O) and standard-dose isosor-bide dinitrate. The patients treated with BiPAP had slowerresolution of clinical variables, and a higher rate of intu-bation (80% vs 20%, P � .001), myocardial infarction(55% vs 10%, P � .006), and death (10% vs 0%, P � .05).The study was terminated early because of these findings,and the investigators concluded that “BiPAP…is detri-mental to patients with severe pulmonary edema.” How-ever, several serious concerns about the study limit theability to draw that conclusion. Randomization was notconcealed, the treatment was delivered pre-hospital bymobile intensive care units, and the intubation criteriawere vague. Most importantly, very low inspiratory andexpiratory pressures were applied (mean 9/4 cm H2O),which suggests that the ventilatory assistance may have



been insufficient. Finally, the enrolled patients were quiteill, and may not have been appropriate for NIV, as evi-denced by the very high intubation rate (80%) in the BiPAPgroup.

Compared to CPAP, NIV has been variably found tomore rapidly improve clinical and physiologic variables,but NIV does not reduce intubation rate or mortality. Sys-tematic reviews have not found NIV superior to CPAP inavoiding intubation or lowering mortality (see Figs. 1 and 2)or myocardial infarction rate.24-26

Summary

In patients with ACPE, numerous randomized trials havecompared CPAP to conventional therapy, NIV to conven-tional therapy, and CPAP to NIV. There were importantmethodological limitations in many of the trials, such aslack of blinding and inclusion of only a proportion ofpatients who presented with ACPE, which restricts thegeneralizability of these results to all patients with ACPE.Nevertheless, based on these trials, noninvasive PSV andCPAP equally and safely improve vital signs and gas ex-change, and the systematic reviews found that noninvasivePSV and CPAP reduce the intubation rate in patients withACPE, compared to conventional therapy. To date, no trialhas been sufficiently powered to confirm a mortality ben-efit from either technique. The suggestion, in the system-atic reviews, of a mortality benefit from CPAP but notNIV must be considered cautiously, as those analyses arebased on small trials with few events (deaths), and thetrials were conducted over almost 2 decades, during whichtreatment for ACPE has evolved. It remains to be deter-mined in the current era of ACPE therapy whether CPAPor NIV confer a mortality benefit. At this time we cannotconclude that NIV offers any advantages over CPAP.

REFERENCES

1. Mehta S, Hill NS. Noninvasive ventilation. Am J Respir Crit CareMed 2001;163(2):540-577.

2. Penuelas O, Frutos-Vivar F, Esteban A. Noninvasive positive-pres-sure ventilation in acute respiratory failure. CMAJ 2007;177(10):1211-1218.

3. Hill NS, Brennan J, Garpestad E, Nava S. Noninvasive ventilation inacute respiratory failure. Crit Care Med 2007;35(10):2402-2407.

4. Hoffmann B, Welte T. The use of noninvasive pressure supportventilation for severe respiratory insufficiency due to pulmonaryoedema. Intensive Care Med 1999;25(1):15-20.

5. Kelly AM, Georgakas C, Bau S, Rosengarten P. Experience with theuse of continuous positive airway pressure (CPAP) therapy in theemergency management of acute severe cardiogenic pulmonary oe-dema. Aust N Z J Med 1997;27(3):319-322.

6. Minuto A, Giacomini M, Giamundo B, Tartufari A, Denkewitz T,Marzorati S, et al. Non-invasive mechanical ventilation in patientswith acute cardiogenic pulmonary edema. Minerva Anestesiol 2003;69(11):835-840.

7. Rusterholtz T, Kempf J, Berton C, Gayol S, Tournoud C, ZaehringerM, et al. Noninvasive pressure support ventilation (NIPSV) with facemask in patients with acute cardiogenic pulmonary edema (ACPE).Intensive Care Med 1999;25(1):21-28.

8. Sacchetti AD, Harris RH, Paston C, Hernandez Z. Bi-level positiveairway pressure support system use in acute congestive heart failure:preliminary case series. Acad Emerg Med 1995;2(8):714-718.

9. Vaisanen IT, Rasanen J. Continuous positive airway pressure andsupplemental oxygen in the treatment of cardiogenic pulmonaryedema. Chest 1987;92(3):481-485.

10. Bellone A, Barbieri A, Ricci C, Iori E, Donateo M, Massobrio M,Bendinelli S. Acute effects of non-invasive ventilatory support onfunctional mitral regurgitation in patients with exacerbation of con-gestive heart failure. Intensive Care Med 2002;28(9):1348-1350.

11. Chadda K, Annane D, Hart N, Gajdos P, Raphael JC, Lofaso F.Cardiac and respiratory effects of continuous positive airway pres-sure and noninvasive ventilation in acute cardiac pulmonary edema.Crit Care Med 2002;30(11):2457-2461.

12. Rasanen J, Heikkila J, Downs J, Nikki P, Vaisanen I, Viitanen A.Continuous positive airway pressure by face mask in acute cardio-genic pulmonary edema. Am J Cardiol 1985;55(4):296-300.

13. Bersten AD, Holt AW, Vedig AE, Skowronski GA, Baggoley CJ.Treatment of severe cardiogenic pulmonary edema with continuouspositive airway pressure delivered by face mask. N Engl J Med1991;325(26):1825-1830.

14. Lin M, Yang YF, Chiang HT, Chang MS, Chiang BN, Cheitlin MD.Reappraisal of continuous positive airway pressure therapy in acutecardiogenic pulmonary edema. Short-term results and long-term fol-low-up. Chest 1995;107(5):1379-1386.

15. Takeda S, Takano T, Ogawa R. The effect of nasal continuous pos-itive airway pressure on plasma endothelin-1 concentrations in pa-tients with severe cardiogenic pulmonary edema. Anesth Analg 1997;84(5):1091-1096.

16. Kelly CA, Newby DE, McDonagh TA, Mackay TW, Barr J, BoonNA, et al. Randomised controlled trial of continuous positive airwaypressure and standard oxygen therapy in acute pulmonary oedema;effects on plasma brain natriuretic peptide concentrations. Eur Heart J2002;23(17):1379-1386.

17. L’Her E, Duquesne F, Girou E, de Rosiere XD, Le Conte P, RenaultS, et al. Noninvasive continuous positive airway pressure in elderlycardiogenic pulmonary edema patients. Intensive Care Med 2004;30(5):882-888.

18. Park M, Lorenzi-Filho G, Feltrim MI, Viecili PR, Sangean MC,Volpe M, et al. Oxygen therapy, continuous positive airway pres-sure, or noninvasive bilevel positive pressure ventilation in the treat-ment of acute cardiogenic pulmonary edema. Arq Bras Cardiol 2001;76(3):221-230.

19. Crane SD, Elliott MW, Gilligan P, Richards K, Gray AJ. Random-ised controlled comparison of continuous positive airways pressure,bilevel non-invasive ventilation, and standard treatment in emer-gency department patients with acute cardiogenic pulmonary oe-dema. Emerg Med J 2004;21(2):155-161.

20. Park M, Sangean MC, Volpe Mde S, Feltrim MI, Nozawa E, LeitePF, et al. Randomized, prospective trial of oxygen, continuous pos-itive airway pressure, and bilevel positive airway pressure by facemask in acute cardiogenic pulmonary edema. Crit Care Med 2004;32(12):2407-2415.

21. Gray A, Goodacre S, Newby DE, Masson M, Sampson F, Nicholl J;3CPO Trialists. Noninvasive ventilation in acute cardiogenic pulmo-nary edema. N Engl J Med 2008;359(2):142-151.

22. Takeda S, Nejima J, Takano T, Nakanishi K, Takayama M, Saka-moto A, Ogawa R. Effect of nasal continuous positive airway pres-sure on pulmonary edema complicating acute myocardial infarction.Jpn Circ J 1998;62(8):553-558.



23. Masip J, Roque M, Sanchez B, Fernandez R, Subirana M, ExpositoJA. Noninvasive ventilation in acute cardiogenic pulmonary edema:systematic review and meta-analysis. JAMA 2005;294(24):3124-3130.

24. Peter JV, Moran JL, Phillips-Hughes J, Graham P, Bersten AD.Effect of non-invasive positive pressure ventilation (NIPPV) on mor-tality in patients with acute cardiogenic pulmonary oedema: a meta-analysis. Lancet 2006;367(9517):1155-1163.

25. Winck JC, Azevedo LF, Costa-Pereira A, Antonelli M, Wyatt JC.Efficacy and safety of non-invasive ventilation in the treatment ofacute cardiogenic pulmonary edema–a systematic review and meta-analysis. Crit Care 2006;10(2):R69.

26. Ho KM, Wong K. A comparison of continuous and bi-level positiveairway pressure non-invasive ventilation in patients with acute car-diogenic pulmonary oedema: a meta-analysis. Crit Care 2006;10(2):R49.

27. Collins SP, Mielniczuk LM, Whittingham HA, Boseley ME, SchrammDR, Storrow AB. The use of noninvasive ventilation in emergencydepartment patients with acute cardiogenic pulmonary edema: a sys-tematic review. Ann Emerg Med 2006;48(3):260-269.

28. Masip J, Betbese AJ, Paez J, Vecilla F, Canizares R, Padro J, et al.Non-invasive pressure support ventilation versus conventional oxy-gen therapy in acute cardiogenic pulmonary oedema: a randomisedtrial. Lancet 2000;356(9248):2126-2132.

29. Levitt MA. A prospective, randomized trial of BiPAP in severe acutecongestive heart failure. J Emerg Med 2001;21(4):363-369.

30. Nava S, Carbone G, DiBattista N, Bellone A, Baiardi P, Cosentini R,et al. Noninvasive ventilation in cardiogenic pulmonary edema: amulticenter randomized trial. Am J Respir Crit Care Med 2003;168(12):1432-1437.

31. Ferrer M, Esquinas An, Leon M, Gonzalez G, Alarcon A, Torres A.Noninvasive ventilation in severe hypoxemic respiratory failure: arandomized clinical trial. Am J Respir Crit Care Med 2003;168(12):1438-1444.

32. Mehta S, Jay GD, Woolard RH, Hipona RA, Connolly EM, CiminiDM, et al. Randomized, prospective trial of bilevel versus continu-ous positive airway pressure in acute pulmonary edema. Crit CareMed 1997;25(4):620-628.

33. Cross AM, Cameron P, Kierce M, Ragg, Nelly AM. Non-invasiveventilation in acute respiratory failure: a randomized comparison ofcontinuous positive airway pressure and bi-level positive airway pres-sure. Emerg Med J 2003;20(6):531-534.

34. Bellone A, Monari A, Cortellaro F, Vettorello M, Arlati S, Coen D.Myocardial infarction rate in acute pulmonary edema: noninvasivepressure support ventilation versus continuous positive airway pres-sure. Crit Care Med 2004;32(9):1860-1865.

35. Bellone A, Vettorello M, Monari A, Cortellaro F, Coen D. Nonin-vasive pressure support ventilation vs. continuous positive airwaypressure in acute hypercapnic pulmonary edema. Intensive Care Med2005;31(6):807-811.

36. Ferrari G, Olliveri F, De Filippi G, Milan A, Apra F, Boccuzzi A, etal. Noninvasive positive airway pressure and risk of myocardialinfarction in acute cardiogenic pulmonary edema: continuous posi-tive airway pressure vs noninvasive positive pressure ventilation.Chest 2007;132(6):1804-1809.

37. Moritz F, Brousse B, Gellee B, Chajara A, L’Her E, Hellot MF,Benichou J. Continuous positive airway pressure versus bilevel non-invasive ventilation in acute cardiogenic pulmonary edema: a ran-domized multicenter trial. Ann Emerg Med 2007;50(6):666-675.

38. Rusterholtz T, Bollaert P, Feisssel M, Romano-Girard F, Harlay M,Zaehringer M, Dusang B, Sauder P. Continuous positive airwaypressure vs. proportional assist ventilation for noninvasive ventila-tion in acute cardiogenic pulmonary edema. Intensive Care Med2008;34(5):840-846.

39. Sharon A, Shpirer I, Kaluski E, Moshkovitz Y, Milovanov O, PolakR, et al. High-dose intravenous isosorbide-dinitrate is safer and bet-ter than Bi-PAP ventilation combined with conventional treatmentfor severe pulmonary edema. J Am Coll Cardiol 2000;36(3):832-837.

Discussion

Pierson: Has anyone looked at pa-tient preferences for NIV versusCPAP?

Mehta: I haven’t seen any trials onthat.

Gay: The studies you’ve describedspan a couple of decades. How muchdo you think the findings might beconfounded by the fact that—and I’mnot as cynical as this may sound—wehaven’t gotten that much better at us-ing NIV but the cardiologists have got-ten a lot better at treating ACPE?

Benditt: MI [myocardial infarc-tion], you mean?

Gay: Well, the reduction in mortal-ity when they aggressively treat ACPE

now, I think, is just as impressive aswhen they attack acute ischemia.

Mehta: So, you’re asking whethermedical management has changed inthe last 30 years?

Gay: Hasn’t better ACPE therapymore likely influenced the outcomethan intervention with NIV?

Mehta: I can’t tell you whetherthat’s the case. I assume so, but it isdifficult to compare doses or medica-tions between trials, and some acuteinterventions, such as angioplasty,have become more common recently.I think most of the cardiologists I workwith tend to intubate people veryquickly, and they don’t use or are veryreluctant to use NIV, and they’re un-comfortable with it. NIV doesn’t seemto be highlighted at the cardiology

conferences the way it is at the pul-monary and critical care conferences.

Hill: Or in the emergency depart-ment.

Mehta: Right.

Hill: It’s the emergency doctors whoare on the front lines for this. You’reright about the cardiologists; they seemto be more at ease with the EMT [emer-gency medical technician] crew thanwith a CPAP mask.

Benditt: Maybe this is just at myinstitution, but our cardiologists use alot of CPAP. The other day there were4 people in the ward on CPAP. I am afirm believer that CPAP is the betterchoice, because physiologically itmakes a lot more sense to me. OnCPAP, as our mentor John Butler



taught us, the endothoracic pressure ishigher, and the pre-load and after-loadare reduced quite effectively withCPAP, so it makes great physiologicsense to me, and I think the data showthat.

Mehta: But BiPAP offers all of thesebenefits as soon as you have a highenough PEEP [positive end-expiratorypressure].

Benditt: But I’m not sure how thecardiac cycle is interacting with therespiratory cycle; it may change; Idon’t know. With CPAP you knowthe pressure you’re at, and that’s whatyou’re going to get.

Mehta: I agree. In patients withACPE there are many reasons to useCPAP: it’s a lot easier to use, there’sno titration involved, anyone can ap-ply it, and it’s easier to understand.However, even though I’ve tried toconvince people at my institution touse CPAP, they hardly even seem toknow what CPAP is any more. Anytime I go down to the emergency de-partment or the critical care unit orstep-down unit they’re using BiPAPregardless of the etiology of the respi-ratory failure.

Benditt: And that’s the problem, Ithink.

Hill: When I give NIV talks, I ofteninformally survey the audience abouttheir choice of BiPAP versus CPAPfor ACPE. Usually they are over-whelmingly in favor of BiPAP. I’mnot disturbed by this though, because,as you said, Geeta, it really doesn’tmake any difference. You increase in-trathoracic pressure either way. As bestwe can tell, the studies show that theyhave essentially the same outcomes inthe end. I think there are some earlyphysiologic benefits of BiPAP overCPAP, as suggested by our study,1

especially for patients who are initiallyhypercapnic or very dyspneic, in whichcase BiPAP can reduce work of breath-

ing and PaCO2a little more quickly.

Crane et al2 also saw some of theseearlier physiologic benefits. The rea-son I’m not disturbed, though, is thatthe BiPAP equipment is there in theemergency department anyway, sothere’s no cost advantage to CPAP.Furthermore, it’s not that hard to ti-trate BiPAP, and people are usuallycomfortable with it.

1. Mehta S, Jay GD, Woolard RH, HiponaRA, Connolly EM, Cimini DM, et al. Ran-domized, prospective trial of bi-level ver-sus continuous positive airway pressure inacute pulmonary edema. Crit Care Med1997;25(4):620-628.

2. Crane SD, Elliott MW, Gilligan P, Rich-ards K, Gray AJ. Randomised controlledcomparison of continuous positive airwayspressure, bi-level noninvasive ventilation,and standard treatment in emergency de-partment patients with acute cardiogenicpulmonary oedema. Emerg Med J 2004;21(2):155-161.

Kacmarek: Stefano, in your sub-group analysis, did you find a benefitfrom NIV? Although I would agreethat it probably doesn’t make a differ-ence in the vast majority of patients, itseems reasonable to start out withCPAP, unless the patient is hypercar-bic. You can use the same ventilatorto do either CPAP or NIV. This couldbe more of an academic argument thana real discussion on managing patients,unless there is substantial hypercar-bia, in which case my bias is to pro-vide ventilation, as opposed to simpleCPAP.

Nava: Bellone et al1 excluded pa-tients with preexisting chronic respi-ratory disorders, and they found nodifference between CPAP and NIV.In our study2 the large majority of pa-tients with hypercapnia also had achronic respiratory disorder. So prob-ably what Bob said is very true onlywhen the main cause of hypercapniamay be muscle distress or something.Maybe hypercapnia in ACPE has dif-ferent meanings.

I think the helmet is one of the big-gest advances for ACPE, because it

allows treating the patient inside a pro-tected environment. For example, ifACPE occurs in the dialysis unit, yougo there with the helmet, you connectthe device into the centralized oxygenflow, and you can easily treat ACPEin patients on dialysis. The same ap-plies in an ambulance. Even the car-diologists have changed their mind;they seem to like the helmet a lot.

1. Bellone A, Vettorello M, Monari A, Cor-tellaro F, Coen D. Noninvasive pressuresupport ventilation vs continuous positiveairway pressure in acute hypercapnic pul-monary edema. Intensive Care Med 2005;31(6):807-811.

2. Nava S, Carbone G, DiBattista N, BelloneA, Baiardi P, Cosentini R, et al. Noninva-sive ventilation in cardiogenic pulmonaryedema: a multicenter randomized trial.Am J Respir Crit Care Med 2003;168(12):1432-1437.

Epstein: In scoring the trials, yougave a check when intubation criteriawere specified. How often do the in-vestigators go back to see whether theintubation criteria were accurately ap-plied? In my experience it’s still a judg-ment call. We tend to rate those trialshigher, but is that really the right thingto do?

Mehta: I agree. Obviously we haveto look for objective intubation crite-ria when we evaluate trial quality, butI’m not sure how much it actuallymeans to have objective intubationcriteria. For example, one of the stud-ies had “deterioration in gas exchangeor clinical status” as the intubationcriteria. So even if intubation criteriaare defined, the intubation decision isgenerally based on the clinical evalu-ation.

Hess: Regarding CPAP versus NIVfor hypercapnic ACPE, a 1991 ran-domized controlled trial by Berstenet al1 found a significant reduction inPaCO2

with CPAP.

1. Bersten AD, Holt AW, Vedig AE, Skow-ronski GA, Baggoley CJ. Treatment of se-vere cardiogenic pulmonary edema withcontinuous positive airway pressure deliv-



ered by face mask. N Engl J Med 1991;325(26):1825-1830.

Keenan: In your literature summaryyou said the systematic reviews iden-tified a mortality benefit from CPAP,but not for NIV. That finding is de-pendent on the trials available. I thinkwe can all accept the strong rationalethat NIV or CPAP should prevent in-tubation in patients with ACPE. I amless certain that NIV or CPAP willhave as direct an impact on mortalityas we might expect in patients withCOPD.

The reason for better survival in pa-tients with COPD is that NIV obvi-ates intubation. Intubated patients withCOPD usually require longer durationventilation than do those with ACPE.The latter generally either require briefventilation or do very poorly becauseof their underlying cardiac disease. Anintubated patient who is on the venti-lator longer is at higher risk of pneu-monia and its associated mortality.

Several studies on CPAP were con-ducted before the NIV studies, and theCPAP studies are driving the apparentmortality benefit. I believe that mor-tality in these patients is presentlylinked to their underlying cardiac dis-ease, and treatment for the same. Otherthan patients with comorbid COPD, Iam uncertain whether the apparentmortality benefit of CPAP versus NIVreported in the systematic reviews isrelevant today.

Hill: One of the limitations in someof the more recent trials that looked atnoninvasive positive pressure on theMI rate was the exclusion of patientswith acute coronary syndromes. Onecannot determine whether NIV has ad-verse effects on that population ifthey’re excluded from the studies. Didyou take that into consideration? Hassomebody looked at NIV in patientswith infarcts?

Nava: Takeda et al,1 in Japan, didone study. A Russian group was sched-uled to present another study at theEuropean Respiratory Society meeting2 years ago, but they didn’t show up.

1. Takeda S, Nejima J, Takano T, NakanishiK, Takayama M, Sakamoto A, Ogawa R.Effect of nasal continuous positive pres-sure on pulmonary edema complicatingacute myocardial infarction. Jpn Circ J1998;62(8):553-5558.

Mehta: Takeda et al randomized asmall number of patients who pre-sented with ACPE due to acute MI toeither CPAP or standard therapy. TheCPAP group had a lower intubationrate and hospital mortality than thestandard-therapy group.

Hill: Is that a concern in exoneratingthe modes contributing to infarction ifyou exclude the people at risk? Don’tyou consider that a problem?

Mehta: Of course, yes.

Doyle:* Is the current standard ofcare for ACPE endotracheal intuba-tion, CPAP, or standard O2 therapy?

Mehta: I don’t think that’s clear. Itdisturbs me that trials are still beingdone that include an oxygen arm asstandard therapy. Oxygen alone shouldnot be considered standard therapy forpatients with ACPE. I’m sure all ofyou here agree that CPAP or NIVshould be the standard therapy, but Idon’t think any recent surveys havelooked at what the first-line ACPEtherapy tends to be.

Nava: Park et al1 found several im-portant things. In their patient popu-lation only 30% of those with hyper-capnic ACPE fit their criteria, whichmeans that 7 out of 10 patients whowere only in the hospital for ACPEdid not need any mechanical ventila-tion.

1. Park M, Sangean MC, Volpe MS, FeltrimMI, Nozawa E, Leite PF, et al. Random-ized, prospective trial of oxygen, contin-uous positive airway pressure, and bi-levelpositive airway pressure by face mask inACPE. Crit Care Med 2004;32(12):2407-2415.

Mehta: Yes, many of the trials areplagued with the problem of “cherry-picking” and non-consecutive patientenrollment.

* Peter Doyle RRT, Respironics, Carlsbad, Cal-ifornia.



noninvasive ventilation in patients with acute cardiogenic...

Documents