history and epidemiology of noninvasive...

Conference Proceedings

History and Epidemiology of Noninvasive Ventilationin the Acute-Care Setting

David J Pierson MD FAARC

IntroductionEvolution of Ventilatory Support in Acute Respiratory FailureHistory of Noninvasive Ventilation in the Acute-Care SettingEpidemiology of Noninvasive Ventilation in the Acute-Care Setting

Noninvasive Ventilation Outside the Setting of Clinical Trials:Efficacy Versus Effectiveness

Data From Surveys: What Clinicians Say They DoData From Observational Studies of Actual NIV Use

Problems With the Accurate Assessment of Current NIV UseSummary

Although noninvasive ventilation (NIV) was first used to treat patients with acute respiratoryfailure in the 1940s, the history of this mainstay of today’s respiratory care armamentarium hasmainly been written in the last 20 years. There is now a robust evidence base documenting theefficacy of NIV in exacerbations of chronic obstructive pulmonary disease, cardiogenic pulmonaryedema, and acute respiratory failure in immunocompromised patients, and evidence in support ofNIV in other settings, such as hypoxemic acute respiratory failure and the management of patientswho decline endotracheal intubation, is accumulating rapidly. Efficacy as demonstrated in clinicaltrials does not necessarily translate to clinical effectiveness in practice, however, and importantbarriers need to be overcome if NIV is to realize for the average patient the potential it has shownin research studies. However, although the expansion of its use in everyday patient care has laggedbehind the growth of its evidence base, an increasing number of studies document the steadilyexpanding use of NIV in the acute-care setting. This article reviews the history of NIV as appliedin acutely ill patients and summarizes the studies of NIV outside the research setting during the lastdecade. Key words: noninvasive ventilation, NIV, epidemiology, history, clinical practice, acute respi-ratory failure, chronic obstructive pulmonary disease, COPD, acute care. [Respir Care 2009;54(1):40–50. © 2009 Daedalus Enterprises]

Introduction

Noninvasive ventilation (NIV) has become a requiredcomponent of the clinician’s armamentarium in the acute-

care setting. NIV is now the standard of care in acuterespiratory failure (ARF) due to chronic obstructive pul-monary disease (COPD),1-3 evidence is strong for NIV’sbenefits in at least some patients with cardiogenic pul-

David J Pierson MD FAARC is affiliated with the Division of Pulmonaryand Critical Care Medicine, Department of Medicine, Harborview Med-ical Center, and the University of Washington, Seattle, Washington.

Dr Pierson presented a version of this paper at the 42nd RESPIRATORY

CARE Journal Conference, “Noninvasive Ventilation in Acute Care: Con-troversies and Emerging Concepts,” held March 7-9, 2008, in Cancun,Mexico.

The author reports no conflict of interest related to the content of thispaper.

Correspondence: David J Pierson MD FAARC, Division of Pulmonaryand Critical Care Medicine, Harborview Medical Center, 325 Ninth Av-enue, Box 359762, Seattle WA 98104. E-mail: [email protected].

40 RESPIRATORY CARE • JANUARY 2009 VOL 54 NO 1

monary edema,4-6 and a rapidly evolving literature doc-uments its use in numerous other clinical settings.7-11

Table 1 lists the most prominent of these settings forNIV, in relation to the strength of the supporting evi-dence in each.10 The table refers mainly to the use ofNIV in adult patients, although this therapy is also be-ing used with increasing frequency in infants and chil-dren.12-14 To the clinical settings listed under the table’sthird category (that is, those supported by the least firmevidence at this point) can be added acute neuromus-cular disease,15-17 pre-hospital and emergency-depart-ment use for patients with acute respiratory distress,18-20

use during the performance of tracheotomy,21 and acuteapplication in palliative care.22-24

The literature on NIV consists primarily of the results oftechnical assessments and reports of clinical investigations.Much less has been written about the extent and nature ofNIV use in everyday patient care. As the first of the seriesof reviews developed from the conference, “NoninvasiveVentilation in Acute Care: Controversies and EmergingConcepts,” this article first traces the historical develop-ment of NIV as an intervention in managing acutely illpatients, and then reviews what is known about the clinicaluse of this therapy outside the research setting.

Evolution of Ventilatory Supportin Acute Respiratory Failure

Noninvasive methods for supporting ventilation havefeatured prominently throughout the history of respiratorycare, which in turn has been determined in large measureby the need to support the failing respiratory system (Ta-

ble 2). Although supplemental oxygen was used clinicallyin a few hospitals in the 1920s, the first feasible means forsustaining life in patients who were unable to breathe forthemselves came with the introduction of the tank venti-lator (iron lung) at the end of that decade.25-28 The emer-gence of mechanical ventilation in its modern sense wasspurred by the devastating polio epidemics of the 1950s,when experience in Denmark,29 and subsequently in theUnited States and elsewhere, demonstrated that lives couldbe saved acutely, and apneic patients supported virtuallyindefinitely with tracheostomy and positive-pressure ven-tilation.30-32 Thereafter, once mid-20th century medicineevolved from a home-based activity to an institution-fo-cused enterprise taking place primarily in hospitals,33 ad-vances in the understanding of normal and abnormal re-spiratory physiology combined with new devices and othertechnology to create the first intensive care units (ICUs),whose emergence was driven in large part by the need tosupport and monitor ventilation, oxygenation, and airwaycare.

By the early 1970s virtually every American acute-carehospital had an ICU, and a respiratory therapy departmentwhose members were becoming specialists in invasive me-chanical ventilation. Ventilators rapidly became more ca-pable and more sophisticated, with a plethora of new modesand other features, whose use was guided by blood gasanalysis and other new ways of physiologic monitoring.Soon, however, awareness of the complications of inva-sive mechanical ventilation34,35 and artificial airways,36

and subsequently of ventilator-induced lung injury,37,38 ledto renewed interest in less aggressive, potentially less in-jurious ventilatory support.

History of Noninvasive Ventilationin the Acute-Care Setting

The application of intermittent positive inspiratory pres-sure via an anesthesia mask in the treatment of acute re-spiratory illness was studied by Motley and colleagues atBellevue Hospital in the 1940s.39 These clinician-investi-gators used the apparatus shown in Figure 1 to deliverintermittent positive-pressure ventilation to patients withpneumonia, pulmonary edema, near-drowning, Guillain-Barre syndrome, and acute severe asthma.39 However, thisapproach to life support in the acute-care setting took aback seat to invasive mechanical ventilation as the latteremerged and was refined during the next 2 decades.

Noninvasive positive-pressure ventilation did not dis-appear from the scene, however; it found wide use bothin acute-care hospitals and for outpatient treatments inthe form of intermittent positive-pressure breathing(IPPB).40 So widespread did the use of IPPB become bythe early 1970s—administered to 10% or more of allhospitalized patients, with each respiratory therapist typ-

Table 1. Noninvasive Ventilation in the Acute Care Setting:Clinical Conditions and Strength of Supporting Evidence

Evidence from multiple randomized controlled trials and meta-analysesExacerbation of chronic obstructive pulmonary diseaseCardiogenic pulmonary edemaAcute respiratory failure in immunocompromised patientsPrevention of weaning failure in high-risk patientsNot effective in established extubation failure

Consistent findings in more than one published clinical trial, case-control series, or cohort studyPostoperative respiratory failureOxygenation prior to endotracheal intubationSupport during endoscopy

Case series or conflicting findings in other types of studiesAcute lung injury and acute respiratory distress syndromeExtubation failureAcute severe asthmaPneumoniaAcute respiratory failure in patients who do not wish to be intubated

(Adapted in part from Reference 10.)

HISTORY AND EPIDEMIOLOGY OF NONINVASIVE VENTILATION IN THE ACUTE-CARE SETTING

RESPIRATORY CARE • JANUARY 2009 VOL 54 NO 1 41

Table 2. Evolution of Ventilatory Support in the Acute-Care Setting, Particularly With Respect to Noninvasive Ventilation

Pre-1930sFirst clinical use of supplemental oxygen in hospitalsNo practical means for supporting ventilation

1930s-1940sIntroduction of tank ventilatorsSupport of apneic patient possible for first time

1950sPolio epidemics in Europe and United StatesIntroduction of positive-pressure ventilation via tracheostomyDevelopment of special cadre of hospital workers for caring for patients with respiratory problems (inhalation therapists)Use of supplemental oxygen and IPPB in aviation

1960sMajor progress in understanding pulmonary gas exchangeWidespread use of IPPB in United States hospitals for “breathing treatments”Experience with IPPB in acute respiratory insufficiencyWidespread introduction of volume ventilatorsAvailability of improved endotracheal tubesUse of arterial blood gases in patient assessmentFirst dedicated ICUsRecognition of ARDSFirst use of PEEP to treat hypoxemia in ARDS

1970sMajor progress in understanding lung physiology and pathologyUse of CPAP in neonatesPresence of ICUs in virtually all acute-care hospitalsMore sophisticated and capable ICU ventilatorsIntroduction of intermittent mandatory ventilation and other new ventilation modesIncreasing awareness of complications of invasive mechanical ventilationSugarloaf conference; de-emphasis of IPPB

1980sIncreasing focus on respiratory muscle function in acute care settingsInvasive mechanical ventilation as initial approach in virtually all settings of acute respiratory failureWidespread use of pulse oximetry and other noninvasive respiratory monitoringIncreasing computerization of ventilators and other respiratory care equipmentIntroduction of nasal CPAP for treating obstructive sleep apneaIncreasing experience with long-term NPPV in settings other than polioFirst reports of use of NPPV in acute hypercapnic respiratory failure in COPDIntroduction of pressure supportIntroduction of modern bi-level pressure-targeted ventilators for NPPV

1990sIncreasing reported experience with NPPV in acute-care settings other than COPDFirst randomized controlled trials of NPPV in acute respiratory failureIncorporation of FIO2

control and better monitoring into bi-level ventilators for NPPVIncreasing variety of patient interfaces for NPPVRESPIRATORY CARE consensus conference on NPPV in the acute care settingRapid worldwide dissemination of research findingsRise of evidence-based medicineIncreasing focus on ventilator-induced lung injury and concept of lung-protective ventilationConcept of NPPV as bridge to weaningVentilator-associated pneumonia and its relationship to intubationIncreased focus on DNAR/DNI and withdrawal of life support

2000sRich database on efficacy of NPPV: multiple RCTs; meta-analyses; evidence-based clinical practice guidelinesNPPV as standard of care for COPD exacerbationIncreasing use of NPPV in other settingsIncreased focus on DNI and palliative care in the acute-care settingIncreasing focus on knowledge-transfer and addressing the gap between efficacy and effectiveness

IPPB � intermittent positive-pressure breathing; ICU � intensive care unit; ARDS � acute respiratory distress syndrome; PEEP � positive end-expiratory pressure; CPAP � continuous positiveairway pressure; NPPV � noninvasive positive-pressure ventilation; COPD � chronic obstructive pulmonary disease; FIO2 � fraction of inspired oxygen; DNAR � do not attempt resuscitation;DNI � do not intubate; RCT � randomized controlled trial



ically giving 150 –200 “treatments” per month,41 at anannual cost to the United States health-care system ofmore than $400 million42—that the National Institutesof Health and the American Thoracic Society conveneda special conference (the “Sugarloaf Conference”) toreview the issue.43 In large part because of the dearth ofscientific evidence to support IPPB at that conference,its use subsequently decreased.

Although it was first tried as early as the 1950s, and wassubsequently used in a few centers of special expertise,44

long-term support of ventilation via NIV only became wide-spread starting in the 1980s.45 Continuous positive airwaypressure (CPAP), delivered via nasal mask to patients withobstructive sleep apnea, had been introduced by Sullivanet al in 1981.46 In 1987 Sullivan’s group reported thesuccessful use of NIV via nasal mask in 3 patients withpost-infection muscle weakness and 2 with muscular dys-trophy.47 Several other reports quickly followed and dem-onstrated that NIV could be effective in various long-termsettings and diagnoses.48-52

Stimulated by the successful application of nasal CPAPin sleep apnea, the availability of improved patient inter-faces, an increasing desire to avoid the complications ofinvasive mechanical ventilation, and the refusal of somepatients to be intubated, there followed a renewed interestin NIV for managing ARF.45 In 1989, Meduri and col-leagues reported the successful application of NIV viafull-face mask in 10 patients, and the avoidance of intu-bation in 8 of them (4 of 6 with COPD, 2 of 2 withcongestive heart failure, and 2 of 2 with pneumonia).53 Anumber of other studies confirmed the efficacy of NIV in

COPD exacerbations, using both nasal54-57 and full-facemasks.58,59

The increased use of NIV in the ICU and in other acute-care settings was facilitated by the introduction of im-proved bi-level ventilators that have effective compensa-tion for air leaks, such as the Respironics BIPAP ST/D,which replaced an earlier home-care model in the early1990s (personal communication, Derek Glinsman RRTFAARC, Respironics, June 10, 2008). In a 1995 review,Sassoon summarized the subsequent rapid increase in re-ported experience with NIV in various forms of ARF(Fig. 2).60 The years since 1995 have brought an avalancheof clinical investigations and other publications on the useof NIV in ARF (Fig. 3).10

As noted in a previous review of the history of NIV, thisform of ventilatory support has been called different thingsby different researchers, clinicians, and manufacturers,which led to confusion on the part of clinicians and inves-tigators alike.45 To some extent this diversity of terminol-ogy persists. However, as familiarity with NIV has in-creased, the resulting confusion may now be less. Theterm noninvasive positive-pressure ventilation (abbrevi-ated NPPV or NIPPV) was formerly used to distinguish itfrom noninvasive negative-pressure ventilation, althoughconsidering the rarity of the latter today, the simpler termNIV is more convenient. Because a number of bi-levelventilators are now available for NIV (and also because ofits use by one European manufacturer of ICU ventilatorsfor one of its modes), colloquial use of the term BIPAP (aproprietary product name) as a generic term for NIV shouldbe discouraged.

Fig. 1. Apparatus used by Motley and associates in the mid-1940sto deliver intermittent positive-pressure ventilation, with or withoutpositive end-expiratory pressure, to patients with acute respira-tory failure. A corrugated rubber hose (A) connected a Bendixpressure demand regulator (B) to a Bennett Clinical ResearchModel X-2 respirator (C), from which air or oxygen was deliveredto the patient by means of a Bennett face mask (D). (From Refer-ence 39, with permission.)

Fig. 2. Increasing published experience with noninvasive ventila-tion in patients with acute respiratory failure, 1989–1994, as com-piled by Sassoon. The majority consisted of patients with chronicobstructive pulmonary disease and patients ventilated postoper-atively. Numbers above bars indicate total number of patients whoreceived noninvasive ventilation in each year. (From Reference 60,with permission.)



Epidemiology of Noninvasive Ventilationin the Acute-Care Setting

Noninvasive Ventilation Outside the Setting ofClinical Trials: Efficacy Versus Effectiveness

As indicated in Table 1, evidence supporting the use ofNIV, particularly in some settings, is now plentiful andcompelling. However, such evidence has been gained pri-marily in the context of clinical research rather than fromeveryday clinical practice. Both anecdotal observation ofNIV use and a large body of literature on other health-careinterventions suggest that both utilization and outcomesmay be very different in these 2 settings. A main reason isthe distinction between efficacy, which is what is demon-strated under the structured conditions of a clinical study,and clinical effectiveness, which is what happens in ordi-nary, everyday practice (Table 3).

Regardless of the evidence supporting it in the researchsetting, for any new procedure or treatment approach to besuccessfully implemented in an institution, a number ofconditions must be met and important barriers overcome.

This has been amply demonstrated with weaning proto-cols,61,62 other aspects of ventilator management,63 andother respiratory care64 and acute-care interventions.65,66

As emphasized by several of the other presentations at thisJournal Conference, NIV is as much an art as a science,with a substantial learning curve and important prerequi-sites for successful implementation at a particular hospital.

Evidence for the current use of NIV outside the settingof clinical research comes from 2 kinds of studies: sur-veys, in which institutions or individual practitioners whocare for patients who are potential candidates for NIV arequeried about their use of it; and observational studies thatdocument actual utilization in specific clinical settings.Published evidence available at the time of writing foreach of these contexts is summarized below.

Data From Surveys: What Clinicians Say They Do

Seven studies have characterized the use of NIV in theacute-care setting, as determined by survey data.23,67-72

Table 4 summarizes those studies’ participants, clinicalcontexts, patient populations, and main findings, in theorder in which they were carried out, in the decade be-tween 1997 and 2006. Three of these surveys69,70,72 soughtinformation on all NIV use in acute-care settings, whereas3 others67,68,71 dealt only with the management of COPDexacerbations, and one23 was restricted to do-not-intubatepatients. Five67-71 sought information on institutional avail-ability and use of NIV, two70,72 queried individual physi-cians about their personal practices and attitudes, and one23

included both physicians and respiratory therapists. One ofthe studies72 surveyed individual physician attitudes andexperience rather than the practice of the institutions withwhich they were affiliated, whereas another study68 dealtonly with the use of NIV in the emergency department.

Figure 4, from the study by Devlin et al,72 shows thefrequency of NIV use in different types of ARF, as re-ported by 623 North American and European critical-carephysicians. The respondents indicated that they used NIVmost frequently in patients with obesity hypoventilationsyndrome, COPD exacerbations, and cardiogenic pulmo-

Fig. 3. Increase in the number of articles on noninvasive ventilation(shaded bars) and in the use of NIV in acute respiratory failure(white bars) since 1983, based on citations retrieved via PubMedas of December 2007. (From Reference 10, with permission.)

Table 3. Important Distinctions Between Efficacy (as Demonstrated in Clinical Trials) and Clinical Effectiveness (as Experienced in EverydayPractice)

Efficacy Effectiveness

Results under research conditions Results obtained in real-world, everyday clinical practicePatients carefully selected Unselected patientsNo comorbidities or other interfering problems Many patients have other medical conditions and other problems

that complicate managementRigidly controlled protocol for management and monitoring Techniques and protocol may or may not match what was done in

the clinical trialOverseen by investigators and dedicated research staff No special oversight of the intervention



Tab

le4.

Rep

orte

dR

esul

tsof

Surv

eys

Abo

utth

eU

seof

Non

inva

sive

Ven

tilat

ion

inth

eA

cute

-Car

eSe

tting

Stud

yW

hen

Perf

orm

edSe

tting

and

Res

pons

eR

ate

Dia

gnos

esIn

clud

edPr

inci

pal

Find

ings

Com

men

ts

Doh

erty

6719

9726

8B

ritis

hho

spita

lsw

ithre

spir

ator

yco

nsul

tant

sav

aila

ble;

in-p

atie

ntun

itson

ly.

98%

resp

onde

d

CO

PDex

acer

batio

nN

IVav

aila

ble

inon

ly48

%of

hosp

itals

.W

here

avai

labl

e,us

edin

�10

patie

nts/

yin

42%

ofho

spita

ls,

in�

60pa

tient

s/y

inon

ly7%

.

Mar

ked

regi

onal

diff

eren

ces

inN

IVav

aila

bilit

yan

dus

e.R

espo

ndin

gho

spita

lsid

entif

ied

lack

oftr

aini

ng(o

fph

ysic

ians

in53

%,

ofot

her

staf

fin

63%

)an

dfi

nanc

ial

limita

tions

ofac

quir

ing

equi

pmen

tin

63%

,as

barr

iers

toN

IVim

plem

enta

tion.

Mos

tho

spita

lsre

port

edpl

ans

toof

fer

NIV

with

in2

y.

Van

pee6

820

0114

5E

Ds

inB

elgi

um.

68%

resp

onde

dC

OPD

exac

erba

tion

NIV

avai

labl

ein

only

49%

ofde

part

men

ts(6

7%of

univ

ersi

tyho

spita

ls,

45%

ofge

nera

lho

spita

ls).

Use

din

�10

patie

nts/

yin

37%

;in

�50

patie

nts/

yin

45%

ofho

spita

ls.

Am

ong

resp

ondi

ngE

Ds,

72%

also

repo

rted

usin

gN

IVin

card

ioge

nic

pulm

onar

yed

ema,

and

45%

inpn

eum

onia

.C

ited

reas

ons

for

non-

use

ofN

IVw

ere

lack

ofeq

uipm

ent

(cos

t)in

71%

,la

ckof

clin

icia

nex

peri

ence

in33

%,

and

too

time-

cons

umin

gfo

rph

ysic

ians

and

nurs

esin

22%

.

Mah

eshw

ari6

920

02-2

003

RT

dire

ctor

sof

all

81ac

ute-

care

hosp

itals

inst

ates

ofM

assa

chus

etts

and

Rho

deIs

land

.88

%re

spon

ded

All

acut

eap

plic

atio

nsN

IVav

aila

ble

in98

%of

hosp

itals

.A

mon

gve

ntila

ted

patie

nts,

NIV

was

initi

ally

used

in20

%.

56%

ofre

spon

dent

hosp

itals

had

prot

ocol

sfo

rN

IVus

e.

Mar

ked

vari

abili

tyam

ong

resp

ondi

ngho

spita

lsin

NIV

use.

Lar

ger

hosp

itals

and

teac

hing

hosp

itals

used

NIV

mor

eof

ten.

NIV

initi

ated

inIC

Uin

55%

ofca

ses,

inE

Din

26%

,an

don

gene

ral

war

din

18%

.M

ost

com

mon

lyci

ted

barr

ier

toN

IVus

ew

asph

ysic

ian

lack

ofkn

owle

dge,

follo

wed

byin

adeq

uate

equi

pmen

t,la

ckof

prev

ious

expe

rien

ce,

and

inad

equa

teR

Ttr

aini

ng.

Bur

ns70

2003

808

atte

ndin

gan

dre

side

ntph

ysic

ians

at15

teac

hing

hosp

itals

inO

ntar

io.

48%

resp

onde

d

AR

F(a

llca

uses

)63

%of

phys

icia

nsre

port

edus

ing

NIV

inA

RF.

12of

15ho

spita

lsha

dpr

otoc

ols,

guid

elin

es,

orot

her

NIV

polic

ies.

Gre

ater

use

ofN

IVam

ong

resp

irat

ory

and

criti

cal

care

phys

icia

ns,

bym

ore

rece

ntly

trai

ned

phys

icia

ns,

and

inho

spita

lsw

ithm

ore

vent

ilato

rsfo

rN

IV.

NIV

was

used

inal

lho

spita

lar

eas:

mos

tof

ten

inIC

Uan

dE

D.

CO

PDan

dC

HF

wer

em

ost

com

mon

diag

nose

sfo

rN

IVus

e..

Sinu

ff23

2003

-200

5In

tens

ivis

ts,

pulm

onol

ogis

ts,

and

RT

sat

18C

anad

ian

and

2U

nite

dSt

ates

hosp

itals

.57

%of

phys

icia

nsan

d61

%of

RT

sre

spon

ded

DN

Ian

dC

MO

patie

nts

only

57%

ofph

ysic

ians

used

NIV

atle

ast

som

etim

esin

DN

Ipa

tient

sw

ithA

RF.

NIV

disc

usse

dat

leas

tso

met

imes

with

DN

I(6

2%)

and

CM

O(4

9%)

patie

nts.

Pulm

onol

ogis

tsm

ore

likel

yth

anin

tens

ivis

tsto

use

NIV

inD

NI

patie

nts.

Phys

icia

nsw

ere

mor

elik

ely

than

RT

sto

belie

veth

atN

IVre

lieve

sdy

spne

aan

dfa

cilit

ates

com

mun

icat

ion

inD

NI

and

CM

Opa

tient

s.

Dru

mm

ond7

120

04A

ll33

Can

adia

nho

spita

lsw

ith�

200

beds

and

with

pulm

onar

ytr

aini

ngpr

ogra

ms.

100%

resp

onde

d

CO

PDex

acer

batio

nN

IVav

aila

ble

atal

lin

stitu

tions

but

stan

dard

-of-

care

inon

lyha

lf.

70%

used

NIV

only

inE

Dor

ICU

setti

ngs.

18%

used

NIV

onge

nera

lw

ards

.U

sed

in�

3pa

tient

s/m

oin

24%

ofho

spita

ls,

and

�5

patie

nts/

mo

in52

%of

hosp

itals

.

Mar

ked

regi

onal

vari

abili

tyin

NIV

use.

Use

d“r

outin

ely”

in61

%(r

ange

byre

gion

40–

100%

).In

DN

Ipa

tient

s,N

IVw

asra

rely

orne

ver

offe

red

in32

%of

hosp

itals

.

Dev

lin72

2006

Cro

ss-s

ectio

nal

Web

-bas

edsu

rvey

of2,

985

inte

nsiv

ist

phys

icia

nsin

AC

CP

and

ER

Sre

gist

ries

.27

%re

spon

ded:

41%

inE

urop

e,an

d19

%In

Nor

thA

mer

ica

All

acut

eap

plic

atio

ns44

%re

port

edus

ing

NIV

�25

%of

the

time

inpa

tient

sad

mitt

edw

ithA

RF.

Eur

opea

nsm

ore

likel

yth

anN

orth

Am

eric

anph

ysic

ians

tous

eN

IVin

�25

%of

AR

Fpa

tient

s(6

8%vs

37%

,P

�.0

1).

Mar

ked

regi

onal

vari

atio

nin

stat

edus

eof

NIV

.N

IVm

ost

likel

yto

beus

edin

CO

PDex

acer

batio

n,C

HF,

and

obes

ity-h

ypov

entil

atio

nsy

ndro

me.

Nor

thA

mer

ican

phys

icia

nsus

edse

dativ

es(4

1%vs

24%

),an

alge

sics

(48%

vs35

%),

and

hand

rest

rain

ts(2

7%vs

16%

)m

ore

ofte

nth

anE

urop

ean

phys

icia

ns(P

�.0

1fo

rea

chco

mpa

riso

n).

CO

PD�

chro

nic

obst

ruct

ive

pulm

onar

ydi

seas

eN

IV�

noni

nvas

ive

vent

ilatio

nE

D�

emer

genc

yde

part

men

tR

T�

resp

irat

ory

ther

apis

tIC

U�

inte

nsiv

eca

reun

itA

RF

�ac

ute

resp

irat

ory

failu

re

CH

F�

cong

estiv

ehe

art

failu

reD

NI

�do

not

intu

bate

CM

O�

com

fort

mea

sure

son

lyA

CC

P�

Am

eric

anC

olle

geof

Che

stPh

ysic

ians

ER

S�

Eur

opea

nR

espi

rato

rySo

ciet

y



nary edema, and least often in failed extubation and pa-tients who do not wish to be intubated.

As valuable as surveys can be for indicating awareness,access, and attitudes about NIV in the different contexts inwhich it is carried out, a number of shortcomings of suchstudies should be mentioned.73 The reported results aretaken only from the surveys that were returned (which inthe studies summarized in Table 4 ranged from 100%down to 27% of those sought), and may not reflect whatthe non-responders know, think, or do. Because they re-port data from individual practitioners and institutions,such surveys may or may not be relevant to other clini-cians, in different practice contexts, for different types orsizes of institutions, or in other geographic or culturalareas. And, importantly, these studies can only tell us whatthe institutions and individuals surveyed say they do, notwhat they actually do. Reported and actual policies andpractices may be quite different, as has been documentedfor ventilator charting and other respiratory care practices.74

Data From Observational Studies of Actual NIV Use

Observational cohort studies of the use of NIV in theacute-care setting get around at least some of the problems

inherent to surveys. They document actual practice in theinstitutions in which they are performed, at least at thetime of the study, for the patients included in the cohort,and in the clinical setting evaluated. Seven such studieshave been published as full peer-reviewed articles,75-81 andan eighth was recently reported in abstract form82 (Ta-ble 5).

The reported studies differ considerably in design andsample size. Three75,78,79 report single-center cohorts,whereas the rest are multicenter studies, including datafrom 42 ICUs76 to as many as 361 separate ICUs.77 Twoof them80,81 are follow-up studies in which current (or atleast more recent) NIV use is compared to the results ofprevious cohorts76,77 from the same groups of investiga-tors. In 7 of the 8 studies summarized in Table 5 theauthors included all acute-care use of NIV in adult pa-tients, and reported usage rates and outcomes among pa-tients with COPD, congestive heart failure, and hypox-emic ARF.

In a study aimed at detecting temporal trends in ICU-related pneumonia and other hospital-acquired infections,Girou and associates75 tracked NIV use in the managementof COPD exacerbations and cardiac pulmonary edema intheir 26-bed medical ICU from 1994 through 2001. As

Fig. 4. Reported frequency of noninvasive ventilation (NIV) use in different clinical settings among 790 intensivist physicians from theAmerican College of Chest Physicians’ Critical Care Network and the European Respiratory Society’s Assembly of Critical Care whoresponded to a Web-based survey on sedation practices during NIV for acute respiratory failure. COPD � chronic obstructive pulmonarydisease. pts � patients. (From Reference 72, with permission.)



Tab

le5.

Rep

orte

dR

esul

tsof

Obs

erva

tiona

lSt

udie

sof

the

Use

ofN

onin

vasi

veV

entil

atio

nin

Acu

teC

are

Out

side

the

Setti

ngof

Clin

ical

Tri

als

Stud

yW

hen

Perf

orm

edSe

tting

and

Stud

yD

esig

nD

iagn

oses

Prin

cipa

lFi

ndin

gsC

omm

ents

Gir

ou75

1994

–200

1M

edic

alIC

Uof

aFr

ench

univ

ersi

tyho

spita

l.R

etro

spec

tive

exam

inat

ion

ofpr

ospe

ctiv

ely

colle

cted

data

onm

echa

nica

llyve

ntila

ted

patie

nts.

CO

PDex

acer

batio

nan

dC

HF

Am

ong

479

patie

nts

vent

ilate

ddu

ring

the

6-y

peri

od,

313

(65%

)re

ceiv

edN

IV,

ofw

hom

35(1

1%)

subs

eque

ntly

requ

ired

intu

batio

n.N

IVus

epr

ogre

ssiv

ely

incr

ease

das

apr

opor

tion

ofal

lpa

tient

sve

ntila

ted

duri

ngth

est

udy

peri

od.

ICU

mor

talit

yde

crea

sed

from

21%

in19

94to

7%in

2001

.T

hera

teof

ICU

-acq

uire

dpn

eum

onia

decr

ease

dpr

ogre

ssiv

ely

from

20%

to8%

over

the

sam

epe

riod

.Pa

tient

str

eate

dw

ithN

IVha

dsh

orte

rIC

Ust

ayth

anth

ose

who

rece

ived

inva

sive

vent

ilatio

n(m

ean

10d

vs8

d,P

�.0

2).

Car

lucc

i76

1997

42IC

Us

inFr

ance

,Sw

itzer

land

,B

elgi

um,

Spai

n,an

dT

unis

ia.

Pros

pect

ive

stud

yw

ith3-

wk

obse

rvat

ion

peri

od.

NIV

used

asin

itial

vent

ilatio

nap

proa

chin

108

(16%

)of

689

patie

nts.

Hyp

oxem

icA

RF

(48%

),hy

perc

apni

cA

RF

(15%

),co

ma

(30%

),C

HF

(7%

)N

IVsu

cces

sful

(no

need

for

intu

batio

n)in

65(6

0%)

of10

8pa

tient

s.N

IVus

edin

50%

ofpa

tient

sw

ithhy

perc

apni

cA

RF,

27%

with

CH

F,14

%w

ithhy

poxe

mic

AR

F,an

d0%

with

com

a.

Mea

ndu

ratio

nof

NIV

was

5.6

din

hype

rcap

nic

AR

F,2.

4d

inC

HF,

and

6.3

din

hypo

xem

icA

RF.

Mea

nho

urs

ofN

IVpe

rda

yw

as�

9h

atal

ltim

esin

all

grou

ps.

Est

eban

7719

9836

1IC

Us

in20

coun

trie

s.Pr

ospe

ctiv

eco

hort

stud

yof

adul

tpa

tient

sve

ntila

ted

for

�12

hdu

ring

a28

-dpe

riod

.N

IVus

edas

initi

alve

ntila

tion

appr

oach

in25

6(4

.9%

)of

5,18

3pa

tient

s.

Hyp

oxem

icA

RF

(69%

,in

clud

ing

CH

F10

%),

com

a(1

7%),

CO

PDex

acer

batio

n(1

0%)

85C

OPD

patie

nts

rece

ived

NIV

,of

who

m22

(26%

)w

ere

subs

eque

ntly

intu

bate

d.54

(36%

)of

148

patie

nts

with

hypo

xem

icA

RF

who

rece

ived

NIV

wer

esu

bseq

uent

lyin

tuba

ted.

Mor

talit

yin

CO

PDpa

tient

sw

as14

%w

hen

NIV

was

succ

essf

ulan

d42

%w

hen

intu

batio

nw

assu

bseq

uent

lyre

quir

ed.

Inot

her

patie

nts

trea

ted

initi

ally

with

NIV

,m

orta

lity

was

high

erif

intu

batio

nw

asre

quir

ed,

com

pare

dto

patie

nts

initi

ally

intu

bate

d(4

8%vs

31%

).

Paus

-Jen

ssen

7820

01Pr

ospe

ctiv

eco

hort

stud

yof

all

NIV

use

ina

Can

adia

nte

achi

ngho

spita

lov

era

5-m

ope

riod

.75

patie

nts

wer

ein

clud

ed:

64N

IV,

11C

PAP

only

.

Shor

tnes

sof

brea

th(2

4%),

CO

PDex

acer

batio

n(1

7%),

hypo

xem

icA

RF

(17%

),C

HF

(13%

),ot

her

(29%

)N

IVin

itiat

edin

ED

(32%

),IC

U(2

7%),

war

dob

serv

atio

nun

it(2

3%),

orge

nera

lm

edic

alw

ard

(18%

).13

%of

patie

nts

requ

ired

intu

batio

nan

d24

%di

ed(1

6%w

ithD

NA

Rst

atus

).

Stud

yho

spita

lha

dno

NIV

prot

ocol

orpo

licy

othe

rth

anre

quir

ing

aph

ysic

ian’

sor

der.

Stud

yda

taw

ere

reco

rded

byth

eR

Ts

who

prov

ided

the

care

.

Sche

ttino

7920

01Pr

ospe

ctiv

eco

hort

stud

yin

teac

hing

hosp

ital

ofN

IVus

edu

ring

1-y

peri

odA

llad

ult

patie

nts

who

rece

ived

NIV

orC

PAP

for

anac

ute

indi

catio

nan

ywhe

rein

the

hosp

ital.

DN

Ipa

tient

sex

clud

ed.

Hyp

oxem

icA

RF

(60%

),po

st-e

xtub

atio

nA

RF

(40%

),hy

perc

apni

cA

RF

non-

CO

PD(3

8%),

CO

PDex

acer

batio

n(2

4%),

CH

F(1

8%)

458

epis

odes

in44

9pa

tient

s.N

IVin

itiat

edin

ICU

in47

%,

onge

nera

lm

edic

alw

ard

in33

%,

inE

Din

20%

.O

vera

llm

orta

lity

21%

(47%

whe

nN

IVw

asun

succ

essf

ulan

din

tuba

tion

was

carr

ied

out)

.49

%of

NIV

patie

nts

man

aged

inan

ICU

wer

esu

bseq

uent

lyin

tuba

ted,

vs27

%on

the

gene

ral

war

d.

53%

ofpa

tient

sw

ere

man

aged

inan

ICU

afte

rN

IVin

itiat

ion,

35%

inge

nera

lm

edic

al-s

urgi

cal

war

ds,

and

12%

excl

usiv

ely

inth

eE

D.

RT

/pat

ient

ratio

1.6:

1.8

and

nurs

e/pa

tient

ratio

1.4:

1.6

outs

ide

the

ICU

setti

ng.

Of

the

97pa

tient

sw

ithC

HF,

60%

wer

em

anag

edw

ithC

PAP

alon

e,an

don

ly18

%of

them

requ

ired

intu

batio

n;m

ost

ofth

emw

ere

man

aged

inth

eE

Dw

ithou

tad

mis

sion

toth

eIC

U.

Dem

oule

8020

02Fo

llow

-up

stud

y5

yaf

ter

1997

coho

rt.7

6

Obs

erva

tiona

lco

hort

stud

yof

all

vent

ilate

dpa

tient

sin

70Fr

ench

ICU

s(3

2un

iver

sity

,38

non-

univ

ersi

ty;

28w

ere

also

in19

97st

udy)

over

a3-

wk

peri

od.

All

adul

tpa

tient

sve

ntila

ted

inIC

U.

Hyp

oxem

icA

RF

(42%

),ac

ute-

on-c

hron

icre

spir

ator

yfa

ilure

(16%

:C

OPD

11%

,re

stri

ctiv

edi

seas

e5%

),co

ma

(34%

),C

HF

(8%

).

Of

1,07

6pa

tient

sve

ntila

ted,

249

(23%

)re

ceiv

edN

IVas

initi

alsu

ppor

t,co

mpa

red

to16

%in

1997

coho

rt.

Inpa

tient

sno

tin

tuba

ted

prio

rto

adm

issi

on,

NIV

was

used

in52

%,

vs35

%in

1997

.38

%of

NIV

patie

nts

subs

eque

ntly

requ

ired

intu

batio

n.

The

prop

ortio

nof

all

patie

nts

who

rece

ived

vent

ilato

rsu

ppor

tw

hoha

dN

IVsu

cces

sful

lyap

plie

dw

ithou

tth

ene

edfo

rin

tuba

tion

incr

ease

dfr

om9%

to13

%,

com

pare

dto

the

1997

coho

rt.

Est

eban

8120

04Fo

llow

-up

stud

yof

1998

coho

rt.7

7

1-m

oob

serv

atio

nal

coho

rtst

udy

ofal

lpa

tient

sve

ntila

ted

in34

9IC

Us

in23

coun

trie

s.

All

adul

tpa

tient

sve

ntila

ted

inIC

Ufo

r�

12h.

Hyp

oxem

icA

RF

(72%

,in

clud

ing

CH

F6%

),co

ma

(19%

),C

OPD

exac

erba

tion

(5%

).

4,96

8pa

tient

sw

ere

incl

uded

,of

who

m1,

675

wer

em

anag

edin

107

ICU

sth

atal

sopa

rtic

ipat

edin

the

1998

stud

y.11

%of

vent

ilate

dpa

tient

sre

ceiv

edN

IV,

vs4%

in19

98.

NIV

was

used

in48

of10

9pa

tient

sw

ithC

OPD

and

in10

9of

1,08

3pa

tient

sw

ithpr

imar

yA

RF:

each

prop

ortio

nsi

gnif

ican

tlym

ore

than

inth

e19

98co

hort

.

Nei

ther

the

requ

irem

ent

for

intu

batio

n(3

5%)

nor

mor

talit

y(2

4%)

was

diff

eren

tin

the

2004

coho

rt,

com

pare

dto

the

patie

nts

stud

ied

in19

98.

Orz

sanc

ak82

2007

Surv

eyof

NIV

prac

tice

in8

“low

-util

izat

ion”

hosp

itals

inM

assa

chus

etts

and

Rho

deIs

land

,as

iden

tifie

din

prev

ious

surv

ey.6

9

Pros

pect

ive

1-m

oco

hort

stud

y.

All

acut

eap

plic

atio

nsof

NIV

inad

ults

.C

OPD

(25%

),C

HF

(26%

),pn

eum

onia

(18%

),ot

her

(31%

).24

4(4

2%)

of58

1pa

tient

sbe

gun

onm

echa

nica

lve

ntila

tion

rece

ived

NIV

.E

xclu

ding

patie

nts

who

wer

ein

tuba

ted

for

airw

aypr

otec

tion,

NIV

was

initi

alap

proa

chin

81%

ofC

OPD

patie

nts,

73%

with

CH

F,49

%w

ithpn

eum

onia

,an

d66

%in

othe

rca

uses

ofA

RF.

Ove

rall

succ

ess

rate

with

NIV

was

71%

,an

dm

orta

lity

was

16%

.

Rep

orte

don

lyin

abst

ract

form

asof

the

time

ofth

isw

ritin

g.

ICU

�in

tens

ive

care

unit

CO

PD�

chro

nic

obst

ruct

ive

pulm

onar

ydi

seas

eC

HF

�co

nges

tive

hear

tfa

ilure

NIV

�no

ninv

asiv

eve

ntila

tion

AR

F�

acut

ere

spir

ator

yfa

ilure

DN

I�

dono

tin

tuba

teE

D�

emer

genc

yde

part

men

tD

NA

R�

dono

tat

tem

ptre

susc

itatio

nR

T�

resp

irat

ory

ther

apis

tC

PAP

�co

ntin

uous

posi

tive

airw

aypr

essu

re



shown in Figure 5, NIV use, as a proportion of all patientswith these diagnoses who received mechanical ventilationin the unit, increased steadily throughout that period.

Schettino et al79 prospectively documented non-inves-tigational NIV use during the year 2001 in a large teachinghospital with extensive experience with this therapy. Theyexcluded do-not-intubate patients but included all otherapplications of NIV in patients � 18 y old, in all areas ofthe hospital. Figure 6 shows the outcomes of the patientsin 5 different diagnostic categories. The rates of NIV fail-ure (that is, the need to intubate and invasively ventilate)and fatal outcome among patients who failed NIV differedconsiderably in the patient categories.

In a study not summarized in Table 5, Bruge et al18

recently reported the results of a 2-year prospective ob-servational investigation of NIV in pre-hospital care, inemergency-response vehicles equipped with bi-level ven-

tilators for NIV, operated by their institution in France.During the observation period, out-of-hospital NIV wasattempted in 138 patients with congestive heart failure(56%), COPD exacerbation (28%), or primary ARF (16%).NIV was deemed successful (ie, intubation was not re-quired either in the field or in the emergency department)in 102 patients (74%). Patients with congestive heart fail-ure were more likely to be managed successfully withNIV, and major air leaks that signified inability to achievea satisfactory mask seal predicted subsequent need forintubation. As of the time of this writing, no other reportsof NIV use in pre-hospital emergency care have been pub-lished.

Problems With the AccurateAssessment of Current NIV Use

Although they are designated as evaluations of practice“outside the research setting,” the studies summarized hereall involved the collection of data in “real time” in allinstances of NIV use. Thus, the practices documented were,to a degree, observed in a research setting. However, ex-cept for investigations that involved large administrativedatabases, which are necessarily limited in what they canreveal about institutional practice and clinician behavior,this approach to studying current NIV use is probably theonly practical way to address the issue.

The epidemiology of NIV use in the acute-care settingis, however, a moving target. Although, as in other areasof medicine, practice appears to have lagged behind theevidence base by several years, it is apparent that NIV isbeing used by more and more clinicians and is now avail-able in most if not all acute-care institutions. Assessmentof historical trends and current use has been complicatedby different definitions of NIV (eg, the inclusion of CPAPin some studies), the variety of locations in which NIV hasbeen used (eg, ICU vs general ward vs emergency depart-ment), the sometimes vague criteria for patient inclusion,the definitions of NIV success and failure, and the uncleardenominators from which included patients were drawn insome series.

Summary

Since the 1940s, NIV has evolved in parallel with in-vasive mechanical ventilation in the care of patients withARF. With the explosion of reported studies on NIV use indifferent patient populations and clinical contexts in thelast 20 years has come a steady (if belated) increase in theuse of this therapy in everyday practice. Although theyprobably do not reflect the dimensions of practice in 2009very accurately, numerous studies of reported and actualNIV use in acute-care settings show that this modality is

Fig. 5. Increasing use of noninvasive ventilation, as a proportion ofall uses of mechanical ventilation, in the management of 479 pa-tients with exacerbations of chronic obstructive pulmonary dis-ease or acute cardiogenic pulmonary edema during a 6-year pe-riod in the 26-bed intensive care unit of a French university hospital.The vertical lines represent the 95% confidence limits. (From Ref-erence 75, with permission.)

Fig. 6. Proportions of 449 patients, in a cohort of acute-care pa-tients who received noninvasive ventilation in a major teachinghospital, who required intubation (black bars) and, once intubated,the proportion who died (white bars), in different diagnostic groups.COPD � chronic obstructive pulmonary disease. (From Refer-ence 79, with permission.)



now widely available and routinely used by increasingnumbers of clinicians.

REFERENCES

1. MacIntyre N, Huang YC. Acute exacerbations and respiratory failurein chronic obstructive pulmonary disease. Proc Am Thorac Soc 2008;5(4):530-535.

2. Quon BS, Gan WQ, Sin DD. Contemporary management of acuteexacerbations of COPD: a systematic review and meta analysis.Chest 2008;133(3):756-766.

3. Ward NS, Dushay KM. Clinical concise review: mechanical venti-lation of patients with chronic obstructive pulmonary disease. CritCare Med 2008;36(5):1614-1619.

4. Mehta S, Al-Hashim AH, Keenan S. Noninvasive ventilation in pa-tients with acute cardiogenic pulmonary edema. Respir Care 2009;54(2): in press.

5. 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 200810;359(2):142-151.

6. Masip J. Noninvasive ventilation in acute cardiogenic pulmonaryedema. Curr Opin Crit Care 2008;14(5):531-535.

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

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

9. Hess DR, Fessler HR. Should NPPV be used in all forms of respi-ratory failure? Respir Care 2007;52(5):568-578.

10. Ambrosino N, Vagheggini G. Noninvasive positive pressure venti-lation in the acute care setting: where are we? Eur Respir J 2008;31(4):874-886.

11. Kallet RH. Noninvasive ventilation in acute care: controversies andemerging concepts. Respir Care 2009;54(2):(in press)

12. Deis JN, Abramo TJ, Crawley L. Noninvasive respiratory support.Pediatr Emerg Care 2008;24(5):331-338.

13. Levine DA. Novel therapies for children with severe asthma. CurrOpin Pediatr 2008;20(3):261-265.

14. Askin DF. Noninvasive ventilation in the neonate. J Perinat NeonatalNurs 2007;21(4):349-358.

15. Mehta S. Neuromuscular disease causing acute respiratory failure.Respir Care 2006;51(9):1016-1023.

16. Hess DR. Noninvasive ventilation in neuromuscular disease: equip-ment and application. Respir Care 2006;51(8):896-912.

17. Seneviratne J, Mandrekar J, Wijdicks EF, Rabinstein AA. Noninva-sive ventilation in myasthenic crisis. Arch Neurol 2008;65(1):54-58.

18. Bruge P, Jabre P, Dru M, Jbeili C, Lecarpentier E, Khalid M, et al.An observational study of noninvasive positive pressure ventilationin an out-of-hospital setting. Am J Emerg Med 2008;26(2):165-169.

19. Taylor DM, Bernard SA, Masci K, MacBean CE, Kennedy MP.Prehospital noninvasive ventilation: a viable treatment option in theurban setting. Prehosp Emerg Care 2008;12(1):42-45.

20. Weitz G, Struck J, Zonak A, Balnus S, Perras B, Dodt C. Prehospitalnoninvasive pressure support ventilation for acute cardiogenic pul-monary edema. Eur J Emerg Med 2007;14(5):276-279.

21. Orlikowski D, Prigent H, Gonzales-Bermejo J, Aubert P, Lofaso F,Raphael JC, Clair B. Noninvasive ventilation as an alternative toendotracheal intubation duringtracheotomy in advanced neuromus-cular disease. Respir Care 200752(12):1728-1733.

22. Cuomo A, Delmastro M, Ceriana P, Nava S, Conti G, Antonelli M,Iacobone E. Noninvasive mechanical ventilation as a palliative treat-ment of acute respiratory failure in patients with end-stage solidcancer. Palliat Med 2004;18(7):602-610.

23. Sinuff T, Cook DJ, Keenan SP, Burns KEA, Adhikari NKJ, RockerGM, et al. Noninvasive ventilation for acute respiratory failure nearthe end of life. Crit Care Med 2008;36(3):789-794.

24. Curtis JR, Cook DJ, Sinuff T, White DB, Hill N, Keenan SP, et al.Noninvasive positive pressure ventilation in critical and palliativecare settings: understanding the goals of therapy. Crit Care Med2007;35(3):932-939.

25. Drinker P, Shaw LA. An apparatus for the prolonged administrationof artificial respiration: I. A design for adults and children. J ClinInvest 1929;7(2):229-247.

26. Comroe JH Jr. Retrospectoscope: man-cans. Am Rev Respir Dis1977;116(5):945-950.

27. Drinker PA, McKhann CF III. The iron lung: first practical means ofrespiratory support. JAMA 1986;55:1476-1480.

28. Hill NS. Clinical application of body ventilators. Chest 1986;90(6):897-905.

29. Lassen HCA. Management of life-threatening poliomyelitis. Edin-burgh: E&S Livingstone; 1956.

30. Baker AB. Artificial respiration, the history of an idea. Med Hist1971;15(4):336-351.

31. Morch ET. History of mechanical ventilation. In: Kirby RR, SmithRA, Downs JB. Clinical applications of ventilatory support. NewYork: Churchill-Livingstone; 1990:1-61.

32. Colice GL. Historical perspective on the development of mechanicalventilation. In: Tobin MJ, editor. Principles and practice of mechan-ical ventilation. New York: McGraw-Hill; 1994:1-35.

33. Beeson P. Changing times: reflections on a professional lifetime: aninterview with Paul Beeson. Interview by Richard V Lee. Ann InternMed 2000;132(1):71-79.

34. Zwillich CW, Pierson DJ, Creagh CE, Sutton FD Jr, Schatz E, PettyTL. Complications of assisted ventilation: a prospective study of 354consecutive episodes. Am J Med 1974;57(2):161-70.

35. Pierson DJ. Complications associated with mechanical ventilation.Crit Care Clin 1990;6:711-24.

36. Stauffer J, Silvestri RC. Complications of endotracheal intubation,tracheostomy, and artificial airways. Respir Care 1982;27(4):417-434.

37. Slutsky AS. Ventilator-induced lung injury: from barotrauma to bio-trauma. Respir Care 2005;50(5):646-659.

38. Tremblay LN, Slutsky AS. Ventilator-induced lung injury: from thebench to the bedside. Intensive Care Med 2006;32(1):24-33.

39. Motley HL, Werko L, Cournand A, Richards DW. Observations onthe clinical use of intermittent positive pressure. J Aviation Med1947;18(5):417-435.

40. Murray JF. Review of the state of the art in intermittent positivepressure breathing therapy. Am Rev Respir Dis 1974;110(6 Pt 2):193-199.

41. Baker JP. Magnitude of usage of intermittent positive pressure breath-ing. Am Rev Respir Dis 1974;110(6 Pt 2):170-177.

42. McConnell DH, Maloney JV Jr, Buckberg GD. Postoperative inter-mittent positive-pressure breathing treatments. Physiological consid-erations. J Thorac Cardiovasc Surg 1974;68(6):944-952.

43. Conference on the scientific basis of respiratory therapy. Am RevRespir Dis 1974;110(6 Pt 2):1-204.

44. Bach JR, Alba AS, Saporito LR. Intermittent positive pressure ven-tilation via the mouth as an alternative to tracheostomy for 257 users.Chest 1993;103(1):174-182.

45. Pierson DJ. Noninvasive positive pressure ventilation: history andterminology. Respir Care 1997;42(3):370-379.

46. Sullivan CE, Berthon-Jones M, Issa FG, Eves L. Reversal of ob-structive sleep apnea by continuous positive airway pressure appliedthrough the nose. Lancet 1981;1(8225):862-865.

47. Ellis ER, Bye PTP, Bruderer JW, Sullivan CE. Treatment of respi-ratory failure during sleep in patients with neuromuscular disease:



positive-pressure ventilation through a nose mask. Am Rev RespirDis 1987;135(1):148-152.

48. Kerby GR, Mayer LS, Pingleton SK. Nocturnal positive-pressureventilation via nasal mask. Am Rev Respir Dis 1987;135(3):738-740.

49. Bach JR, Alba AS, Mosher R, Delaubier A. Intermittent positivepressure ventilation via nasal access in the management of respira-tory insufficiency. Chest 1987;92(1):168-170.

50. Ellis ER, Grunstein RR, Chan S, Bye PT, Sullivan CE. Noninvasiveventilatory support during sleep improves respiratory failure in ky-phoscoliosis. Chest 1988;94(4):811-815.

51. Carroll N, Branthwaite MA. Control of nocturnal hypoventilation bynasal intermittent positive pressure ventilation. Thorax 1988;43(5):49-3453.

52. Leger P, Jennequin J, Gerard M, Robert D. Home positive pressureventilation via nasal mask for patients with neuromuscular weaknessor restrictive lung or chest-wall disease. Respir Care 1989;34(2):73-77.

53. Meduri GU, Conoscenti CC, Menashe P, Nair S. Noninvasive facemask ventilation in patients with acute respiratory failure. Chest1989;95(4):865-870.

54. Foglio C, Vitacca M, Quadri A, Scalvini S, Marangoni S, AmbrosinoN. Acute exacerbations in severe COPD patients. Treatment usingpositive pressure ventilation by nasal mask. Chest 1992;101(6):1533-8.

55. Bott J, Carroll MP, Conway JH, Keilty SE, Ward EM, Brown AM,et al. Randomised controlled trial of nasal ventilation in acute ven-tilatory failure due to chronic obstructive airways disease. Lancet1993;341(8860):1555-1557.

56. Conway JH, Hitchcock RA, Godfrey RC, Carroll MP. Nasal inter-mittent positive pressure ventilation in acute exacerbations of chronicobstructive pulmonary disease: a preliminary study. Respir Med 1993;87(5):387-394.

57. Meecham-Jones DJ, Paul EA, Grahame-Clarke C, Wedzicha JA.Nasal ventilation in acute exacerbations of chronic obstructive pul-monary disease: effect of ventilator mode on arterial blood gas ten-sions. Thorax 1994;49(12):1222-1224.

58. Brochard L, Isabey D, Piquet J, Amaro P, Mancebo J, Messadi A, etal. Reversal of acute exacerbations of chronic obstructive lung dis-ease by inspiratory assistance with a face mask. J Engl J Med 1990;323(22):1523-1530.

59. Meduri GU, Abou-Shala N, Fox RC, Jones CB, Leeper KV, Wun-derink RG. Noninvasive face mask mechanical ventilation in patientswith acute hypercapnic respiratory failure. Chest 1991;100(2):445-454.

60. Sassoon CSH. Noninvasive positive-pressure ventilation in acuterespiratory failure: review of reported experience with special atten-tion to use during weaning. Respir Care 1995;40(2):282-288.

61. Ely EW, Bennett PA, Bowton DL, Murphy SM, Florance AM, Ha-ponik EF. Large scale implementation of a respiratory therapist-driven protocol for ventilator weaning. Am J Respir Crit Care Med1999;159(2):439-446.

62. Chatburn RL, Deem S. Respiratory controversies in the critical caresetting. Should weaning protocols be used with all patients whoreceive mechanical ventilation? Respir Care 2007;52(5):609-621.

63. Rubenfeld GD. Implementing effective ventilator practice at the bed-side. Curr Opin Crit Care 2004;10(1):33-39.

64. Ford RM, Phillips-Clar JE, Burns DM. Implementing therapist-drivenprotocols. Respir Care Clin N Am 1996;2(1):51-76.

65. Kahn JM, Matthews FA, Angus DC, Barnato AE, Rubenfeld GD.Barriers to implementing the Leapfrog Group recommendations for

intensivist physician staffing: a survey of intensive care unit direc-tors. J Crit Care 2007;22(2):97-103.

66. Carlbom DJ, Rubenfeld GD. Barriers to implementing protocol-basedsepsis resuscitation in the emergency department: results of a na-tional survey. Crit Care Med 2007;35(11):2525-2532.

67. Doherty MJ, Greenstone MA. Survey of non-invasive ventilation(NIPPV) in patients with acute exacerbations of chronic obstructivepulmonary disease (COPD) in the U.K. Thorax 1998;53(10):863-866.

68. Vanpee D, Delaunois L, Lheureux P, Thys F, Sabbe M, MeulemansA, et al. Survey of non-invasive ventilation for acute exacerbation ofchronic obstructive pulmonary disease patients in emergency depart-ments in Belgium. Eur J Emerg Med 2002;9(3):217-224.

69. Maheshwari V, Paioli D, Rothaar R, Hill NS. Utilization of nonin-vasive ventilation in acute-care hospitals: a regional survey. Chest2006;129(5):1226-1233.

70. Burns KE, Sinuff T, Adhikari NK, Meade MO, Heels-Ansdell D,Martin CM, Cook DJ. Bilevel noninvasive positive pressure venti-lation for acute respiratory failure: survey of Ontario practice. CritCare Med 2005;33(7):1477-1483.

71. Drummond J, Rowe B, Cheung L, Mayers I. The use of noninvasivemechanical ventilation for the treatment of acute exacerbations ofchronic obstructive pulmonary disease in Canada. Can Respir J 2005;12(3):129-133.

72. Devlin JW, Nava S, Fong JJ, Bahhady I, Hill NS. Survey of sedationpractices during noninvasive positive-pressure ventilation to treatacute respiratory failure. Crit Care Med 2007;35(10):2298-2302.

73. Rubenfeld GD. Surveys: an introduction. Respir Care 2004;49(10):1181-1185.

74. Akhtar SR, Weaver J, Pierson DJ, Rubenfeld GD. Practice variationin respiratory therapy documentation during mechanical ventilation.Chest 2003;124(6):2275-2282.

75. Girou E, Brun-Buisson C, Taille S, Lemaire F, Brochard L. Seculartrends in nosocomial infections and mortality associated with non-invasive ventilation in patients with exacerbation of COPD and pul-monary edema. JAMA 2003;290(22):2985-2991.

76. Carlucci A, Richard JC, Wysocki M, Lepage E, Brochard L; SRLFCollaborative Group on Mechanical Ventilation. Noninvasive versusconventional mechanical ventilation: an epidemiologic survey. Am JRespir Crit Care Med 2001;163(4):874-80.

77. Esteban A, Anzueto A, Frutos F, Alía I, Brochard L, Stewart TE, etal; for the Mechanical Ventilation International Study Group. Char-acteristics and outcomes in adult patients receiving mechanical ven-tilation. JAMA 2002;287(3):345-355.

78. Paus-Jenssen ES, Ried JK, Cockroft DW, Laframboise K, Ward HA.The use of noninvasive ventilation in acute respiratory failure at atertiary care center. Chest 2004;126(1):165-172.

79. Schettino G, Altobelli N, Kacmarek RM. Noninvasive positive-pres-sure ventilation in acute respiratory failure outside clinical trials:experience at the Massachusetts General Hospital. Crit Care Med2008;36(2):441-447.

80. Demoule A, Girou E, Richard JC, Taille S, Brochard L. Increaseduse of noninvasive ventilation in French intensive care units. Inten-sive Care Med 2006;32(11):1747-1755.

81. Esteban A, Ferguson ND, Meade MO, Frutos-Vivar F, ApezteguiaC, Brochard L, et al; VENTILA Group. Evolution of mechanicalventilation in response to clinical research. Am J Respir Crit CareMed 2008;177(2):170-177.

82. Ozsancak A, Alkana P, Khodabandeh A, Maheshwari V, Hill NS.Increasing utilization of non-invasive positive pressure ventilation inacute care hospitals in Massachusetts and Rhode Island (abstract).Am J Respir Crit Care Med 2008;177(Abstracts Issue):A283.



Discussion

Hill: You alluded to the difficultyin getting accurate information on thistopic, and I entirely concur, havingdone some of the work that you pre-sented. One of the problems is thatmany of these studies, including ourown, relied on questionnaires andlacked validation. Also, very few ofthe studies defined how the types ofrespiratory failure were diagnosed.

We’ve been doing follow-up workon-site—not relying on question-naires, but documenting actual NIVutilization. And one of the challengesis defining the target population. Thereare patients who go on mechanical ven-tilation because of procedures, sur-gery, or general anesthesia, who arenot relevant to what we’re talkingabout. And there are patients who getintubated for airway protection, whoare comatose or have severe swallow-ing or secretion problems and shouldnot be on NIV in any case.

We looked at the hospitals that werelow NIV utilizers—less than 15% ofthe initial ventilator starts were withNIV in the Maheshwari et al1 surveyin early 2003, but now NIV is beingused in more than 50% of initial ven-tilator starts. I think we’re seeing moreacceptance of NIV over the last half-a-dozen years. I don’t think the stud-ies you showed are up-to-date enoughto reflect that trend. Parenthetically, Ithink the Europeans were ahead of theNorth Americans on this, but the NorthAmericans are catching up.

1. Maheshwari V, Paioli D, Rothaar R, HillNS. Utilization of noninvasive ventilationin acute care hospitals: a regional survey.Chest 2006;129(5):1226-1233.

Pierson: Your point is a good one.The goal is not necessarily to increaseNIV use in and of itself. There hasbeen a disturbing tendency in some ofthe units in which I attend that anypatient who develops an acute respi-ratory problem is immediately slappedon NIV, without as much consider-ation as perhaps there should be about

the patient’s ability to protect the air-way and clear secretions, the patient’smental status, and so forth. Our goalshouldn’t just be to have the largestpossible number of patients who haverespiratory problems getting NIV.

Sean Keenan’s group published astudy1 in RESPIRATORY CARE a few yearsago in which they looked at NIV inmild exacerbations of COPD andfound that (1) NIV didn’t seem to haveany benefit over standard therapy, and(2) they couldn’t get them to do itanyway. So I think it’s an oversimpli-fication to say that we should just sim-ply increase the number or proportionof patients with acute respiratory prob-lems who are administered NIV.

1. Keenan SP, Powers CE, McCormack DG.Noninvasive positive-pressure ventilationin patients with milder chronic obstructivepulmonary disease exacerbations: a ran-domized controlled trial. Respir Care 2005;50(5):610-616.

Epstein: You both have good points.Do we use NIV when it should beused? Do we use it when it should notbe used? And when we use it in thecorrect patient, do we use it correctly?To create a quality NIV program thoseare important questions. Do we haveany data on whether NIV is being usedin the right patients and with the rightsettings?

Kacmarek: Clearly we do not havethose data. We did not get into the“whys and wherefores” of not usingNIV; Dave simply looked at NIV uti-lization. I still run into practitionerswho say they rarely do NIV. NIV hasbecome a more refined technique inthe institutions that use it frequently.It’s a learned process that requires alot of“art” in addition to science to be suc-cessful. Unless you’re really luckywhile initiating an NIV program, thewhole process can take on a negativetone in an institution, and early fail-ures can adversely affect the futureuse of NIV at any institution.

Hill: I think I’m hearing less of that,Bob, than I did, say, half a dozen yearsago.

Kacmarek: I agree.

Hill: I think we’re moving in thedirection of using it more. My groupis trying to get at the issue of appro-priate utilization. Is it being used whenit’s supposed to be used, and not be-ing used when it shouldn’t? Ideallywe want to optimize utilization, notjust increase utilization. That’s a dif-ficult issue to get at because when it’sused is a judgment call, and unlessyou’re right there at the bedside it canbe very hard to know in “gray area”cases whether NIV is appropriate. Butat least we will be able to pick upgross outliers, and we haven’t gonethrough the data on that, but we’readdressing it.

Benditt: At our hospital one of thehardest things has been to convincethe emergency-department physiciansthat NIV is effective. I would say thatmost of the appropriate NIV starts arein the emergency department, forCOPD, where the effect seems to begreatest. We’ve started a liaison be-tween our respiratory care departmentand our emergency department, withteaching sessions and so forth to try toincrease early use of NIV. One of themajor stumbling blocks is that maybethe pulmonologists and the intensiv-ists are thinking about it, but maybewe are looking in the wrong place.

Pierson: Josh, you and I both prac-tice in Seattle, which is world-renowned for its Medic One system.Not only do I agree that in the emer-gency department it’s very importantto use the right decision-making, but,also, increasingly, outside the hospitalin pre-hospital conditions. I’ve heardit said that Seattle is an ideal place tohave a catastrophic event out on thestreet, but it’s not a very good place toexperience a simple faint, because ifyou do, you’re going to wind up in-



tubated and in the emergency depart-ment at Harborview! Many of our pa-tients who are said to have had acuterespiratory distress, when they first en-ter the emergency department, are al-ready intubated.

Nava: Timing is important, espe-cially when you take into consider-ation the epidemiology of NIV. An-dres Esteban found that COPDadmission to the ICU dropped from10% to 5% in the last 6 years.1 COPDexacerbations did not decrease, so thatmeans that most of the patients weretreated outside the ICU. And most ofthe surveys were performed in theICUs, including those by Demoule,2

Esteban,1 and Hill.3 It’s important tounderstand where and how we treatthose patients—not only in the ICU.In North America it’s a bit different,and Nick pointed out the differencebetween Europe and the States. In Eu-rope we do a lot more NIV than theydo in North America, including out-side the ICU, in the emergency de-partment and pulmonology ward,which is not an ICU. I think NIV ismore popular in Europe because wehave 2 big fields of application, de-pending on timing.

1. Esteban A, Ferguson ND, Meade MO, Fru-tos-Vivar F, Apezteguia C, Brochard L,et al. Evolution of mechanical ventilationin response to clinical research. Am J Re-spir Crit Care Med 2008;177(2):170-177.

2. Demoule A, Girou E, Richard JC, Taille S,Brochard L. Increased use of noninvasiveventilation in French intensive care units.Intensive Care Med 2006;32(11):1747-1755.

3. Maheshwari V, Paioli D, Rothaar R, HillNS. Utilization of noninvasive ventilationin acute care hospitals: a regional survey.Chest 2006;129(5):1226-1233.

Hess: Addressing the internationalaspects of NIV, I’ll share an anecdote.Some time ago I was at a conferencein Southeast Asia. I gave a lecture onNIV and how to initiate NIV in a hos-pital. A physician came up to me afterthat lecture and said that in his coun-try he was certain there had never beena single use of NIV.

Mehta: We in this room, and oth-ers, are responsible for making thefield of NIV overwhelming for manypractitioners. There have been numer-ous trials in the last 10 years, and wecan’t expect the average clinician tobe aware of all those trials. So thegoal of having all physicians, respira-tory therapists, and institutions famil-iar and comfortable with NIV may beoverly ambitious. And I’m not surethat we should be encouraging everysingle clinician to be comfortable withNIV, in that I think it might be dan-gerous.

We know that there are certain pa-tients we can harm by using NIV toolong, starting it too late, or delayingnecessary intubation. Over the last yearwe’ve started outreach “access” teamsthroughout most of Canada—it’s be-come a government mandate—to havea dedicated team in the hospital whoare extremely comfortable with NIV.

Also, the number of types of ven-tilators and masks and modes is over-whelming. And I don’t think we can

expect the residents or average clini-cian to be comfortable with this. Inmy hospital the people who are mostcomfortable and have the expertisewith NIV are the respiratory therapists,and they are present everywhere, in-cluding in the emergency departmentwhen patients arrive. Maybe theyshould be the group leading NIV usein many institutions?

Doyle:* In the surveys on NIV andCPAP did they survey any neonatalgroups? It seems to me that a lot ofneonatal patients are not being intu-bated, but instead are placed on whatI would call high-flow systems. Arethose included in the surveys’ defini-tions of NIV? It also appears to methere’s a proliferation of use of high-flow (30-40 L/min) oxygen nasal-prongs systems in adult patients, whichI suspect deliver some level of CPAP.Are such high-flow oxygen system in-cluded in the definitions of NIV?

Pierson: In my literature search,without any restriction on patient age,nothing popped up on any of the in-quiries with respect to the neonataland pediatric populations. The firstNIV support of acute respiratory fail-ure probably was in neonates—at leastone of the early instances I heard aboutwas—but I am not aware of any or-ganized documentation of what thepractice has been.

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



history and epidemiology of noninvasive...

Documents