Article review

Article review

This assignment requires you to review the provided article (uploaded document) and need to be completed in three sections (summary, critique, and discussion) as it is shown below.

Please follow strictly the following format:
1.    The title of the research review should be different from the title of the paper being reviewed. ?
2.    Summary (400 words max)
In this section where you would describe in details the provided article, through giving a summary of the aims of the study, the setting, design, sample (including participants details), methods (including The research process used, including any interventions and outcomes measured), results and conclusion of the published study.
*** Students Please Note that Reproduction of the article abstract is not permitted.
3.    Critique (400 words max)
This is where you would critique the article… so do some research on the type of method they have used – what sort of measures of rigour apply to this method and how have they have fulfilled these measures of rigour? Are the findings and conclusions they have drawn reliable?
Include in your critique the methodological quality, ethical considerations and limitations.
4.    Discussion and conclusion (400 words max)
In this section you will discuss this chosen article in relation to other articles that have investigated this topic. Does this article support findings in other similar studies or do the findings differ? If they differ, can you think of any explanation for this?
**You should finish your discussion section with a conclusion – after looking at this article in the context of past research, what is the significance of the article’s findings and how do you think clinicians should use these findings in their clinical practice?
NOTE:
1.  You need to use at least 15 references to support your argument.
2. The bulk of this review should be dedicated to a thorough and considered critique of the paper, both in relation to the quality of the research, and where this research is positioned in relation to other published work in the area. Please note that, for this assessment only, you should follow the referencing convention of the provided journal article.

new england
journal of medicine
The

established in 1812

february 28, 2013

vol. 368  no. 9

High-Frequency Oscillation in Early Acute Respiratory
Distress Syndrome
Niall D. Ferguson, M.D., Deborah J. Cook, M.D., Gordon H. Guyatt, M.D., Sangeeta Mehta, M.D., Lori Hand, R.R.T.,
Peggy Austin, C.C.R.A., Qi Zhou, Ph.D., Andrea Matte, R.R.T., Stephen D. Walter, Ph.D., Francois Lamontagne, M.D.,
John T. Granton, M.D., Yaseen M. Arabi, M.D., Alejandro C. Arroliga, M.D., Thomas E. Stewart, M.D.,
Arthur S. Slutsky, M.D., and Maureen O. Meade, M.D., for the OSCILLATE Trial Investigators
and the Canadian Critical Care Trials Group*

A bs t r ac t
Background

Previous trials suggesting that high-frequency oscillatory ventilation (HFOV) reduced
mortality among adults with the acute respiratory distress syndrome (ARDS) were limited by the use of outdated comparator ventilation strategies and small sample sizes.
Methods

In a multicenter, randomized, controlled trial conducted at 39 intensive care units
in five countries, we randomly assigned adults with new-onset, moderate-to-severe
ARDS to HFOV targeting lung recruitment or to a control ventilation strategy targeting
lung recruitment with the use of low tidal volumes and high positive end-expiratory
pressure. The primary outcome was the rate of in-hospital death from any cause.
Results

On the recommendation of the data monitoring committee, we stopped the trial after
548 of a planned 1200 patients had undergone randomization. The two study groups
were well matched at baseline. The HFOV group underwent HFOV for a median of
3 days (interquartile range, 2 to 8); in addition, 34 of 273 patients (12%) in the
control group received HFOV for refractory hypoxemia. In-hospital mortality was
47% in the HFOV group, as compared with 35% in the control group (relative risk
of death with HFOV, 1.33; 95% confidence interval, 1.09 to 1.64; P?=?0.005). This
finding was independent of baseline abnormalities in oxygenation or respiratory
compliance. Patients in the HFOV group received higher doses of midazolam than
did patients in the control group (199 mg per day [interquartile range, 100 to 382]
vs. 141 mg per day [interquartile range, 68 to 240], P<0.001), and more patients in the
HFOV group than in the control group received neuromuscular blockers (83% vs.
68%, P<0.001). In addition, more patients in the HFOV group received vasoactive
drugs (91% vs. 84%, P?=?0.01) and received them for a longer period than did patients in the control group (5 days vs. 3 days, P?=?0.01).
Conclusions

In adults with moderate-to-severe ARDS, early application of HFOV, as compared with
a ventilation strategy of low tidal volume and high positive end-expiratory pressure, does
not reduce, and may increase, in-hospital mortality. (Funded by the Canadian Institutes of Health Research; Current Controlled Trials numbers, ISRCTN42992782 and
ISRCTN87124254, and ClinicalTrials.gov numbers, NCT00474656 and NCT01506401.)

From the Interdepartmental Division of
Critical Care Medicine (N.D.F., S.M., J.T.G.,
T.E.S., A.S.S.), Departments of Medicine
and Physiology (N.D.F.), the Department
of Medicine, Division of Respirology (S.M.,
J.T.G., T.E.S.), and the Departments of
Medicine, Surgery, and Biomedical Engineering (A.S.S.), University of Toronto, University Health Network and Mount Sinai
Hospital (N.D.F., J.T.G.), Mount Sinai Hospital (S.M., T.E.S.), St. Michael’s Hospital
(A.S.S.), and the Critical Care Program, University Health Network (A.M.), Toronto; the
Interdepartmental Division of Critical Care,
Hamilton Health Sciences (D.J.C., M.O.M.),
the Departments of Medicine and Clinical
Epidemiology and Biostatistics (D.J.C.,
G.H.G., S.D.W., M.O.M.), and the CLARITY
Research Centre (D.J.C., G.H.G., S.D.W.,
M.O.M., L.H., P.A., Q.Z.), McMaster University, Hamilton, ON; and the Centre de
Recherche Clinique Étienne-Le Bel, Université de Sherbrooke, Sherbrooke, QC
(F.L.) — all in Canada; the Intensive Care
Department, King Saud bin Abdulaziz
University for Health Sciences, Riyadh,
Saudi Arabia (Y.M.A.); and the Department of Medicine, Scott and White Healthcare and Texas A&M Health Science Center College of Medicine, Temple (A.C.A.).
Address reprint requests to Dr. Meade at
1280 Main St. W., Hamilton, ON L8N 3Z5,
Canada, or at meadema@hhsc.ca.
*    A complete list of the investigators in the
Oscillation for Acute Respiratory Distress
Syndrome Treated Early (OSCILLATE)
trial is provided in the Supplementary
Appendix, available at NEJM.org.
This article was published on January 22,
2013, at NEJM.org.
N Engl J Med 2013;368:795-805.
DOI: 10.1056/NEJMoa1215554
Copyright © 2013 Massachusetts Medical Society.

n engl j med 368;9  nejm.org  february 28, 2013

The New England Journal of Medicine
Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission.
Copyright © 2013 Massachusetts Medical Society. All rights reserved.

795

The

n e w e ng l a n d j o u r na l

T

he acute respiratory distress syndrome (ARDS) is a common complication
of critical illness.1,2 Mortality is high, and
survivors often have long-term complications.3,4
Although mechanical ventilation is life-sustaining
for patients with ARDS, it can perpetuate lung
injury. Basic research suggests that repetitive
overstretching or collapse of lung units with each
respiratory cycle can generate local and systemic
inflammation, contributing to multiorgan failure and death.5 Consistent with these findings
are data from clinical trials that support the use
of smaller tidal volumes (6 vs. 12 ml per kilogram of predicted body weight)6 and higher levels of positive end-expiratory pressure (PEEP).7-10
Mortality remains high, however, and additional
therapies are needed to protect the lung in cases
of severe ARDS.11,12
One such approach is high-frequency oscillatory ventilation (HFOV), which delivers very small
tidal volumes (approximately 1 to 2 ml per kilogram13) at very high rates (3 to 15 breaths per
second).14-19 Previous randomized trials of the use
of HFOV in adults with ARDS have suggested that
this strategy results in improvements in oxygenation and survival, but the trials were limited by
small sample sizes and outdated ventilation strategies for the control group.20-22 Consequently, despite the frequent use of HFOV in patients who
do not have an adequate response to conventional
mechanical ventilation and the increased use of
HFOV earlier in the course of the disease, this
approach remains an unproven therapy for adults
with ARDS.23-26 We therefore compared HFOV
with a conventional ventilation strategy that used
low tidal volumes and high levels of PEEP in patients with new-onset, moderate-to-severe ARDS.

Me thods
Study Oversight

For the pilot phase of the study, we enrolled patients at 11 centers in Canada and 1 in Saudi
Arabia from July 2007 through June 2008; for the
main trial, we enrolled patients at the same centers and at an additional 27 centers in Canada, the
United States, Saudi Arabia, Chile, and India from
July 2009 through August 2012 (see the Supplementary Appendix, available with the full text of
this article at NEJM.org). The trial protocol, which
is available at NEJM.org, was approved by the research ethics board at each participating site.
796

of

m e dic i n e

The first and last author vouch for the accuracy
and completeness of the reported data and for
the fidelity of this report to the study protocol.
For HFOV, we used the SensorMedics 3100B
High-Frequency Oscillatory Ventilator (CareFusion);
the manufacturer loaned nine ventilators and
provided technical support but had no role in the
design of the study, the collection or analysis of
the data, or the preparation of the manuscript.
Patients

Patients were eligible for inclusion if they had
had an onset of pulmonary symptoms within the
previous 2 weeks, had undergone tracheal intubation, had hypoxemia (defined as a ratio of the
partial pressure of arterial oxygen [Pao2] to the
fraction of inspired oxygen [Fio2] of =200, with
an Fio2 of =0.5), and had bilateral air-space opacities on chest radiography. Patients were excluded
if they had hypoxemia primarily related to left
atrial hypertension, suspected vasculitic pulmonary hemorrhage, neuromuscular disorders that
are known to prolong the need for mechanical
ventilation, severe chronic respiratory disease, or
preexisting conditions with an expected 6-month
mortality exceeding 50%; if they were at risk for
intracranial hypertension; if there was a lack of
commitment to life support; if the expected duration of mechanical ventilation was less than 48
hours; if they were younger than 16 years of age
or older than 85 years of age; or if their weight
was less than 35 kg or more than 1 kg per centimeter of height. We did not enroll patients who
had already met the eligibility criteria for more
than 72 hours, those who were already receiving
HFOV, or those whose physicians declined to enroll them.
After enrollment, standardized ventilator settings were used for all the patients: pressurecontrol mode, a tidal volume of 6 ml per kilogram, and an Fio2 of 0.60 with a PEEP level of
10 cm of water or higher if needed for oxygenation. After 30 minutes, if the Pao2:Fio2 ratio
remained at 200 or lower, patients underwent
randomization; otherwise the standardized ventilator settings were maintained, and the patients
were reassessed at least once daily for up to
72 hours. Eligible patients were randomly assigned in a 1:1 ratio to the HFOV group or to
the conventional-ventilation group. Randomization was performed in undisclosed block sizes of
2 and 4 with the use of a central Web-based

n engl j med 368;9  nejm.org  february 28, 2013

The New England Journal of Medicine
Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission.
Copyright © 2013 Massachusetts Medical Society. All rights reserved.

High-Frequency Oscillation in Early ARDS

Table 1. Ventilator Protocols.*
Component Variable
Ventilator mode

HFOV

Control Ventilation

High-frequency oscillatory
ventilation

Pressure control

Tidal volume target (ml/kg of predicted body weight)

NA

6

Tidal volume range (ml/kg of predicted body weight)

NA

4–8

Plateau airway pressure (cm of water)

NA

=35

Positive end-expiratory pressure (cm of water)

NA

Adjusted according to
oxygenation†

Adjusted according to
oxygenation†

Measured but not adjusted

3–12 Hz

=35 breaths/min

Mean airway pressure (cm of water)
Respiratory frequency
Pressure amplitude target (cm of water)

90

NA

Partial pressure of arterial oxygen (mm Hg)

55–80

55–80

Oxygen saturation by pulse oximetry (%)

88–93

88–93

7.25–7.35

7.30–7.45

Ratio of inspiratory-to-expiratory time

Arterial blood pH

1:2

1:1–1:3

Recruitment maneuvers

Yes

Yes

*    The full version of the study protocol is available at NEJM.org. HFOV denotes high-frequency oscillatory ventilation, and NA
not applicable.
†    For more information on the protocol for adjustment, see Table 2.

randomization system, stratified according to
center. All patients or their legal surrogates provided written informed consent for participation
in the study.

Table 2. Usual Combinations of the Fraction of Inspired
Oxygen (Fio2) and Positive End-Expiratory Pressure (PEEP)
or Mean Airway Pressure Used to Adjust Ventilators.
HFOV

HFOV Protocol

The HFOV protocol was designed on the basis of
the results of pilot testing and consensus guidelines.24,27 We first conducted a recruitment maneuver, by applying 40 cm of water pressure for
40 seconds to the airway opening in an effort to
reopen closed lung units. We then initiated HFOV
with a mean airway pressure of 30 cm of water,
adjusting the pressure thereafter according to the
protocol, targeting a Pao2 of 55 to 80 mm Hg (Tables 1 and 2). We minimized HFOV tidal volumes
by using the highest possible frequency that
would maintain arterial blood pH above 7.25.13,28
After 24 hours of HFOV, conventional ventilation could be resumed if the mean airway pressure was 24 cm of water or less for 12 hours.
This transition was mandatory when airway
pressures reached 20 cm of water. Thereafter,
mechanical ventilation followed the control protocol. Over the next 48 hours, if an Fio2 of more
than 0.4 or a PEEP level of more than 14 cm of
water was required for more than 1 hour to
achieve oxygenation targets, HFOV was resumed.

Fio2

Control Ventilation

Mean Airway
Pressure

Fio2

cm of water

PEEP
cm of water

0.4

20

0.3

0.4

22

0.3

5
8

0.4

24

0.3

10

0.4

26

0.4

10

0.4

28

0.4

12

0.4

30

0.4

14

0.5

30

0.4

16

0.6

30

0.4

18

0.6

32

0.5

18

0.6

34

0.5

20

0.7

34

0.6

20

0.8

34

0.7

20

0.9

34

0.8

20

1.0

34

0.8

22

1.0

36

0.9

22

1.0

38

1.0

22

1.0

24

n engl j med 368;9  nejm.org  february 28, 2013

The New England Journal of Medicine
Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission.
Copyright © 2013 Massachusetts Medical Society. All rights reserved.

797

The

n e w e ng l a n d j o u r na l

m e dic i n e

Control Ventilation Protocol

Statistical Analysis

The control ventilation protocol, which was adapted from an earlier trial,9 called for a target tidal
volume of 6 ml per kilogram, with plateau airway
pressure of 35 cm of water or less and high levels
of PEEP. After an initial recruitment maneuver
(the same as that used for the HFOV group), clinicians applied ventilation using pressure-control
mode with a PEEP level of 20 cm of water and
then adjusted the PEEP level and the Fio2 according to the protocol (Tables 1 and 2). The protocol
permitted the use of volume-assist control mode
or pressure-support mode with the same limits
for tidal volumes and airway pressures. For patients receiving pressure support with PEEP levels of 10 cm of water or less and an Fio2 of 0.4 or
less, there were no limits on tidal volume or airway pressures. The weaning protocol, which has
been published previously, included daily trials of
spontaneous breathing.9,29

We anticipated that mortality in the control group
would be 45%. Assuming a two-sided alpha level
of 0.05, we calculated that enrollment of 1200
patients would provide at least 80% power to detect a relative-risk reduction with HFOV of 20%,
even if mortality in the control group was as low
as 37%.
Investigators reviewed feasibility data from the
pilot phase, which involved 94 patients, but remained unaware of the clinical outcomes. The
independent data monitoring committee reviewed
the clinical outcomes from the pilot phase and
recommended that the trial continue to the next
phase. As originally planned, data from the patients involved in the pilot phase were included
in the current analyses. In addition to an interim
analysis after 800 patients had undergone randomization, safety analyses of physiological data
at the initiation of the study were planned after
300, 500, and 700 patients had undergone randomization. After reviewing these safety data, the
data monitoring committee could request analyses of in-hospital mortality, which they did after
both the 300-patient and 500-patient safety
analyses. With plans to stop the study early only
in response to a strong signal of harm in association with the use of HFOV, we used the
O’Brien–Fleming method to calculate alpha
spending and generated one-sided P values for
considering early stopping after random assignment of 300 patients (P=0.00001), 500 patients
(P=0.0001), and 700 patients (P=0.0064).
We used SAS software, version 9.2, for the
statistical analyses. We summarized data using
means with standard deviations, medians and
interquartile ranges, or proportions. Normally
distributed data were compared with the use of
Student’s t-test, nonnormally distributed data
with the use of the Wilcoxon rank-sum test, and
proportions with the use of the Mantel–Haenszel
chi-square test, with stratification according to
center. We analyzed data from all patients according to their assigned group.
The primary outcome was in-hospital mortality, with the outcome compared between the two
groups stratified according to center. Other than
recording whether death occurred as a result of
withdrawal of life support, we did not record
specific causes of death. As a sensitivity analysis, we used logistic regression to adjust the
treatment effect for prespecified baseline vari-

Procedures in Both Groups

When hypoxemia persisted despite increases in
PEEP or mean airway pressure, or when, on the
basis of radiographic or clinical evidence, physicians judged that the lungs were over-distended,
they could reduce PEEP or mean airway pressure
to a level below that indicated in the assigned
protocol (Table 2).
For patients with hypoxemia who required an
Fio2 of 0.9 or greater, clinicians could institute
therapies for hypoxemia (e.g., prone positioning
or inhaled nitric oxide) that did not interfere
with the assigned ventilator protocols. Physicians
could institute any alternative therapy (including
HFOV in the control group) for patients who met
any one of the following criteria: refractory hypoxemia (Pao2 <60 mm Hg for 1 hour with an
Fio2 of 1.0 and neuromuscular blockade), refractory barotrauma (persistent pneumothorax or increasing subcutaneous emphysema despite two
thoracostomy tubes on the involved side), or refractory acidosis (pH of =7.05 despite neuromuscular blockade).
Physicians prescribed fluids, sedatives, and
neuromuscular blockers at their discretion. We
recorded cardiorespiratory variables daily as well
as data on cointerventions applied while patients
were undergoing mechanical ventilation for up to
60 days. Intensivists reviewed chest radiographs
for evidence of new barotrauma. Patients were
followed until their discharge from the hospital.
798

of

n engl j med 368;9  nejm.org  february 28, 2013

The New England Journal of Medicine
Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission.
Copyright © 2013 Massachusetts Medical Society. All rights reserved.

High-Frequency Oscillation in Early ARDS

ables: age, the Acute Physiology Score component of the Acute Physiology and Chronic Health
Evaluation (APACHE) II score,30 the presence or
absence of sepsis, and the duration of hospitalization before randomization.9 To compare the
two groups with respect to the time to death, we
used a survival analysis, in which patients who
were discharged alive from the hospital were assumed to be alive at day 60.
We conducted prespecified subgroup analyses
to determine whether there were interactions of
the treatment effect with baseline severity of lung
injury (in quartiles of the Pao2:Fio2 ratio) or with
center experience with HFOV and study protocols (in thirds of number of patients recruited).
In addition, we studied interactions of the treatment effect with baseline dynamic compliance
measured from tidal breaths during conventional
ventilation (in quartiles), baseline body-mass index (in quartiles), and receipt or no receipt of
vasopressors at baseline — all post hoc analyses.

R e sult s
Early Termination of the Trial

After the 500-patient analysis, the steering committee terminated the trial, acting on a unanimous recommendation from the data monitoring committee, although the threshold P value
for stopping had not been reached. At the time of
termination, 571 patients had been enrolled, of
whom 548 had undergone randomization: 275 to
the HFOV group and 273 to the control-ventilation group (Fig. 1). Important prognostic factors
were similar in the two groups at baseline (Table 3,
and Table S1 in the Supplementary Appendix).
Mortality

A total of 129 patients (47%) in the HFOV group,
as compared with 96 patients (35%) in the control
group, died in the hospital (relative risk of death
with HFOV, 1.33; 95% confidence interval, 1.09 to
1.64; P?=?0.005) (Table 4 and Fig. 2). The results
were consistent in a multivariable analysis (Table
S2 in the Supplementary Appendix), in an analysis of mortality in the intensive care unit (ICU),
and in an analysis of 28-day mortality. Subgroup
analyses showed no interaction of mortality with
baseline severity of hypoxemia, respiratory compliance, body-mass index, or use or nonuse of
vasopressors or with center experience in the
trial (Fig. S1 in the Supplementary Appendix).

Early Physiological Responses to Ventilation

Table S3 in the Supplementary Appendix shows
early physiological responses to HFOV and to
control ventilation. The use of vasopressors was
similar in the HFOV and control groups before
the initiation of ventilation (66% and 61%, respectively; P?=?0.24) but increased in the HFOV
group as compared with the control group within 4 hours after initiation (73% vs. 62%, P?=?0.01)
and increased even more in the HFOV group by
the following day (78% vs. 58%, P<0.001). The
use of neuromuscular blockers followed a similar
pattern: 27% of patients in the HFOV group and
29% of those in the control group received neuromuscular blockers before the initiation of ventilation (P?=?0.66), 46% as compared with 31%
received them within 4 hours after initiation
(P<0.001), and 46% as compared with 26% received them the next day (P<0.001). The mean Fio2
at these time points decreased to a similar extent
in both groups: the Fio2 was 0.75 in the HFOV
group and 0.73 in the control group before initiation (P?=?0.93); 0.62 and 0.64 in the two groups, respectively, 4 hours after initiation (P?=?0.94); and
0.51 and 0.50, respectively, the next day (P?=?0.97).
Cardiorespiratory Results

Table S4 in the Supplementary Appendix shows
cardiorespiratory data from the first week of the
study. On day 1, the mean (±SD) of the mean airway pressure in the HFOV group was 31±2.6 cm
of water, with a frequency of 5.5±1.0 Hz; patients
in the control group underwent ventilation with
a tidal volume of 6.1±1.3 ml per kilogram, PEEP
of 18±3.2, and plateau pressure of 32±5.7 cm
of water. The mean Fio2 in the control group
was similar to or lower than that in the HFOV
group, despite lower mean airway pressures. The
net fluid balance was higher in the HFOV group
than in the control group, but the difference was
not significant. In the HFOV group, 270 of the
275 patients (98%) underwent HFOV for a median
of 3 days (interquartile range, 2 to 8); a total of
222 patients (81%) survived and were transitioned
to conventional ventilation for a further 5 days
(interquartile range, 2 to 7). In the control group,
34 patients (12%) crossed over to HFOV (31 according to protocol and 3 in violation of protocol) for
7 days (interquartile range, 5 to 15), beginning
2 days (interquartile range, 1 to 4) after randomization; 24 of those 34 patients (71%) died in the
hospital.

n engl j med 368;9  nejm.org  february 28, 2013

The New England Journal of Medicine
Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission.
Copyright © 2013 Massachusetts Medical Society. All rights reserved.

799

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

2720 Patients met inclusion criteria

1571 Were excluded (some patients may
have more than 1 exclusion criterion)
411 Had primary cardiac failure
316 Had chronic respiratory disease
259 Had condition with expected >50%
6-mo mortality
193 Were considering palliation rather than
aggressive care
118 Were at risk for intracranial hypertension
103 Had vasculitic pulmonary hemorrhage
373 Had other reasons

1149 Were eligible

574 Were not enrolled
254 Did not provide consent
130 Were withdrawn by physician
82 Were in ICU >72 hr
75 Were already undergoing HFOV
24 Were enrolled in related trial
9 Had other reasons

571 Were enrolled

23 Did not undergo randomization
19 Had Pao2:FIo2 ratio >200 mm Hg
on standard settings
4 Had other reasons

548 Underwent randomization

273 Were assigned to receive control
ventilation
273 Received assigned intervention

275 Were assigned to receive HFOV
270 Received assigned intervention
2 Died before HFOV could be started
2 Withdrew consent after randomization
1 Had approval withdrawn by physician

1 Had premature termination of assigned
strategy after withdrawal of consent

3 Had premature termination of assigned
strategy after withdrawal of consent

273 Were included in primary analysis

275 Were included in primary analysis

Figure 1. Screening, Randomization, and Follow-up.
HFOV denotes high-frequency oscillatory ventilation, ICU intensive care unit, and Pao2:Fio2 the ratio of the partial
pressure of arterial oxygen to the fraction of inspired oxygen.

800

n engl j med 368;9  nejm.org  february 28, 2013

The New England Journal of Medicine
Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission.
Copyright © 2013 Massachusetts Medical Society. All rights reserved.

High-Frequency Oscillation in Early ARDS

Table 3. Baseline Characteristics of the Patients.*
HFOV Group
(N?=?275)

Characteristic
Age — yr
Female sex — no. (%)

Control Group
(N?=?273)

Patients Eligible but
Not Enrolled
(N?=?472)†

P Value‡

55±16

54±16

53±16

0.18

108 (39)

120 (44)

198 (42)

0.42

26±8

<0.001

APACHE II score§

29±8

29±7

Duration of hospital stay — days

5.6±8.0

4.9±8.0

Duration of mechanical ventilation — days

2.5±3.3

1.9±2.3

Risk factors for ARDS — no. of patients (%)
Sepsis

128 (47)

130 (48)

193 (41)

0.01

Pneumonia

155 (56)

164 (60)

289 (61)

0.37

49 (18)

44 (16)

51 (11)

0.02

Gastric aspiration
Trauma

10 (4)

5 (2)

24 (5)

0.07

Other

71 (26)

67 (25)

137 (29)

0.34

Tidal volume — ml/kg of predicted body weight

7.2±1.9

7.1±1.8

Plateau pressure — cm of water

29±6

29±7

27±7

<0.001

11±4

<0.001

13±3

13±4

Minute ventilation — liters/min

Set PEEP — cm of water

11.3±3.1

11.2±3.3

Oxygenation index

19.6±11.2

19.9±9.3

17.8±10.2

0.002

Pao2:Fio2 ratio — mm Hg

121±46

114±38

118±47

0.17

PaCo2 — mm Hg
Arterial pH
Barotrauma — no. of patients (%)

46±13

47±14

45±14

0.01

7.32±0.10

7.31±0.10

7.32±0.12

0.06

19 (7)

14 (5)

184 (67)

171 (63)

Cointerventions — no. of patients (%)
Inotropes or vasopressors
Renal-replacement therapy

29 (11)

28 (10)

Glucocorticoids

93 (34)

96 (35)

Neuromuscular blockers

84 (31)

94 (34)

*    Plus–minus values are means ±SD. There were no significant differences between the two study groups in any of the baseline characteristics listed here, with the exception of duration of mechanical ventilation, for which P?=?0.003. ARDS denotes
acute respiratory distress syndrome, and Pao2 partial pressure of arterial oxygen.
†    Not all centers had approval from an ethics committee to collect data on patients who were eligible but not enrolled in
the study.
‡    The P values are for the comparison of patients who were eligible but not enrolled with all patients who underwent randomization, with adjustment for stratification according to center.
§     Scores on the Acute Physiology and Chronic Health Evaluation II (APACHE II) range from 0 to 71, with higher scores
indicating greater severity of illness.

Cointerventions

During the course of the study, larger proportions
of patients in the HFOV group than in the control
group received vasoactive drugs (91% vs. 84%,
P?=?0.01) and neuromuscular blockers (83% vs.
68%, P<0.001); vasoactive drugs were administered for an average of 2 days longer in the HFOV
group than in the control group, and neuromus-

cular blockers were administered for an average
of 1 day longer in the HFOV group (Table S5 in the
Supplementary Appendix). Sedatives and opioids
(most commonly midazolam and fentanyl) were
administered for the same duration in the two
groups (median, 10 days [interquartile range, 6 to
18] and 10 days [interquartile range, 6 to 17], respectively; P?=?0.99), but during the first week the

n engl j med 368;9  nejm.org  february 28, 2013

The New England Journal of Medicine
Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission.
Copyright © 2013 Massachusetts Medical Society. All rights reserved.

801

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

Table 4. Outcomes.
HFOV Group
(N = 275)

Control Group
(N = 273)

Relative Risk
(95% CI)

P Value

Death in hospital — no. (%)

129 (47)

96 (35)

1.33 (1.09–1.64)

0.005

Death in intensive care unit — no. (%)

123 (45)

84 (31)

1.45 (1.17–1.81)

0.001

Death before day 28 — no. (%)

111 (40)

78 (29)

1.41 (1.12–1.79)

0.004

New barotrauma — no./total no. (%)*

46/256 (18)

34/259 (13)

1.37 (0.91–2.06)

0.13

New tracheostomy — no./total no. (%)†

59/273 (22)

66/267 (25)

0.87 (0.64–1.19)

0.39

19 (7)

38 (14)

0.50 (0.29–0.84)

0.007

Outcome

Refractory hypoxemia — no. (%)
Death after refractory hypoxemia — no./total no. (%)

15/19 (79)

25/38 (66)

1.20 (0.87–1.66)

0.31

Refractory acidosis — no. (%)

9 (3)

8 (3)

1.12 (0.44–2.85)

0.82

Refractory barotrauma — no. (%)

2 (<1)

2 (<1)

0.99 (0.14–7.00)

0.99

Use of mechanical ventilation, among survivors
— days
Median
Interquartile range

0.59
11

10

7–19

6–18

15

14

9–25

9–26

Stay in intensive care, among survivors — days
Median
Interquartile range

0.93

Length of hospitalization, among survivors — days
Median
Interquartile range

0.74
30

25

16–45

15–41

*    Barotrauma was defined as pneumothorax, pneumomediastinum, pneumopericardium, or subcutaneous emphysema
occurring spontaneously or after a recruitment maneuver. Excluded from this category were patients who had barotrauma
at baseline.
†    Excluded from this category were patients who had a tracheostomy at baseline.

median doses of midazolam were significantly
higher in the HFOV group than in the control
group (199 mg per day [interquartile range, 100
to 382] vs. 141 mg per day [interquartile range,
68 to 240], P<0.001), and there was a trend toward higher doses of fentanyl equivalents in the
HFOV group (2980 µg per day [interquartile range,
1258 to 4800] vs. 2400 µg per day [interquartile
range, 1140 to 4430], P?=?0.06) (for daily doses of
selected sedative and analgesic drugs, see Fig. S2
in the Supplementary Appendix). The rates of use
of other cointerventions, including glucocorticoids, renal-replacement therapy, and prone positioning, were similar in the two groups (Table S5
in the Supplementary Appendix).
Other Outcomes

Refractory hypoxemia developed in significantly
more patients in the control group than in the
HFOV group; however, the total number of deaths
after refractory hypoxemia was similar in the
802

two groups (Table 4). The proportion of deaths
after withdrawal of life support was similar in
the two groups (55% [71 of 129 patients] in the
HFOV group and 49% [47 of 96 patients] in the
control group, P?=?0.12). The rate of new-onset
barotrauma was higher in the HFOV group than
in the control group, but the difference was not
significant (18% and 13%, respectively; P?=?0.13).
Among survivors, the duration of ventilation and
the length of stay in the ICU were similar in the
two groups (Table 4).

Discussion
The main finding of this multicenter, randomized trial is that among patients with moderateto-severe ARDS, early application of HFOV was
associated with higher mortality than was a ventilation strategy that used small tidal volumes
and high PEEP levels, with HFOV used only in
patients with severe refractory hypoxemia. HFOV

n engl j med 368;9  nejm.org  february 28, 2013

The New England Journal of Medicine
Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission.
Copyright © 2013 Massachusetts Medical Society. All rights reserved.

High-Frequency Oscillation in Early ARDS

1.0
0.9

Probability of Survival

was associated with higher mean airway pressures
and with greater use of sedatives, neuromuscular
blockers, and vasoactive drugs.
We stopped the trial early on the basis of a
strong signal for increased mortality with HFOV,
even though the prespecified stopping thresholds
had not been reached. Studies that are stopped
early on the basis of harm (or benefit) typically
overestimate the magnitude of effect.31 We chose
to terminate the study for three reasons: there
was a consistent finding of increased mortality
with HFOV in three consecutive analyses that were
conducted after enrollment of 94, 300, and 500
patients; the increased need for vasoactive drugs
in the HFOV group suggested a mechanism of
harm that was not offset by better oxygenation
and lung recruitment; and the effect size was
sufficiently large that we concluded that even if
early HFOV did not increase mortality, it would
be very unlikely to decrease mortality. We believe that continued enrollment would have put
patients at risk with little likelihood of benefit.
Our results are inconsistent with the physiological rationale for HFOV and with the results
of studies in animals. In studies in animals in
which benefits of HFOV were observed, lung
injury was induced with the use of saline lavage
— a highly recruitable model of surfactant deficiency — which our results suggest does not
translate directly to human adults with ARDS, in
whom recruitability can be heterogeneous.32 Our
results also contrast with those of prior randomized trials involving adults.22 A possible explanation, which provided motivation for our trial, is
that prior studies used control ventilation strategies that are now known to be potentially harmful.20,21 We found no benefit with HFOV when a
current ventilation strategy was used as a control. This finding of no benefit with respect to
mortality is consistent with the results of another trial now reported in the Journal; in that
trial, conducted in the United Kingdom, current
standards for lung protection were suggested
but not mandated.33 More surprising was our
finding of harm. Several plausible mechanisms
may contribute to increased mortality with
HFOV. Higher mean airway pressures may result
in hemodynamic compromise by decreasing venous return or directly affecting right ventricular
function.34 Increased use of vasodilating sedative agents may also contribute to hemodynamic
compromise. Moreover, we cannot exclude the

0.8
0.7

Control

0.6
0.5
0.4

HFOV

0.3
0.2
P=0.004 by log-rank test

0.1
0.0

0

15

30

45

60

54
54

26
39

Days since Randomization
No. at Risk
HFOV
Control

275
273

169
181

98
92

Figure 2. Probability of Survival from the Day of Randomization to Day 60
in the HFOV and Control Groups.

possibility of increased barotrauma in association with HFOV.
The HFOV strategy that we chose, which was
supported by preclinical data15,16 and a prospective physiological study,24 aimed to adjust mean
airway pressure on the deflation limb of the
volume-pressure curve and use the highest frequency possible to limit oscillatory volumes.
This approach led to relatively high mean airway
pressures, even considering that when mean airway pressures are delivered with a ratio of inspiratory-to-expiratory time of 1:2, as in our study,
the pressures measured at the airway opening
during HFOV are somewhat higher than those
measured in the trachea.35-37 It is possible that an
HFOV protocol that uses lower mean airway pressures, a different ratio of inspiratory-to-expiratory
time, or a lower oscillatory frequency might have
led to different results.
The strengths of this trial include its methodologic rigor, the application of protocols designed
to open lung units in patients in both groups on
the basis of the best available evidence, and enrollment at centers in several countries, which
enhances the generalizability of our findings.
Because we were cognizant that there is a learning curve associated with the use of HFOV,38,39
we enrolled most patients at centers that were
experienced with HFOV, and we did not detect
an interaction between treatment effect and the
number of enrolled patients per site.

n engl j med 368;9  nejm.org  february 28, 2013

The New England Journal of Medicine
Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission.
Copyright © 2013 Massachusetts Medical Society. All rights reserved.

803

The

n e w e ng l a n d j o u r na l

Our results raise serious concerns about the
early use of HFOV for the management of ARDS
in adults. The results of this study increase the
uncertainty about possible benefits of HFOV even
when applied in patients with life-threatening
refractory hypoxemia.
In conclusion, in adults with moderate-tosevere ARDS, the early application of HFOV
targeting lung recruitment — as compared with
a ventilation strategy that uses low tidal volume
and high PEEP and that permits HFOV only in
cases of refractory hypoxemia — does not reduce mortality and may be harmful.
Supported by the Canadian Institutes of Health Research,
Randomized Controlled Trial (RCT) Program (Ottawa) and the
King Abdullah International Medical Research Center (Riyadh,

of

m e dic i n e

Saudi Arabia). Dr. Ferguson is supported by a Canadian Institutes of Health Research New Investigator Award; Dr. Cook
holds a Canada Research Chair; Dr. Lamontagne is supported by
a Fonds de Recherche de Québec–Santé Research Career Award;
and Drs. Ferguson and Meade and Drs. Lamontagne and Meade
were supported by Canadian Institutes of Health Research RCT
Mentorship awards.
Dr. Ferguson reports receiving grant support through his institution from the Physicians Services Incorporated Foundation;
Dr. Granton, receiving consulting fees from Ikaria, lecture fees
from Actelion and Eli Lilly, grant support through his institution from Pfizer, Actelion, Eli Lilly, GlaxoSmithKline, and Bayer, payment through his institution from Telus for the sale of a
software site license, and support from Actelion for his hospital
foundation for pulmonary hypertension research and providing
expert testimony for Pfizer regarding patent legal action; and
Dr. Slutsky, receiving consulting fees from Maquet Medical,
Novalung, Gambro, Ikaria, and Hemodec. No other potential
conflict of interest relevant to this article was reported.
Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.

References
1.     ARDS Definition Task Force, Ranieri

VM, Rubenfeld GD, et al. Acute respiratory distress syndrome: the Berlin Definition. JAMA 2012;307:2526-33.
2.     Rubenfeld GD, Caldwell E, Peabody E,
et al. Incidence and outcomes of acute
lung injury. N Engl J Med 2005;353:
1685-93.
3.     Phua J, Badia JR, Adhikari NKJ, et al.
Has mortality from acute respiratory distress syndrome decreased over time? Am J
Respir Crit Care Med 2009;179:220-7.
4.     Herridge MS, Tansey CM, Matté A, et
al. Functional disability 5 years after acute
respiratory distress syndrome. N Engl J
Med 2011;364:1293-304.
5.     Tremblay LN, Slutsky AS. Ventilatorinduced lung injury: from the bench to
the bedside. Intensive Care Med 2006;32:
24-33.
6.     The Acute Respiratory Distress Syndrome Network. Ventilation with lower
tidal volumes as compared with traditional tidal volumes for acute lung injury and
the acute respiratory distress syndrome.
N Engl J Med 2000;342:1301-8.
7.     Brower RG, Lanken PN, MacIntyre N,
et al. Higher versus lower positive endexpiratory pressures in patients with the
acute respiratory distress syndrome.
N Engl J Med 2004;351:327-36.
8.     Mercat A, Richard J-CM, Vielle B, et
al. Positive end-expiratory pressure setting in adults with acute lung injury and
acute respiratory distress syndrome: a
randomized controlled trial. JAMA 2008;
299:646-55.
9.     Meade MO, Cook DJ, Guyatt GH, et al.
Ventilation strategy using low tidal volumes, recruitment maneuvers, and high
positive end-expiratory pressure for acute
lung injury and acute respiratory distress
syndrome: a randomized controlled trial.
JAMA 2008;299:637-45.

804

10.     Briel M, Meade M, Mercat A, et al.

Higher vs lower positive end-expiratory
pressure in patients with acute lung injury
and acute respiratory distress syndrome:
systematic review and meta-analysis.
JAMA 2010;303:865-73.
11.     Terragni PP, Rosboch G, Tealdi A, et
al. Tidal hyperinflation during low tidal
volume ventilation in acute respiratory
distress syndrome. Am J Respir Crit Care
Med 2007;175:160-6.
12.     Terragni PP, Del Sorbo L, Mascia L, et
al. Tidal volume lower than 6 ml/kg enhances lung protection: role of extracorporeal carbon dioxide removal. Anesthesiology 2009;111:826-35.
13.     Hager DN, Fessler HE, Kaczka DW, et
al. Tidal volume delivery during high-frequency oscillatory ventilation in adults
with acute respiratory distress syndrome.
Crit Care Med 2007;35:1522-9.
14.     Ferguson ND, Slutsky AS, Kacmarek
R. High-frequency ventilation is/is not the
optimal physiological approach to ventilate ARDS patients. J Appl Physiol 2008;
104:1230-1.
15.     Froese AB. High-frequency oscillatory
ventilation for adult respiratory distress
syndrome: let’s get it right this time. Crit
Care Med 1997;25:906-8.
16.     Rimensberger PC, Pache JC, McKerlie
C, Frndova H, Cox PN. Lung recruitment
and lung volume maintenance: a strategy
for improving oxygenation and preventing lung injury during both conventional
mechanical ventilation and high-frequency
oscillation. Intensive Care Med 2000;26:
745-55.
17.     Imai Y, Nakagawa S, Ito Y, Kawano T,
Slutsky AS, Miyasaka K. Comparison of
lung protection strategies using conventional and high-frequency oscillatory
ventilation. J Appl Physiol 2001;91:
1836-44.

18.     Muellenbach RM, Kredel M, Said HM,

et al. High-frequency oscillatory ventilation reduces lung inflammation: a largeanimal 24-h model of respiratory distress.
Intensive Care Med 2007;33:1423-33.
19.     Sedeek KA, Takeuchi M, Suchodolski
K, et al. Open-lung protective ventilation
with pressure control ventilation, highfrequency oscillation, and intratracheal
pulmonary ventilation results in similar
gas exchange, hemodynamics, and lung
mechanics. Anesthesiology 2003;99:
1102-11.
20.     Derdak S, Mehta S, Stewart TE, et al.
High frequency oscillatory ventilation for
acute respiratory distress syndrome: a
randomized controlled trial. Am J Respir
Crit Care Med 2002;166:801-8.
21.     Bollen CW, van Well GT, Sherry T, et
al. High frequency oscillatory ventilation
compared with conventional mechanical
ventilation in adult respiratory distress
syndrome: a randomized controlled trial.
Crit Care 2005;9:R430-R439.
22.     Sud S, Sud M, Friedrich JO, et al. High
frequency oscillation in patients with
acute lung injury and acute respiratory
distress syndrome (ARDS): systematic review and meta-analysis. BMJ 2010;340:
c2327.
23.     Fort P, Farmer C, Westerman J, et al.
High-frequency oscillatory ventilation for
adult respiratory distress syndrome — a
pilot study. Crit Care Med 1997;25:937-47.
24.     Ferguson ND, Chiche J-D, Kacmarek
RM, et al. Combining high-frequency oscillatory ventilation and recruitment maneuvers in adults with early acute respiratory distress syndrome: the Treatment
with Oscillation and an Open Lung Strategy (TOOLS) Trial pilot study. Crit Care
Med 2005;33:479-86.
25.     Mehta S, Granton J, MacDonald RJ, et
al. High-frequency oscillatory ventilation

n engl j med 368;9  nejm.org  february 28, 2013

The New England Journal of Medicine
Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission.
Copyright © 2013 Massachusetts Medical Society. All rights reserved.

High-Frequency Oscillation in Early ARDS
in adults: the Toronto experience. Chest
2004;126:518-27.
26.     Adhikari NKJ, Bashir A, Lamontagne
F, et al. High-frequency oscillation in
adults: a utilization review. Crit Care Med
2011;39:2631-44.
27.     Fessler HE, Derdak S, Ferguson ND,
et al. A protocol for high-frequency oscillatory ventilation in adults: results from a
roundtable discussion. Crit Care Med
2007;35:1649-54.
28.     Fessler HE, Hager DN, Brower RG.
Feasibility of very high-frequency ventilation in adults with acute respiratory distress syndrome. Crit Care Med 2008;36:
1043-8.
29.     Esteban A, Frutos F, Tobin MJ, et al. A
comparison of four methods of weaning
patients from mechanical ventilation.
N Engl J Med 1995;332:345-50.
30.     Knaus WA, Draper EA, Wagner DP,
Zimmerman JE. APACHE II: a severity of

disease classification system. Crit Care
Med 1985;13:818-29.
31.     Montori VM, Devereaux PJ, Adhikari
NK, et al. Randomized trials stopped
early for benefit: a systematic review.
JAMA 2005;294:2203-9.
32.     Gattinoni L, Caironi P, Cressoni M, et
al. Lung recruitment in patients with the
acute respiratory distress syndrome.
N Engl J Med 2006;354:1775-86.
33.     Young D, Lamb SE, Shah S, et al. Highfrequency oscillation for acute respiratory
distress syndrome. N Engl J Med 2013;
368:806-13.
34.     Guervilly C, Forel J-M, Hraiech S, et
al. Right ventricular function during
high-frequency oscillatory ventilation in
adults with acute respiratory distress syndrome. Crit Care Med 2012;40:1539-45.
35.     Hatcher D, Watanabe H, Ashbury T,
Vincent S, Fisher J, Froese A. Mechanical
performance of clinically available, neo-

natal, high-frequency, oscillatory-type
ventilators. Crit Care Med 1998;26:1081-8.
36.     Pillow JJ, Neil H, Wilkinson MH,
Ramsden CA. Effect of I/E ratio on mean
alveolar pressure during high-frequency
oscillatory ventilation. J Appl Physiol
1999;87:407-14.
37.     Mentzelopoulos SD, Malachias S,
Kokkoris S, Roussos C, Zakynthinos SG.
Comparison of high-frequency oscillation
and tracheal gas insufflation versus standard high-frequency oscillation at two
levels of tracheal pressure. Intensive Care
Med 2010;36:810-6.
38.     The HiFi Study Group. High-frequency oscillatory ventilation compared with
conventional mechanical ventilation in the
treatment of respiratory failure in preterm
infants. N Engl J Med 1989;320:88-93.
39.     Bryan AC, Froese AB. Reflections on
the HIFI trial. Pediatrics 1991;87:565-7.
Copyright © 2013 Massachusetts Medical Society.