&get_box_var;

ORIGINAL ARTICLE

Randomized Trial of Apneic Oxygenation during Endotracheal

Intubation of the Critically Ill
Matthew W. Semler1, David R. Janz2, Robert J. Lentz1, Daniel T. Matthews1, Brett C. Norman1, Tuk R. Assad1,
Raj D. Keriwala1, Benjamin A. Ferrell1, Michael J. Noto1, Andrew C. McKown1, Emily G. Kocurek1, Melissa A. Warren1,
Luis E. Huerta1, and Todd W. Rice1; for the FELLOW Investigators and the Pragmatic Critical Care Research Group
1
Division of Allergy, Pulmonary, and Critical Care Medicine, Vanderbilt University School of Medicine, Nashville, Tennessee; and 2Section
of Pulmonary/Critical Care and Allergy/Immunology, Louisiana State University School of Medicine, New Orleans, Louisiana

Abstract Measurements and Main Results: Median lowest arterial oxygen

saturation was 92% with apneic oxygenation versus 90% with usual
Rationale: Hypoxemia is common during endotracheal intubation care (95% condence interval for the difference, 21.6 to 7.4%;
of critically ill patients and may predispose to cardiac arrest and P = 0.16). There was no difference between apneic oxygenation and
death. Administration of supplemental oxygen during laryngoscopy usual care in incidence of oxygen saturation less than 90% (44.7 vs. 47.2%;
(apneic oxygenation) may prevent hypoxemia. P = 0.87), oxygen saturation less than 80% (15.8 vs. 25.0%; P = 0.22),
or decrease in oxygen saturation greater than 3% (53.9 vs. 55.6%;
Objectives: To determine if apneic oxygenation increases the lowest P = 0.87). Duration of mechanical ventilation, intensive care unit length
arterial oxygen saturation experienced by patients undergoing of stay, and in-hospital mortality were similar between study groups.
endotracheal intubation in the intensive care unit.
Conclusions: Apneic oxygenation does not seem to increase lowest
Methods: This was a randomized, open-label, pragmatic trial in arterial oxygen saturation during endotracheal intubation of critically ill
which 150 adults undergoing endotracheal intubation in a medical patients compared with usual care. These ndings do not support routine
intensive care unit were randomized to receive 15 L/min of 100% use of apneic oxygenation during endotracheal intubation of critically ill
oxygen via high-ow nasal cannula during laryngoscopy (apneic adults.
oxygenation) or no supplemental oxygen during laryngoscopy (usual
Clinical trial registered with www.clinicaltrials.gov (NCT 02051816).
care). The primary outcome was lowest arterial oxygen saturation
between induction and 2 minutes after completion of endotracheal Keywords: intratracheal intubation; airway management;
intubation. pulmonary ventilation

Hypoxemia is the most common death (5). The traditional approach to abnormalities render preoxygenation less
complication of endotracheal intubation in avoiding desaturation during intubation is effective (7) and often insufcient to
the critically ill (14) and the strongest risk preoxygenation (610). However, in prevent desaturation during even short
factor for periprocedural cardiac arrest and critically ill patients, acute physiologic periods of apnea (11).

( Received in original form September 6, 2015; accepted in final form September 30, 2015 )
Supported by NHLBI T32 award (HL087738 09). Data collection used the Research Electronic Data Capture (REDCap) tool developed and maintained with
Vanderbilt Institute for Clinical and Translational Research grant support (UL1 TR000445 from National Center for Advancing Translational Sciences/National
Institutes of Health). The funding institutions had no role in conception, design, or conduct of the study; collection, management, analysis, interpretation, or
presentation of the data; or preparation, review, or approval of the manuscript.
Authors Contributions: Study concept and design, M.W.S., D.R.J., and T.W.R. Acquisition of data, M.W.S., D.R.J., R.J.L., D.T.M., B.C.N., T.R.A., R.D.K.,
B.A.F., M.J.N., A.C.M., E.G.K., M.A.W., and L.E.H. Analysis and interpretation of data, M.W.S., D.R.J., R.J.L., D.T.M., B.C.N., T.R.A., R.D.K., B.A.F., M.J.N.,
A.C.M., E.G.K., M.A.W., and L.E.H. Drafting of the manuscript, M.W.S., D.R.J., and T.W.R. Critical revision of the manuscript for important intellectual content,
D.R.J., R.J.L., D.T.M., B.C.N., T.R.A., R.D.K., B.A.F., M.J.N., A.C.M., E.G.K., M.A.W., and L.E.H. Statistical analysis, M.W.S, D.R.J., and T.W.R. Study
supervision, M.W.S., D.R.J., R.J.L., and T.W.R. M.W.S. and D.R.J. had full access to all the data in the study and take responsibility for the integrity of the data
and the accuracy of the data analysis. M.W.S. and D.R.J. conducted and are responsible for the data analysis.
Correspondence and requests for reprints should be addressed to Matthew W. Semler, M.D., 1161 21st Avenue South, T-2220 MCN, Nashville, TN 37232-
2650. E-mail: [email protected]
This article has an online supplement, which is accessible from this issues table of contents at www.atsjournals.org
Am J Respir Crit Care Med Vol 193, Iss 3, pp 273280, Feb 1, 2016
Copyright 2016 by the American Thoracic Society
Originally Published in Press as DOI: 10.1164/rccm.201507-1294OC on October 1, 2015
Internet address: www.atsjournals.org

Semler, Janz, Lentz, et al.: Apneic Oxygenation for Emergent Intubation 273
 ORIGINAL ARTICLE

oxygenation in out-of-operating room (intervention) or usual care (control). Per

At a Glance Commentary intubations (22, 23), the effectiveness of the factorial design, patients were also
apneic oxygenation in this context remains simultaneously randomized to either
Scientic Knowledge on the unclear. We conducted a prospective, video or direct laryngoscopy, details of
Subject: Hypoxemia is the most randomized trial comparing the which will be reported separately. The
common complication of emergent impact of apneic oxygenation with usual sequence of study group assignments was
endotracheal intubation. Provision of care on lowest arterial oxygen saturation generated via a computerized algorithm
supplemental oxygen by nasal cannula during endotracheal intubation of critically using permuted blocks of 4, 8, and 12.
during laryngoscopy (apneic ill adults. We hypothesized that the Study group assignments were placed in
oxygenation) has been shown in small lowest arterial oxygen saturation during sequentially numbered opaque envelopes
randomized trials to prevent intubation would be higher with apneic that remained sealed until the decision
desaturation during elective intubation oxygenation. had been made that a patient required
of healthy preoperative patients and intubation and was enrolled in the study.
has been recommended during
intubation of the acutely ill, although it Methods Study Treatments
has never been tested in this setting. For all patients, study protocol governed
Study Design only provision of supplemental oxygen
What This Study Adds to the The FELLOW (Facilitating EndotracheaL during apnea and laryngoscopy device
Field: This randomized clinical trial intubation by Laryngoscopy technique used on the rst laryngoscopy attempt.
found no difference between apneic and apneic Oxygenation Within the Decisions regarding intubation, approach to
oxygenation and usual care in the intensive care unit) Study was a randomized, preoxygenation, patient positioning,
lowest oxygen saturation experienced open-label, parallel-group, pragmatic trial medications for induction and
by critically ill adults undergoing comparing apneic oxygenation with usual neuromuscular blockade, ventilation
emergent endotracheal intubation. care during endotracheal intubation of between induction and laryngoscopy,
critically ill adults. The trial was factorialized choice of laryngoscope blade type and
to also compare direct with video size, and use of additional airway
Apneic oxygenation is the delivery laryngoscopy, the details of which will management equipment were made by the
of supplemental oxygen to the be reported separately. The study protocol clinical team. Intubation practices in the
nasopharynx in the absence of ventilation was approved by the institutional review study environment are detailed in the
(12). Even without lung expansion, board at Vanderbilt University with online supplement. Operator compliance
alveolar oxygen diffuses into the waiver of informed consent. The trial with general best-practices in airway
bloodstream and is consumed into was registered online before initiation management was prospectively collected
carbon dioxide. Because of the high (NCT02051816), and the statistical from a convenience sample of consecutive
afnity of carbon dioxide for hemoglobin analysis plan was made publically intubations.
and effective buffering, the volume of available before completion of enrollment Although preoxygenation was allowed,
carbon dioxide returned to the alveoli (https://starbrite.vanderbilt.edu/rocket/ patients in the usual care group were
is less than the volume of oxygen page/FELLOW; available in the online intubated without supplemental oxygen
removed. As a result, alveolar pressure data supplement). during laryngoscopy. For patients in the
decreases and gas is drawn down from apneic oxygenation group, a high-ow
the nasopharynx (10). By increasing the Study Participants nasal cannula (Comfort Soft Plus;
fraction of oxygen in the gas moving from From February 13, 2014 to February 11, Westmed, Inc., Tucson, AZ) set to 15 L/min
the nasopharynx to the lungs, apneic 2015 we enrolled patients undergoing ow of 100% oxygen was placed in the
oxygenation aims to prevent arterial endotracheal intubation in the medical patients nares before induction and
desaturation. ICU at Vanderbilt University Medical kept in place until intubation was complete.
Apneic oxygenation has been used Center. All patients 18 years or older being Because of the nature of the study
to prevent desaturation in patients intubated by a pulmonary and critical intervention, clinicians and study
undergoing brain death examination (13), care medicine fellow were eligible. Patients personnel were aware of study group
bronchoscopy (14), endoscopy (15), were excluded if awake intubation was assignments after enrollment.
and even elective endotracheal intubation planned, if intubation was required so
for general anesthesia (1619). Although emergently that randomization could not be Data Collection
administration of oxygen by nasal achieved, or if the treating clinicians believed To minimize observer bias, data
cannula (20) during laryngoscopy has been a specic approach to intraprocedural collection during intubation was
adopted in many emergency departments oxygenation or a specic laryngoscopy device performed by independent observers
(21) and intensive care units (ICUs) (22), was mandated for the safe performance of the unaware of the study hypothesis and not
there are signicant differences between procedure (Figure 1). involved in the performance of the
intubating electively in the operating procedure. To conrm the accuracy of
room and urgently in out-of-operating Randomization data collected by the independent observers,
room settings. Despite two recent before- Eligible patients were randomly assigned in the primary investigators concurrently
after studies suggesting benet for apneic a 1:1 ratio to receive apneic oxygenation assessed the same outcomes for a

274 American Journal of Respiratory and Critical Care Medicine Volume 193 Number 3 | February 1 2016
ORIGINAL ARTICLE

196 patients met all inclusion criteria

46 were excluded
23 required intubation too urgently to obtain envelope
13 for cardiac arrest
8 for respiratory arrest
2 for acute hypoxic respiratory failure
18 were felt to require video or fiberoptic intubation
1 was felt to require direct laryngoscopy
1 was felt to require apneic oxygenation
3 were excluded for unknown reasons

150 underwent randomization

73 were assigned to usual care 77 were assigned to apneic oxygenation

68 received usual care 76 received apneic oxygenation
5 received apneic oxygenation 1 received usual care

73 were included in intention-to-treat 77 were included in intention-to-treat

analysis for primary outcome analysis for primary outcome

Figure 1. Enrollment, randomization, intervention, and analysis. Of 196 adults intubated by pulmonary and critical care medicine fellows during the study
period, 46 were excluded and 150 were randomized, followed, and included in the intention-to-treat analysis.

convenience sample of around 10% of study additional airway equipment or operators, We performed four prespecied
intubations. and incidence of nonhypoxemia secondary analyses: (1) the effect of the
Subjective assessments of Cormack- complications. Tertiary outcomes included intervention on secondary and tertiary
Lehane grade of view (24), difculty of duration of mechanical ventilation, ICU outcomes, (2) the effect of the intervention
intubation, and airway complications length of stay, and in-hospital mortality. on the primary outcome in prespecied
during the procedure were self-reported by patient and procedural subgroups,
the operator. All other data on baseline Statistical Analysis (3) per-protocol analyses comparing
characteristics, prelaryngoscopy and Anticipating a SD of 10% in lowest lowest arterial oxygen saturation between
postlaryngoscopy management, and arterial oxygen saturation (6), enrollment of patients who received apneic oxygenation
clinical outcomes were collected from the 150 patients would provide 80% statistical with those who did not, and (4) linear
medical record by study personnel. All power (at a two-sided a level of 0.05) to regression for the outcome of lowest arterial
patients were followed until the rst of detect a difference between groups in oxygen saturation in which the exposure
hospital discharge, death, or 28 days after mean lowest arterial oxygen saturation of variable of randomized group assignment
enrollment. 4.6%, within the 5% minimum difference was accompanied rst by just the covariate
considered clinically meaningful in prior of oxygen saturation at induction and
Study Outcomes studies (see online supplement) (6, 9, 17, then by potential baseline confounders.
The primary outcome was the lowest 18, 22). Subgroup analyses were performed
arterial oxygen saturation measured by Analyses were conducted according using logistic regression with heterogeneity
continuous pulse oximetry (SpO2) between to a statistical analysis plan that was of treatment effect determined on the
induction and 2 minutes after successful publically available before completion of basis of statistical test for interaction
endotracheal tube placement (lowest enrollment. Continuous variables were between treatment assignment and
arterial oxygen saturation). Secondary reported as mean 6 SD or median and subgrouping variable. A two-sided P value
efcacy outcomes included incidence of interquartile range (IQR); categorical less than 0.05 was used to determine
hypoxemia (SpO2 ,90%), severe hypoxemia variables as frequencies and proportions. signicance. All analyses were performed
(SpO2 ,80%), desaturation (decrease in Between-group differences were using SPSS Statistics v.22 (IBM Corp.,
SpO2 .3%), and change in saturation analyzed with the Mann-Whitney rank Armonk, NY) or R version 3.2.0 (R
from baseline. Secondary safety outcomes sum test for continuous variables, Fisher Foundation for Statistical Computing,
included Cormack-Lehane grade of glottic exact test for categorical variables, and Vienna, Austria).
view (24), incidence of successful Spearman rank correlation coefcient for
intubation on the rst laryngoscopy correlation between two continuous
attempt (placement of an endotracheal tube variables. The primary analysis was an Results
in the trachea during the rst insertion of unadjusted, intention-to-treat comparison
the laryngoscope into the oral cavity of patients randomized to apneic Enrollment and Baseline
without the use of any other devices), oxygenation versus usual care with regard Characteristics
number of laryngoscopy attempts, time to the primary outcome of lowest arterial Of 196 medical ICU patients intubated
from induction to intubation, need for oxygen saturation. by fellows during the study period,

Semler, Janz, Lentz, et al.: Apneic Oxygenation for Emergent Intubation 275
 ORIGINAL ARTICLE

Table 1. Patient and Operator Characteristics at Baseline

Usual Care Apneic Oxygenation

(n = 73) (n = 77)

Patient characteristics
Age, median (IQR), yr 60 (5067) 60 (5168)
Male, n (%) 46 (63.0) 45 (58.4)
White, n (%) 62 (84.9) 63 (82.9)
BMI, median (IQR), kg/m2 28.6 (23.432.8) 28.6 (23.332.8)
APACHE II score, median (IQR) 22 (1727) 22 (1627)
Vasopressors, n (%) 9 (12.3) 11 (14.3)
Lowest MAP in prior 6 h, median (IQR), mm Hg 68 (5780) 65 (5779)
Lowest oxygen saturation in prior 6 h, median (IQR), % 91 (8893) 92 (8895)
Highest FIO2 in prior 6 h, median (IQR) 0.40 (0.300.80) 0.40 (0.270.60)
BiPAP use in prior 6 h, n (%) 31 (42.5) 26 (33.5)
Reintubation within 24 h of extubation, n (%) 11 (15.1) 9 (11.7)
Intensive care unit diagnoses, n (%)
Sepsis 50 (68.5) 49 (63.6)
Septic shock 14 (19.2) 20 (26.0)
Hemorrhagic shock 3 (4.1) 6 (7.8)
Cardiogenic shock 1 (1.4) 2 (2.6)
Myocardial infarction 9 (12.3) 4 (5.2)
COPD exacerbation 8 (11.0) 4 (5.2)
Hepatic encephalopathy 10 (13.7) 10 (13.0)
Delirium 35 (50.0) 33 (43.4)
Indication for intubation, n (%)
Hypoxic or hypercarbic respiratory failure 42 (57.5) 43 (55.8)
Altered mental status or encephalopathy 18 (24.7) 21 (27.3)
Other 13 (17.8) 13 (16.8)
Comorbidities complicating intubation, n (%)
BMI .30 kg/m2 23 (31.5) 25 (32.5)
Upper gastrointestinal bleeding 7 (9.6) 6 (7.8)
Limited mouth opening* 3 (4.1) 3 (3.9)
Limited neck mobility* 3 (4.1) 2 (2.6)
Head or neck radiation 1 (1.4) 0 (0.0)
Airway mass or infection 0 (0.0) 1 (1.3)
Witnessed aspiration 1 (1.4) 0 (0.0)
Epistaxis or oral bleeding 0 (0.0) 0 (0.0)
Preoxygenation, n (%)
Nonrebreather mask 32 (43.8) 25 (32.5)
BiPAP 23 (31.5) 23 (29.9)
Bag-valve-mask ventilation 31 (42.5) 33 (42.9)
Standard nasal cannulax 2 (2.7) 6 (7.8)
Other 1 (1.4) 0 (0.0)
Oxygen saturation at induction, median (IQR), % 98 (9499) 99 (96100)
Operator characteristics
Total number of prior intubations, median (IQR) 56 (4069) 68 (5269)
Months of fellowship training, median (IQR) 21.5 (14.429.5) 22.9 (15.431.6)

Definition of abbreviations: APACHE II = Acute Physiology and Chronic Health Evaluation II, ranging from 0 to 71 with higher scores indicating higher
severity of illness; BiPAP = bilevel positive airway pressure; BMI = body mass index; COPD = chronic obstructive pulmonary disease; IQR = interquartile
range; MAP = mean arterial pressure; shock = MAP less than 65 mm Hg or vasopressor use.
Noninvasively measured oxygen saturation at the time of induction was higher in the apneic oxygenation arm (P = 0.03).
*As reported by the fellow performing the intubation.

Patients could receive more than one method of preoxygenation.

Bag-valve-mask ventilation was routinely accompanied by use of a positive end-expiratory pressure valve set to 510 cm H2O.
x
Standard nasal cannula delivered ,6 L/min of nonhumidified oxygen.

150 met no exclusion criteria and no differences in method of time of induction was 99% (IQR,
were enrolled (Figure 1). Patients preoxygenation (Table 1), choice of 96100%) with apneic oxygenation
randomized to receive apneic induction agent or neuromuscular compared with 98% (IQR, 9499%) with
oxygenation (n = 77) and usual care blocker, ventilation between induction usual care (P = 0.03).
(n = 73) were similar at baseline (Table 1). and laryngoscopy, or laryngoscope type,
There was no difference between the except for higher propofol use for Airway Management
two arms in the prior airway management induction in the usual care arm (13.7 vs. Five patients (6.8%) in the usual care arm
experience of the fellow performing 2.5%; P = 0.02) (see Table E1 in the online received apneic oxygenation during
the intubation (Table 1). There were supplement). Oxygen saturation at the intubation, and two patients (2.6%) in the

276 American Journal of Respiratory and Critical Care Medicine Volume 193 Number 3 | February 1 2016
ORIGINAL ARTICLE

Table 2. Study Outcomes

Usual Care Apneic Oxygenation

(n = 73) (n = 77) P Value

Oxygenation outcomes
Lowest oxygen saturation, median (IQR), % 90 (8096) 92 (8499) 0.16
Lowest oxygen saturation ,90%, n (%) 34 (47.2) 34 (44.7) 0.87
Lowest oxygen saturation ,80%,* n (%) 18 (25.0) 12 (15.8) 0.22
Decrease in oxygen saturation, median (IQR), % 4.5 (114) 4.0 (012) 0.60
Decrease in oxygen saturation .3%, n (%) 40 (55.6) 41 (53.9) 0.87
Procedural outcomes
Intubation on the rst laryngoscopy attempt, n (%) 49 (67.1) 52 (67.5) 0.96
Number of laryngoscopy attempts, median (IQR) 1 (12) 1 (11) 0.60
Time from induction to secured airway, median (IQR), s 150 (102245) 132 (88205) 0.31
Clinical outcomes
Duration of mechanical ventilation, median (IQR), d 3 (27) 3 (110) 0.73
Intensive care unit length of stay, median (IQR), d 7 (310) 4 (29) 0.24
Died within 1 h of intubation, n (%) 1 (2.8) 0 (0.0) .0.99
Died before hospital discharge, n (%) 36 (49.3) 27 (35.1) 0.10

Definition of abbreviation: IQR = interquartile range.

There were no differences between the study groups in the primary outcome of lowest arterial oxygen saturation between induction and 2 minutes after
completion of the procedure, secondary procedural outcomes, or clinical outcomes.
*Lowest oxygen saturation ,80% was added to the analysis post hoc.

apneic oxygenation arm did not oxygen in the prior 6 hours, laryngoscope In an a priori dened per-protocol
(Figure 1). There was no difference choice, and operator experience, apneic analysis comparing patients who received
between those randomized to oxygenation did not impact the lowest apneic oxygenation (n = 80) with those
apneic oxygenation and usual care in arterial oxygen saturation during the who did not (n = 68), there was no
the rate of successful intubation on the procedure (see Table E3). difference in lowest arterial oxygen
rst laryngoscopy attempt (67.5 vs. saturation (92% [IQR, 8498%] vs. 90%
67.1%; P = 0.96), time from induction Secondary Analyses [IQR, 8096%], respectively; P = 0.21) or
to secured airway (132 vs. 150 s; P = 0.31), There was no difference in lowest arterial in any other clinical outcome (see Tables
or any other recorded aspect of the oxygen saturation between apneic E4 and E5 and Figure E3).
performance of the procedure (Table 2; see oxygenation and usual care in any of The values for lowest arterial
Table E2). the subgroups examined (Figure 3). oxygen saturation recorded concurrently
Specically, apneic oxygenation was not by independent observers and
Main Outcomes signicantly more effective for patients at the primary investigators were
There was no signicant difference potentially greater risk for hypoxemia strongly correlated (Spearmen R2 = 0.893;
between apneic oxygenation and usual based on higher FIO2 requirement, lower P , 0.001). Operators were frequently
care with regard to the primary outcome oxygen saturation at induction, lower compliant with general best-practices
of median lowest arterial oxygen ratio of oxygen saturation to FIO2 (SpO2/FIO2 in airway management including
saturation during the procedure: 92% (IQR, ratio [25]) in the prior 6 hours, higher body preoxygenation, equipment preparation,
8499%) versus 90% (IQR, 8096%), mass index, more difcult intubation, or end-tidal carbon dioxide detector
respectively (P = 0.16) (Figure 2). Apneic longer duration of laryngoscopy (see availability, and presence of a second
oxygenation did not impact the proportion Figure E1). The type of laryngoscopy operator (see Table E6).
of patients who experienced an oxygen device assigned did not modify the effect
saturation less than 90%, less than 80%, or of apneic oxygenation on lowest arterial
a desaturation greater than 3% during the oxygen saturation (P value for the Discussion
procedure (Table 2). There were no interaction = 0.15) (see Figure E2).
differences in duration of mechanical One patient in each arm of the study This randomized trial comparing
ventilation, ICU length of stay, or was missing a value for lowest arterial apneic oxygenation with usual care during
in-hospital mortality (Table 2). oxygen saturation. In sensitivity analyses endotracheal intubation of critically ill
In multivariable linear regression imputing both values (by carrying adults found that apneic oxygenation did
adjusting for saturation at induction alone forward the saturation at induction or by not increase the lowest arterial oxygen
or with age, body mass index, Acute assigning apneic oxygenation the highest saturation. There were no signicant
Physiology and Chronic Health Evaluation possible saturation and usual care the differences between apneic oxygenation and
II score, shock, prior noninvasive lowest), there remained no difference usual care in any primary or secondary
ventilation, highest fraction of inspired between apneic oxygenation and usual care. outcome, overall or in any subgroup.

Semler, Janz, Lentz, et al.: Apneic Oxygenation for Emergent Intubation 277
 ORIGINAL ARTICLE

A Miguel-Montanes and coworkers (22) delivery in the control arm. Median lowest
100
observed higher oxygen saturation during arterial oxygen saturations for those
Lowest oxygen saturation (%)

intubation after their ICU switched from receiving 60 L/min apneic oxygenation
80
apneic oxygenation at 6 L/min to 60 L/min versus none were identical to our study at
and Wimalasena and coworkers (23) 92 and 90%, respectively.
reported a 6% decrease in the incidence Our study has several strengths. It is
60 of desaturation after their helicopter the rst randomized trial specically
emergency medical service adopted comparing apneic oxygenation with usual
apneic oxygenation at 15 L/min care during intubations outside the
40 by nasal cannula. operating room and is ve times larger than
P = .16 In contrast to prior studies, our trial any prior trial. The primary outcome, lowest
0 showed no difference between apneic arterial oxygen saturation during
Usual Care Apneic Oxygenation
oxygenation and usual care. There are intubation, is of interest to clinicians; has
B several potential explanations for this been used in prior airway management
100 100
discordance. Prior reports of apneic trials; and is linked to patient-centered
Oxygen saturation (%)

oxygenations use outside the operating outcomes, such as cardiac arrest and
80 80 room were before-after designs in which death. Collection of study endpoints
other changes over time may have by an independent observer and
confounded the perceived impact of apneic contemporaneous validation of these data
60 60
oxygenation. Self-reported outcomes in by the primary investigators reduces
prior studies may have predisposed to potential for observer bias. The limited
40 40 observer bias. In contrast to healthy exclusion criteria and relatively small
0 0
patients undergoing elective anesthesia number excluded promote generalizability.
Induction Lowest Induction Lowest (1619) and patients intubated primarily Our study also has limitations. Conduct
Figure 2. Lowest arterial oxygen saturation by for traumatic, hemodynamic, or neurologic in one medical ICU at a single academic
study group. (A) The primary outcome of lowest conditions (22, 23), most patients in our center may limit generalizability. High
arterial oxygen saturation between induction and study were intubated for respiratory failure. compliance with preoxygenation (including
2 minutes after completion of endotracheal For patients with pulmonary function so noninvasive ventilation for patients with
intubation (lowest oxygen saturation) is displayed abnormal that provision of oxygen by hypoxemia), patient positioning, and
for patients randomized to apneic oxygenation mask or noninvasive ventilation was equipment preparation best-practices may
(squares) and usual care (circles). Horizontal bars insufcient to avert intubation, providing have reduced the potential additive
represent median and interquartile range. (B) The 15 L/min by nasal cannula during impact of apneic oxygenation (6, 28, 29).
relationship between oxygen saturation at
intubation might be expected to be Had we used a standardized intubation
induction and lowest oxygen saturation is
displayed for each patient in the usual care (left)
similarly ineffective. Although our protocol (6) or a highly uniform group of
and apneic oxygenation (right) groups. analyses did not suggest efcacy for operators (7), we might have reduced
apneic oxygenation in any subgroup, practice-related variation in lowest arterial
whether apneic oxygenation could be oxygen saturation, making any effect of
Hypoxemia is the most common effective in patients with normal apneic oxygenation easier to detect.
complication of endotracheal intubation pulmonary function being intubated for Comparing apneic oxygenation with a nasal
(13) and the most closely linked to cardiac other reasons requires further study. cannula delivering ambient air (placebo)
arrest and death (5). Preoxygenation is Finally, whether the dose of apneic could have allowed blinding, but would
often insufcient to prevent desaturation oxygenation delivered was adequate have inaccurately represented usual care,
during intubation (3, 11) and the provision is important. Our use of 15 L/min via obscured complications related to
of supplemental oxygen during apnea has high-ow nasal cannula was based on the delivery of the intervention
been advocated as a safe and inexpensive expert recommendation (10) and is a (e.g., disruption of mask seal for
intervention to improve periintubation higher ow rate than prior trials in the bag-valve-mask ventilation by the nasal
oxygenation (10, 26, 27). The use of apneic operating room (1619) but lower than cannula itself), and created a safety
oxygenation has been reported in four the 60 L/min delivered in a recent hazard by providing a false source of
small randomized trials in the operating observational study (22). It seems unlikely oxygen to teams conducting emergent
room (1619) and two before-after that a higher ow rate would improve intubation. Our study was powered to
studies of emergent intubation (22, 23). results, however, based on the recent detect the 5% difference in lowest
Trials of apneic oxygenation during PREOXYFLOW trial (9). Although the arterial oxygen saturation that has been
elective anesthesia ranged in size from PREOXYFLOW trial focused on considered clinically meaningful in prior
12 to 34 patients, all without acute preoxygenation with high-ow nasal trials (6, 9, 17, 18), but a smaller difference
pulmonary dysfunction (1619). cannula versus face mask, saturation at might have been missed. Exclusion of
Provision of 35 L/min of oxygen induction was the same in both arms and patients clinically determined to require
nasally signicantly prolonged the the high-ow nasal cannula continued video laryngoscopy may limit applicability
duration of apnea without desaturation to deliver 60 L/min of oxygen during of our results to patients with abnormal
(1619). Outside the operating room, laryngoscopy compared with no oxygen upper airway anatomy at risk for prolonged

278 American Journal of Respiratory and Critical Care Medicine Volume 193 Number 3 | February 1 2016
ORIGINAL ARTICLE

Favors Apneic Oxygenation sensitive world of emergent endotracheal

intubation and clinicians should focus
their resources on interventions that
Lowest Oxygen Saturation

10
prevent complications (e.g., effective
Mean Difference in

preoxygenation [6]). Future research

should use rigorously designed trials
0 either to identify populations who do
benet from apneic oxygenation or shift
focus to other aspects of airway
management with potential to improve
10 patient outcomes.
Favors Usual Care In summary, the results of this
Patients: 150 94 34 62 86 56 90 112 36 116 32 71 75 clinical trial suggest that apneic

ba < 1 tion
oxygenation during endotracheal

13 ec
5%

BM 50

at 0
BM 30

tu ts
pt

to bat intu n

c
0
in rior %
sa tio LL

se
tio
pr 0.5

p
m

s
0.
95

m
I>
ra > 9

ba
A

Ti to in ficu ba

35
te
intubation of critically ill adults does not

5
I

e
r<
R

Ea att
VE

io

increase lowest arterial oxygen saturation

in
M ne
tio

le
O

lt

n
p
a

sy

n
O

tip

io
ur

tio
6h
6h

compared with usual care. Routine use of

tu

ul
t
sa

f
di
in

tu

tu
e/
apneic oxygenation during emergent
In tion

2
2
n

in
O

t
O
io

intubation cannot be recommended. n

Fi

er
Fi
ct
c
du

od
du

e
m
In

m
M

Ti

Figure 3. Subgroup analyses. The mean difference in lowest arterial oxygen saturation (%)
Author disclosures are available with the text
between apneic oxygenation and usual care is given for patients in prespecified subgroups of this article at www.atsjournals.org.
present at the time of induction (circles) and arising after procedure initiation (squares). Vertical
bars represent the 95% confidence interval around the mean difference. BMI = body mass index in
kg/m2; FIO2 in 6 hours prior = the highest fraction of inspired oxygen in the 6 hours before the Acknowledgment: The authors thank
intubation; time to intubation = time from induction until successful endotracheal intubation. the nurses, respiratory therapists, residents,
and attending physicians of the Vanderbilt
intubation times. Although not observed in The results of our trial suggest that, for Medical Intensive Care Unit for making this
our analyses, benet in specic subgroups patients being intubated in the medical ICU, study possible. Additionally, they appreciate
the critical review of the manuscript
of patients (e.g., severe hypoxic respiratory routine use of apneic oxygenation is safe but provided by Wesley H. Self, M.D.; James R.
failure, preserved pulmonary function) ineffective. Safety without efcacy is Sheller, M.D.; and Arthur P. Wheeler, M.D. at
cannot be excluded. insufcient in the high-stakes, time- Vanderbilt University Medical Center.

References 7. Mort TC. Preoxygenation in critically ill patients requiring

emergency tracheal intubation. Crit Care Med 2005;33:
1. Jaber S, Amraoui J, Lefrant J-Y, Arich C, Cohendy R, Landreau L, 26722675.
Calvet Y, Capdevila X, Mahamat A, Eledjam J-J. Clinical 8. Mort TC, Waberski BH, Clive J. Extending the preoxygenation period
practice and risk factors for immediate complications of from 4 to 8 mins in critically ill patients undergoing emergency
endotracheal intubation in the intensive care unit: a intubation. Crit Care Med 2009;37:6871.
prospective, multiple-center study. Crit Care Med 2006;34: 9. Vourch M, Asfar P, Volteau C, Bachoumas K, Clavieras N, Egreteau P-Y,
23552361. Asehnoune K, Mercat A, Reignier J, Jaber S, et al. High-ow nasal
2. Griesdale DEG, Bosma TL, Kurth T, Isac G, Chittock DR. Complications cannula oxygen during endotracheal intubation in hypoxemic patients:
of endotracheal intubation in the critically ill. Intensive Care Med 2008; a randomized controlled clinical trial. Intensive Care Med 2015;41:
34:18351842. 15381548.
3. Simpson GD, Ross MJ, McKeown DW, Ray DC. Tracheal intubation in 10. Weingart SD, Levitan RM. Preoxygenation and prevention of
the critically ill: a multi-centre national study of practice and desaturation during emergency airway management. Ann Emerg
complications. Br J Anaesth 2012;108:792799. Med 2012;59:16575.e1.
4. De Jong A, Molinari N, Terzi N, Mongardon N, Arnal J-M, Guitton C, 11. Farmery AD, Roe PG. A model to describe the rate of oxyhaemoglobin
Allaouchiche B, Paugam-Burtz C, Constantin J-M, Lefrant J-Y, desaturation during apnoea. Br J Anaesth 1996;76:284291.
et al.; AzuRea Network for the Frida-Rea Study Group. Early 12. Frumin MJ, Epstein RM, Cohen G. Apneic oxygenation in man.
identication of patients at risk for difcult intubation in the Anesthesiology 1959;20:789798.
intensive care unit: development and validation of the MACOCHA 13. Perel A, Berger M, Cotev S. The use of continuous ow of oxygen and
score in a multicenter cohort study. Am J Respir Crit Care Med PEEP during apnea in the diagnosis of brain death. Intensive Care
2013;187:832839. Med 1983;9:2527.
5. Mort TC. The incidence and risk factors for cardiac arrest during 14. Pathak V, Welsby I, Mahmood K, Wahidi M, MacIntyre N, Shofer S.
emergency tracheal intubation: a justication for incorporating the Ventilation and anesthetic approaches for rigid bronchoscopy. Ann
ASA guidelines in the remote location. J Clin Anesth 2004;16: Am Thorac Soc 2014;11:628634.
508516. 15. Lee SC. Improvement of gas exchange by apneic oxygenation with
6. Baillard C, Fosse J-P, Sebbane M, Chanques G, Vincent F, nasal prong during beroptic intubation in fully relaxed patients.
Courouble P, Cohen Y, Eledjam J-J, Adnet F, Jaber S. J Korean Med Sci 1998;13:582586.
Noninvasive ventilation improves preoxygenation before 16. Teller LE, Alexander CM, Frumin MJ, Gross JB. Pharyngeal insufation
intubation of hypoxic patients. Am J Respir Crit Care Med 2006; of oxygen prevents arterial desaturation during apnea.
174:171177. Anesthesiology 1988;69:980982.

Semler, Janz, Lentz, et al.: Apneic Oxygenation for Emergent Intubation 279
 ORIGINAL ARTICLE

17. Taha SK, Siddik-Sayyid SM, El-Khatib MF, Dagher CM, Hakki MA, 23. Wimalasena Y, Burns B, Reid C, Ware S, Habig K. Apneic oxygenation
Baraka AS. Nasopharyngeal oxygen insufation following pre- was associated with decreased desaturation rates during rapid
oxygenation using the four deep breath technique. Anaesthesia sequence intubation by an Australian helicopter emergency medicine
2006;61:42743. service. Ann Emerg Med 2015;65:371376.
18. Baraka AS, Taha SK, Siddik-Sayyid SM, Kanazi GE, El-Khatib MF, 24. Cormack RS, Lehane J. Difcult tracheal intubation in obstetrics.
Dagher CM, Chehade J-MA, Abdallah FW, Hajj RE. Supplementation Anaesthesia 1984;39:11051111.
of pre-oxygenation in morbidly obese patients using nasopharyngeal 25. Rice TW, Wheeler AP, Bernard GR, Hayden DL, Schoenfeld DA, Ware
oxygen insufation. Anaesthesia 2007;62:769773. LB; National Institutes of Health, National Heart, Lung, and Blood
19. Ramachandran SK, Cosnowski A, Shanks A, Turner CR. Apneic Institute ARDS Network. Comparison of the SpO2/FIO2 ratio and the
oxygenation during prolonged laryngoscopy in obese patients: a PaO2/FIO2 ratio in patients with acute lung injury or ARDS. Chest
randomized, controlled trial of nasal oxygen administration. J Clin 2007;132:410417.
Anesth 2010;22:164168. 26. Weingart SD. Preoxygenation, reoxygenation, and delayed sequence
20. Ward JJ. High-ow oxygen administration by nasal cannula for adult intubation in the emergency department. J Emerg Med 2011;40:
and perinatal patients. Respir Care 2013;58:98122. 661667.
21. Weingart S. EMCrit Call/Response Intubation Checklist. EMCrit 27. Bair A. Rapid sequence intubation in adults. In: Post T, editor. UpToDate.
February 2013 [accessed 2015 May 22]. Available from: http://emcrit. Topic 270. Version 16.0. Waltham, MA: UpToDate; 2015.
org/wp-content/uploads/2013/02/emcrit-airway-checklist-2013-02- 28. Jaber S, Jung B, Corne P, Sebbane M, Muller L, Chanques G, Verzilli D,
05.pdf Jonquet O, Eledjam J-J, Lefrant J-Y. An intervention to decrease
22. Miguel-Montanes R, Hajage D, Messika J, Bertrand F, Gaudry S, complications related to endotracheal intubation in the intensive care
Rafat C, Labbe V, Dufour N, Jean-Baptiste S, Bedet A, et al. Use unit: a prospective, multiple-center study. Intensive Care Med 2010;
of high-ow nasal cannula oxygen therapy to prevent 36:248255.
desaturation during tracheal intubation of intensive care patients 29. Flavin K, Hornsby J, Fawcett J, Walker D. Structured airway
with mild-to-moderate hypoxemia. Crit Care Med 2015;43: intervention improves safety of endotracheal intubation in an
574583. intensive care unit. Br J Hosp Med Lond Engl 2012;73:341344.

280 American Journal of Respiratory and Critical Care Medicine Volume 193 Number 3 | February 1 2016