The Laryngoscope

V C 2012 The American Laryngological,

Rhinological and Otological Society, Inc.

Laryngeal Mask Airway Versus Endotracheal Tube in Pediatric

Adenotonsillectomy

David I. Sierpina, MD; Hamad Chaudhary, MD; David L. Walner, MD, FAAP; Dana Villines, MA;
Karen Schneider, MD; Marissa Lowenthal, MD; Yuri Aronov, MD

Objectives/Hypothesis: Evaluation of safety and postoperative outcomes of the laryngeal mask airway (LMA) during
pediatric tonsil surgery compared to use of the endotracheal tube (ETT).
Study Design: Randomized controlled trial.
Methods: A population-based sample of 117 patients ages 2 to 18 years requiring adenotonsillectomy, adenoidectomy, or
tonsillectomy was studied. Evaluation forms covering 36 safety, surgery duration, and patient comfort variables were given to
the surgeon, anesthesiologist, and phase I and phase II recovery nurses to collect data on the intra- and postoperative
course. A phone survey was conducted 24 hours after surgery.
Results: At the a level following Bonferroni correction, LMA showed less coughing or gagging during the anesthesia
phase for all surgeries combined (48% for ETT vs. 20% for LMA; v 2 10.153, P .002), and for ETT nontonsillectomy vs.
LMA nontonsillectomy (48% for ETT vs. 3% for LMA; v 2 15.196, P .000), spontaneous ventilation was used more often in
the LMA group when comparing all surgeries (v2 19.493, P .000), and when comparing ETT tonsillectomy and LMA
tonsillectomy (v2 11.131, P .000).
Conclusions: Use of the LMA during pediatric tonsil surgery does not appear to have any major disadvantages
compared to use of the ETT. In fact, analysis of safety, comfort, complications, and postoperative problems suggests that
LMA may be superior for some outcome variables such as coughing and gagging. Use of spontaneous ventilation is more
common among LMA patients, although the significance of this finding is uncertain.
Key Words: Laryngeal mask airway, endotracheal tube, laryngeal mask, airway, pediatric, adenotonsillectomy,
tonsillectomy, adenoidectomy, laryngospasm, recovery.
Level of Evidence: 1b.
Laryngoscope, 122:429-435, 2012

INTRODUCTION
In the United States, the number of ambulatory
surgical visits for tonsillectomies and/or adenoidectomies
is approximately 530,000 per year. 1 As with any surgical
procedure, better outcomes, less pain, quicker recovery,
less need for postoperative pain medication, and a safer,
shorter surgery time are ultimate goals. Use of the rein-
forced laryngeal mask airway (LMA) and endotracheal
intubation (ETT) are both accepted standards of care
used to ensure adequate ventilation during a variety of
surgical procedures.2 However, numerous differences
From the MacNeal Hospital ( D.I.S.), Berwyn, Illinois; and Henry
Ford Clinic (H.C.), Detroit, Michigan, Advocate Health Care ( D.V.), Chicago,
Illinois, Division of Pediatric Otolaryngology ( D.L.W.), Lutheran General
Childrens Hospital (K.S., M.L., Y.A.), Park Ridge, Illinois, U.S.A.
Editors Note: This Manuscript was accepted for publication
October 24, 2011.
Presented as a poster at the American Society of Pediatric Otolar-
yngology Annual Meeting, Chicago, Illinois, U.S.A., April 28-May 1,
2011; and at the Lutheran General Research Consortium, Park Ridge,
Illinois, U.S.A., May 5, 2011.
Thirty LMA Flexible reinforced laryngeal mask airways were pro-
vided by LMA North America (San Diego, CA). The authors have no
other funding, financial relationships, or conflicts of interest to disclose.
Send correspondence to David L. Walner, MD, Lutheran General
Childrens Hospital, Division of Pediatric Otolaryngology, 1675 Dempster
St., 3rd Floor, Park Ridge, IL 60068. E-mail: dwalner@comcast.net
DOI: 10.1002/lary.22458
between these two modalities exist and have been char-
acterized in the literature.
For outpatient procedures, the reinforced LMA
decreases the use of nondepolarizing muscle relaxants
and as a result may avoid possible side effects, such as
prolonged muscle paralysis or the need for reversal
agents such as atropine, glycopyrrolate, or neostigmine.
Other advantages of the LMA include avoidance of stim-
ulation of the larynx and vocal cords, which reduces
cardiovascular and respiratory reflexes, 3,4 obviation of
the need for laryngoscopy, elimination of risks of endo-
bronchial or esophageal intubation, decreased use of
intraoperative fentanyl,5 and improvement in various
postoperative outcome parameters, such as sore
throat,5,6 cough,3,4,6-9 desaturation,4,8,9 bronchospasm,3
postoperative nausea, 5 laryngospasm, 3,6,8,9 pain,7 stri-
dor,3,7,9 and hoarse voice.6 Some studies have also
suggested lower cost,5 reported decreases in intraopera-
tive anesthetic requirements,10 and shown decreased
time required in the operating room, 7 during recovery,5
during intubation,4 and during extubation.9,11
Disadvantages of the LMA have also been reported,
including trouble visualizing the surgical field, as well
as leaking or kinking, which lead to difficulty ventilat-
ing, problems with oxygenation, and high rates of
abandonment of the LMA in favor of the ETT.7,11,12 An

Laryngoscope 122: February 2012 Sierpina et al.: LMA vs. ETT in Pediatric Adenotonsillectomy
429
 TABLE I.
Pain and Safety Comparisons for Endotracheal Tube Cuffed
Versus Endotracheal Tube Uncuffed.
t Test Test
or v2 Statistic P Value
Surgery duration t 0.682
Anesthesia induction to surgery end t 1.033
Anesthesia induction to OR exit t 1.238
Phase I pain t 0.378
Phase II pain t 1.054
24-hour pain t 0.161
Laryngospasm v2 None*
Phase I trouble ventilating v2 None*
from the anesthesia form
Phase I breathing problems v2 0.549
Phase I desaturations v2 3.793
Bonferroni corrected a .006.
*No incidence of target variable.
OR operating room.
increased incidence of laryngospasm has also been
observed.13
Although the literature is replete with examples of
studies comparing the LMA and the ETT, few exist that
focus primarily on their use in pediatric adenotonsillec-
tomy. The goal of this study was to contribute to our
current understanding of the differences between these air-
way devices in general, and to more specifically
understand how the two compare in pediatric adenotonsil-
lectomy with respect to multiple safety and patient comfort
variables. We hypothesized that many of the observations
made in other patient populations will be borne out by this
study, namely that the LMA is a safe and well-tolerated de-
vice that can be used as an alternative to the ETT in
children undergoing adenotonsillectomy.
MATERIALS AND METHODS
After receiving institutional review board approval and
informed consent, a total of 117 pediatric patients from a hospi-
tal and an ambulatory surgery center scheduled for airway
surgery from December 2009 to March 2011 were randomized
by birth date to undergo intubation using either the LMA or
ETT; patients with birth dates in odd-numbered months
received the LMA, and those with birth dates in even-numbered
months received the ETT. Approval was given for patients
between the ages of 2 and 18 years, and all had an American
Society of Anesthesiologists (ASA) score of I or II. Two patients
were excluded from the final analysis. One patient was random-
ized to the LMA group and was found to have an upper
respiratory tract infection (URI) with copious airway secretions
on the day of surgery. The other was randomized to the ETT
group and suffered an episode of laryngospasm prior to tube
insertion. The final analysis included 50 patients in the ETT
group, 27 cuffed ETTs and 23 uncuffed, and 65 patients in the
LMA group.
Anesthesia was provided by attending pediatric anesthesi-
ologists who were with the patients for the entirety of the
surgeries, and an attending pediatric otolaryngologist per-
formed all of the operations. A McIVOR mouth gag was slid in
over the flexible LMA, with Surgilube used occasionally. If the
patient was not ventilating well, adjustments were made after
Laryngoscope 122: February 2012
430
suspending the mouth gag from a Mayo stand. All tonsillecto-
mies were done with coblation, and all adenoidectomies were
done with a microdebrider followed by suction electrocautery for
hemostasis. Adenotonsillectomies were all performed with cobla-
tion alone, and no local anesthetic was used for any case.
A standard anesthesia protocol was used for all patients.
.499
The preoperative protocol included 0.3 mg/kg of midazolam, and
.307 no preoperative Tylenol was given. During the surgery, no ni-
.222 trous oxide was used; only sevoflurane, oxygen, or air was used
.707 for maintenance. Other intraoperative medications included
.297 dexamethasone 0.5 mg/kg to maximum 10 mg, ondansetron 0.1
mg/kg to maximum 4 mg, fentanyl 0.5 to 1 lg/kg, and at least
.872
20 mL/kg of fluids. Patients either underwent tonsillectomy, ad-
enoidectomy, adenotonsillectomy, tonsillectomy with
myringotomy and tube placement, adenotonsillectomy with
myringotomy and tube placement, or adenoidectomy with myr-
.588 ingotomy and tube placement.
.285 Surveys were given to the anesthesiologist, otolaryngolo-
gist, phase I recovery nursing staff, and phase II recovery
nursing staff, and an additional phone survey was completed 24
hours following surgery by phase II nursing staff. Blinding was
limited to nursing staff in recovery phases and to phone sur-
veyors. Data analysis was performed using SPSS 18 (SPSS Inc.,
Chicago, IL). First, to assess the degree of homogeneity between
patients with cuffed versus uncuffed ETTs and the feasibility of
combining these groups for later tests, an independent samples
t test was performed on 10 variables relating to safety and sur-
gery duration (Table I). Next, LMA subjects were compared to
ETT subjects using v2 for dichotomous and categorical data and
analysis of variance (ANOVA) for continuous data. There were
36 variables analyzed describing safety, surgery duration, com-
plications, and patient comfort during, immediately after, and
24 hours after the surgery. Methods of ventilation including
spontaneous, controlled, and assisted were recorded as well.
Finally, to compare intubation methods in groups of patients
with similar surgery types, LMA subjects whose surgery
included tonsillectomy were compared to ETT subjects whose
surgery included tonsillectomy, and LMA subjects whose sur-
gery did not include tonsillectomy were compared to ETT
subjects whose surgery did not include tonsillectomy. The same
36 variables were analyzed using v2 for categorical data and
ANOVA for continuous data (Tables I-V).
The Bonferroni correction for multiple comparisons was
applied to each group of tests. Statistically significant values at a
.05 as well as at the determined Bonferroni corrections are
identified to provide the most stringent tests possible.
RESULTS
Sample demographics are displayed in Table VI by
ETT and LMA surgery groups. Both groups were similar
on all variables reported. The mean age for the ETT sur-
gery group was 5.88 years (SD, 2.33; range, 2-13 years),
and the mean age for the LMA surgery group was 5.02
years (SD, 2.31; range, 2-13 years). The male to female
ratio in both groups was 3:2, as was the ratio of ASA I
to ASA II patients, and the site distribution was approx-
imately 2:1 in both groups. Surgery types were equally
distributed in both groups, with 54% ETT tonsillecto-
mies compared to 52% LMA tonsillectomies, and 46%
ETT nontonsillectomies compared to 48% ETT nontonsil-
lectomies. No adverse outcomes were seen in either
group.
Surgery variables were collected as three groups:
ETT cuffed, ETT uncuffed, and LMA; however, the
Sierpina et al.: LMA vs. ETT in Pediatric Adenotonsillectomy
primary focus of this study was to compare ETT versus
LMA on relevant study variables. To assess the degree
of homogeneity between patients with cuffed versus
uncuffed ETTs and the feasibility of combining these
groups for later tests, 10 variables relating to safety and
surgery duration were compared. There were no statisti-
cally significant differences between the groups (Table
I); therefore, ETT cuffed and ETT uncuffed were com-
bined for the remainder of the analysis.
Next, patient safety and comfort variables as well
as duration variables were compared for ETT versus
LMA during the anesthesia, surgery, immediately after
surgery, and 24 hours after surgery. Analysis was per-
formed for all ETT procedures versus all LMA surgeries,
for ETT tonsillectomy versus LMA tonsillectomy sur-
geries, and for ETT nontonsillectomy versus LMA
nontonsillectomy surgeries. Overall, ETT and LMA were
comparable with few statistically significant differences
(Tables II-V).
When comparing all ETT surgeries to all LMA sur-
geries during surgery, only two of the 11 safety, comfort,
and duration variables were statistically significant at
the Bonferroni correction level or at the .05 level (Table
II). Subjects on whom the ETT was used were found to
cough or gag more during the anesthesia phase than
subjects in whom the LMA was used (48% for ETT vs.
20% for LMA; v2 10.153, P .002). This difference
was also statistically significant for ETT nontonsillec-
tomy versus LMA nontonsillectomy surgeries (48% for
ETT vs. 3% for LMA; v 2 15.196, P .000), but there
was no statistical difference for coughing or gagging for
ETT tonsillectomy versus LMA tonsillectomy surgeries.
The method of ventilation was significantly differ-
ent for all ETT surgeries versus all LMA surgeries (ETT
spontaneous 57%, assisted 16%, controlled 27% vs. LMA
spontaneous 84%, assisted 16%, controlled 0%: v2
19.493, P .000), and for ETT tonsillectomy versus
LMA tonsillectomy (ETT spontaneous 58%, assisted
15%, controlled 27% vs. LMA spontaneous 88%, assisted
12%, controlled 0%: v 2 11.131, P .000) at the Bonfer-
roni correction level. Although ventilation type for ETT
nontonsillectomy versus LMA nontonsillectomy was sig-
nificant at the .05 level (ETT spontaneous 57%, assisted
17%, controlled 26% vs. LMA spontaneous 79%, assisted
21%, controlled 0%: v 2 8.6, P .014), this difference
failed to achieve statistical significance at the Bonferroni
correction level of .006. Additionally, surgery time to
operating room (OR) exit approached significance at the
.05 level for all surgeries combined and for ETT nonton-
sillectomy versus LMA nontonsillectomy surgeries. The
mean differences in surgery time were approximately 90
seconds and 2 minutes longer, respectively, for ETT than
LMA. All three comparison groups were similar for the
other duration variables, the number of placement
attempts and gap adjustments, occurrence of laryngo-
spasm or ventilation trouble, and ease of wake-up.
In general, the three comparison groupings were
very similar for the 10 safety, comfort, and duration var-
iables during phase I recovery (Table III). There were no
statistically significant differences for surgery duration,
coughing, gagging, vomiting, pain, breathing problems,
Laryngoscope 122: February 2012
fentanyl dose, production, desaturation, and hoarse
voice/stridor for all three comparison groups. The differ-
ence between desaturation for the ETT nontonsillectomy
versus LMA nontonsillectomy surgeries approached sig-
nificance when compared to the .05 criteria but failed to
approach the Bonferroni corrected criteria of .002 (v2
5.823, P .054). In the ETT nontonsillectomy group,
82% of the patients had no observed desaturations, 9%
had one observed desaturation, and 9% had two
observed desaturations, whereas no subjects in the LMA
nontonsillectomy group experienced a desaturation
event.
Results for phase II recovery and 24 hours following
surgery were similar to those for phase I recovery
(Tables IV and V). Coughing, hoarse voice, and stridor
immediately after surgery were higher in the ETT non-
tonsillectomy group versus LMA nontonsillectomy group
(coughing: 35% in ETT, 7% in LMA; hoarse voice/stridor:
17% in ETT, 0% in LMA). Although both of these differ-
ences were statistically significant when compared to
the .05 criteria, both failed to approach the Bonferroni
corrected criteria of .002 (coughing: v2 7.024, P .012;
hoarse voice/stridor: v 2 5.823, P .028). No other com-
parisons on phase II recovery variables were statistically
significant: duration, vomiting, pain, breathing prob-
lems, production, bleeding severity, and oral fluid
intake. All 24-hour postsurgery variables were statisti-
cally insignificant for all ETT versus LMA surgeries,
ETT tonsillectomy versus LMA tonsillectomy surgeries,
and ETT nontonsillectomy versus LMA nontonsillectomy
surgeries when applying the Bonferroni correction. One
variable, noisy breathing, was statistically significant for
all ETT versus LMA surgeries when compared to the .05
criteria (12% for ETT vs. 2% for LMA; v2 5.411, P
.042). Hoarseness, coughing, production, gagging, vomit-
ing, pain, swallowing soft foods, and drinking were not
statistically significant at either a criterion.
DISCUSSION
The LMA was introduced into clinical practice in
1988 and was first implemented for adenotonsillectomy
in 1993.2,4 Despite the documentation of numerous
advantages of LMA use in this and other settings,
widespread adoption of the device for pediatric adenoton-
sillectomy has been stalled by concerns over limitation
of surgical access, aspiration, laryngospasm, and
obstruction of the tube caused by kinking, which may
lead to difficulty ventilating and decreased oxygenation.
The results of the current study add to a growing body
of evidence supporting LMA use in general, and to a
small number of reports supporting LMA use in pediat-
ric adenotonsillectomy.
There is impressive agreement in the literature on
the point that postoperatively, the LMA is associated
with increased patient comfort in studies comparing it to
the ETT. This is reflected in multiple independent meas-
urements as outlined above. Decrease in cough is a
common finding in such studies,3,4,5,7-9 as was the case
with our data set. More specifically, we found that
coughing or gagging was significantly increased to the
Sierpina et al.: LMA vs. ETT in Pediatric Adenotonsillectomy
431
 TABLE II.
Comparisons for Endotracheal Tube Versus Laryngeal Mask Airway During Surgery.
All Surgeries Surgeries Surgeries Not
Combined, Test Including Tonsillectomy, Including Tonsillectomy,
t Test or v2 Statistic (P Value) Test Statistic (P Value) Test Statistic (P Value)

No. of placement attempts t 0.159 (.874) 1.278 (.206) 0.857 (.395)

Trouble ventilating v2 1.566 (.504) None* 1.541 (.502)
Ventilation type v2 19.493 (.000) 11.131 (.004) 8.6 (.014)
Laryngospasm v2 1.566 (.504) 1.642 (.498) None*
Coughing or gagging v2 10.153 (.002) 1.028 (.432) 15.196 (.000)
Ease of wake up v2 2.724 (.111) 0.856 (.422) 4.281 (.071)
Surgery duration t .128 (.898) 1.140 (.259) 0.858 (.395)
Anesthesia-surgery end time t 0.412 (.681) 1.272 (.208) 1.114 (.270)
Anesthesia to OR exit t 0.823 (.412) 0.981 (.330) 1.465 (.149)
Surgery end to OR exit t 1.913 (.058) 0.729 (.469) 1.98 (.053)
No. of gag adjustments t 1.744 (.084) 0.385 (.702) 1.751 (.086)
Bonferroni corrected a .006.
*No incidence of target variable.

Significant at Bonferroni corrected a .006; endotracheal tube showed more coughing/gagging and less spontaneous ventilation.

Significant at a .05; endotracheal tube showed increased controlled ventilation.

Bonferroni a level in the anesthesia phase in ETT

patients whose adenoids, but not tonsils, were operated
on. While in recovery and at 24 hours after surgery this
level of significance was not observed. We also found an
increase in cough, hoarse voice, and stridor in phase II
recovery in ETT patients not undergoing tonsillectomy,
and an increase in noisy breathing at 24 hours following
surgery when the ETT was used, comparing all LMA to
all ETT patients. Although these findings were signifi-
cant at the .05 level, they failed to reach significance at
the Bonferroni a level used to correct for multiple com-
parisons. Webster et al. compared the LMA and the ETT
in pediatric adenotonsillectomy and also found an
increase in stridor in ETT patients. 4 In that study, ade-
noidectomy and adenotonsillectomy patients were
grouped together for analysis. A systematic review by
Yu et al. examined 29 randomized control trials (RCTs)
comparing ETT and LMA in general anesthesia and
found an increase in hoarse voice, coughing, and sore
throat associated with use of the ETT.6 These findings
are logical in light of the fact that the LMA avoids con-
tact with the larynx, vocal cords, and trachea, thus
decreasing irritation and trauma to the airway. More-
over, a decrease in symptoms of airway irritation with
the use of the LMA suggests that concerns over laryn-
geal soiling with blood and secretions through leaks in
the seal the LMA forms with the laryngeal inlet may be
unfounded. Peng et al. compared the LMA and the ETT
in 132 children undergoing adenotonsillectomy, and
found soiling of the laryngeal inlet in only one of 48
LMA patients,11 and Webster et al. found no laryngeal
soiling or blood on the laryngeal surface of the LMA in
their series of 55 children.4
Despite various reports in small patient populations
(<150) describing high rates of conversion (8.1%-16.7%)
from LMA to ETT due to obstruction from kinking and
trouble visualizing the surgical field,4,7,11,12 our analysis
of 65 cases of LMA use did not result in a single
 TABLE III.
Comparisons for Endotracheal Tube Versus Laryngeal Mask Airway During Phase I Recovery.
Surgeries Including Surgeries Not
All Surgeries Combined, Tonsillectomy, Test Including Tonsillectomy,
2
t Test or v Test Statistic (P Value) Statistic (P Value) Test Statistic (P Value)

Duration t 0.501 (.617) 0.320 (.750) 0.385 (.702)

Coughing v2 3.550 (.169) 2.209 (.331) 1.789 (.409)
Gagging v2 1.385 (.239) 1.354 (.392) None*
Vomiting None* None* None*
Pain score t 0.470 (.639) 0.793 (.432) 1.740 (.880)
Breathing problems v2 .810 (.510) 0.789 (.448) 0.111 (1.000)
Fentanyl doses v2 2.829 (.587) 2.931 (.402) 5.775 (.217)
Production v2 1.006 (.800) 1.466 (.690) 2.878 (.237)
Desaturation v2 1.484 (.686) 1.840 (.606) 5.823 (.054)
Hoarse voice/stridor v2 4.238 (.120) 1.880 (.391) 2.523 (.283)
Bonferroni corrected a .002.
*No incidence of target variable.

Laryngoscope 122: February 2012 Sierpina et al.: LMA vs. ETT in Pediatric Adenotonsillectomy
432
 TABLE IV.
Comparisons for Endotracheal Tube Versus Laryngeal Mask Airway During Phase II Recovery.
All Surgeries Surgeries Surgeries Not
Combined, Test Including Tonsillectomy, Including Tonsillectomy,
t Test or v2 Statistic (P Value) Test Statistic (P Value) Test Statistic (P Value)

Duration t 0.296 (.768) 0.598 (.552) 0.024 (.981)

Coughing v2 1.952 (.162) 0.339 (.740) 7.024 (.012)*
Vomiting v2 1.217 (.327) 0.547 (.647) 0.683 (.640)
Pain score t 0.733 (.465) 0.263 (.794) 1.709 (.093)
Breathing problems None None None
2
Production v 0.647 (.412) 0.807 (1.000) 2.799 (.177)
Bleeding severity v2 1.942 (.379) 0.644 (.623) 1.460 (.482)
Oral fluid intake v2 0.561 (.755) 0.340 (.844) 0.888 (.642)
Hoarse voice/stridor v2 2.932 (.124) 0.123 (1.000) 5.823 (.028)*
Bonferroni corrected a .002.
*Significant at a .05; endotracheal tube subjects experienced more coughing, hoarse voice, and stridor than laryngeal mask airway subjects.

No incidence of target variable.

instance of conversion, although there were two cases of
difficult ventilation that resolved with gag adjustment.
Our findings are in agreement with larger studies such
as that performed by Gravningsbra ten et al., in which
1,126 children underwent adenotonsillectomy with an
LMA in place, and conversion occurred in only seven
patients (0.6%) due to six cases of air leak around the
LMA and one case of atelectasis. 14 Joshi et al. similarly
showed a 1% conversion rate in 341 adult outpatients
but did not report reasons for LMA abandonment, 5 and
Verghese and Brimacombe reported a 0.19% conversion
rate in 11,910 adults and children due to leaks and
placement failure.15 These numbers compare to Rose
and Cohens report of 0.3% ETT failure in a study of
18,500 patients, and in that same study an LMA failure
rate of 4.7% is given for 634 patients, with reasons for
failure not reported.16 Considering our experience and
taking the results of prior studies into account with
greater weight on studies with larger patient popula-
tions, it is reasonable to conclude that high rates of
conversion associated with kinking may be a result of
technique rather than the fault of the LMA per se. Web-
ster et al. pointed out that in their series the LMA was
abandoned in five patients, all of which occurred during
the first 15 patients of the trial. Three of these patients
showed signs of obstruction on opening of the Boyle-
Davis gag, a problem that was ameliorated later in the
study with increased experience and anesthesia use.
This led the authors to conclude that the likeliest cause
of airway obstruction on opening the B-D gag with the
LMA was inadequate anaesthesia with reflex laryngeal
closure.4 The other conversions occurred due to a leak
around the LMA, and due to bronchospasm that was
likely related to a recent URI.
Reports regarding conversion due to difficulty with
surgical field visualization are similarly disparate, with an
apparently equal number of studies in both adults and
children reporting no difficulty with access3,4,14 and access
troubles extreme enough to require abandonment of the LMA
in favor of the ETT.6,11,12 Additionally, Hern et al. found that
the weight of excised tonsillar tissue was less in patients
where the LMA was used. In our experience, use of the
LMA with a McIVOR gag offered an excellent view of the
surgical field.
It has been noted in prior studies that LMA use is
associated with an increase in spontaneous ventilation
 TABLE V.
Comparisons for Endotracheal Tube Versus Laryngeal Mask Airway 24 Hours After Surgery.
All Surgeries Surgeries Including Surgeries Not
Combined, Test Tonsillectomy, Test Including Tonsillectomy,
2
T Test or v Statistic (P Value) Statistic (P Value) Test Statistic (P Value)

Hoarseness v2 1.143 (.298) 0.108 (.786) 1.563 (.322)

Coughing v2 1.643 (.440) 1.707 (.426) 0.541 (.560)
Production v2 0.125 (1.000) 1.300 (.371) 0.753 (.569)
Gagging v2 0.001 (1.000) 0.040 (1.000) 0.047 (1.000)
Vomiting v2 0.125 (1.000) 2.054 (.214) 2.799 (.177)
Pain t 0.936 (.352) 0.470 (.640) 1.324 (.191)
Noisy breathing v2 5.411 (.042)* 2.820 (.161) 2.799 (.177)
Swallowing soft foods v2 0.210 (.900) 0.321 (.852) 0.000 (1.000)
Drinking v2 0.281 (.869) 0.164 (.921) 1.983 (.371)
Bonferroni corrected a .002.
*Significant at a .05; endotracheal tube subjects experienced more noisy breathing than laryngeal mask airway subjects.

Laryngoscope 122: February 2012 Sierpina et al.: LMA vs. ETT in Pediatric Adenotonsillectomy
 TABLE VI.
Sample Characteristics.
ETT, n 50 LMA, n 65
Age, yr, mean (SD) 5.88 (2.33) 5.02 (2.31)
Gender, no. (%)
Male 30 (60) 38 (59)
Female 20 (40) 27 (41)
Site, no. (%)
Hospital surgery center 33 (66) 44 (67)
Ambulatory surgery center 17 (34) 21 (32)
Surgeries including tonsillectomy 27 (54) 34 (52)
Surgeries not including tonsillectomy 23 (46) 31 (48)
ASA, no. (%)
I 30 (60) 38 (59)
II 20 (40) 27 (41)
ETT endotracheal tube, LMA laryngeal mask airway; SD
standard deviation; ASA American Society of Anesthesiologists.
compared to ETT use.4,17 Webster et al. reported that in
addition to decreased use of assisted ventilation, when
assisted ventilation was used in LMA patients it was of
shorter duration.4 In our analysis, this effect reached
significance at the Bonferroni a level of .006 for all
surgeries combined and for surgeries including tonsillec-
tomy, and reached significance at the a .05 level for
surgeries not including tonsillectomy. Our anesthesiolo-
gists routinely use assisted ventilation in patients with
an ETT to overcome the resistance of the tube. Further-
more, assisted ventilation is often avoided in LMA
patients due to concerns over gastric insufflation, aspira-
tion, and nausea, although Verghese and Brimacombe
found that there was no significant reduction in critical
incidents in LMA patients when spontaneous ventilation
was used, including regurgitation, vomiting, aspiration
of gastric contents, bronchospasm, laryngospasm, gastric
dilation, hypotension, arrhythmias, or arrest. 15 Sponta-
neous ventilation has not been shown to decrease
anesthetic requirements in these patients, 18 but it is
possible that emergence from anesthesia is expedited in
patients breathing on their own compared to those de-
pendent on the ventilator. Beyond this observation, the
clinical significance of increased use of spontaneous ven-
tilation is uncertain.
In our series, two patients in the LMA group expe-
rienced brief episodes of laryngospasm on removal of the
LMA that were easily controlled using face mask venti-
lation for <1 minute with no sequelae. There were no
episodes of laryngospasm in the ETT group. However,
this difference failed to reach statistical significance.
Most RCTs conducted to date comparing the ETT to the
LMA in adenotonsillectomy or for use in other surgery
types have shown either an increased incidence of lar-
yngospasm when ETTs are used,3,6,8,9 or no difference
was detected.4,11,12 Moreover, the systematic review by
Yu et al. outlined above included seven studies that
dealt with this topic. Their analysis showed an increased
incidence of laryngospasm in surgeries where the ETT
was used compared to those in which the LMA was
used. However, a 2:1 case-control study including 130
Laryngoscope 122: February 2012
434
cases of laryngospasm in children under 18 years old
conducted by Flick et al. identified LMA use as an inde-
pendent risk factor for laryngospasm in general
anesthesia.13 In balance, and taking our experience into
consideration, it appears that the weight of the evidence
suggests against an increased risk of laryngospasm with
use of the LMA.
Given the current prevalence of pediatric adenoton-
sillectomy,1 any decrease in the cost of this procedure
would represent a large total cost savings. To this end,
the effects of LMA use on time spent in the OR and in
recovery have been investigated by this and other
reports. Doksrd et al. showed in a population of 134
children undergoing adenotonsillectomy that the
patients in the LMA group spent significantly less time
in the OR after surgery (4.2 minutes), but total OR time
failed to reach significance. 7 The authors pointed out
that in their study patients were moved to recovery with
the LMA still in place. This finding was replicated by
Webster et al. in their 1999 study on intranasal surgery
in adults and children, which reported a decreased time
from discontinuation of anesthesia to OR exit in LMA
patients compared to ETT patients who were extubated
awake when the LMA was removed in recovery. 9 In their
recent study of 131 children undergoing adenotonsillec-
tomy, Peng et al. also found a decreased time between
surgery end and extubation in the LMA group, but
found no significant difference in total anesthesia time,
in their judgment owing to either time devoted to LMA
tube adjustments or to fiberoptic laryngoscopy performed
in the LMA group.11 Joshi et al. failed to note a signifi-
cant difference between the LMA and the ETT with
respect to time from surgery end to device removal in a
study of 381 adult patients undergoing elective ambula-
tory surgery.5 Our findings are in agreement with Joshi
et al., although the time between surgery end and OR
exit was shorter in the LMA group and approached sig-
nificance for all surgeries combined. This result may
have reached significance had we brought patients to re-
covery with the LMA in place like Doksrd and Webster.
However, whether or not this would have translated to a
significant reduction in OR time and ultimately cost sav-
ings is doubtful considering that an actual reduction in
OR time has not been shown in any study to date.
Our study also failed to demonstrate significant
reductions in time spent in recovery, which agrees with
Peng et al.11 Although Joshi did report less time in the
PACU and a decreased time to ambulation, this did not
translate to home readiness.5 The differences between
these results may be because a standardized anesthesia
protocol was put in place in our study and that of Peng
et al., whereas Joshi et al. used no standardization, lead-
ing to an increase in fentanyl usage in the ETT group.
Our investigation was limited in several ways.
First, our findings cannot be applied to patients with
abnormal airways, because all of our enrolled subjects
were ASA category I or II. Second, the homogeneity of
the groups under study was limited. Various surgery
types were included, which we attempted to address by
categorizing surgeries involving tonsillectomy and those
not involving tonsillectomy. In addition, both cuffed and
Sierpina et al.: LMA vs. ETT in Pediatric Adenotonsillectomy