Full text of DOI:10.1007/BF03017423

CARDIOTHORACIC ANESTHESIA, RESPIRATION AND AIRWAY
871
Laryngeal mask cuff inflation at removal does not
affect early postoperative laryngopharyngeal
morbidity
[Un ballonnet encore gonflé lors du retrait du masque laryngé ne modifie pas la
morbidité laryngo-pharyngée postopératoire]
Brian O’ Brien FCARCSI, Dominic Harmon FCARCSI, Michelle Duggan FFARCSI, Noel Flynn MD
Purpose: We assessed the effect of cuff inflation of the laryngeal
mask airway at removal on sore throat, pharyngeal morbidity and
airway complications.
tial. Les épisodes de toux, de haut-le-cœur, laryngospasme et efforts
pour vomir, et les symptômes de maux de gorge et de raucité de la
voix ont été notés par le même observateur.
Methods: In a prospective randomized trial, we used a standard-
ized technique of anesthesia and of laryngeal mask insertion in 126
consecutive day-case patients. Postoperatively, on eye opening, the
masks were removed either inflated (Group A) or deflated (Group
B) and examined for blood by a blinded observer. Episodes of
coughing, gagging, laryngospasm, hiccups and retching, and symp-
toms of sore throat and hoarseness were recorded by the same
observer.
Résultats : Les données démographiques des patients étaient com-
parables. Des traces de sang ont été observées chez 21 % des
patients du Groupe A (n = 63) vs 13 % du Groupe B (n = 63; P =
0,23); l’incidence de mal de gorge a été identique (19 %). On a noté
plus d’enrouement chez les patients du Groupe A (22 % vs 9 %; P =
0,05). L’ensemble des complications n’a pas présenté de différence
intergroupe (19 % vs 11 %; P = 0,21).
Conclusion : Le retrait du masque laryngé dont le ballonnet est
encore gonflé ne réduit pas l’incidence de mal de gorge, de morbidité
pharyngée ou de complications des voies respiratoires.
Results: Demographics were similar. Bloodstaining occurred in
2
1% of patients in Group A (n = 63) vs 13% in Group B (n = 63;
P = 0.23); the incidence of sore throat was identical (19%). Group
A experienced more hoarseness (22% vs 9%; P = 0.05). Overall
airway complications did not differ between groups (19% vs 11%;
P = 0.21).
Conclusion: We conclude that removal of the laryngeal mask air-
way inflated does not reduce the incidence of sore throat, pharyn-
geal morbidity or airway complications.
HE laryngeal mask airway (LMA) is widely
used in anesthesia to maintain a patent air-
way in fasting patients undergoing day-case
T
and routine surgery. During anesthesia the
1
cuff is usually kept inflated; at the conclusion of the
procedure practices differ as some anesthesiologists
deflate it prior to removal while others remove the
device with the cuff inflated. Advocates of the former
approach suggest that the smaller deflated cuff should
cause less trauma to the pharynx, while the latter
approach, it is argued, by bearing out more secretions
should evoke less coughing, and other complications.
Studies have mostly evaluated the effect of initial and
intraoperative cuff volumes on laryngopharyngeal mor-
Objectif : Évaluer l’effet du gonflement du ballonnet du masque
laryngé lors du retrait sur l’incidence de maux de gorge, sur la mor-
bidité pharyngée et sur les complications touchant les voies respira-
toires.
Méthode : Au cours d’un essai prospectif et randomisé, nous avons
utilisé une technique normalisée d’anesthésie et d’insertion du masque
laryngé chez 126 patients consécutifs en chirurgie d’un jour. Après
l’opération, à l’ouverture des yeux, le masque était enlevé, soit en gar-
dant le ballonnet gonflé (Groupe A), soit non gonflé (Groupe B) et
examiné pour la présence possible de sang par un observateur impar-
2
bidity, though it has been established that the removal
of the device in an inflated state removes a greater mass
From the Department of Anesthesia, University College Hospital, Galway, Ireland.
Address correspondence to: Dr. Brian O’Brien, Department of Intensive Care Medicine, The Alfred Hospital, Commercial Rd., Prahran
181, Victoria, Australia. Phone: 613-92763036; Fax: 613-92763780; E-mail: drbobrien@hotmail.com
3
Note: No grants or financial support were received by the authors in the conduct of this study.
Accepted for publication April 11, 2002.
Revision accepted July 10, 2002.
CAN J ANESTH 2002 / 49: 8 / pp 871–873

872
CANADIAN JOURNAL OF ANESTHESIA
of secretions.3 Whether this is clinically significant
remains unclear. Since sore throat following LMA usage
is experienced by about one third of patients in the
Group A patients (n = 63) had their LMAs
removed inflated, and Group B theirs removed deflat-
ed (n = 63). LMAs were removed by the attending
anesthesiologist when the patient was able to open
his/her mouth to command and were inspected by a
blinded observer (the recovery room nurse) for the
presence of blood. All masks were presented inflated
for examination. The same nurse observer then asked
the patient about the symptoms of sore throat and
hoarseness and recorded episodes of hiccup, retching,
laryngospasm and coughing/gagging while in the
recovery room. Nurses were randomly allocated to
patients, as is our usual hospital practice. Continuous
data were analyzed using student’s t test; frequency
data were analyzed using the Chi square test with 95%
confidence intervals (CI). A P value < 0.05 was con-
sidered statistically significant.
4
postanesthesia care unit, it seems appropriate to study
the clinical impact cuff inflation during removal may
have on this complication. The aim of this study then
was to assess the influence of the inflationary state of the
LMA at removal on clinical postoperative pharyngeal
morbidity and airway complications.
Methods
Following local Ethics Committee approval, 126 con-
secutive adult patients undergoing general anesthesia
for minor surgery using the LMA were included. A pri-
ori power analysis determined that a sample size of 54
patients in each group would be able to detect a 25%
absolute difference (from 35% to 10%) between groups
of sore throat following LMA use during general anes-
thesia (α = 0.05), with a power of 80%. Patients were
excluded from the analysis if more than one attempt
was required to insert the LMA, there was history of
recent sore throat or of respiratory infection.
Results
One hundred and twenty-six patients participated in
the study over a nine-month period, from March until
December of 2000. The groups did not differ signifi-
cantly in their demographics, duration of surgery or
smoking behaviour (Table I). Intraoperative airway
complications of hiccupping, coughing or gagging
were seen in 13 (10%) patients. Five (8%) patients in
Group A had intraoperative airway complications (i.e.,
three episodes of hiccupping and two of gagging and
coughing) as did eight (12%) patients in Group B (six
episodes of coughing and gagging, and two of hic-
cups). These differences were not statistically signifi-
cant (P = 0.38).
Postoperative airway complications did not differ
significantly between groups (Group A having 12
cases (19%) vs seven in Group B (11%; P = 0.21, 95%
CI, 0.20–0.04). In this group, there was one postop-
erative laryngospasm - the other 11 events were com-
binations of coughing and gagging. In Group B, six of
the seven postoperative complications were of cough-
ing and gagging, while one was a case of hiccups
(Table II). Thirteen (21%) patients had blood stained
masks in Group B (n = 63) vs eight (13%) patients in
Group A (n = 63; P = 0.23, 95% CI, 0.04–0.20). The
incidence of sore throat was the same in both groups
(19%). There was an increased incidence of hoarseness
in Group A i.e., 14 (22%) vs six (9%), (P = 0.05, 95%
CI, 0.01–0.25).
Premedication was not given, anticholinergic
agents were avoided and anesthetic management was
standardized. Standard monitors of electrocardio-
gram, pulse oximetery, capnography and non-invasive
blood pressure were applied prior to induction of
anesthesia. Patients were preoxygenated for three
minutes. Anesthesia was induced with propofol 2.5
–1
–1
mg·kg and fentanyl 1.5 µg·kg followed by N O
2
and sevoflurane in oxygen. All patients had an LMA
inserted using a standardized approach (with the pos-
terior aspect of the LMA against the hard palate and a
single twisting movement) with the cuff partly inflat-
ed (10 mL in size 3, 20 mL in size 4) and lubricant on
the lateral and posterior aspects. Size 3 masks were
used in women and size 4 masks in men. The masks
were assessed clinically by easy ventilation of the lungs
without a significant leak at 15 cm H O and by a non-
2
obstructed capnography trace. Masks were placed by
two experienced anesthesiologists unaware of patient
group (this was only decided at the conclusion of the
procedure by opening of sealed envelopes). No
attempt was made to ventilate the patients’ lungs by
face mask or to instrument the airway with suction
catheters at any time during the anesthetic. Diclofenac
100 mg pr was administered with the patients’ prior
consent and the wound infiltrated with local anesthet-
ic where appropriate, to provide postoperative analge-
sia. Intraoperative airway complications were
recorded, the intraoperative period being considered
to end when the sevoflurane and nitrous oxide were
turned off at the conclusion of surgery.
Discussion
In the present study we found a low incidence of symp-
tomatic pharyngeal morbidity and airway complications
in a group of 126 patients as would be anticipated. The
incidence of sore throat associated with LMA usage has

O’Brien et al.: LMA REMOVAL
873
TABLE I Patient characteristics and durations of surgical
procedures
more common practice in our institution, although
increasingly the larger sized LMAs are now being
used. For the purposes of the study we chose to use
the more familiar, smaller sized masks, however.
Whilst initially it was recommended that the LMA
be deflated prior to removal, subsequent guidelines
suggested that the device be removed with the cuff
moderately inflated “to aid more complete removal of
Group A
Group B
P value
Age (yr)
41.6 (18–72) 38.7 (18–83) 0.33
Weight (kg)
Duration (min)
Gender (m/f)
Smoker (y/n)
72.9 (15.3)
33 (26.0)
30/33
71 (15.2)
38 (30.3)
26/37
0.49
0.49
0.47
0.72
30/33
28/35
6
salivary secretions”. It is indeed established that a
Values are means (standard deviation or range). Age, weight and
duration were compared using Student’s unpaired t test, and gen-
der and smoking using the Chi squared test.
3
greater quantity of secretions is thus removed; how-
ever it appears from our findings that no clinical ben-
efit accrues form this approach. This is in keeping with
3
the suggestion of Deakin and colleagues that while
TABLE II The incidence of intraoperative and postoperative air-
way complications in Group A (with cuff inflated at removal) and
Group B (deflated at removal)
“
the difference in the mass of secretions removed is
statistically significant, we do not believe (it) is likely
to be clinically significant”. We have found the patient
outcome to be similar and suggest that the decision
whether to remove the LMA inflated or deflated is not
clinically important, as airway and laryngopharyngeal
morbidity are uncommon, and appear to be unaffect-
ed by the decision.
Group A (n = 63) Group B (n = 63)
Intraoperative complications
Gagging/coughing
Hiccupping
Laryngospasm
Retching
5
2
3
0
0
8
6
2
0
0
Recovery room complications 12
7
Gagging/coughing
Hiccupping
Laryngospasm
Retching
Blood on LMA
Sore throat
Hoarseness
11
0
1
0
8
6
0
0
0
13
12
6
References
1
2
3
Morris GN, Marjot R. Laryngeal mask airway perfor-
mance: effect of cuff deflation during anaesthesia. Br J
Anaesth 1996; 76: 456–8.
Asai T, Brimacombe J. Cuff volume and size selection
with the laryngeal mask. Anaesthesia 2000; 55:
12
14*
1179–84.
Deakin CD, Diprose P, Majumdar R, Pulletz M. An
investigation into the quantity of secretions removed
by inflated and deflated laryngeal mask airways.
Anaesthesia 2000; 55: 478–80.
The operative period is taken to end when the delivery of anes-
thetic vapour ceases. Values shown are patient numbers. Statistical
significance (where P = 0.05, in this case) is denoted as *. LMA =
laryngeal mask airway.
4
Brimacombe J, Holyoake L, Keller C, et al. Emergence
characteristics and postoperative laryngopharyngeal
morbidity with the laryngeal mask airway: a compari-
son of high versus low initial cuff volume. Anaesthesia
2000; 55: 338–43.
5
been found to vary widely from 0 to 50%. That of early
postoperative sore throat (i.e., in the recovery room)
has been found to be in the region of 35%; though our
4
5 Keller C, Sparr HJ, Brimacombe JR. Laryngeal mask
lubrication. A comparative study of saline versus 2% lig-
nocaine gel with cuff pressure control. Anaesthesia
1997; 52: 592–602.
power analysis was based on this figure, our observed
rate of sore throat was a little lower. Our results are also
compatible with the low rate of clinical complications
observed in the 100 million or more uses of the device
6 Brimacombe JR, Brain AIJ, Berry AM. The Laryngeal
2
th
worldwide since its introduction. We found no signifi-
Mask Airway Instruction Manual. 4 ed. London:
cant relationship between the presence of air in the cuff
at the time of removal and patient outcome or blood-
staining of the device.
Intavent Research Ltd., 1999.
Our protocol involved the use of LMAs one size
smaller than recommended by a variety of investiga-
2
tors, as discussed in a recent review. However, as
those authors observed, “the size 3 mask seems to be
used frequently in females and the size 4 in males. The
reasons for this practice are not clear”. This remains