The endotracheal tube (ETT) has traditionally been considered to be the safest device in preventing aspiration and to ensure free airway and unobstructed ventilation in adenotonsillectomies.1–3 The laryngeal mask airway (LMA) was introduced in 1983 as an alternative to the endotracheal tube (ETT).4 In 1990, the reinforced laryngeal mask (RLMA) was presented, intended for use during surgical procedures involving the mouth and throat.5 Although not providing an airway seal within the trachea, the RLMA has several potential advantages over ETT.1 It is usually inserted blindly, without use of the laryngoscope. This leads to reduced oropharyngeal or laryngeal stimulation and injury6 and, therefore, minimal activation of respiratory and circulatory reflexes.7 RLMA is always placed without the use of muscle relaxants, also avoiding need of reversal agents and potential side-effects of these drugs. In addition, the RLMA can remain in the throat until the patient is fully emerged, thus reducing the danger of postoperative aspiration.8 In a study of Boisson-Bertrand9 of 100 children the RLMA was associated with less airway problems compared with the ETT. The potential problems with RLMA may be dislocation during the procedure, risk of ventilation into oesophagus and ventricle, and increased risk of aspiration into the airways during the procedure. For adenotonsillectomies in particular, the LMA has also been associated with reduced surgical access to the lower parts of the tonsils.
The goals of our study were to examine whether the use of RLMA resulted in less overall pain, nausea and respiratory irritation compared with ETT for adenotonsillectomies. We were also interested in whether RLMA would reduce the time spent in the operating room after surgery.
The protocol for this randomised single-centre study was approved by the Regional Ethics Committee in Southern Norway (REK SØR, Blindern, NO-0318 Oslo, ID:S-07031b) and reported to the Norwegian Social Science Data Services (NSD) and www.clinicaltrials.gov (NCT-00570414). Written information and consent forms were sent in advance to the children's parents and guardians. Children over the age of 12 also received their own written information about the study.
One hundred and thirty-four children between the ages of 3 and 16 years due for adenoidectomy and/or tonsillectomy at Telemark Hospital, Porsgrunn, were included and randomised to either RLMA (n = 69) or ETT (n = 62) between April 2007 and June 2008.
Randomisation was done according to instructions in sequentially numbered sealed envelopes, based on a computer-generated list of random numbers. The first or second author enrolled the patients after written consent forms had been obtained.
Exclusion criteria were ASA class 3 or higher, weight less than 10 kg or more than 60 kg and congenital deformities of the mouth or throat. Demographic data, risk factors for postoperative nausea and vomiting (PONV) and preexisting obstructive lung disease were all registered prior to surgery. The patients and their parents or guardians were instructed in the use of the Faces Pain Scale/VAS10 0–10, where 0 is no pain and 10 is worst pain imaginable.
The patients were premedicated with midazolam mixture (0.4 mg kg−1, maximum 15 mg) and application of a local anaesthesia cream with occlusion (EMLA; Astra Zeneca AS, Oslo, Norway) on the dorsum of both hands. Anaesthesia was induced through an intravenous (i.v.) line with running saline-glucose (Salidex) solution and bolus doses of alfentanil 20 μg kg−1 and propofol 3–5 mg kg−1. In case of problems with establishing an i.v. line, an alternative induction technique with sevoflurane mask induction and 50% nitrous oxide in oxygen was used. The anaesthesia was maintained with sevoflurane 1–2% (end-tidal) and N2O/O2 2: 1 in both groups. Alfentanil 5–10 μg kg−1 i.v. was administered every 8–10 min, or when there were signs of clinical stress (i.e movement, high BP or heart rate). Patients were given a single dose of dexamethasone 0.15 mg kg−1 i.v. along with paracetamol 40 mg kg−1 rectally. During the final phase of the surgical procedure, lidocain 1 mg kg−1 i.v. and morphine 100–200 μg kg−1 i.v. were administered.
Both groups were mechanically normoventilated in pressure-controlled mode11 with Datex Ohmeda S/5 anaesthesia apparatus. Monitoring of each patient included ECG, SpO2, end-tidal CO2 and noninvasive blood pressure.
In Group RLMA a laryngeal mask (LMA-Flexible, The Laryngeal Mask Company UK, Ltd) of suitable size (#2 <25 kg, #2.5 25–30 kg, #3 >30 kg) was blindly inserted into the pharynx, guided by the anaesthesiologist's fingers. If there was leakage the RLMA was reinserted (a second time) or it was replaced with one of larger size. Placement was determined to be correct if ventilation was easily possible and if measurement of leakage in the patient system was less than 40 ml/min, and less than 20% of the tidal volume. Russel Davis gag or Boyle Davis gag were used according to the discretion of the surgeon. Care was taken to secure the airway during gag insertion. If there was acceptable ventilation and the operator had a good view, the surgery could begin.
In Group ETT the patients were intubated with a Mallinckrodt Hi-contour cuffed tube. The endotracheal tube was selected from the following formula: Tube size = [(age/4) + 3.5)]. The surgeon placed the tongue blade when sufficient anaesthetic depth had been achieved.
In Group RLMA the children were moved to the recovery unit next door to the operating room with the RLMA in place as soon as they were spontaneously breathing with a frequency of more than 10 min−1. They had an oxygen mask with 6–9 l flow placed adjacent to the proximal opening of the RLMA, which was removed by the recovery staff when the child was fully awake and not tolerating the RLMA anymore. In Group ETT the children were extubated in the operating room when they were spontaneously breathing but still in anaesthesia, and then transferred to the recovery unit with 6–9 l O2 through an open mask when a free airway was established and the respiratory rate and SpO2 was more than 10 l min−1 and 95%, respectively.
The patients were monitored with pulse-oxymetry in the recovery unit. Morphine 50 μg kg−1 i.v. was administered if Faces Pain Scale/VAS was at least 3. PONV was registered on a scale from 0 to 3 (0 = none, 3 = severe). Ondansetron 0.1 mg/kg was given in cases of retching or vomiting. When the child was awake, had satisfactory Spo2 (>95%) without O2 administration, there was no bleeding problem and pain was stable VAS = 0–2, the child was allowed to drink clear fluids and the i.v. line was removed. Patients stayed in the recovery unit around 4 h routineously before discharge.
All types of perioperative complications, including airway irritation (i.e. coughing or stridor) and any oxygen saturation lower than 92% were registered.
Prior to discharge, the parents or guardians were informed about the use and dosage of paracetamol for the next 24 h (25 mg kg−1 × 4 supp). They were also provided with written instructions and a hospital telephone number. The following day, a telephone interview was carried out with the patient and/or parents/guardians, including questions about pain and nausea. The communication also addressed particular problems, such as bleeding or difficulty with the intake of fluids. After the study period, we examined the patient files to check out whether the patients had been in contact with any hospital for postoperative problems.
The primary endpoint of this study was postoperative pain, characterized by VAS and rescue analgesic doses. The secondary endpoints were postoperative nausea, incidence of anaesthesiological complications and procedure duration.
The sample size was calculated based on our own previous experience of a 75% incidence of need for postoperative opioids after use of endotracheal tube. We thought that a reduction in opioid need to 50% after RLMA would be of interest to reveal. With 80% statistical power and a level of statistical significance of 0.05, 60 patients would be needed in each group.
Intention to treat was used as the basic principle, and all patients randomised to RLMA were primarily analysed in this group. Any conversions to ETT were treated as side-effect. In order to examine whether ETT per se resulted in airway irritation, a secondary analyses was also done on those patients who received ETT regardless of initial group allocation. All continuous variables were tested for normal distribution with the Kolmogorov–Smirnov test. If normally distributed the T-test was used, if not Mann–Whitney U test. The Fisher's Exact test was used for comparison of dichotomous variables. Comparison of categorical variables with greater than two categories was performed with the chi-squared test. All data were coded and entered into an Excel arithmetical sheet (Microsoft, Redmond, Washington, USA) and processed in the SPSS statistical software version 15.0 (SPSS Inc, Chicago, Illinois, USA).
One hundred and seventy patients were considered for inclusion in this study. One hundred and thirty-four patients were randomised and followed up according to protocol throughout the planned perioperative hospital period, one was lost for follow-up after discharge (Fig. 1). The demographic data, peroperative vital signs and peroperative medication in the two groups were similar (Table 1). In five RLMA patients the airway management was converted to ETT, all due to inferior surgical access. These patients were primarily analysed with the RLMA group for all data, according to the intention to treat principle.
There was significantly less maximal pain in Group RLMA during the first 4 h postoperative in the recovery unit (P = 0.015), whereas there was a tendency (ns) also of lower pain scores at recovery admission and at 24 h postoperatively. However, there were no significant differences in need of rescue opioid medication between the two groups (Table 2).
The incidence of nausea was high and similar in both groups (Table 3), 73 (56%) of the children complained of nausea during their stay in the recovery unit and 56 (43%) children vomited.
There was a nonsignificant tendency towards lower incidence of perioperative airway irritations (i.e. coughing or stridor) in Group RLMA (4 vs. 8, P = 0.15). When analysed per protocol two of the patients with RLMA during surgery had airway irritations compared with 10 (including two out of five patients with conversion from RLMA to ETT) of those with ETT (P < 0.02). There was one case of bronchospasm in Group ETT after intubation with the lowest SpO2 at 65%. This patient was given aminophylline and racemic adrenaline, and the operation was cancelled. The child was transferred to the paediatric ward for observation, but discharged the same evening (Table 4). The surgical procedures and duration of surgery were similar between the two groups, but the ETT patients spent significantly more time in the operating room after end of surgery, mean difference of 4.2 min, (Table 4).
Three patients (2.3%) were readmitted and operated on the same day due to bleeding; one from Group ETT and two from Group RLMA. The patient from Group ETT was reoperated a second time on day 2 due to bleeding. Another three children were in contact with the hospital due to concerns about bleeding (one from Group ETT and two from Group RLMA), but these were only under observation for a few hours and were not reoperated. Two children in Group ETT had contact with the hospital the following day due to pain. They were given a prescription for another medication (NSAID) and sent home.
Three children were in contact with the hospital with concerns about dehydration. Two were from Group RLMA and one from Group ETT. One of the children from Group RLMA was admitted for 4 days and needed i.v. fluid infusion. The others were able to return home after a visit to the outpatient clinic. The data from the routine postoperative telephone interview on day one showed similar results of pain control, rescue medication and side-effects between the two groups (data not shown).
The Group RLMA scored significantly lower for maximal pain during the first 4 h postoperatively (P = 0.015). Hohlrieder et al.12 found a similar difference between proSeal LMA and ETT for a population of women undergoing gynaecological laparoscopic surgery. One reason for less pain being reported in the RLMA group may be that intubation in the ETT group in itself leads to more trauma and postoperative pain. Another explanation is that the laryngeal mask allows the child to sleep through the first postoperative phase because of less irritation in the throat. Whereas crying and restlessness may be associated with ETT and extubation, the RLMA removal is more delayed and smooth with potentially less stress in the wound area. Although the tendency toward less pain after RLMA, measured with VAS, remained throughout the postoperative period in the hospital and until next day, the difference was not significant (P = 0.084). Also, the amount of morphine administered to the two groups was similar. This is somewhat surprising, as the staff in the postoperative ward was instructed to administer opioids according to the VAS. Still, there is always a delay from perceived pain to rescue opioid administration is given which may be perceived differently on the VAS. Also the nurses are used to, and may be biased to, give opioids to some of these patients anyway without using the VAS criteria strictly. Further, as surgery is done in the same area as the trauma of airway device, it may be that subtle differences in pain due to different airway device are concealed by the stronger pain from surgery.
We registered a high incidence of nausea in the study population, with no significant differences between the two groups. Several factors may contribute to the high incidence of nausea. The surgical procedure in itself is associated with PONV13,14 and the use of morphine13 and nitrous oxide15 can further increase the incidence. The RLMA may theoretically lead to less risk of blood and mucus aspiration into the stomach, as it obstructs the entrance of the oesophagus. With ETT no such obstruction is provided, and it may be speculated that more blood and mucus will distend or irritate the stomach, increasing the incidence of PONV. Still, Anderson et al.13 has previously shown in a retrospective study, with the aid of a logistic regression model, that LMA was similar to ETT in terms of incidence of PONV after surgery and morphine. Although we used dexamethasone as routine nausea prophylaxis in all patients, we think the incidence of 56% nausea still is too high, and have now added ondansetron nausea prophylaxis to our routine care in these patients.
As in previous studies,1,6,7,9,16–20 our data demonstrates that the use of a laryngeal mask airway is well tolerated during adenotonsillectomies in terms of perioperative respiratory complications. In fact, our material indicates a tendency toward fewer respiratory complications in the RLMA group compared with the ETT group when analysed as intention to treat. However, five of the patients in the RLMA group (7.2%) were converted to ETT. In all five cases this was due to reduced surgical access. In four of the cases the problem was due to inferior view of the lower tonsillar pole, and in one case there were difficulties performing the adenoidectomy with the RLMA in place. When these are pooled with the other ETT patients, there is a significant higher incidence of airway irritations (i.e. <0.02). This because two of the RLMA group patients with respiratory complications had complications after conversion to endotracheal intubation.
We provided a prestudy training period of 1 year where surgeons, anaesthesiologists and anaesthesiology nurses were made familiar with the RLMA method. Five surgeons and nine anaesthesiologists, all with different levels of experience, participated. The percentage of completion rate for RLMA during operations was 92.8%, which indicates that this can also be considered a reliable method in smaller hospitals. Still, there is a potential for even higher completion rate as shown by Gravningsbraaten et al.20 In their study of 1126 patients in an experienced single surgeon practice, only 0.5% of the LMAs were converted to ETT.
The total time spent in the operating room was 3.7 min shorter in Group RLMA (P = 0.057). The surgical duration was similar in the two groups suggesting that surgical accessibility and ease were not reduced by the RLMA. However, in the RLMA group the patient could be transported into postoperative ward as soon as adequate self-respiration had been achieved, thus significantly reducing the time spent in the operating room with mean 4.2 min after end of surgery. In the ETT group it was necessary to ensure that the patient was extubated and then had a free airway before transport. There were no reports on problems with observation and removal of the RLMA by the nurses in the recovery. Thus, the shorter operating room time may result in an overall more effective service, as delays in the operating room are far more expensive than the potential extra work spent on removing the RLMA in the postoperative ward. A prerequisite of such a practice is to have continuous surveillance and easy access to personnel and equipment in postoperative ward to intervene in case of very rare airway complications after LMA removal.21 Three patients were admitted (2.3%) and reoperated, two from the RLMA group and one from the ETT group. Postoperative bleeding has recently been associated with the use of corticosteroids for pain and nausea prophylaxis.22 In that study an incidence of reoperation of 5.0% after corticosteroid medication vs. none in the placebo group was found. Still, our incidence is lower and quite comparable with other studies not using corticosteroids.23
We conclude that the RLMA technique, when feasible, is a well tolerated and effective alternative to the ETT for use during adenotonsillectomies in children, with beneficial effects on respiratory complications, operating room efficiency and early postoperative pain.
This work was supported by the Departement of Emergency Medicine and the Departement of Science and Development, Telemark Hospital, Skien, Norway.
There are no conflicts of interest.
1 Webster AC, Morley-Forster PK, Dain S, et al
. Anaesthesia for adenotonsillectomy: a comparison between tracheal intubation
and the armoured laryngeal mask
airway. Can J Anaesth 1993; 40:1171–1177.
2 Hern JD, Jayaraj SM, Sidhu VS, et al
. The laryngeal mask
airway in tonsillectomy
: the surgeon's perspective. Clin Otolaryngol Allied Sci 1999; 24:122–125.
3 Bishop P, Patel A. Laryngeal mask
or tube for tonsillectomy
? Hosp Med 2003; 64:440.
4 Brain AIJ. The laryngeal mask
airway-a new concept in airway management. In: 1983: 801–806.
5 Alexander CA. A modified Intavent laryngeal mask
for ENT and dental anaesthesia. Anaesthesia 1990; 45:892–893.
6 Heumann H, Reimann B, Stelzner J, et al
. Using the laryngeal mask
in adenotomy and tonsillectomy
. HNO 2001; 49:664–669.
7 Aziz L, Bashir K. Comparison of armoured laryngeal mask
airway with endotracheal tube for adenotonsillectomy. J Coll Physicians Surg Pak 2006; 16:685–688.
8 John RE, Hill S, Hughes TJ. Airway protection by the laryngeal mask
. A barrier to dye placed in the pharynx. Anaesthesia 1991; 46:366–367.
9 Boisson-Bertrand D. Tonsillectomies and the reinforced laryngeal mask
. Can J Anaesth 1995; 42:857–861.
10 Bieri D, Reeve RA, Champion GD, et al
. The Faces Pain Scale for the self-assessment of the severity of pain experienced by children
: development, initial validation, and preliminary investigation for ratio scale properties. Pain 1990; 41:139–150.
11 Bordes M, Semjen F, Degryse C, et al
. Pressure-controlled ventilation is superior to volume-controlled ventilation with a laryngeal mask
airway in children
. Acta Anaesthesiol Scand 2007; 51:82–85.
12 Hohlrieder M, Brimacombe J, Eschertzhuber S, et al
. A study of airway management using the ProSeal LMA laryngeal mask
airway compared with the tracheal tube on postoperative analgesia requirements following gynaecological laparoscopic surgery. Anaesthesia 2007; 62:913–918.
13 Anderson BJ, Pearce S, McGann JE, et al
. Investigations using logistic regression models on the effect of the LMA on morphine induced vomiting after tonsillectomy
. Paediatr Anaesth 2000; 10:633–638.
14 Furst SR, Rodarte A. Prophylactic antiemetic treatment with ondansetron in children
. Anesthesiology 1994; 81:799–803.
15 Apfel CC, Korttila K, Abdalla M, et al
. A factorial trial of six interventions for the prevention of postoperative nausea and vomiting. N Engl J Med 2004; 350:2441–2451.
16 Williams PJ, Bailey PM. Comparison of the reinforced laryngeal mask
airway and tracheal intubation
for adenotonsillectomy. Br J Anaesth 1993; 70:30–33.
17 Wehrle HJ, Gottstein P. Experiences with use of the laryngeal mask
with flexible, wire reinforced tube for ENT interventions in childhood. Anasthesiol Intensivmed Notfallmed Schmerzther 1997; 32:151–154.
18 Kretz FJ, Reimann B, Stelzner J, et al
. The laryngeal mask
in pediatric adenotonsillectomy. A meta-analysis of medical studies. Anaesthetist 2000; 49:706–712.
19 Heath ML, Sinnathamby SW. The reinforced laryngeal mask
airway for adenotonsillectomy. Br J Anaesth 1994; 72:728–729.
20 Gravningsbraten R, Nicklasson B, Raeder J. Safety
of laryngeal mask
airway and short-stay practice in office-based adenotonsillectomy. Acta Anaesthesiol Scand 2009; 53:218–222.
21 Pappas AL, Sukhani R, Lurie J, et al
. Severity of airway hyperreactivity associated with laryngeal mask
airway removal: correlation with volatile anesthetic choice and depth of anesthesia. J Clin Anesth 2001; 13:498–503.
22 Czarnetzki C, Elia N, Lysakowski C, et al
. Dexamethasone and risk of nausea and vomiting and postoperative bleeding after tonsillectomy
: a randomized trial. JAMA 2008; 300:2621–2630.
23 Marret E, Flahault A, Samama CM, Bonnet F. Effects of postoperative, nonsteroidal, antiinflammatory drugs on bleeding risk after tonsillectomy
: meta-analysis of randomized, controlled trials. Anesthesiology 2003; 98:1497–1502.