AURAONCETM LARYNGEAL MASK FOR GENERAL ANAESTHESIA WITH CONTROLLED VENTILATION IN PARALYZED PATIENTS

a Department of Anaesthesia, Antrim Area Hospital, 45 Bush Road, BT41 2RL, Antrim, United Kingdom/ Northern Ireland b Department of Anaesthesia, Royal Victoria Hospital, 274 Grosvenor Road, BT12 6BA Belfast, United Kingdom/ Northern Ireland c Department of Anaesthesiology and Intensive Care, 1 Faculty of Medicine Charles University and General University Hospital, U Nemocnice 2, 128 08, Prague 2, Czech Republic E-mail: pavel.michalek@vfn.cz


INTRODUCTION
Disposable supraglottic airway devices (SADs) are used for airway maintenance in the majority of cases where tracheal intubation is not clearly indicated.During the last decade, there have been various SADs developed and clinically investigated.
The i-gel TM airway [Intersurgical Ltd., Wokingham, UK] is a disposable SAD which was introduced into clinical practice in the United Kingdom in January 2007.It is composed of a stem with an integral buccal stabilizer and a cuff which is made from styrene ethylene butadiene styrene.The cuff does not require inflation.The device also incorporates a channel for communication with the oesophagus (Fig. 1).The i-gel TM has been studied in patients undergoing elective surgery and under both spontaneous and controlled ventilation 1,2 and was found to perform well under these conditions.The AuraOnce TM [Ambu A/S, Ballerup, Denmark] laryngeal mask is another disposable SAD introduced into practice in the United Kingdom in February 2004 (Fig. 1).It is based on the classic laryngeal mask airway (LMA Classic TM ) which was conceived over twenty-five years ago by Brain 3 and which entered the clinical arena in 1988.The classic laryngeal mask was a reusable device and because of the concerns about the safety of resterilization, several similar disposable supraglottic devices have been invented.Differences between the AuraOnce TM and LMA Classic TM include an angulated stem and a soft, thin cuff that is designed to minimize both pressure on tissues and leakage from around the device.It also has no grilles.The AuraOnce TM device has also been tested under conditions of both spontaneous 4 and controlled ventilation 5,6 .
We designed a study to compare the i-gel TM with the AuraOnce TM under conditions of controlled ventilation.Due to the fact that some of the procedures required muscle relaxation 7,8 , and to provide homogenous conditions for device insertion, we decided to administer a nondepolarizing muscle blocker to all patients.The primary outcome measure was seal pressure.Secondary outcome measures involved peak pressures, ease and timing of insertion and postoperative complications.The null hypotheses for the study were that there was no difference in the insertion characteristics, total success rate of the two devices or in the incidence of postoperative complications.

MATERIAL AND METHODS
Following both local and regional ethics committee approvals (Office for Research Ethics Committees Northern Ireland number 07/NIR01/124) a randomized, controlled, single-blinded study was established and registered (ClinicalTrials.govidentifier NCT00668278).We were quite explicit with the ethics committee about the use of muscle relaxants, and the committee stipulated that the patient information sheet should mention that such agents were not necessarily a part of the anaesthetic regimen they would have undergone if they were not enrolled in the study.The potential risks associated with the use of these drugs were also outlined.Patients undergoing elective procedures under general anaesthesia who would normally have been managed with a supraglottic airway were invited to participate in the study.
Inclusion criteria were: males or females, aged 18-89 years, American Society of Anesthesiologist (ASA) status I-III, scheduled for elective procedures in general surgery, vascular surgery, urology and gynaecology.
Exclusion criteria were: age less than 18 years or more than 89 years, emergency surgical procedures, surgical procedures inside the abdominal cavity, surgery requiring prone or steep head-down positioning.Patients at increased risk of aspiration were also excluded.
A two-arm table with random numbers was created prior to the study using a randomization freeware (www.graphpad.com).After giving informed consent and prior to induction to general anaesthesia, patients were randomized to receive either the i-gel TM or the AuraOnce TM using sealed envelopes.All data were recorded in a case record sheet.
Patients were fasted according to local guidelines (6 h solids, 2 h clear fluids).Data were recorded including age, sex and weight, ASA status and routine airway assessment -modified Mallampatti score, mouth opening, Mandibular protrusion score and thyromental distance.The type and duration of procedure were also recorded.
Standard anaesthetic monitoring was applied.After obtaining intravenous access, 1 μg/kg of fentanyl was administered and preoxygenation was commenced.After induction with 1-3 mg/kg propofol to loss of verbal contact, atracurium 0.5 mg/kg was administered.Anaesthesia was maintained with sevoflurane in an air-oxygen mixture and analgesia provided using morphine sulphate, paracetamol and/or regional anaesthesia techniques.The "Train Of Four" (TOF) count was measured every 10 s using the MechanoSensor in the neuromuscular transmission module of the anaesthetic machines (GE Healthcare, Helsinki, Finland) until no twitches were present, at which point the device was deemed ready for insertion.
The devices were prepared and lubricated using Aquagel TM jelly [Ecolab, Leeds, UK].Selection of the size of each device was on a weight basis as recommended by each of the manufacturers.In the case of the AuraOnce TM the device was inflated with air, once inserted, according to the manufacturer's guidelines.Insertion of the device was timed from the moment the operator received the device to the first effective breath as confirmed by a capnographic trace.The timing was carried out by an independent operator using a stopwatch, which was timed to the nearest 0.1 s.Where there were multiple attempts The number of attempts and any repositioning maneuvers during the course of the procedure (jaw thrust, head flexion, neck extension) were also recorded.Operators experienced with the insertion of both devices inserted the devices.
Manual ventilation was carried out and adequacy of chest wall movement and end-tidal carbon dioxide waveforms were checked.Adequate oxygenation was checked and noted.Adequate oxygenation was defined as oxygen saturation (SpO 2 ) above 94%.Suboptimal oxygenation was defined as a SpO 2 reading between 94% and 91%.The fractional inspired concentration of oxygen was set on 50%.Failure to achieve adequate oxygenation was defined as a SpO 2 reading of 90% or below 9 .
If insertion proved difficult, or ventilation or oxygenation deemed suboptimal then the protocol allowed for a change of device to the other arm of the trial.Results were for the airway device that was inserted successfully although a note was made of these crossovers.If the oxygenation or ventilation were deemed as failed even after the insertion of the other study device then the patient was to undergo endotracheal intubation.If the patient required endotracheal intubation, for whatever reason, they were excluded from the postoperative questionnaire and reported separately.
Once successfully inserted, the seal pressure was assessed: total gas flow was set at 5 l.min -1 , the Adjustable Pressure Limiting (APL) valve was set to 40 cmH 2 O.The operator waited until an audible leak was detected from around the device.The pressure at which the leak was noted was recorded.If the pressure reached 40 cmH 2 O before a leak was detected the leak pressure was recorded as 40 cmH 2 O.The pressure gauge on the anaesthetic machine was used to record the pressure.
Intermittent Positive Pressure Ventilation (IPPV) was commenced with the tidal volume set to 8-10 ml.kg -1 .The tidal volume was recorded and at this tidal volume the peak airway pressure was noted and recorded.The respiratory rate was adjusted to maintain the ETCO 2 between 4.1 kPa and 6.0 kPa.Adequate ventilation was defined as an end-tidal carbon dioxide level (ETCO 2 ) in the range 4.1-  The data is expressed as mean (±SD).At the end of the procedure, the sevoflurane was discontinued and devices removed when the patients were breathing spontaneously and able to open their eyes and mouth on command.Reversal of neuromuscular blockade was administered as appropriate.Upon removal, the presence of any blood on the device was noted as was whether the inner aspect of the bowl was dry.If the bowl of the device was not dry, the contents were noted (bile stained, clear fluid, sputum etc.).The patients were then transferred to the postoperative recovery ward.
Postoperative questionnaires were used to assess the presence and severity (none/mild/moderate/severe) of the following symptoms: sore throat, difficulty swallowing, pain swallowing, difficulty speaking, pain in the mouth, neck pain, tongue numbness, ear pain, difficulty hearing, nausea, cough and hiccups.These were assessed at one hour postoperatively.A follow-up visit on the ward at 24 h repeated the same questions.Some patients were followed up with a telephone conversation at 24 h.Data were recorded using a proforma and the two postoperative questionnaires.

Statistics
Sample size calculation was based on the results of a previous study published by Richez et al 1 .A power analysis showed that at least 87 patients in each arm would be required to detect a 10% difference in the primary outcome (seal pressure) at a significance level of 5% and a power of 80%.One-hundred patients were selected for each arm to compensate for potential cross-overs and lost patients.
The data were entered into a database (Microsoft Excel for Mac, Microsoft WA 98052, USA).
The results were expressed as mean ± standard deviations (SD) for demographic data and as mean and 95% confidence interval (CI) for the rest of the data.The twosample t-test was used for quantitative data and Fischer's exact test was used for qualitative data.P values less than 0.05 were considered significant.

RESULTS
Demographic data between the two groups were similar (Table 1, 2).There was a greater proportion of female patients in each group.256 patients in total were asked to participate in the study, 46 of them (18%) refused to take part and another 16 patients (2.3%) were excluded for other reasons -change of surgical procedure or cancellation of surgery (Fig. 2).
In total, 204 patients participated in the study.103 patients were initially randomized to the i-gel TM arm and in 101 an AuraOnce TM was initially inserted.Insertion failed in three patients in the i-gel TM group (3.7%); one of them had successful insertion of AuraOnce TM , the other two required endotracheal intubation (Fig. 2).In the AuraOnce TM group, 97 patients were ventilated without any complication, while in three of them cross-over to the I-gel TM vs. AuraOnce TM laryngeal mask for general anaesthesia with controlled ventilation in paralyzed patients  i-gel TM group was needed and in one case, endotracheal intubation had to be performed.The average insertion times were 10.6 s (i-gel) and 12.0 s (AuraOnce) (P=0.19) and the average number of attempts was 1.18 (i-gel) and 1.21 (AuraOnce) (P=0.73).First/Second/Third/Fourth attempt insertions were 89/12/1/1 for the i-gel TM and 81/15/2/0 for the AuraOnce TM .First time insertion rates were 85.6% for the i-gel TM and 82% for the AuraOnce TM with overall success rates of 96.2% for the i-gel TM and 94.2% for the AuraOnce TM (P=0.54).All patients were included in the overall success rates but those who either crossed to the other arm, or who were intubated, were excluded from further statistical analysis.Repositioning or manipulation with the devices at any stage of procedure was required in 39 patients: 15 (14.6%) with the igel TM and 24 (24.5%) with the AuraOnce TM (P=0.11).The main reason for manipulation or repositioning was minor leak, often associated with a change of surgical position.Oxygenation in all patients was adequate and ventilation was rated as inadequate in three -two in the i-gel TM group and one in the AuraOnce TM group.Those in whom ventilation was inadequate had the alternative device inserted.The average seal pressure for the i-gel TM was 30.3, Fig. 3).
The incidence of blood on the device was 6/103 in the i-gel TM group (5.8%) and 2/98 in the AuraOnce TM group (2%) (P=0.28).In two cases (2%) bile-stained contents were noted on the inner aspect of the devices upon re-moval, both in the AuraOnce TM group.Neither of these patients exhibited symptoms or signs of pulmonary aspiration.
All except one postoperative questionnaire were completed and analyzed.The overall incidence of postoperative complications was low in each of the groups (Table 4, Table 5).The only areas that achieved statistical significance were those of sore throat and difficulty swallowing: at 24 h 12 patients complained of sore throat in the i-gel TM group (12%) and 22 patients (22.7%) in the AuraOnce TM group (P=0.03), and at 24h 2 patients in the i-gel TM group (2%) complained of difficulty swallowing with 9 patients (9.3%) in the AuraOnce TM group (P=0.02).

DISCUSSION
The results of this study showed that both devices performed without any significant difference apart from higher seal pressures in the i-gel TM group and lower incidence of postoperative sore throat and difficulty swallowing at 24 h also in the i-gel TM group.There was a preponderance of female subjects in our study (74.5%).This was due to the high proportion of gynaecological and breast surgery in our institution -surgical procedures for which a supraglottic airway may be suitable.(Table 2) A previous study involving the i-gel TM only included female patients 3 and it may be the case that women are more likely to have such devices inserted.We did not choose a crossover design for this study, although most studies comparing two airway devices employ this design.However, crossover design I-gel TM vs. AuraOnce TM laryngeal mask for general anaesthesia with controlled ventilation in paralyzed patients does not allow comparison of the incidence of postoperative complications for each device, and we feel that this outcome is very important.

Comparison with other studies
Insertion attempts for the AuraOnce TM are similar to those found in other studies 4,5 .10][11][12] ).
First time success rates for the i-gel TM found in our study was 85.6% which is lower than the results published in recent cohorts.The i-gel was inserted in several studies on the first attempt in more than 90% of patients 1,2,[13][14][15] .This finding could have been caused by an initial relative lack of experience in some operators.Overall success rate of the i-gel TM insertion in these studies varied between 93-100% and our success rate of 96.3% compares well with these findings.At least one study has demonstrated a shallow learning curve before attainment of high success rates with the i-gel TM (ref. 16).The effective airway time (time to connection of the breathing circuit) was shorter and the average number of insertions was slightly less in the i-gel TM group, although neither was found to be significant.The i-gel TM was found to be the easiest of several SADs to insert in one manikin-based study 17 .Time taken to insert the AuraOnce TM was shorter in our study than in previous studies 5,6,10 and comparable to the results of another 12 which may be explained by a greater experience of all our operators with insertion of the AuraOnce TM device.
Three patients required endotracheal intubation.Each of these patients had some degree of abnormity at preoperative airway assessment.The incidence of blood on the device was higher in the i-gel TM group.The technique of insertion is not the same as that for laryngeal masks and the manufacturer recommends repositioning maneuvers in cases of resistance to insertion, such as deep rotation.Insertion of the i-gel TM has led to trauma to the frenulum in our institution 18 .The deep bowl of the i-gel TM , which is designed so as not to fold the epiglottis, can fold the tongue in on insertion.If this is not recognized then injury is likely.There have been only two observational studies The results expressed as none/mild/moderate/severe; total incidence (total incidence in %).(*) statistically significant.Fischer's exact test used.evaluating traumatic insertion of the i-gel TM (ref. 1,2) and its incidence is reported to be as low as 1%.The other studies had crossover design and thus occurrence of airway trauma could not be attributed to any one particular device.The incidence of traumatic insertion with the AuraOnce TM has been shown to be as high as 20% (ref. 5).Our incidence of 2% for this device is much lower.The i-gel TM was found to have a higher seal pressure than the AuraOnce TM .A recent study comparing the i-gel TM to an endotracheal tube showed a comparable seal (in terms of leak fraction) -albeit under pressure controlled ventilation 19 .However, in another study using a cadaver model, the oesophageal seal pressure of the i-gel TM was found to be significantly lower than with the Classic TM and ProSeal TM laryngeal masks 20 .
In initial investigations performed on cadavers it was found that the i-gel TM effectively mirrored the perilaryngeal anatomy and that it consistently achieved positioning for supraglottic ventilation 21 .The device was designed after investigation using cadaveric dissections, endoscopies and radiological imaging in an attempt to ensure the cuff mirrored the pharyngeal, perilaryngeal and laryngeal anatomy.It may be also used as a conduit for fibreoptic intubation 22 and has been used in the management of a patient with subglottic stenosis 23 .
The postoperative complaints in our study demonstrated a lower incidence of postoperative sore throat and difficulty swallowing at 24h in the i-gel TM group.Very similar results were experienced in another study where the i-gel TM caused lower incidence of postoperative sore throat and neck complaints 24 .This fact may be attributed to the different pressures exerted on pharyngeal mucosa by the inflated cuff of the laryngeal mask compared with the cufless anatomical shape of the i-gel TM .
The incidence of soiling on the inner aspect of both devices was low overall, being lower in the i-gel TM group.The phenomenon of visible regurgitation with the i-gel TM is now well documented 25,26 .The gastric channel allows early recognition of regurgitant gastric contents when placed correctly, although Gibbison's case series points out that this may not prevent ingression of gastric contents into the airway.We experienced two cases of gastric content contamination to the inner aspect of AuraOnce TM , however no patient exhibited symptoms of aspiration.The incidence of aspiration associated with laryngeal mask use is estimated as low as 0.2% (ref. 27).
The overall postoperative complication rate was low, comparable to those found in other studies of these devices.Supraglottic airways are frequently used as conduits for positive pressure ventilation, and avoid the potential pitfalls and postoperative complications of endotracheal intubation.

Study limitations
There are several limitations to our study: The study was performed in patients under confirmed neuromuscular blockade which was not necessary for all procedures.However, there is some evidence that muscu-lar tone in the pharynx is not influenced by neuromuscular blockade 28 , and that neuromuscular blockade does not alter the incidence of sore throat with the LMA-classic™ (ref. 29) but the results of our study cannot be applied to those in whom neuromuscular blocking drugs are not administered.
Secondly, the fact that we did not stipulate the checking of cuff pressures with the AuraOnce TM device may have prejudiced against that device in terms of its intraoperative performance and the postoperative course of patients in whom it was inserted.The changes in the cuff pressure of a laryngeal mask may have influence on the postoperative airway morphology 30 .The lack of an inflatable cuff in the i-gel TM obviates the need for checking of cuff pressures, and given our experience that checking of pressures is a variable practice this may be an advantage in itself.
Thirdly, the comparison between a device with a gastric channel and one without may disadvantage the device without for reasons explained above.The AuraOnce TM is the most common disposable SAD in use in our region and its cost is comparable with the i-gel TM .The LMA-Supreme TM a newer disposable device that possesses a drain tube may have been a better candidate, but it was not available at the time the study was designed and its cost is several times higher than the cost of the i-gel TM .

CONCLUSIONS
We have shown that these two disposable supraglottic airway devices perform almost equally under conditions of controlled ventilation, and that they have a low postoperative complication rate.The i-gel TM provided better seal and caused less postoperative sore throat and swallowing difficulties at 24 h.

Fig. 2 .
Fig. 2. Diagram describing the flow of the study participants through each stage of the randomized trial.

Fig. 3 .
Fig. 3. Seal and peak pressures in the i-gel TM and AuraOnce TM groups.▲ = means, ■ = upper and lower limits, the boxes represent the interquartile ranges 25-75%.
4 cmH 2 O and 27.8 cmH 2 O for the AuraOnce TM (P=0.007).Average peak airway pressure for the i-gel TM was 15.3 cmH 2 O and 15.6 cmH 2 O (P=0.57) in the AuraOnce TM group (Table

Table 1 .
Demographic data of the patients participating in the study.

Table 2 .
Types of surgery the patients in the trial were scheduled to undergo.
volumes) were automatically recorded.The operators assessed the SADs continuously for an audible leak.Minor leak was defined as any audible leak, without significant decrease in the expiratory volumes and requested repositioning of the device and/or patients´ head.Major leak was defined as an audible leak associated with significant decrease in expiratory volume, deformation of etCO 2 trace and/or hypoxaemia.

Table 3 .
Insertion data, pressure data, device removal data.

Table 4 .
Postoperative questionnaire results taken immediately after the procedure.
The results expressed as none/mild/moderate/severe; total incidence (total incidence in %).(*) statistically significant.Fischer's exact test used.

Table 5 .
Postoperative questionnaire results taken at 24h after the procedure.