Provisional data from this study were presented to the Difficult Airway Society Annual Meeting 2007, Portsmouth November 2007.
Evaluation of the size 4 i-gel™ airway in one hundred non-paralysed patients*
Article first published online: 9 JUL 2008
© 2008 The Authors. Journal compilation © 2008 The Association of Anaesthetists of Great Britain and Ireland
Volume 63, Issue 10, pages 1124–1130, October 2008
How to Cite
Gatward, J. J., Cook, T. M., Seller, C., Handel, J., Simpson, T., Vanek, V. and Kelly, F. (2008), Evaluation of the size 4 i-gel™ airway in one hundred non-paralysed patients. Anaesthesia, 63: 1124–1130. doi: 10.1111/j.1365-2044.2008.05561.x
- Issue published online: 5 SEP 2008
- Article first published online: 9 JUL 2008
- Accepted: 12 March 2008
We studied the i-gel™ in 100 elective, anaesthetised patients (55 : 45 male:female, median age 53 years) assessing: ease of use, airway quality, positioning, seal and complications. First insertion attempt was successful in 86 patients, second attempt in 11 patients, and third attempt in three patients. Fifty three manipulations were required in 26 patients (median 1) to achieve a clear airway. Median insertion time was 15 s. During ventilation, expired tidal volume of 7 ml.kg−1 was achieved in 96 (96%) patients. Median airway leak pressure was 24 cmH2O. On fibreoptic examination via the device, vocal cords were visible in 87 patients (91%). During maintenance, six patients (6%) required 12 airway manipulations. There was one episode of regurgitation, without aspiration. Other complications and patient side-effects were mild and few. The i-gel is easily and rapidly inserted, providing a reliable airway in over 90% of cases. Further studies are indicated to assess safety and performance compared to other supraglottic airway devices.
The i-gel (Intersurgical, Wokingham, UK) is a relatively new, CE marked, single-use, supraglottic airway device (SAD) for use during anaesthesia (Fig. 1). It has design similarities to a classic laryngeal mask airway (cLMA, Intavent Orthofix, Maidenhead, UK) in that it has a stem and an anatomically designed cuff. In contrast to the cLMA, its cuff is filled with a medical grade thermoplastic elastomer gel (styrene ethylene butadene styrene), so that, unlike most other SADs, it does not require inflation. Its manufacturers describe it as creating an anatomical seal with pharyngeal, laryngeal and perilaryngeal structures . The device has a firm tube section with a rigid proximal section acting as an integral bite block. The tube is elliptical in cross sectional and curved longitudinally; features designed to facilitate insertion and minimise axial rotation once inserted. Like the cLMA, once in place its distal tip is positioned at the top of the oesophagus. The tip incorporates the distal end of an integrated drainage tube, a design similarity with the Proseal LMA (PLMA, Intavent Orthofix). This is termed a ‘gastric channel’ by the manufacturers but we consider this a misnomer and will refer to it as a ‘drain tube’. The drain tube runs within the device and exits next to the 15-mm airway connector. The drain tube is designed to separate the gastrointestinal and respiratory tracts, to prevent gastric inflation during ventilation, allow venting of gastric contents and enable the passage of a gastric tube. In contrast to the PLMA, which has been studied extensively , none of these features has yet been tested. There are no clinical papers published on the i-gel to date. We present a cohort study of 100 anaesthetised patients in order to evaluate its performance.
Local ethics committee approval was obtained for the study. Adult patients attending for elective surgery who were American Society of Anesthesiologists (ASA) 1–3 were considered for inclusion. Patients were asked to participate in the study if they were expected to undergo anaesthesia with airway management using a cLMA or PLMA. Patients were excluded if they had pathology of the neck, upper respiratory or upper alimentary tracts, previous or anticipated airway problems, or were at risk of regurgitation or pulmonary aspiration of gastric contents.
Patients were given written and verbal information by the anaesthetist before signing a consent form. All anaesthetists participating in the study had at least 5 years experience of airway management and were experienced with both the cLMA and PLMA. Before recruiting patients into the study, each anaesthetist was instructed in the use of the i-gel and used the device to manage the airway of 10 patients. Data from these patients are not included in analysis.
The Samsoon and Young modification  of the Mallampatti classification  was used to assess the airway of all patients. Anaesthetic monitoring conformed to the minimum standard recommended by the Association of Anaesthetists of Great Britain and Ireland . Anaesthesia consisted of induction with fentanyl 1–3 μg.kg−1 and propofol 1% delivered by target controlled infusion (target 4–7 μg.ml−1). Neuromuscular blocking drugs were not used. Target levels of propofol were increased if there were difficulties in mouth opening, coughing, gagging or if body movement occurred during the insertion of the i-gel. Maintenance of anaesthesia was with a target controlled infusion of propofol 1%. Anaesthesia was considered adequate for device insertion when the patient was unresponsive, had lost their eye lash reflex, did not respond to anterior jaw thrust and the effect site concentration exceeded 2 μg.ml−1 .
The i-gel was inserted according to the manufacturer’s instructions. The front, back and sides of the cuff are lubricated with a thin layer of water based lubricant. The device is then grasped by the integral bite block and its tip directed towards the hard palate. It is then inserted with a continuous but gentle push until resistance is felt . A size 4 i-gel (designed for patients with body mass between 50–90 kg) was used in all patients. This was the only size available from the manufacturers at the time of commencement of the study and the manufacturers were of the opinion that this size would be effective in the vast majority of cases.
The broad aims of the study were to assess the ease of use of the device, the quality of airway achieved and any associated complications. Ease of device insertion was recorded as ‘no or minimal resistance’, ‘significant resistance’, or ‘impossible to pass without excessive force’. The number of attempts at insertion and the time until successful insertion were recorded. If it was not possible to insert the device, the following adjustments were allowed: neck extension or flexion, chin lift, jaw thrust and gentle pushing or pulling of the device. The number of airway manipulations required to establish the airway was recorded. Gentle manual ventilation of the patient’s lungs was attempted before formal evaluation. Successful ventilation was defined as visible chest movement, square wave capnogram, stable arterial oxygen saturation above 95% and the ability to achieve an expired tidal volume of 7 ml.kg−1 (with an initial tidal volume of 7 ml.kg−1 increased if necessary up to a maximum of 10 ml.kg−1). If it was not possible to ventilate the lungs, the same adjustments as for insertion were allowed. If ventilation was still not possible, further attempts at insertion of the device were allowed to a maximum of three attempts. Airway quality was classified as clear, partially obstructed or obstructed.
If insertion or ventilation failed, the use of the i-gel was abandoned and the airway maintained as clinically indicated (i.e. with a cLMA or PLMA). The i-gel was used as the primary airway throughout anaesthesia unless complications necessitated its removal. Any complications during insertion (defined before the study) were recorded.
The anatomical position of the i-gel was assessed clinically (by observing a filling of the anterior neck and that the midline of the device remained in the anatomical midline) and by fibreoptic examination via both channels of the device. The view via the airway channel was scored as follows: Grade 1, clear view of the vocal cords; Grade 2, view of the arytenoids only; Grade 3, view of the epiglottis only; Grade 4, no laryngeal structures visible . The view via the drain tube was recorded as: mucosa; closed upper oesophagus; open upper oesophagus; larynx; drain tube occluded (if the fibrescope could not be passed).
The anaesthetic circuit was then closed to atmosphere to assess airway leak pressure. Ventilation was stopped momentarily and using a fresh gas flow of 5 l.min−1, the pressure in the airway circuit was recorded until an audible leak occurred or airway pressure plateaued. Airway pressure was not allowed to exceed 40 cmH2O. The ability to pass a well-lubricated size French 12 gauge gastric tube though the drain tube was recorded. Two attempts were allowed.
During maintenance of anaesthesia, the quality of the airway (clear; intermittent partial obstruction; intermittent complete obstruction; complete obstruction) and the number and type of airway manipulations required to maintain airway patency were recorded. Complications (previously defined) encountered during maintenance of anaesthesia were recorded. Finally, the anaesthetist was asked to rate the overall airway quality, ease of hands-free anaesthesia and usefulness of the device using a five-point subjective scale (excellent; good; fair; poor; inadequate).
The use of the i-gel was initially evaluated during controlled ventilation; however, during maintenance, as neuromuscular blocking drugs were not used, the patients were able to resume spontaneous ventilation. This occurred at varying stages of maintenance and occupied a variable proportion of the case due to widely varying durations of surgery. However, the methodology allowed evaluation of the i-gel during both controlled and spontaneous ventilation.
On emergence from anaesthesia, the device was removed by recovery staff when the patient was able to open their mouth to command. During emergence and removal, airway complications and the presence or absence of blood or troublesome secretions were recorded.
Patients underwent a structured interview before leaving the recovery area and again at 24 h to elicit the presence of nausea, sore throat, neck, jaw or ear pain, pain on speaking or altered sensation or swelling in the tongue or oropharynx. All positive responses were rated by the patient as mild, moderate or severe.
Data were analysed using microsoft excel (Microsoft Corporation, Redmond, WA, 98052-6399, USA).
We studied 55 males and 45 females, with a median age of 53 years (range 18–84 years), a median weight of 73 kg (range 42–113 kg) and median body mass index 25 kg.m−2 (range 20–30 kg.m−2). The ASA physical status grading was Grade 1 (n = 62), Grade 2 (n = 36) and Grade 3 (n = 2). The modified Mallampati classification was class 1 (n = 61), class 2 (n = 32), class 3 (n = 4) and class 4 (n = 3). The surgical speciality was orthopaedics for 39 patients, general surgery for 35, urology for 17, gynaecology for six and otorhinolaryngology for three. The median duration of procedures was 45 min (range 8–145 min).
Insertion of the i-gel with a degree of subsequent ventilation was possible in all 100 patients. Successful insertion was after the 1st attempt in 86 patients, the 2nd attempt in 11 and the third attempt in three. There was ‘no or minimal resistance’ to insertion in 94 patients, and ‘significant resistance’ in six patients. An increase in the target level of propofol was required during insertion in 10 cases. Securing an effective airway took 30 s or less in 92 cases, with an overall median duration of 15 s (range 5–120 s). Following successful insertion, including airway manipulations if required, the airway was clear in 99 patients and partially obstructed in one. Twenty-six patients required a total of 53 airway manipulations (median 1, range 1–5) to establish a patent airway. Excluding airway manipulations and re-insertions, the airway was established uneventfully (no complications) in 95 patients. There were early complications in five patients. One patient developed low oxygen saturations (91%) during a re-insertion. In two cases, an adequate airway seal could not be obtained. In both cases, some chest wall movement was achieved and arterial oxygen saturation remained stable but the leak was regarded as excessive and the i-gel was removed. A cLMA was inserted and ventilation established successfully. In one of these patients, there was a small amount of blood present on the i-gel on its removal. One patient coughed during insertion and another developed hiccoughs. These were both temporary complications; the lowest oxygen saturations in these patients were 95% and 94% respectively. On attempting to ventilate the patient, chest wall movement was seen in all patients, although this was deemed inadequate in the two cases of poor airway seal. An expired tidal volume of 7 ml.kg−1 was achieved in 96 patients. A square wave capnogram was seen in 99 patients, and arterial oxygen saturation was stable in 99 patients. Median [interquartile (range)] airway leak pressure was 24 cmH2O [18–30 (7–40)] and was greater than 20 cmH2O in 61 patients.
After insertion of the i-gel, filling of the anterior neck was observed in all patients. The device midline remained in the anatomical midline in 97 patients. Fibreoptic inspection through the airway channel was performed in 96 patients and through the drain tube in 91 patients (some patients were not examined because the fibrescope was unavailable). The vocal cords were visible in 91%; details of fibreoscopy are recorded in Table 1. Passage of a gastric tube via the drain tube was attempted in 98 patients. This was successful at the first attempt in 90 patients (92%) and the second attempt in one patient (1%). It was not possible in seven patients (7%).
|View via airway channel (n = 96)||No (%)|
|Grade 1||87 (91)|
|Grade 2||4 (4)|
|Grade 3||4 (4)|
|Grade 4||1 (1)|
|View via drain tube (n = 91)||No (%)|
|Oesophagus closed||8 (9)|
|Oesophagus open||11 (12)|
|Drain tube occluded||1 (1)|
The i-gel was used throughout maintenance of anaesthesia in 98 patients; with the only two failures being those described above in whom there was a clear airway, but inadequate seal for ventilation. During maintenance, the airway remained patent in 95 patients (97%). There was intermittent partial obstruction in three patients (3%) and intermittent complete obstruction in one patient (1%). Six patients (6%) required a total of 12 manipulations of the airway during maintenance. Overall, arterial oxygen saturation remained above 90% in 99 patients and above 95% in 94 patients.
There was one episode of regurgitation during the study. An 83 year-old male patient who was ASA Grade 1 and had a BMI of 27 kg.m−2 was undergoing surgery to release a Dupuytren’s contracture. I-gel insertion was uneventful on the first attempt, as was ventilation. During maintenance, the patient coughed and shortly afterwards, gastric contents were seen to exit from the drain tube. The device was removed and the airway maintained by facemask ventilation, with a brief period of desaturation to 86%. The airway channel of the i-gel was visually inspected and was free of gastric contents. Fibreoptic examination of the airway channel and trachea revealed no evidence of aspiration. The i-gel was therefore replaced without incident and the procedure was completed uneventfully.
During emergence from anaesthesia, the i-gel was tolerated well in 94 cases (96%) and poorly in four cases (4%). The device was removed from two patients before emergence. Three patients (3%) coughed (arterial oxygen saturations remained above 94% in all cases) and in one patient (1%) there was a brief episode of laryngeal spasm (arterial oxygen saturations remained above 99%). Upon removal of the i-gel, secretions were troublesome in two patients (2%) and blood was visible on the device in one patient (1%).
Anaesthetist-rated quality of airway during maintenance, ease of hands-free anaesthesia and overall usefulness of the device are shown in Table 2.
|Airway quality during maintenance||Ease of hands-free anaesthesia||Overall usefulness of the device|
The results of the postoperative questionnaire to detect potential side effects are shown in Table 3.
|Symptom||Recovery area||24 h|
|Mild||14 (14)||7 (7)|
|Moderate||2 (2)||2 (2)|
|Pain on swallowing|
|Mild||4 (4)||3 (3)|
|Moderate||1 (1)||1 (1)|
|Pain in jaw||0||0|
|Pain in mouth||1 (1)||0|
|Pain on speaking||1 (1)||0|
|Ear pain||0||1 (1)|
|Hearing changes||2 (2)||0|
The results of this small study are very encouraging. The i-gel was successfully inserted in all patients and allowed effective controlled ventilation in 98%. Rates of failure, manipulations required and complications were very low compared to other devices we have studied with similar methodology [8–13]. The device was well tolerated throughout anaesthesia and emergence, with few sequelae reported by patients. The i-gel therefore appears to be a suitable device for anaesthesia using controlled ventilation and in spontaneously breathing patients. Our study can only examine efficacy and others are required to determine safety, but we encountered no major complications.
Insertion of the i-gel airway was easy in the vast majority of cases. The first time insertion success rate was 86%, which is comparable to the PLMA, though somewhat lower than the cLMA . Overall insertion success (within three attempts or less) was 100%, similar to the cLMA and the PLMA . Securing an effective airway was rapid in most cases. The median time to ventilation was 15 s, which compares favourably with the cLMA (9–34 s) and the PLMA (15–41 s) [8, 14–16]. The i-gel is a new device and the anaesthetists involved had only performed 10 insertions before commencing the study. For the PLMA, it has been suggested that 20–30 insertions may be necessary to achieve competence . It is therefore possible that, with more practice, first-time insertion rates may have been higher and insertion times faster. On the other hand, there are data to suggest that the i-gel is easily inserted by novices and experienced users alike. A study conducted in our hospital in 40 patients found that the device can be placed reliably and quickly by non-anaesthetic medical staff and medical students . A recent manikin study also showed that experience had no effect on insertion; neither experience with the device itself nor overall anaesthetic experience improved time to ventilation . The i-gel, once lubricated, is often inserted with remarkably little friction between it and the tissues. The fact that there is no cuff to inflate also speeds insertion.
Its relative ease of insertion and design features that aim to reduce the risk of aspiration have already led to suggestions that the i-gel may have a role in airway management during cardiopulmonary resuscitation (CPR) by novices . In a recent study, albeit in a manikin, the i-gel was found to be 50% faster to insert than the cLMA or tracheal tube and 25% faster than the PLMA, both with and without concurrent chest compressions being performed . The only SADs currently recommended by the European Resuscitation Council for use in CPR are the cLMA, the Combitube™ (Kendall-Sheridan, Argyll, NY, USA) and the Laryngeal Tube (LT, VBM GmBbH Sulz, Germany) . This is because these are the only SADs to be studied in this setting. The cLMA is used most commonly, and has been shown to achieve better ventilation than bag-mask ventilation during CPR . No such data are available for the i-gel and further studies are required using the device during CPR in patients .
The median seal pressure achieved with the i-gel was 24 cmH2O. This is similar to a previous case series , and whilst being superior to the cLMA (16–20 cmH2O), falls short of the 32 cmH2O leak pressure which can be achieved with the PLMA . In contrast to the cLMA, where the leak pressure rarely exceeds 30 cmH2O , in six patients in this study, the i-gel leak pressure exceeded our upper limit of 40 cmH2O. Also of note is that only the size 4 i-gel was used during this study. This size is marketed for use in patients weighing 50–90 kg. Nineteen of the patients in our study weighed more than 90 kg. Unexpectedly, however, the median seal pressure in these patients was higher than in patients weighing less than 90 kg (27 cmH2O vs 23 cmH2O).
We achieved controlled ventilation to a tidal volume of 7 ml.kg−1 in 96 (96%) patients without the use of neuromuscular blocking drugs. This is comparable with data on the cLMA and the PLMA . In two patients in whom this was not achieved (one weighing 66 kg, one 106 kg), chest expansion was observed and deemed adequate to continue with anaesthesia and surgery. In two patients (both weighing less than 90 kg and with BMI < 28 kg.m−2), there was inadequate airway seal to enable controlled ventilation, despite a clear airway, and the device was removed. The lack of a larger sized i-gel (which is now available) may have contributed to these four failures, but this is speculative.
Anatomical positioning of the i-gel appears to compare well with other SADs. Filling of the anterior neck occurred with all insertions and the midline of the device stayed in the midline in 97% of cases. On fibreoptic examination, a Grade 1 view of the larynx was obtained in 91% of cases, which is very similar to the cLMA and PLMA . This suggests that the device may be of use as a conduit in difficult airway management and failed intubation. Other features of the device that may theoretically be of benefit in this regard are the short (192 mm for an i-gel size 4 compared to 205 mm for a cLMA size 4) and wide airway channel (allowing for the passage of a standard tracheal tube), the large bowl and the absence of grilles or bars at the distal end. There is one case report of the successful use of the i-gel for rescue intubation in the literature . A recent abstract reported the i-gel to be better than the LMA-Classic™ for Aintree Intubation Catheter placement in a manikin (Maties O, Heath M. Comparison of trachea intubation using the Aintree Intubation Catheter via the LMA Classic™ and the i-gel airway: a manikin study. Difficult Airway Society Annual meeting, Portsmouth 2007). It is quite possible that this was due to differences in ease of insertion of the device into a manikin as the i-gel is much easier to insert in a manikin than the cLMA [24, 25]. Further work, in patients, would be welcome.
An appropriately sized gastric tube was successfully passed via the drain tube in 93% of cases, which is similar to success rates achieved with most studied SADs with a drain tube [2, 13]. We made no attempt, however, to assess the efficacy of the drain tube in preventing inflation of the stomach or aspiration of regurgitated gastric contents. In the size 4 i-gel, the drain tube is significantly smaller than in a size 4 PLMA (French gauge 12 compared to French gauge 16). The PLMA has been investigated in a cadaveric study and found to offer significant protection against aspiration . No such study has been performed for the i-gel and whether the smaller drain tube is adequate remains unproven. We identified one episode of regurgitation in our study, in which gastric contents were vented out of the drain tube, probably preventing aspiration. The incidence of regurgitation and aspiration with the i-gel is unknown. Two other cases of regurgitation whilst using the device have occurred in our hospital (patients not in this trial) in approximately 280 uses. In one case, the i-gel fully protected the airway; in the other case there was evidence of partial aspiration, though the patient made an uneventful recovery . The incidence of regurgitation associated with the device is, however, almost certainly lower than this case series would suggest, as no other case reports have been published despite over 200 000 uses of the device (David Chapman, Intersurgical, personal communication). Of note, even if protection is incomplete, the presence of a drain tube allows early identification of regurgitation and prompt responses to prevent or minimise aspiration. The incidence of regurgitation and aspiration with the cLMA remains controversial. It has been quoted as 1 in 4–11 000 in starved, elective patients . The PLMA has been associated with three confirmed cases of aspiration . No comparative trial has been conducted between the PLMA and the cLMA, but the evidence strongly supports the improved protective abilities of the PLMA.
During maintenance of anaesthesia, the i-gel performed well. Only a small number of patients required airway manipulations to maintain a patent airway, and in the vast majority of cases, ease of hands-free anaesthesia was rated as excellent. The incidence of complications with the i-gel was generally low; being seen in less than 10% of patients at all three phases during which they were recorded (induction, maintenance and emergence from anaesthesia). They were generally self-limiting and mild.
Side effects were mostly restricted to sore throat. Eighteen patients (18%) developed a sore throat and/or pain on swallowing, which persisted until 24 h in eleven patients (11%). This is similar to the cLMA, where the incidence of sore throat has variously been reported to be 12–28% [29, 30]. This may be related to the low incidence of injury to the oropharyngeal mucosa in our study: blood was only detected on one i-gel (1%), compared to 12–15% for the cLMA [31, 32] and 10% for the PLMA . This may be due to the gel filled cuff causing less direct trauma and/or pressure damage to the oropharyngeal mucosa, though further study is necessary to confirm this.
There are several limitations of this study. Firstly, we studied only low risk patients who had normal airways and were mostly not obese. We cannot extrapolate from these data to performance in other groups. Secondly, we did not compare performance with likely competitors of the i-gel. This study can be regarded as a feasibility study and randomised controlled studies comparing performance are required. Thirdly, this study has not examined the performance of the i-gel in unstarved patients or those who may have a full stomach. The results do not show that use of the i-gel in such groups would be appropriate. Finally, this study is relatively small and while it shows that the i-gel appears to be efficacious, it offers almost no useful evidence of the safety of the device, which requires data from a considerably larger cohort in routine practice.
The i-gel has been designed to achieve a peri-laryngeal airway seal without the use of an inflatable cuff, whilst offering other features aimed at separating the gastrointestinal and respiratory tracts and potentially improving stability. Our small study has found that the airway seal offered by the i-gel may be inferior to its most likely competitor, the PLMA, although still sufficient for controlled ventilation in the vast majority of patients. The efficacy of the drain tube has also yet to be demonstrated. Insertion of the device into the correct functional and anatomical position was easy and rapid. The incidence of airway obstruction, airway irritation, oropharyngeal trauma and other complications was low. These attributes would suggest potential roles in anaesthesia, management of the difficult airway and airway management during CPR. Further studies comparing the i-gel with its likely clinical competitors, including the cLMA, PLMA and Supreme™ airway (Intavent Orthofix) during anaesthesia are now indicated.
Potential conflicts of interest
The i-gels used in this study were donated free by the manufacturers (Intersurgical, Wokingham, UK). TMC has received payment for lecturing from Intavent Orthofix and the LMA Company, both of which manufacture laryngeal mask airways, competitors of the i-gel.
- 1Intersurgical. iGEL Users Guide. Available from http://www.intersurgical.com/utils/GetFile.aspx?FileID=29456 (accessed January 2008).
- 5Association of Anaesthetists of Great Britain and Ireland. Recommendations for Standards of Monitoring During Anaesthesia and Recovery, 4th edn. London: AAGBI, 2007.