A randomised trial comparing the i-gelTM with the LMA ClassicTM in children


correspondence to: Dr Jin-Tae Kim


We performed a prospective, randomised trial comparing the i-gelTM with the LMA ClassicTM in children undergoing general anaesthesia. Ninety-nine healthy patients were randomly assigned to either the i-gel or the LMA Classic. The outcomes measured were airway leak pressure, ease of insertion, time taken for insertion, fibreoptic examination and complications. Median (IQR [range]) time to successful device placement was shorter with the i-gel (17.0 (13.8−20.0 [10.0−20.0]) s) compared with the LMA Classic (21.0 (17.5−25.0 [15.0−70.0]) s, p = 0.002). There was no significant difference in oropharyngeal leak pressure between the two devices. A good fibreoptic view of the glottis was obtained in 74% of the i-gel group and in 43% of the LMA Classic group (p < 0.001). There were no significant complications. In conclusion, the i-gel provided a similar leak pressure, but a shorter insertion time and improved glottic view compared with the LMA Classic in children.

The use of supraglottic airway devices in paediatric anaesthesia has increased recently [1]. The most commonly used supraglottic airway device is the laryngeal mask airway (LMA) Classic™ (Laryngeal Mask Company Ltd., Henley-on-Thames, UK), and nearly all supraglottic airway devices, including the LMA Classic, are equipped with an inflatable cuff for sealing the perilaryngeal space. Accurate positioning and adequate pressure and volume within the cuff are fundamental to achieving optimal function as well as reducing the likelihood of complications.

The i-gelTM (Intersurgical, Wokingham, UK) is a newly developed disposable supraglottic airway device. The cuff of the i-gel is constructed from medical-grade thermoplastic elastomer (styrene ethylene butadiene styrene) [2] which eliminates the need to inflate the cuff or adjust intracuff pressure. The potential advantages of the i-gel are easy and rapid insertion and a reduction in the risk of pharyngeal tissue compression due to high cuff pressure. The stem of the i-gel is less flexible than that of the LMA Classic and has an integral bite-block. In addition, the shape of the stem is wide and oval, which increases buccal stabilisation and may reduce rotation of the i-gel within the oropharynx. Several studies in adults have shown that the clinical performance of the i-gel is comparable with, or better than, other devices with respect to insertion time, ease of insertion, oropharyngeal leak pressure and vocal cord view via a fibreoptic scope [3–9].

The paediatric i-gels (sizes 1, 1.5, 2 and 2.5) have recently become available commercially and their feasibility should be confirmed before introduction into widespread clinical use, and comparison with established devices should be demonstrated. These types of studies are particularly important in children because they are known to be more vulnerable to complications related to the use of supraglottic airway devices [10, 11], however, there are few published studies to date that have evaluated the paediatric i-gel [12–14], especially in small children.

We performed this study to evaluate the clinical performance of paediatric-sized i-gels in children by comparing their efficacy with the LMA Classic in terms of insertion time, sealing pressures, glottic view via a fibreoptic scope and incidence of complications.


This study was approved by the local research ethics committee and written informed consent was obtained from the parents of all patients. Children aged 1–108 months, weighing between 6 and 30 kg, of ASA physical status 1–2 who were scheduled for elective surgery were considered eligible and were randomly assigned to either the i-gel or LMA Classic. The size of i-gel or LMA Classic selected for insertion was based on the patients weight and according to the manufacturer’s recommendations (i-gel: size 1.5, 5–12 kg; size 2, 12–25 kg; size 2.5, 25–30 kg. LMA Classic: size 1.5, 5–10 kg; size 2, 10–20 kg; size 2.5, 20–30 kg.). Exclusion criteria were inability of parents to understand the study or consent process and children with previous or anticipated airway problems.

Standard monitoring was instituted according to the Association of Anaesthetists of Great Britain and Ireland guidelines [15] and venous access was obtained. Anaesthesia was induced with an intravenous injection of 3 mg.kg−1 propofol or 6 mg.kg−1 thiopental. Once adequate depth of anaesthesia was achieved (lack of motor response to jaw thrust) 0.3 mg.kg−1 rocuronium was given intravenously. Ventilation of the lungs was continued manually with oxygen and sevoflurane. Three minutes after administration of rocuronium, one or other device was inserted by a single investigator who had experience with more than 20 i-gel and 200 LMA Classic insertions. Each device was lubricated with a water-based agent and inserted according to the manufacturer’s recommendations. With the patient’s head in the neutral position, the time to successful insertion was measured from the moment the facemask was removed until the first capnography upstroke after insertion of the device. If chest movement was not adequate and/or the capnography wave was not square-shaped after insertion, ventilation was considered inadequate and manipulations were allowed in the following sequence: gentle pushing or pulling of the device; changing head position by extension or flexion; and jaw thrust. After confirming adequate ventilation, all devices were fixed in place with tape. The ease of insertion was assessed using a subjective scale of 1–4 (1, no resistance; 2, mild resistance; 3, moderate resistance; 4, inability to place the device). In the case of grade 4, exclusion of the participant and tracheal intubation was planned. The number of manipulations required to obtain adequate ventilation was also recorded.

Airway sealing was examined by measuring oropharyngeal leak pressure and peak inspiratory pressure at tidal volume of 10 ml.kg−1. To determine the leak pressure, the expiratory valve was closed and a fresh gas flow of 3 l.min−1 was set until equilibrium was reached (airway pressure was not allowed to exceed 40 cmH2O) and then released completely. The PrimusTM Drager anaesthesia machine (Dräger Medical GmbH, Lübeck, Germany) with a built-in pressure gauge and spirometer attachment was used. A flexible fibreoptic scope (external diameter 3.1 mm; Olympus Optical Co., Tokyo, Japan) was then used to view and record the anatomical alignment of the device in relation to the larynx. The images were recorded as follows [16, 17]: grade 1, clear view of the vocal cords; grade 2, only arytenoids visible; grade 3, only epiglottis visible; grade 4, larynx not visible.

Anaesthesia was maintained with sevoflurane in an oxygen/air mixture and positive pressure mechanical ventilation of the lungs. When the operation was completed, atropine and neostigmine were administered for the reversal of neuromuscular paralysis and sevoflurane was discontinued. Both devices were removed when an adequate respiratory pattern was observed, together with spontaneous eye opening, attempted vocalisation, facial grimacing and evidence of appropriate motor activity.

Throughout anaesthetic maintenance and recovery the occurrence of any complications, including hypoxia (SpO2 < 93%), airway obstruction, suspicion of laryngospasm, sustained coughing, vomiting, or blood stains on the removed device, were recorded.

The null hypothesis was that there was no difference in oropharyngeal leak pressure between the i-gel and LMA Classic. Based on previous studies demonstrating a mean (SD) leak pressure with the i-gel of 24.9 (5.8) cmH2O [12] and that of the LMA Classic of 19 (5.8) cmH2O [18] we calculated the sample size. Given a type-1 error of 0.05 and a power of 0.8, it was determined that 15 patients were required in each group for statistical significance. Considering a 20% drop-out rate and the inclusion of three device sizes (1.5, 2, 2.5) of each device, the number of subjects for each device size was calculated to be 36. All frequency data were compared with the chi-squared test and continuous data were analysed with student t-tests or Mann–Whitney tests for the detection of differences between the i-gel and LMA Classic groups. All data were analysed with spss version 18 (SPSS, Inc., Chicago, IL, USA), and a value of p < 0.05 was considered statistically significant.


One hundred and eight patients were initially recruited to the study. Five children were not studied due to intercurrent upper respiratory tract infection, one was considered at risk of tooth loss, and three children were withdrawn following a change to their surgical plan. The results for 99 children were included in the study. Patients’ characteristics are presented in Table 1.

Table 1.   Baseline characteristics of patients undergoing anaesthesia with the i-gel or LMA Classic supraglottic airway devices. Values are median (IQR [range]) or number (proportion).
 i-gel (n = 50)LMA Classic (n = 49)p value
Age; months36 (10–60 [2–108])41 (10–67 [1–108])0.973
Sex; male20 (40%)35 (71%)0.003
Height; cm95 (77–119 [56–136])97 (76–115 [57–130])0.709
Weight; kg16.1 (10.6–26.7 [6.2–30.0])16.3 (9.8–20.5 [6.0–28.0])0.259
Device size
 1.518 (36%)16 (33%)0.801
 216 (32%)17 (34%)
 2.516 (32%)16 (33%)

The insertion features for each device are described in Table 2. The insertion time for the i-gel was significantly shorter than that for the LMA Classic although ease of insertion did not differ between the two and both devices were inserted without any difficulty (grade 1) in 78% and 76% of children, respectively. Most children did not require further manipulation for adequate positioning, although 20% and 35% of children in the i-gel group and LMA Classic group, respectively, required some manipulation. The i-gel provided a better view of the glottis compared with the LMA Classic; a good fibreoptic view (grade 1 or 2) of the glottis was obtained in 74% of the i-gel group and in 43% of the LMA Classic group, p < 0.001. Median oropharyngeal leak pressure was approximately 20 cmH2O and was similar with either device.

Table 2.   Comparative data for the i-gel and LMA Classic. Values are number (proportion) or median (IQR [range]).
 i-gel (n = 50)LMA Classic (n = 49)p value
  1. *The glottic view via fibreoptic examination was scored using the following: score 1, clear view of the vocal cords; score 2, only arytenoids visible; score 3, only epiglottis visible; score 4, larynx not visible.

Insertion success
 1st attempt48 (96%)45 (92%)0.385
 2nd attempt2 (4%)4 (8%) 
Insertion time; s17.0 (13.8-20.0 [10.0-40.0])21.0 (17.5-25.0 [15.0-70.0])0.002
Peak inspiratory pressure; cmH2O14.0 (12.0-16.5 [10.0-26.0])13.0 (11.0-18.0 [10.0-27.0])0.909
Leak pressure; cmH2O22 (20–25 [15–30])20 (18–23 [10–38])0.117
View of glottis*
 116 (32%)4 (8%)< 0.001
 221 (42%)17 (35%)
 38 (16%)16 (33%)
 45 (10%)12 (24%)

The incidence of complications was low in both groups. Bloodstaining was identified in two of the i-gel group vs five in the LMA Classic group. Seven patients in the i-gel group and six in the LMA Classic group displayed sustained cough, and breath-holding occurred in one i-gel patient and in two LMA Classic patients. The device was displaced during maintenance in one child in each group, both with the size 1.5 in situ. There were no oxygen desaturations or laryngospasm in any patients.


Our results demonstrate that the paediatric i-gel has a similar performance in terms of ease of insertion, sealing quality and the number of manipulations required for adequate positioning compared with the LMA Classic airway. However, the i-gel has advantages over the LMA Classic in terms of shorter insertion times and improved fibreoptic views of the glottis.

Although the i-gel has been widely studied in adults and is generally accepted as a useful alternative to the LMA Classic [3–9], it has only recently become available in paediatric sizes and its clinical performance has yet to be studied extensively in children. Beylacq et al. [12] conducted the first observational study of the i-gel in children and demonstrated that it is a reliable supraglottic airway device in terms of ease of insertion, adequate seal pressure (mean value of 25 cmH2O), and a low complication rate. In addition, the i-gel was inserted on the first attempt in all patients and the authors concluded that there was no learning curve required to become proficient at i-gel insertion. However, the study was conducted with adult-sized i-gels in a group of patients with an average age of 12 years. An entirely paediatric-sized i-gel study was recently conducted by Theiler et al. [13] in which the i-gel was compared with the Ambu® AuraOnce laryngeal mask device in children with an average age of 6 years. The i-gel performance data from this study were comparable to those of the AuraOnce although only three small children (less than 10 kg body weight) were included in the study and smaller children are known to be more vulnerable to complications arising from supraglottic airway devices [10, 11]. Furthermore, there are few clinical data to support the efficacy or safety of modified LMAs, including the AuraOnce laryngeal mask, in children [1], and in the present study we compared the i-gel with the LMA Classic in a relatively large number of infants and children.

Easy insertion and shorter insertion times influence the feasibility of supraglottic airway devices. Our results demonstrate shorter insertion times for the i-gel compared with the LMA Classic, probably because the less flexible stem of the i-gel makes insertion easy and there is no need for cuff inflation. When comparing insertion times it should be remembered that the definition varies between different studies; in a cohort study evaluating the paediatric i-gel [14] the median insertion time for the i-gel was 14 s which was similar to our results and was defined as the time from picking up the device to the first capnography upstroke after insertion, whereas in the study by Theiler et al. [13] insertion time included the time taken to fix the device with tape. Modified laryngeal mask airways LMAs, such as the LMA Supreme or AuraOnce, may be easier to insert than the LMA Classic because of stiffer, angulated stems. In a comparison of the i-gel with LMAs in adults, the time taken to first adequate ventilation through the i-gel was comparable with that through the LMA Classic [6], but longer than with the LMA Supreme [8, 19]. Insertion of the LMA Classic in paediatric patients is known be more difficult than in adults and alternative techniques for insertion have been described, including insertion with partial inflation and lateral, or rotational, insertion [20, 21]. Even when compared with the average time for insertion of the LMA Classic reported in previous studies, the i-gel insertion time in our study can be considered clinically acceptable.

Oropharyngeal leak pressure is one of the characteristics that determine the efficiency of a supraglottic airway device. In our study, the i-gel demonstrated similar oropharyngeal leak pressures and peak inspiratory pressures with 10 ml.kg−1 tidal volume to the LMA Classic, and was comparable with the LMA Classic in other studies [22, 23]. In Beringer et al.’s cohort study [14], the median leak pressure of the i-gel was 20 cmH2O which was comparable to our result of 22 cmH2O. In a study comparing the i-gel with the AuraOnce [13], the average leak pressure was higher with the i-gel (22 cmH2O vs 19 cmH2O), but the difference was not considered clinically important. Therefore, the paediatric-sized i-gel can be considered to have an acceptable pharyngeal seal in children of various ages. Furthermore, the LMA Classic has inherent limitations, including the need to inflate the cuff and adjust the cuff pressure for adequate airway sealing. The LMA Classic cuff pressure should be adjusted after inflation because too low, or too high, a pressure within the cuff significantly affects the quality of airway sealing [24–26] and in this respect the i-gel may be more convenient to use than the LMA Classic.

One of the advantages of the i-gel was that it provided an improved glottic view, and similar results were reported in Beringer et al.’s cohort study as well as a number of adult studies [6, 14, 19]. Although there is little or no correlation between device position as assessed fibreoptically [27, 28], the fibreoptic score confirms that the i-gel occupies a favourable anatomical location to ensure unimpeded ventilation or facilitate the passage of a tracheal tube. Recently, supraglottic airway devices have been recommended as a conduit for tracheal tube insertion in cases of difficult intubation [29], so an improved glottic view is useful.

Several limitations exist with this study. Firstly, only children with normal airway anatomy were included and our findings cannot be extrapolated to different groups. Secondly, unblinded observers collected all the data. Thirdly, the anaesthetist inserting the supraglottic airways had considerably more experience in inserting the LMA Classic than the i-gel and this may have introduced bias into the results. Fourthly, a characteristic of the i-gel is the presence of a gastric channel, but we did not attempt to verify its usefulness with regard to gastric tube insertion.

In conclusion, our study demonstrates that the i-gel is an effective supraglottic airway device for use in children and has few complications. In comparison with the LMA Classic it is quicker to insert and provides better fibreoptic views of the glottis.

Competing interests

No external funding and no competing interests declared.