A cohort evaluation of the paediatric i-gel™ airway during anaesthesia in 120 children†
Presented in part at the Association of Paediatric Anaesthetists of Great Britain and Ireland Annual Scientific Meeting, Torquay, 2011.
We studied the i-gel™ in 120 anaesthetised children (92 boys, 28 girls; median (IQR [range]) age (3 -7 [0.4 -13]) years and weight 19 (15–26 [7–35]) kg) to assess efficacy and usability. Insertion was successful on the first/second/third attempt in 110/8/1 children and failed in one child. Median (IQR [range]) insertion time was 14 (9–16 [6–200]) s. Manual ventilation was possible in all cases, although excess leak precluded a tidal volume above 7 ml.kg−1 in three children. Fibreoptic inspection through the i-gel revealed a clear view of the vocal cords in 40 out of 46 cases (87%). Median (IQR [range]) leak pressure was 20 (16–26 [8–30]) cmH2O. During maintenance of anaesthesia, 16 manipulations were required in 11 children to improve the airway. One child regurgitated without aspirating. Other complications and side effects were infrequent. The i-gel was inserted without complications, establishing a clear airway and enabling spontaneous and controlled ventilation, in 113 (94%) children.
You can respond to this article at http://www.anaesthesiacorrespondence.com
The i-gel™ (Intersurgical, Wokingham, UK) is a single-use supraglottic airway device, which became available in CE marked paediatric sizes in March 2009 and was officially launched for use in January 2010. The i-gel differs from most supraglottic airway devices in that the device ‘mask’ has no inflatable cuff, but is made of an elastomer gel . The stem of the device incorporates a bite block and is elliptical in cross-section, so as to reduce axial rotation. An integral drain tube runs from the tip of the device to its base, similar to the ProSeal™ laryngeal mask airway (PLMA, Intavent Direct, Maidenhead, UK). This is intended to separate the airway from the gastrointestinal tract resulting in three potential advantages over more traditional supraglottic airway devices: allowing venting of regurgitated gastric contents; reducing gastric insufflation during controlled ventilation; and allowing easy insertion of a gastric tube.
Evaluation of the i-gel in adult patients has shown that it is easy to insert and provides an effective airway in the majority of patients [2–4]. We performed a prospective observational study to investigate the efficacy and usability of the i-gel in children. Our primary aim was to determine the proportion of children in whom the i-gel could be inserted without complications, establishing a clear airway and enabling spontaneous and controlled ventilation. Secondary outcomes of interest were ease of insertion, time taken for insertion, incidence of complications, airway seal pressure, fibreoptic view obtained through the device and the incidence and nature of postoperative sequelae.
Local research ethics committee approval was obtained for the study, which was performed at two sites: the Bristol Royal Hospital for Children and the Royal United Hospital, Bath. Written informed consent was obtained in all cases: for older children, from the child and parent and for younger children, from the parents only. Healthy children (ASA physical status 1-2), weighing between 5 and 35 kg, scheduled for elective surgery under general anaesthesia using a supraglottic airway device, were considered eligible. Children at risk of regurgitation or with known airway abnormalities were not studied.
All the investigators were consultant anaesthetists with at least 10 years’ anaesthetic experience. All had previously used the i-gel in children. Gaseous or intravenous induction of anaesthesia was performed by a consultant anaesthetist. Once an adequate depth of anaesthesia was achieved (no eyelash reflex and no response to jaw thrust) , the appropriately sized i-gel was inserted as per the manufacturer’s instructions . A thin layer of water-based lubricant was applied to the back, sides and front of the mask. The child was placed in the ‘sniffing the morning air position’ and whilst grasping the integral bite block, the i-gel was directed posteriorly against the hard palate and advanced with continuous gentle pressure until resistance was felt. Manipulations such as jaw thrust or slight twisting of the device in the oropharynx were performed as required to aid insertion. If an acceptable airway was not obtained after three attempts, an alternative device was used. We collected data relating to the performance of the device during insertion, maintenance of anaesthesia, device removal and following recovery from anaesthesia.
During insertion, we recorded the number of insertion attempts, whether an alternative device was required, the time taken to establish an effective airway (from picking up the device to first capnography upstroke after insertion), the number and type of airway manipulations required to establish an airway (such as neck extension or jaw thrust) and complications of insertion (such as coughing or laryngeal spasm).
Once the i-gel was inserted, we assessed the adequacy of ventilation through the device (demonstrated by chest movement, stable oxygen saturations, square capnograph trace and whether, using a maximum inspired tidal volume of 10 ml.kg−1, it was possible to achieve an expired tidal volume of 7 ml.kg−1). The anatomical position was evaluated by examining the anterior neck for midline swelling, and (where a fibreoptic bronchoscope was available) assessing the fibreoptic view through the airway port. The fibreoptic view was classified as: grade 1, clear view of vocal cords; grade 2, only arytenoids visible; grade 3, only epiglottis visible; grade 4, no laryngeal structures visible . Airway leak pressure was measured using the manometer on the anaesthetic breathing system with the pressure-relieving valve set at 30 cmH2O and a fresh gas flow of 3 l.min−1. The location of any gas leak was recorded.
During maintenance of anaesthesia, we recorded the following: surgical duration; airway manipulations required to maintain an adequate airway (such as neck extension or jaw thrust); airway complications (such as oxygen desaturation below 92% or regurgitation); whether the device was removed during the procedure; and an assessment of the quality of the airway during maintenance (clear throughout, intermittent partial obstruction, intermittent complete obstruction or complete obstruction).
Clinical judgement was used to determine the best time for device removal. We recorded whether the device was tolerated well during emergence, whether there was blood visible on the device, whether secretions were troublesome and any complications of removal.
A structured interview of the parents and child, if appropriate, was performed immediately following recovery and 24 h later, to elicit postoperative side effects such as sore throat, hoarse voice, oropharyngeal swelling, neck and jaw changes, dysphagia and changes in hearing. If present, these were graded as mild, moderate or severe.
Finally, the anaesthetist subjectively scored the overall usefulness of the device for each patient as excellent, good, fair, poor or inadequate.
Results were analysed using Microsoft Excel Spreadsheet (Microsoft Corporation, Redmond, WA, USA). Post hoc analysis examining differences between the three sizes was performed using the chi-squared and Fisher’s exact tests for categorical data, ANOVA for continuous data and Scheffé analysis of variance. A p value of < 0.05 was considered statistically significant.
We studied 92 boys and 28 girls, with a median (IQR [range]) age of 5 (3–7 [0.4–13]) years and weight of 19 (15–26 [7–35]) kg. Ninety-four children were classified as ASA 1, and 26 as ASA 2. The surgical speciality was urology for 55 children, general surgery for 35, orthopaedics for 23, radiology for four and ear, nose and throat for three children. The median (IQR [range]) duration of procedure was 30 (25–45 [8–122]) min. A size 1.5 i-gel was used for 13 children (11%), size 2 for 70 children (58%) and size 3 for 37 children (31%) (Table 1).
Table 1. Results for each size of i-gel showing: insertion data; quality of airway; and leak pressure. Values are number, number (proportion) or mean (IQR [range]).
| 1st attempt||11 (85%)||67 (96%)||32 (86%)||110 (92%)|
| 2nd attempt||1 (8%)||2 (3%)||5 (14%)||8 (7%)|
| 3rd attempt||0||1 (1%)||0||1 (1%)|
|Failed insertion||1 (8%)||0||0||1 (1%)|
|Clear airway||11 (92%)||65 (93%)||37 (100%)||113 (95%)|
|VT > 7 ml.kg−1 achieved||11 (92%)||69 (99%)||35 (94%)||115 (96%)|
|Leak pressure; cmH2O||26 (20–30 [15–30])||20 (15–27 [8–30])||20 (17–24 [12–30])||20 (16–26 [8–30])|
Insertion and subsequent ventilation were successful on the first attempt in 110 children (92%), the second attempt in eight children (7%) and the third attempt in one child. Insertion was unsuccessful after three attempts in one child for whom an alternative supraglottic airway device was used: this child weighed 10 kg. Median (IQR [range]) insertion time was 14 (9–16 [6–200]) s. Twenty-seven children required a total of 35 minor manipulations to achieve a satisfactory airway (median (range) 0 (0–3)). Manipulations required, in descending order of frequency, were jaw thrust, in/out procedure, repositioning, chin lift and neck extension. Complications during insertion were observed in three children: one developed laryngospasm; one developed stridor; and one child moved (Table 2).
Table 2. Number of children experiencing airway complications during insertion, maintenance and removal of the i-gel. Values are number.
|Spo2 < 92%||0||0||0|
|Loss of airway||0||0||2|
|Loss of tooth||0||0||1|
Following successful insertion, all 119 children (100%) had adequate chest movement and stable oxygen saturations. However, it was not possible to generate an expired tidal volume of 7 ml.kg−1 in three children (3%), and seven children (6%) did not have a square capnograph trace. Midline filling of the anterior neck was observed in 110 children. Fibreoptic examination through the i-gel was performed in 46 children (Table 3). The median (IQR [range]) airway leak pressure was 20 (16–26 [8–30]) cmH2O. Thirteen children (16%) had no leak with an airway pressure of 30 cmH2O. For the other children, the gas leaked through the drain tube in 30 children (37%), through the mouth in 43 children (52%) and by a combination of the drain tube and mouth in eight children (10%). Data regarding gas leak were incomplete for the remaining 38 children.
Table 3. Fibreoptic view through each size of i-gel. Values are number (proportion).
|Clear view of vocal cords||4 (80%)||26 (90%)||10 (83%)||40 (87%)|
|Only arytenoids visible||0||2 (7%)||1 (8%)||3 (7%)|
|Only epiglottis visible||1 (20%)||1 (3%)||1 (8%)||3 (7%)|
Three i-gels were removed during the surgical procedure. In one child, regurgitated stomach contents were noted in the drain tube. There was no evidence of laryngeal soiling or aspiration on fibreoptic inspection. The child’s trachea was subsequently intubated and recovery was uneventful. The i-gel was removed in two spontaneously breathing children on surgical request, as partial obstruction was making the surgery more challenging due to increased abdominal excursion.
To improve the quality of the airway during maintenance, 16 airway manipulations were required in 11 children (median (range) number of manipulations per child 0 (0–3). Manipulations required, in descending order of frequency, were jaw thrust, chin lift, neck extension, in/out procedure and neck flexion. In addition to the child with regurgitation, complications during maintenance were observed in three other children: one child coughed; one developed stridor; and one child developed worsening of their chronic wheeze (Table 2). The anaesthetist graded the quality of the airway as ‘clear throughout’ in 109 children (92%) and ‘intermittent partial obstruction’ in 10 children (8%).
Following completion of the surgical procedure, the i-gel was removed before the child was fully awake in 76 children (66%) and left until the child was fully awake in 40 children (34%). During removal of the i-gel, secretions were troublesome in four children and blood was visible on four i-gel cuffs. Complications of removal were experienced in six children (5%): loss of airway in two; laryngospasm in two; coughing in one; and loss of a loose tooth in one (Table 2).
Immediately postoperatively, six side effects were detected in five children. These were mild hoarse voice (one), mild sore throat (three), moderate sore throat (one) and mild dysphagia (one). Twenty-four hours later, 11 adverse effects were detected in 10 children. These were mild sore throat (two), moderate sore throat (two), dry throat (one), mild hoarse voice (two), moderate hoarse voice (one), mild neck ache (two) and mild dysphonia (one).
The overall usefulness of the device was graded as excellent in 52 children (44%), good in 48 (40%), fair in 14 (12%), poor in three (2%) and inadequate in three (2%).
Analysis of the difference among the three sizes of i-gel (1.5, 2 and 2.5) revealed that there was no statistically significant difference in the number of insertion attempts (p = 0.10), fibreoptic view through the device (p = 0.61), leak pressure (p = 0.05), or complications at any time (p = 0.25). The only statistically significant finding was that the duration of surgery was significantly longer for children with the 1.5 i-gel (p = 0.001).
These results demonstrate that the i-gel was ‘fit for purpose’ in this cohort of 120 children. With respect to our primary aim, the i-gel was inserted without complications, establishing a clear airway and enabling spontaneous and controlled ventilation in 113 children (94%). Of the other seven children (6%), there was only one child for whom the i-gel did not provide a useable airway and had to be exchanged for an alternative device. The device was easy to insert in the majority of patients, associated with a small number of minor complications and few postoperative side effects.
At the time of writing, there is one published evaluation of the i-gel in children. Beylacq et al. . studied 50 older children, all weighing over 30 kg with a median age of 12 years. The authors reported first time insertion success of 100% and a mean leak pressure of 25 cmH20 (measured with a fresh gas flow of 6 l.min−1). The i-gel provided a satisfactory airway for all patients with minimal complications. However, it is difficult to compare the results of this study directly with those of our own, as children in our cohort were considerably younger (median age 5 vs 12 years) and weighed less (mean weight 20 vs 44 kg); supraglottic airway device performance is known to differ considerably in smaller children.
By far the most widely studied supraglottic airway devices in children are the classic laryngeal mask airway (cLMA, Intavent Direct) and the PLMA. The cLMA is the original supraglottic airway device, first introduced in 1983 . Devices based on this design are the most widely used supraglottic airway devices in paediatric anaesthetic practice . The PLMA is a modified version of the cLMA, described as a second generation device  due to its separation of the airway and gastrointestinal tracts. In contrast to the i-gel, both the cLMA and the PLMA are reusable following autoclaving.
The insertion success rate in our study of 98% within two attempts is very similar to that in studies of the cLMA and PLMA in children, with reported success rates of 98%  and 99–100% [12–14], respectively. In those children for whom insertion was successful, the i-gel provided a clear airway in 92% of children and a partially obstructed airway in 8% of children. This is similar to the rate reported for the PLMA (95% and 5%, respectively) , but inferior to that reported for the cLMA (98% and 2%, respectively) .
Fibreoptic assessment of the laryngeal view through the i-gel was performed in 46 children (limited availability of equipment unfortunately precluded this assessment for all cases). This assessment showed that the vocal cords were clearly visible in 87% of cases (80%/90%/83% for sizes 1.5/2/2.5, respectively). Although our numbers are small, these results compare favourably with those from a recent study assessing the fibreoptic view through paediatric sized laryngeal masks, which demonstrated a clear view of the glottic aperture in 58%/73%/78% for sizes 1.5/2/2.5, respectively . Goldmann et al. reported obtaining a view of the cords through sizes 1.5/2/2.5 cLMA in 53%/70%/90% of cases, respectively, and through sizes 1.5/2/2.5 PLMA in 83%/87%/100% of cases, respectively [17–19]. Lopez-Gil et al. obtained a view of the vocal cords in 99% of the 80 children anaesthetised with a size 2 or 3 PLMA . As different scoring systems were used for the laryngeal view in the different studies, it is impossible to compare these values directly. However, our results suggest that the fibreoptic view through the i-gel is similar to that obtained through the PLMA and superior to that obtained through the cLMA in children.
Airway leak pressure is often used as a surrogate marker of the quality of airway seal; however, comparison of leak pressures is complicated by differences in methodology and presentation of results. We report median leak pressures of 26, 20 and 20 cmH2O for i-gel sizes 1.5, 2.0 and 2.5, respectively. These are higher than reported for the paediatric cLMA (15–19 cmH2O) [17–19], but lower than reported for the adult i-gel (24–30 cmH2O) [2, 3, 20, 21] and the paediatric PLMA (Kelly et al. used exactly the same methodology in their assessment of the PLMA and determined median leak pressures of 23, 25 and 30 cmH2O for sizes 1.5, 2 and 2.5, respectively) . Goldmann et al. used similar methodology and reported mean leak pressures of 26.7, 18.8 and 22.6 cmH2O for PLMA sizes 1.5, 2 and 2.5, respectively [17–19]). These results would suggest that the seal formed by the i-gel in children is superior to that seen with the cLMA, but inferior to that seen with the PLMA. Adult sizes of the i-gel would seem to form a better airway seal than paediatric sizes.
It is notable that in one case, regurgitation was identified by seeing gastric contents in the drain tube, when there were no other signs of regurgitation. Like other second generation supraglottic airway devices, the drain tube of the i-gel may serve the dual functions of giving an early warning that regurgitation has occurred, as well as potentially preventing aspiration.
Complications were identified in 13 children in our study (11%), which is a similar rate to that reported for the paediatric cLMA and PLMA (12–23% and 6–11%, respectively) [11, 12, 14, 15]. However, anaesthetic technique is known to affect complication rates, making it difficult to draw firm conclusions.
Supraglottic airway devices may cause trauma to the mucosa of the airway, causing bleeding, or postoperative side effects such as sore throat. Interviewing parents and young children may be unreliable, and can underestimate frequency. We identified adverse effects in five children (4%) immediately postoperatively, and in 10 after 24 h (8%). Adult studies consistently report rates of postoperative sore throat of 10–12% for the cLMA, PLMA and i-gel [2, 22, 23]. We reported blood on 3% of i-gels following removal. This compares with 3–6% for the cLMA [17, 18] and 0–3% for the PLMA [13, 17, 18]. Use of the i-gel in adults has been associated with less visible blood than alternative devices , which has led to speculation that the gel-filled cuff is less traumatic to the airway compared with more traditional air-filled cuffs. However, our study does not support this theory in children.
A frequent comment from the investigators in this study was that the stem of the i-gel was too long, making the device awkward and more prone to becoming dislodged. We found that careful taping was required to hold the i-gel in place, so as to maintain a good airway seal. There was a learning curve associated with securing the device proficiently.
Our study has three main limitations. First, as an efficacy study, we did not aim to compare the i-gel with other supraglottic airway devices. Randomised controlled trials will be necessary to make firm conclusions about relative performance. Secondly, we performed this study on healthy children with normal airway anatomy and therefore cannot extrapolate our findings to different groups. Thirdly, we are unable to comment on safety of the device, which will require evaluation in many thousands of children during routine clinical practice.
The frequency with which supraglottic airway devices are used in paediatric practice is not known. In the recent NAP4 study , it was noted that the most common cause of an airway-related death was pulmonary aspiration. Although the incidence was small, almost half of such events occurred during the maintenance phase of anaesthesia with a standard LMA laryngeal mask in place. There were no children among these patients, but the availability of an increasing range of second generation supraglottic airway devices (designed to reduce the risk of aspiration) raises the question as to whether they are genuinely safer and what, if any, disadvantages they have. This small study raises the possibility of potential benefit (better airway seal, presence of a drain tube, good anatomical positioning) and has identified no clear disadvantages to their use. Randomised controlled trials are clearly indicated.
Our study shows that the i-gel is an effective supraglottic airway device for use in children. It is easy to insert and has few complications. The seal pressure is superior to the cLMA and inferior to the PLMA. The incorporation of the gastric drain tube offers potential advantages over more traditional supraglottic airway devices.
The authors thank Rosemary Greenwood (Research design service, University Hospitals Bristol NHS Foundation Trust) for her assistance with statistical analysis.
The i-gels used in this study were donated free of charge by the manufacturers who had no further involvement in the study. 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.