Anaesthesia 2014
doi:10.1111/anae.12746
Review Article A systematic review and meta-analysis of the i-gelâ vs laryngeal mask airway in children G. J. Choi,1 H. Kang,2,3 C. W. Baek,4 Y. H. Jung,4 Y. C. Woo4 and Y. J. Cha5,6 1 Clinical Fellow, 2 Associate Professor, 4 Professor, Department of Anaesthesiology and Pain Medicine, 5 Professor, Department of Laboratory Medicine, 3 Researcher, 6 Manager, Medical Device Clinical Trials Centre, Chung-Ang University College of Medicine, Seoul, Korea
Summary We systematically reviewed randomised controlled trials of the i-gelâ vs different types of laryngeal mask airway in children. We included nine studies. There was no evidence for differences in: rate of insertion at first attempt; insertion time; ease of insertion; or gastric tube insertion. The mean (95% CI) oropharyngeal leak pressure was 3.29 (2.25–4.34) cmH2O higher with the i-gel, p < 0.00001. The relative rate (95% CI) of a good fibreoptic view through the i-gel was 1.10 (1.01–1.19), p = 0.02. There were no significant differences in the rates of complications, except for blood on the airway, relative rate with the i-gel 0.46 (0.23–0.91), p = 0.02. We concluded that the clinical performance of the i-gel and LMA was similar, except for three outcomes that favoured the i-gel. .................................................................................................................................................................
Correspondence to: H. Kang Email: [email protected] Accepted: 19 April 2014
Introduction Supraglottic airway devices are commonly used to manage children’s airways because they are typically easy to use and they avoid problems specific to tracheal intubation [1]. Since the laryngeal mask airway (LMA) was introduced into clinical practice in 1988 [2], various types of LMAs have been developed, the safety and efficacy of which have been established in adults and children. The paediatric i-gelâ is a relatively new and disposable supraglottic airway device for children, which is made of a soft gel-like elastomer with a non-inflatable cuff and a channel for gastric catheter placement [3]. The purpose of this review was to compare systematically the paediatric i-gel with various types of LMA in children. The primary focus was the clinical performance of the airways: the rate of insertion at first © 2014 The Association of Anaesthetists of Great Britain and Ireland
attempt; the ease of insertion; the insertion time; the oropharyngeal leak pressure; the fibreoptic view; and the rate of successful gastric tube insertion. The frequency of other complications, where reported, was also assessed, and the limitations of the data were reviewed.
Methods We used recommended methods for this systematic review [4]. Two authors (GJC and HK) independently searched the Embase, MEDLINE and CENTRAL databases to September 2013 and manually searched the references of retrieved studies. The Appendix lists the search terms used for MEDLINE, with equivalent search strategies used for the other databases. Both authors independently retrieved peer-reviewed randomised controlled trials (RCTs), including crossover designs, that compared the i-gel with LMAs in 1
Anaesthesia 2014
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
193 records identified through database searching (MEDLINE = 42, Embase = 151, CENTRAL = 0)
154 records
after 39 duplicates removed
154 records
139 records excluded
screened with titles and abstracts
15 full-text articles
Excluded (n = 6);
assessed for eligibility
conference proceedings (n = 4) children not studied (n = 2)
Nine studies included for quantitative synthesis (meta-analysis)
Figure 1 Flow diagram showing the number of abstracts and articles identified and evaluated during the review.
Table 1 Characteristics of studies comparing the i-gel with laryngeal mask airways (LMAs). Values are numbers, mean (SD), median (range) or mean [range]. RCT
Airway
n
Age; years
Weight; kg
Airway size
Type of surgery
NBD
Fukuhara et al. 2013 [7] Jagannathan et al. 2013 [10]
i-gel PLMA i-gel SLMA
67 67 83 85
6.9 6.4 5.0 5
(4.6) (4.5) [0.3–11.5] [0.3–11.5]
24 22 20.0 20.0
(13) (13) [5.0–50.6] [6.2–47.4]
1.5, 2, 2.5, 3
No
Tokgoz et al. 2013 [14] Das et al. 2012 [6]
i-gel PLMA i-gel cLMA PLMA i-gel PLMA i-gel cLMA i-gel PLMA i-gel cLMA PLMA i-gel ALMA
95 90 30 30 30 51 51 50 49 30 30 40 40 40 106 102
4.1 4.7 3.2 2.8 2.9 3.6
(3.2) (3.3) [2.0–4.7] [2.2–4.2] [2.0–4.7] [1.5–6.0]
14.2 14.0 14.7 15.5 15.8 16
(6.3) (8.2) [10.0–20.0] [12.0–20.0] [11.0–19.0] [10–25]
3.0 3.4 7.3 8.1 3.8 3.3 3.6 6.3 6.2
[0.2–9.0] [0.1–9.0] [5.0–10.0] [6.0–10.0] [2.3–5.0] [2.0–4.7] [2.2–4.7] (3.7) (4.0)
16.1 16.3 24.7 29.3 14.5 14.2 14.4 24.7 24.7
[6.2–30.0] [6.0–28.0] [20.0–30.0] [25.0–32.0] [11.0–20.0] [10.0–18.0] [10.0–18.0] (11.2) (11.6)
Ophthalmologic, inguinal, urologic, plastic Urologic, radiological, general, ophthalmologic, orthopaedic Ophthalmologic, low abdominal, urogenital Low abdominal, inguinal (herniotomy, circumcision), orthopaedic Inguinal hernia, minor orthopaedic Elective
Gasteiger et al. 2012 Lee et al. 2012 Mitra et al. 2012 Goyal et al. 2011 Theiler 2011 et al. [13]
[8] [11] [12] [9]
1.5, 2, 2.5, 3
1.5, 2, 2.5 2
2 1.5, 2, 2.5 2.5 2
1.5, 2, 2.5, 3
Low abdominal, orthopaedic, inguinal Herniotomy, circumcision, orthopaedic Urologic, orthopaedic, visceral, dermatologic
No
Yes Yes
No Yes Yes No
No
ALMA, Ambu AuraOnce LMA; cLMA, classic LMA; PLMA, ProSeal LMA; SLMA, Supreme LMA; NBD, neuromuscular blocking drug. 2
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
children. We did not restrict language of publication, type or size of LMA. Disagreements were resolved by discussion with two senior authors (YCW and YJC). Two authors (CWB and YHJ) independently extracted: the name of the study’s first author; the year of publication; any commercial sponsorship; whether the trial was registered; the type of surgery; use of neuromuscular blocking drugs; the type and size of airways; the number of children, and their summary age and weight. Outcomes were: insertion time; rate of insertion at first attempt; ease of insertion (‘easy’ or ‘no resistance’ vs other categories); oropharyngeal leak pressure; fibreoptic glottis view; and rate of failed gastric tube insertion. We recorded complications, including: coughing; breath-holding; desaturation; laryngospasm or bronchospasm; dental or lip trauma; regurgitation; blood on the devices; tongue numbness; hoarseness and sore throat; and postoperative nausea and vomiting (PONV). We performed subgroup analyses for different types of LMA. We contacted corresponding authors for missing data. Two authors (GJC and HK) graded risks of bias in seven domains as ‘high’, ‘low’ or ‘unclear’: random sequence generation; allocation concealment; blinding of children and their parents; blinding of outcome assessors during anaesthesia; blinding of outcome assessors after anaesthesia; incomplete outcome data; and selective reporting. We calculated relative rates (RR) for dichotomous outcomes and mean differences for continuous outcomes, both with their 95% confidence intervals (CI) using Review Manager software (version 5.2; The Cochrane Collaboration, Oxford, UK) and Comprehensive Meta-Analysis software (version 2.0; Biostat, Englewood, NJ, USA). We divided i-gel rates by the number of LMA types with which it was compared in a RCT [4]. We defined meta-analysis heterogeneity as a chi-squared test p value < 0.10 and the I2 value > 50%, for which we used a randomeffects model in preference to a fixed-effect model [4, 5]. We used Egger’s test to assess funnel plot asymmetry for small-studies effects [4]. We carried out subgroup analyses for: assessor experience; type of LMA; and use of neuromuscular blocking drug. We computed the correlation coefficient for crossover RCTs, from RCTs excluded from this meta-analysis in the absence of sufficient data in included RCTs: we © 2014 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2014
conducted sensitivity analyses trying different values of correlation coefficients [4]. We also carried out sensitivity analyses to evaluate the influence of a single study on the overall effect estimate by excluding one study at a time in case of heterogeneity > 50%.
Results We included nine RCTs (Fig. 1 and Table 1), all of which were published in English [6–14]. Three RCTs [7, 8, 10] were registered. Three RCTs [6, 9, 11] reported no sponsorship and three RCTs did not mention sponsorship [7, 12, 14]. Commercial companies
Figure 2 Risks of bias summary table: review authors’ judgements about each risk of bias item for each included study. 3
Anaesthesia 2014
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
provided devices free of charge in two RCTs [10, 13]. Two authors of one RCT [8] were consultants for the Laryngeal Mask Company. The i-gel was compared with four LMA types: the ProSealâ in four RCTs [6–8, 12]; the Classicâ (cLMA) in one [11]; the LMA Supremeâ in one [10]; and the AuraOnceâ in one [13]. The LMA sizes were: 2 [6, 8, 9]; 2.5 [12]; and various others [10, 11, 13, 14]. Four RCTs used neuromuscular blocking drugs [6, 11, 12, 14]. Authors provided unpublished mean (SD) values for insertion time and leak pressure for two RCTs [10, 11]. We calculated a correlation coefficient for the crossover study [8] from a similar but excluded RCT [15]. Figure 2 summarises risks of bias. There was no evidence of funnel plot asymmetry for any outcome (graphs not shown). There was no evidence for a difference in: rate of insertion at the first attempt, RR 1.00 (95% CI 0.97–1.04), p = 0.78; rate of easy insertion (Fig. 3), the subgroup analysis of which demonstrated
that observers were more likely than users to categorise i-gel insertion as ‘easy’, p = 0.008; or speed of insertion (Fig. 4), the subgroup analysis of which reduced heterogeneity by grouping RCT results into i-gel vs ProSeal and i-gel vs other LMAs, particularly after the exclusion of one extreme result (from I2 of 98% to 0%) [14]. Oropharyngeal leak pressure was higher with the i-gel but was accompanied by substantial heterogeneity that was not decreased in subgroup analyses based on LMA type (Fig. 5) or use of neuromuscular blockade (data not shown). There was no evidence of interaction in these two subgroup analyses, p = 0.86 and 0.07, respectively. The fibreoptic view was slightly better through the i-gel, a result that was unaffected by exclusion of the most extreme result, although the I2 heterogeneity decreased to 36% (Fig. 6). There was no evidence for a difference in rates of successful gastric tube insertion, RR 0.99 (95% CI 0.97– 1.01), p = 0.30 [6, 8, 10, 12, 14]. There was no evidence
Figure 3 Forest plot of the i-gel vs LMA for the rate of easy insertion (no manipulation or resistance): RCTs were grouped according to whether ease was assessed by an observer or the inserter. The point of interest is that although the subgroup analysis did not expose statistical differences between i-gel and LMA, it did demonstrate that RCTs in which observers scored ease of insertion were more likely to score the i-gel as easy than RCTs in which the user scored ease of insertion. LMA, laryngeal mask airway; RCT, randomised controlled trials; M–H, Mantel–Haenszel. 4
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
Anaesthesia 2014
Figure 4 Forest plot of the i-gel vs LMA for insertion time: RCTs were grouped according to the type of LMA: cLMA, classic; PLMA, ProSeal; SLMA, Supreme; ALMA, Ambu AuraOnce. The point of interest is that the type of LMA studied in each RCT was associated with the mean difference in time taken to insert the i-gel and LMA. IV, inverse of variance; SE, standard error. for differences in rates of: coughing, RR 1.02 (95% CI 0.53–1.94), p = 0.96 [7, 9–13]; breath-holding, RR 0.60 (95% CI 0.15–2.43), p = 0.47 [7, 9, 11]; desaturation, RR 0.91 (95% CI 0.34–2.48), p = 0.86 [7–14]; laryngospasm or bronchospasm, RR 0.62 (95% CI 0.25– 1.54), p = 0.30 [6, 7, 9–14]. Blood was seen less often on the i-gel than on the LMAs (Fig. 7). No participant complained of a numb tongue [6, 9, 12, 13] or hoarseness [6, 7, 9, 12, 13]. There was no evidence for differences in the rates of: sore throat, RR 0.70 (95% CI 0.10–4.72), p = 0.37 [6, 7, 9, 10, 12–14]; PONV, RR 0.75 (95% CI 0.40–1.41), p = 0.38 [7, 13, 14]. One [14] of five RCTs [6, 7, 9, 13, 14] reported dental or lip trauma for one participant with each device. Regurgitation was reported by one [10] of three RCTs [7, 10, 13] for two participants with i-gels.
Discussion Both types of airway exhibited similarly high rates of insertion at the first attempt. A subgroup analysis for ease of insertion showed disparate assessments by © 2014 The Association of Anaesthetists of Great Britain and Ireland
observers (more likely to favour the i-gel) compared with users (more likely to favour the LMA). This significant difference between subgroups might have had nothing to do with the assessor, as it might have been confounded by something else that systematically differed between these two groups of RCTs: researchers should collect assessments by both user and observer for ease of insertion. The heterogeneity for airway insertion time was substantial: the pooled insignificant difference belies two RCTs [10, 13] with a pooled estimate of faster LMA insertion, compared with three RCTs [8, 11, 14] with a pooled estimate of faster i-gel insertion and one RCT [7] that reported the same insertion times. This disparity might be related to slower insertion of the ProSeal LMA compared with other LMAs, although the caveats of subgroup analysis apply. The disparity does not correspond with the different measurements used by these RCTs for insertion time: from facemask removal to ‘adequate’ ventilation [7, 10, 13, 14]; from facemask removal to first capnographic upstroke [11]; 5
Anaesthesia 2014
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
Figure 5 Forest plot of the i-gel vs LMA for oropharyngeal leak pressure: RCTs were grouped according to the type of LMA: cLMA, classic; PLMA, ProSeal; SLMA, Supreme; ALMA, Ambu AuraOnce. The point of interest is that there was no interaction between the type of LMA studied by RCTs and the mean difference in leak pressure. IV, inverse of variance; SE, standard error.
Figure 6 Forest plot of the i-gel vs LMA for the rate of good fibreoptic view through the airway. The point of interest is that a good fibreoptic view was achieved through 93% of i-gels and through 83% of LMAs, with the pooled effect unaltered by exclusion of the most extreme result [11], although I2 heterogeneity was reduced to 36%. M–H, Mantel–Haenszel. or from holding the airway to successful placement [8]. The interpretation of insertion time is further complicated by RCTs’ incorporating different numbers of insertion attempts in the measurement: one [11]; two [8]; or three [7, 10, 13, 14]. 6
There was also substantial heterogeneity for the i-gel’s greater leak pressure: the hypotheses that this effect varied with neuromuscular blockade and LMA type were not supported in subgroup analyses. Higher airway leak pressures may be needed when ventilation © 2014 The Association of Anaesthetists of Great Britain and Ireland
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
Anaesthesia 2014
Figure 7 Forest plot of the i-gel vs LMA for the rate of blood on the airway. M–H, Mantel–Haenszel.
is difficult, for instance with head-down positioning or in obese patients, or those with pulmonary disease [16]. The rate of good fibreoptic view was 93% through the i-gel and 83% through LMAs, but this has not been associated with ease of ventilation [17, 18– 20]. However, the good fibreoptic view provided by igel might be a useful conduit for tracheal intubation in children in whom direct laryngoscopy has failed [21]. We suggest that the higher rate of blood on the LMA was not clinically significant. The present review has several limitations. Blinding of outcome assessment was difficult as the i-gel looks different to LMAs, with five RCTs failing to blind at all [7, 8, 10, 11, 13]. Although some outcomes were more objective than ease of insertion, they were nevertheless vulnerable to bias introduced by knowledge of what airway was being assessed. Conclusions based upon nine RCTs will be less reliable than meta-analyses that include more studies on more participants. In conclusion, the i-gel was similar to LMAs when used in children and delivered ventilation pressures 3 cmH2O higher than LMAs. Few complications were reported with either airway.
Acknowledgement This study was supported by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (A100054). © 2014 The Association of Anaesthetists of Great Britain and Ireland
Competing interests No competing interests declared.
References 1. White MC, Cook TM, Stoddart PA. A critique of elective pediatric supraglottic airway devices. Pediatric Anesthesia 2009; 19 (Suppl 1): 55–65. 2. Brain AI. The development of the Laryngeal Mask–a brief history of the invention, early clinical studies and experimental work from which the Laryngeal Mask evolved. European Journal of Anaesthesiology Supplement 1991; 4: 5–17. 3. Intersurgical. User guide. http://media.intersurgical.com/igel. com/assets/userguide.pdf (accessed 09/01/2013). 4. Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. UK: The Cochrane Collaboration, 2011. 5. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. British Medical Journal 2003; 327: 557–60. 6. Das B, Mitra S, Jamil SN, Varshney RK. Comparison of three supraglottic devices in anesthetised paralyzed children undergoing elective surgery. Saudi Journal of Anaesthesia 2012; 6: 224–8. 7. Fukuhara A, Okutani R, Oda Y. A randomized comparison of the i-gel and the ProSeal laryngeal mask airway in pediatric patients: performance and fiberoptic findings. Journal of Anesthesia 2013; 27: 1–6. 8. Gasteiger L, Brimacombe J, Oswald E, et al. LMA ProSealTM vs. i-GelTM in ventilated children: a randomised, crossover study using the size 2 mask. Acta Anaesthesiologica Scandinavica 2012; 56: 1321–4. 9. Goyal R, Shukla RN, Kumar G. Comparison of size 2 i-gel supraglottic airway with LMA-ProSealTM and LMA-ClassicTM in spontaneously breathing children undergoing elective surgery. Pediatric Anesthesia 2012; 22: 355–9. 10. Jagannathan N, Sommers K, Sohn LE, et al. A randomized equivalence trial comparing the i-gel and laryngeal mask air7
Anaesthesia 2014
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
way Supreme in children. Pediatric Anesthesia 2013; 23: 127–33. Lee JR, Kim MS, Kim JT, et al. A randomised trial comparing the i-gelTM with the LMA ClassicTM in children. Anaesthesia 2012; 67: 606–11. Mitra S, Das B, Jamil SN. Comparison of size 2.5 i-gelTM with proseal LMATM in anaesthetised, paralyzed children undergoing elective surgery. North American Journal of Medical Sciences 2012; 4: 453–7. Theiler LG, Kleine-Brueggeney M, Luepold B, et al. Performance of the pediatric-sized i-gel compared with the Ambu AuraOnce laryngeal mask in anesthetized and ventilated children. Anesthesiology 2011; 115: 102–10. Tokgoz O, Tufek A, Beyaz SG, et al. Comparison of the efficacies of i-gelTM and LMA-ProSealTM for airway management in pediatric patients. Turkish Journal of Medical Sciences 2013; 43: 208–13. Uppal V, Gangaiah S, Fletcher G, Kinsella J. Randomized crossover comparison between the i-gel and the LMA-Unique in anaesthetized, paralysed adults. British Journal of Anaesthesia 2009; 103: 882–5. Theiler L, Gutzmann M, Kleine-Brueggeney M, Urwyler N, Kaempfen B, Greif R. i-gelTM supraglottic airway in clinical practice: a prospective observational multicentre study. British Journal of Anaesthesia 2012; 109: 990–5. Theiler LG, Kleine-Brueggeney M, Kaiser D, et al. Crossover comparison of the laryngeal mask supreme and the i-gel in simulated difficult airway scenario in anesthetized patients. Anesthesiology 2009; 111: 55–62. Francksen H, Renner J, Hanss R, Scholz J, Doerges V, Bein B. A comparison of the i-gel with the LMA-Unique in non-paralysed anaesthetised adult patients. Anaesthesia 2009; 64: 1118–24. Shariffuddin II, Wang CY. Randomised crossover comparison of the Ambu AuraOnce laryngeal mask with the LMA Classic laryngeal mask airway in paralysed anaesthetised patients. Anaesthesia 2008; 63: 82–5. Inagawa G, Okuda K, Miwa T, Hiroki K. Higher airway seal does not imply adequate positioning of laryngeal mask airways in paediatric patients. Pediatric Anesthesia 2002; 12: 322–6. Weiss M, Engelhardt T. Proposal for the management of the unexpected difficult pediatric airway. Pediatric Anesthesia 2010; 20: 454–64.
Appendix Search terms for MEDLINE 1. randomized controlled trial.pt 2. randomized controlled trial$.mp
8
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.
controlled clinical trial.pt controlled clinical trial$.mp random allocation.mp exp double-blind method/ double-blind.mp exp single-blind method/ single-blind.mp or/1–9 clinical trial.pt clinical trial$.mp exp clinical trial/ (clin$ adj25 trial$).mp ((singl$ or doubl$ or tripl$ or trebl$) adj25 (blind $ or mask$)).mp random$.mp exp research design/ research design.mp or/11–18 10 or 19 Case report.tw. Letter.pt. Historical article.pt. Review.pt. or/21–24 20 not 25 exp laryngeal masks/ laryngeal mask airway.mp. (LMA or ProSeal).mp. (mask adj6 airway).mp. or/27–30 i gel.mp. igel.mp. i-gel.mp. or/32–34 31 and 35 26 and 36
© 2014 The Association of Anaesthetists of Great Britain and Ireland
doi:10.1111/anae.12746
Review Article A systematic review and meta-analysis of the i-gelâ vs laryngeal mask airway in children G. J. Choi,1 H. Kang,2,3 C. W. Baek,4 Y. H. Jung,4 Y. C. Woo4 and Y. J. Cha5,6 1 Clinical Fellow, 2 Associate Professor, 4 Professor, Department of Anaesthesiology and Pain Medicine, 5 Professor, Department of Laboratory Medicine, 3 Researcher, 6 Manager, Medical Device Clinical Trials Centre, Chung-Ang University College of Medicine, Seoul, Korea
Summary We systematically reviewed randomised controlled trials of the i-gelâ vs different types of laryngeal mask airway in children. We included nine studies. There was no evidence for differences in: rate of insertion at first attempt; insertion time; ease of insertion; or gastric tube insertion. The mean (95% CI) oropharyngeal leak pressure was 3.29 (2.25–4.34) cmH2O higher with the i-gel, p < 0.00001. The relative rate (95% CI) of a good fibreoptic view through the i-gel was 1.10 (1.01–1.19), p = 0.02. There were no significant differences in the rates of complications, except for blood on the airway, relative rate with the i-gel 0.46 (0.23–0.91), p = 0.02. We concluded that the clinical performance of the i-gel and LMA was similar, except for three outcomes that favoured the i-gel. .................................................................................................................................................................
Correspondence to: H. Kang Email: [email protected] Accepted: 19 April 2014
Introduction Supraglottic airway devices are commonly used to manage children’s airways because they are typically easy to use and they avoid problems specific to tracheal intubation [1]. Since the laryngeal mask airway (LMA) was introduced into clinical practice in 1988 [2], various types of LMAs have been developed, the safety and efficacy of which have been established in adults and children. The paediatric i-gelâ is a relatively new and disposable supraglottic airway device for children, which is made of a soft gel-like elastomer with a non-inflatable cuff and a channel for gastric catheter placement [3]. The purpose of this review was to compare systematically the paediatric i-gel with various types of LMA in children. The primary focus was the clinical performance of the airways: the rate of insertion at first © 2014 The Association of Anaesthetists of Great Britain and Ireland
attempt; the ease of insertion; the insertion time; the oropharyngeal leak pressure; the fibreoptic view; and the rate of successful gastric tube insertion. The frequency of other complications, where reported, was also assessed, and the limitations of the data were reviewed.
Methods We used recommended methods for this systematic review [4]. Two authors (GJC and HK) independently searched the Embase, MEDLINE and CENTRAL databases to September 2013 and manually searched the references of retrieved studies. The Appendix lists the search terms used for MEDLINE, with equivalent search strategies used for the other databases. Both authors independently retrieved peer-reviewed randomised controlled trials (RCTs), including crossover designs, that compared the i-gel with LMAs in 1
Anaesthesia 2014
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
193 records identified through database searching (MEDLINE = 42, Embase = 151, CENTRAL = 0)
154 records
after 39 duplicates removed
154 records
139 records excluded
screened with titles and abstracts
15 full-text articles
Excluded (n = 6);
assessed for eligibility
conference proceedings (n = 4) children not studied (n = 2)
Nine studies included for quantitative synthesis (meta-analysis)
Figure 1 Flow diagram showing the number of abstracts and articles identified and evaluated during the review.
Table 1 Characteristics of studies comparing the i-gel with laryngeal mask airways (LMAs). Values are numbers, mean (SD), median (range) or mean [range]. RCT
Airway
n
Age; years
Weight; kg
Airway size
Type of surgery
NBD
Fukuhara et al. 2013 [7] Jagannathan et al. 2013 [10]
i-gel PLMA i-gel SLMA
67 67 83 85
6.9 6.4 5.0 5
(4.6) (4.5) [0.3–11.5] [0.3–11.5]
24 22 20.0 20.0
(13) (13) [5.0–50.6] [6.2–47.4]
1.5, 2, 2.5, 3
No
Tokgoz et al. 2013 [14] Das et al. 2012 [6]
i-gel PLMA i-gel cLMA PLMA i-gel PLMA i-gel cLMA i-gel PLMA i-gel cLMA PLMA i-gel ALMA
95 90 30 30 30 51 51 50 49 30 30 40 40 40 106 102
4.1 4.7 3.2 2.8 2.9 3.6
(3.2) (3.3) [2.0–4.7] [2.2–4.2] [2.0–4.7] [1.5–6.0]
14.2 14.0 14.7 15.5 15.8 16
(6.3) (8.2) [10.0–20.0] [12.0–20.0] [11.0–19.0] [10–25]
3.0 3.4 7.3 8.1 3.8 3.3 3.6 6.3 6.2
[0.2–9.0] [0.1–9.0] [5.0–10.0] [6.0–10.0] [2.3–5.0] [2.0–4.7] [2.2–4.7] (3.7) (4.0)
16.1 16.3 24.7 29.3 14.5 14.2 14.4 24.7 24.7
[6.2–30.0] [6.0–28.0] [20.0–30.0] [25.0–32.0] [11.0–20.0] [10.0–18.0] [10.0–18.0] (11.2) (11.6)
Ophthalmologic, inguinal, urologic, plastic Urologic, radiological, general, ophthalmologic, orthopaedic Ophthalmologic, low abdominal, urogenital Low abdominal, inguinal (herniotomy, circumcision), orthopaedic Inguinal hernia, minor orthopaedic Elective
Gasteiger et al. 2012 Lee et al. 2012 Mitra et al. 2012 Goyal et al. 2011 Theiler 2011 et al. [13]
[8] [11] [12] [9]
1.5, 2, 2.5, 3
1.5, 2, 2.5 2
2 1.5, 2, 2.5 2.5 2
1.5, 2, 2.5, 3
Low abdominal, orthopaedic, inguinal Herniotomy, circumcision, orthopaedic Urologic, orthopaedic, visceral, dermatologic
No
Yes Yes
No Yes Yes No
No
ALMA, Ambu AuraOnce LMA; cLMA, classic LMA; PLMA, ProSeal LMA; SLMA, Supreme LMA; NBD, neuromuscular blocking drug. 2
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
children. We did not restrict language of publication, type or size of LMA. Disagreements were resolved by discussion with two senior authors (YCW and YJC). Two authors (CWB and YHJ) independently extracted: the name of the study’s first author; the year of publication; any commercial sponsorship; whether the trial was registered; the type of surgery; use of neuromuscular blocking drugs; the type and size of airways; the number of children, and their summary age and weight. Outcomes were: insertion time; rate of insertion at first attempt; ease of insertion (‘easy’ or ‘no resistance’ vs other categories); oropharyngeal leak pressure; fibreoptic glottis view; and rate of failed gastric tube insertion. We recorded complications, including: coughing; breath-holding; desaturation; laryngospasm or bronchospasm; dental or lip trauma; regurgitation; blood on the devices; tongue numbness; hoarseness and sore throat; and postoperative nausea and vomiting (PONV). We performed subgroup analyses for different types of LMA. We contacted corresponding authors for missing data. Two authors (GJC and HK) graded risks of bias in seven domains as ‘high’, ‘low’ or ‘unclear’: random sequence generation; allocation concealment; blinding of children and their parents; blinding of outcome assessors during anaesthesia; blinding of outcome assessors after anaesthesia; incomplete outcome data; and selective reporting. We calculated relative rates (RR) for dichotomous outcomes and mean differences for continuous outcomes, both with their 95% confidence intervals (CI) using Review Manager software (version 5.2; The Cochrane Collaboration, Oxford, UK) and Comprehensive Meta-Analysis software (version 2.0; Biostat, Englewood, NJ, USA). We divided i-gel rates by the number of LMA types with which it was compared in a RCT [4]. We defined meta-analysis heterogeneity as a chi-squared test p value < 0.10 and the I2 value > 50%, for which we used a randomeffects model in preference to a fixed-effect model [4, 5]. We used Egger’s test to assess funnel plot asymmetry for small-studies effects [4]. We carried out subgroup analyses for: assessor experience; type of LMA; and use of neuromuscular blocking drug. We computed the correlation coefficient for crossover RCTs, from RCTs excluded from this meta-analysis in the absence of sufficient data in included RCTs: we © 2014 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2014
conducted sensitivity analyses trying different values of correlation coefficients [4]. We also carried out sensitivity analyses to evaluate the influence of a single study on the overall effect estimate by excluding one study at a time in case of heterogeneity > 50%.
Results We included nine RCTs (Fig. 1 and Table 1), all of which were published in English [6–14]. Three RCTs [7, 8, 10] were registered. Three RCTs [6, 9, 11] reported no sponsorship and three RCTs did not mention sponsorship [7, 12, 14]. Commercial companies
Figure 2 Risks of bias summary table: review authors’ judgements about each risk of bias item for each included study. 3
Anaesthesia 2014
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
provided devices free of charge in two RCTs [10, 13]. Two authors of one RCT [8] were consultants for the Laryngeal Mask Company. The i-gel was compared with four LMA types: the ProSealâ in four RCTs [6–8, 12]; the Classicâ (cLMA) in one [11]; the LMA Supremeâ in one [10]; and the AuraOnceâ in one [13]. The LMA sizes were: 2 [6, 8, 9]; 2.5 [12]; and various others [10, 11, 13, 14]. Four RCTs used neuromuscular blocking drugs [6, 11, 12, 14]. Authors provided unpublished mean (SD) values for insertion time and leak pressure for two RCTs [10, 11]. We calculated a correlation coefficient for the crossover study [8] from a similar but excluded RCT [15]. Figure 2 summarises risks of bias. There was no evidence of funnel plot asymmetry for any outcome (graphs not shown). There was no evidence for a difference in: rate of insertion at the first attempt, RR 1.00 (95% CI 0.97–1.04), p = 0.78; rate of easy insertion (Fig. 3), the subgroup analysis of which demonstrated
that observers were more likely than users to categorise i-gel insertion as ‘easy’, p = 0.008; or speed of insertion (Fig. 4), the subgroup analysis of which reduced heterogeneity by grouping RCT results into i-gel vs ProSeal and i-gel vs other LMAs, particularly after the exclusion of one extreme result (from I2 of 98% to 0%) [14]. Oropharyngeal leak pressure was higher with the i-gel but was accompanied by substantial heterogeneity that was not decreased in subgroup analyses based on LMA type (Fig. 5) or use of neuromuscular blockade (data not shown). There was no evidence of interaction in these two subgroup analyses, p = 0.86 and 0.07, respectively. The fibreoptic view was slightly better through the i-gel, a result that was unaffected by exclusion of the most extreme result, although the I2 heterogeneity decreased to 36% (Fig. 6). There was no evidence for a difference in rates of successful gastric tube insertion, RR 0.99 (95% CI 0.97– 1.01), p = 0.30 [6, 8, 10, 12, 14]. There was no evidence
Figure 3 Forest plot of the i-gel vs LMA for the rate of easy insertion (no manipulation or resistance): RCTs were grouped according to whether ease was assessed by an observer or the inserter. The point of interest is that although the subgroup analysis did not expose statistical differences between i-gel and LMA, it did demonstrate that RCTs in which observers scored ease of insertion were more likely to score the i-gel as easy than RCTs in which the user scored ease of insertion. LMA, laryngeal mask airway; RCT, randomised controlled trials; M–H, Mantel–Haenszel. 4
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
Anaesthesia 2014
Figure 4 Forest plot of the i-gel vs LMA for insertion time: RCTs were grouped according to the type of LMA: cLMA, classic; PLMA, ProSeal; SLMA, Supreme; ALMA, Ambu AuraOnce. The point of interest is that the type of LMA studied in each RCT was associated with the mean difference in time taken to insert the i-gel and LMA. IV, inverse of variance; SE, standard error. for differences in rates of: coughing, RR 1.02 (95% CI 0.53–1.94), p = 0.96 [7, 9–13]; breath-holding, RR 0.60 (95% CI 0.15–2.43), p = 0.47 [7, 9, 11]; desaturation, RR 0.91 (95% CI 0.34–2.48), p = 0.86 [7–14]; laryngospasm or bronchospasm, RR 0.62 (95% CI 0.25– 1.54), p = 0.30 [6, 7, 9–14]. Blood was seen less often on the i-gel than on the LMAs (Fig. 7). No participant complained of a numb tongue [6, 9, 12, 13] or hoarseness [6, 7, 9, 12, 13]. There was no evidence for differences in the rates of: sore throat, RR 0.70 (95% CI 0.10–4.72), p = 0.37 [6, 7, 9, 10, 12–14]; PONV, RR 0.75 (95% CI 0.40–1.41), p = 0.38 [7, 13, 14]. One [14] of five RCTs [6, 7, 9, 13, 14] reported dental or lip trauma for one participant with each device. Regurgitation was reported by one [10] of three RCTs [7, 10, 13] for two participants with i-gels.
Discussion Both types of airway exhibited similarly high rates of insertion at the first attempt. A subgroup analysis for ease of insertion showed disparate assessments by © 2014 The Association of Anaesthetists of Great Britain and Ireland
observers (more likely to favour the i-gel) compared with users (more likely to favour the LMA). This significant difference between subgroups might have had nothing to do with the assessor, as it might have been confounded by something else that systematically differed between these two groups of RCTs: researchers should collect assessments by both user and observer for ease of insertion. The heterogeneity for airway insertion time was substantial: the pooled insignificant difference belies two RCTs [10, 13] with a pooled estimate of faster LMA insertion, compared with three RCTs [8, 11, 14] with a pooled estimate of faster i-gel insertion and one RCT [7] that reported the same insertion times. This disparity might be related to slower insertion of the ProSeal LMA compared with other LMAs, although the caveats of subgroup analysis apply. The disparity does not correspond with the different measurements used by these RCTs for insertion time: from facemask removal to ‘adequate’ ventilation [7, 10, 13, 14]; from facemask removal to first capnographic upstroke [11]; 5
Anaesthesia 2014
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
Figure 5 Forest plot of the i-gel vs LMA for oropharyngeal leak pressure: RCTs were grouped according to the type of LMA: cLMA, classic; PLMA, ProSeal; SLMA, Supreme; ALMA, Ambu AuraOnce. The point of interest is that there was no interaction between the type of LMA studied by RCTs and the mean difference in leak pressure. IV, inverse of variance; SE, standard error.
Figure 6 Forest plot of the i-gel vs LMA for the rate of good fibreoptic view through the airway. The point of interest is that a good fibreoptic view was achieved through 93% of i-gels and through 83% of LMAs, with the pooled effect unaltered by exclusion of the most extreme result [11], although I2 heterogeneity was reduced to 36%. M–H, Mantel–Haenszel. or from holding the airway to successful placement [8]. The interpretation of insertion time is further complicated by RCTs’ incorporating different numbers of insertion attempts in the measurement: one [11]; two [8]; or three [7, 10, 13, 14]. 6
There was also substantial heterogeneity for the i-gel’s greater leak pressure: the hypotheses that this effect varied with neuromuscular blockade and LMA type were not supported in subgroup analyses. Higher airway leak pressures may be needed when ventilation © 2014 The Association of Anaesthetists of Great Britain and Ireland
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
Anaesthesia 2014
Figure 7 Forest plot of the i-gel vs LMA for the rate of blood on the airway. M–H, Mantel–Haenszel.
is difficult, for instance with head-down positioning or in obese patients, or those with pulmonary disease [16]. The rate of good fibreoptic view was 93% through the i-gel and 83% through LMAs, but this has not been associated with ease of ventilation [17, 18– 20]. However, the good fibreoptic view provided by igel might be a useful conduit for tracheal intubation in children in whom direct laryngoscopy has failed [21]. We suggest that the higher rate of blood on the LMA was not clinically significant. The present review has several limitations. Blinding of outcome assessment was difficult as the i-gel looks different to LMAs, with five RCTs failing to blind at all [7, 8, 10, 11, 13]. Although some outcomes were more objective than ease of insertion, they were nevertheless vulnerable to bias introduced by knowledge of what airway was being assessed. Conclusions based upon nine RCTs will be less reliable than meta-analyses that include more studies on more participants. In conclusion, the i-gel was similar to LMAs when used in children and delivered ventilation pressures 3 cmH2O higher than LMAs. Few complications were reported with either airway.
Acknowledgement This study was supported by a grant from the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (A100054). © 2014 The Association of Anaesthetists of Great Britain and Ireland
Competing interests No competing interests declared.
References 1. White MC, Cook TM, Stoddart PA. A critique of elective pediatric supraglottic airway devices. Pediatric Anesthesia 2009; 19 (Suppl 1): 55–65. 2. Brain AI. The development of the Laryngeal Mask–a brief history of the invention, early clinical studies and experimental work from which the Laryngeal Mask evolved. European Journal of Anaesthesiology Supplement 1991; 4: 5–17. 3. Intersurgical. User guide. http://media.intersurgical.com/igel. com/assets/userguide.pdf (accessed 09/01/2013). 4. Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. UK: The Cochrane Collaboration, 2011. 5. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. British Medical Journal 2003; 327: 557–60. 6. Das B, Mitra S, Jamil SN, Varshney RK. Comparison of three supraglottic devices in anesthetised paralyzed children undergoing elective surgery. Saudi Journal of Anaesthesia 2012; 6: 224–8. 7. Fukuhara A, Okutani R, Oda Y. A randomized comparison of the i-gel and the ProSeal laryngeal mask airway in pediatric patients: performance and fiberoptic findings. Journal of Anesthesia 2013; 27: 1–6. 8. Gasteiger L, Brimacombe J, Oswald E, et al. LMA ProSealTM vs. i-GelTM in ventilated children: a randomised, crossover study using the size 2 mask. Acta Anaesthesiologica Scandinavica 2012; 56: 1321–4. 9. Goyal R, Shukla RN, Kumar G. Comparison of size 2 i-gel supraglottic airway with LMA-ProSealTM and LMA-ClassicTM in spontaneously breathing children undergoing elective surgery. Pediatric Anesthesia 2012; 22: 355–9. 10. Jagannathan N, Sommers K, Sohn LE, et al. A randomized equivalence trial comparing the i-gel and laryngeal mask air7
Anaesthesia 2014
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
way Supreme in children. Pediatric Anesthesia 2013; 23: 127–33. Lee JR, Kim MS, Kim JT, et al. A randomised trial comparing the i-gelTM with the LMA ClassicTM in children. Anaesthesia 2012; 67: 606–11. Mitra S, Das B, Jamil SN. Comparison of size 2.5 i-gelTM with proseal LMATM in anaesthetised, paralyzed children undergoing elective surgery. North American Journal of Medical Sciences 2012; 4: 453–7. Theiler LG, Kleine-Brueggeney M, Luepold B, et al. Performance of the pediatric-sized i-gel compared with the Ambu AuraOnce laryngeal mask in anesthetized and ventilated children. Anesthesiology 2011; 115: 102–10. Tokgoz O, Tufek A, Beyaz SG, et al. Comparison of the efficacies of i-gelTM and LMA-ProSealTM for airway management in pediatric patients. Turkish Journal of Medical Sciences 2013; 43: 208–13. Uppal V, Gangaiah S, Fletcher G, Kinsella J. Randomized crossover comparison between the i-gel and the LMA-Unique in anaesthetized, paralysed adults. British Journal of Anaesthesia 2009; 103: 882–5. Theiler L, Gutzmann M, Kleine-Brueggeney M, Urwyler N, Kaempfen B, Greif R. i-gelTM supraglottic airway in clinical practice: a prospective observational multicentre study. British Journal of Anaesthesia 2012; 109: 990–5. Theiler LG, Kleine-Brueggeney M, Kaiser D, et al. Crossover comparison of the laryngeal mask supreme and the i-gel in simulated difficult airway scenario in anesthetized patients. Anesthesiology 2009; 111: 55–62. Francksen H, Renner J, Hanss R, Scholz J, Doerges V, Bein B. A comparison of the i-gel with the LMA-Unique in non-paralysed anaesthetised adult patients. Anaesthesia 2009; 64: 1118–24. Shariffuddin II, Wang CY. Randomised crossover comparison of the Ambu AuraOnce laryngeal mask with the LMA Classic laryngeal mask airway in paralysed anaesthetised patients. Anaesthesia 2008; 63: 82–5. Inagawa G, Okuda K, Miwa T, Hiroki K. Higher airway seal does not imply adequate positioning of laryngeal mask airways in paediatric patients. Pediatric Anesthesia 2002; 12: 322–6. Weiss M, Engelhardt T. Proposal for the management of the unexpected difficult pediatric airway. Pediatric Anesthesia 2010; 20: 454–64.
Appendix Search terms for MEDLINE 1. randomized controlled trial.pt 2. randomized controlled trial$.mp
8
Choi et al. | Comparison of the i-gel and laryngeal mask airway in children
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.
controlled clinical trial.pt controlled clinical trial$.mp random allocation.mp exp double-blind method/ double-blind.mp exp single-blind method/ single-blind.mp or/1–9 clinical trial.pt clinical trial$.mp exp clinical trial/ (clin$ adj25 trial$).mp ((singl$ or doubl$ or tripl$ or trebl$) adj25 (blind $ or mask$)).mp random$.mp exp research design/ research design.mp or/11–18 10 or 19 Case report.tw. Letter.pt. Historical article.pt. Review.pt. or/21–24 20 not 25 exp laryngeal masks/ laryngeal mask airway.mp. (LMA or ProSeal).mp. (mask adj6 airway).mp. or/27–30 i gel.mp. igel.mp. i-gel.mp. or/32–34 31 and 35 26 and 36
© 2014 The Association of Anaesthetists of Great Britain and Ireland
