Secondary Logo

Journal Logo

Airway management

Multicentre validation of manufacturers’ weight-based recommendations for laryngeal mask airway size choice in anaesthetic practice

A retrospective analysis of 20 893 cases

Avidan, Alexander; Eden, Arieh; Reider, Evgeni; Weissman, Charles; Levin, Phillip David

Author Information
European Journal of Anaesthesiology (EJA): June 2015 - Volume 32 - Issue 6 - p 432-438
doi: 10.1097/EJA.0000000000000267



Laryngeal mask airways (LMAs) are commonly used for airway management during general anaesthesia.1,2 Selecting the correct LMA size is important, as this influences the adequacy of the airway seal3 and is related to the incidence of postoperative sore throat.4–6 An initial incorrect size choice might also increase wastage from using more than one LMA per case. Manufacturers’ recommendations for the choice of LMA size have always been based solely on patient weight and have changed over time (Table 1).7–15 Currently, all manufacturers have similar size recommendations15 (Table 2). The only exception is the size 3 LMA that is recommended by different manufacturers either for adults or children of the same weight range. The weight-based manufacturers’ recommendations are not based on published scientific studies.16

Table 1
Table 1:
Timeline of the inventor and manufacturers’ weight-based laryngeal mask airway size recommendations
Table 2
Table 2:
Current weight-based laryngeal mask airway size recommendations by manufacturers, laryngeal mask airway size weight ranges used in the study and study model age and sex-based laryngeal mask airway weight ranges

LMA size recommendations were initially proposed by A.I.J. Brain, the inventor of the LMA,17 based on cadaveric specimens of infants and adults18. Paediatric LMA sizes were scaled down adult sizes and were not even based on cadaveric specimens.10 Today's weight-based manufacturers LMA size recommendations in adults are very approximate guides and not the direct result of research.16,19 Most studies on LMA size selection are related to different sex-based recommendations,5,20–23 which perform better than weight-based ones.24–27 In children, not all weight-based LMA size recommendations seem to be suitable in clinical practice.28–30

The aim of this retrospective study was to examine the practical issue of whether anaesthetists use the LMA size as per manufacturers’ weight-based recommendations or whether other clinical factors (namely gender and age) were associated with the size of the LMA inserted. We hypothesised that anaesthetists adhere to the manufacturers’ weight-based recommendations.

Materials and methods

The study was approved by the appropriate ethics authority of the Hadassah Medical Organization, Jerusalem, Israel (Ethical Committee number 0316-11-HMO, date of approval 30 November 2011, Chairperson Prof. Tova Chajek-Shaul), the Lady Davis Carmel Medical Center, Haifa, Israel (Ethical Committee number 021-38798, date of approval 9 September 2014, Chairperson Dr Ruth Kitzes) and the Sourasky Medical Center, Tel Aviv, Israel (Ethical Committee number TLV-0002-12, date of approval 4 March 2012, Chairperson Prof. Marcel Tupilsky).

Four Israeli hospitals contributed data to the study: Hadassah Medical Center, Ein Karem Jerusalem (data collection period from 02/2007 to 02/2010); Hadassah Medical Center, Mount Scopus, Jerusalem (data collection period from 02/2007 to 02/2010); Lady Davis Carmel Medical Center, Haifa (data collection period from 06/2004 to 06/2010); and Sourasky Medical Center, Tel-Aviv (data collection period from 05/2008 to 07/2010). For all anaesthetics performed using an LMA, the final LMA size used, age, sex and weight of the patient were extracted from the Anaesthesia Information Management System (AIMS) (Metavision; iMDsoft, Tel-Aviv, Israel) of each participating hospital. All hospitals used single use LMAs of the standard design (not specially designed LMAs such as LMA Proseal or LMA-Supreme). LMA size 6 was not available in any of the participating hospitals.

Patients were divided into three age groups:31 adults (>18 years); adolescents (10 to 18 years); and children (newborn to <10 years). In order to examine the study's hypothesis, sexes were examined separately. Data from adults weighing less than 50 or more than 129 kg or recorded with an LMA size smaller than 3, adolescents weighing less than 20 or more than 129 kg or recorded with an LMA size smaller than 2 and children weighing 2 or more than 50 kg or recorded with an LMA size greater than 3 were excluded as reflecting either outliers or charting errors.

All manufacturers have the same weight-based LMA size recommendations that overlap for most LMA sizes at the lower and upper weight borders. In order to assess adherence to manufacturers’ recommendations, the weight limits had to undergo minor adjustments to eliminate overlap (Table 2). Adherence was defined as the proportion of patients in each weight group wherein the LMA size used matched the manufacturers’ recommendations.

In order to discover whether other clinical factors influenced LMA size choice, multinomial logistic regression was used to identify predictors of LMA size as a function of weight, sex and age. These models were created under the assumption that the LMA size recorded reflected the most appropriate size, and thus, the data could be used for predictive purposes. Regression was performed separately for each of the three patient age groups. Using the simplest method of cross-validation (the holdout method), the data sets for each age group were randomly split into a development (70%) and validation set (30%). This ratio was selected as the more data used in creating the model, the better the fit and the greater the accuracy of estimated parameters. The parameters of the model based on the development data set were then assessed using the validation data set. A model for predicting LMA size used was constructed for each age group, based on the predictors that achieved statistical significance in the development set. Concordance between the actual LMA chosen and the LMA size predicted by the model and manufacturers’ recommendations was calculated in the validation data set for each age group. Continuous data were compared using two-tailed Student's t-test. Proportions were compared using the Chi-squared test. A P value of less than 0.05 was defined as statistically significant. Statistical analyses were performed with SAS Version 9.2 (SAS Institute, Cary, North Carolina, USA).


Data from 20 972 operations performed using a LMA were retrieved from the AIMS of the four hospitals, including 13 786 (65.7%) operations in adults, 1816 (8.7%) in adolescents and 5370 (25.6%) in children. Data from 79 (0.4%) cases were excluded due to outlying data [43 adults (0.3%), nine adolescents (0.5%) and 27 children (0.5%)] leaving 20 893 LMA insertions for analysis: 13 743 (65.8%) in adults; 1807 (8.6%) in adolescents; and 5343 (25.6%) in children. Table 3 summarises the demographic data of the study individuals.

Table 3
Table 3:
Demographic data of patients included in the study

The adherence of anaesthetists to manufacturers’ recommendations for LMA size in the three patient groups according to patient weight and sex is summarised in Table 4. Adherence was greater in children (76.3%), than in adolescents (49.0%) and adults (37.8%) (P < 0.001 including all categories and comparison of adults with children and adolescents). There was no difference in adherence according to sex in children (P = 0.152) or adolescents (P = 0.345), while adherence was lower in adult women than in men (P < 0.001).

Table 4
Table 4:
Adherence rate to manufacturers’ laryngeal mask airway size recommendations in the study populationa

The development data set included 14 623 cases and the validation set 6270 cases, with similar distributions of sex, age and weight (Appendix 1, The model set-based parameters and equations for adults, adolescents and children can be found in Appendix 2,

In the adult development data set, both sex (P < 0.001) and weight (P < 0.001) were found to be significant predictors of the LMA size used. Age was not a predictor (P = 0.20). The final model, therefore, analysed men and women separately. For adolescents, sex (P < 0.001), weight (P < 0.001) and age (P < 0.001) were found to be significant predictors of LMA size. However, the contribution of age beyond weight was minimal. Therefore, for the final model, only weight and sex were used as predictors. For children, sex was not found to be a significant predictor of LMA size (P = 0.38). Age and weight were found to be significant in the development data set (P < 0.001), but the contribution of age beyond weight was again minimal. The final model in children used only weight as a predictor. Figure 1 shows the probability scores for LMA sizes by patient weight for each age group. In children, the LMA weight ranges of the model fit well to manufacturers’ recommendations, whereas in adolescents and adults, the weight ranges for the LMA sizes were different for men and women and much wider than the manufacturers’ recommendations (Table 2).

Fig. 1
Fig. 1:
Model-based weight ranges for laryngeal mask airway sizes. Model-based weight probability scores (weight ranges) for LMA sizes for the three age groups and sex (except in children). The vertical lines in the graphs are the model-based borders of the weight ranges (at the interception of the 50% probability score of the LMA sizes).

Table 5 summarises the concordance between the development and validation sets for the model-based and manufacturer-based recommendations. The model performed significantly better than the manufacturers’ recommendation in predicting the LMA used by anaesthetists for adults (P < 0.001) and adolescents (P < 0.001) but not for children (P = 0.126).

Table 5
Table 5:
Concordance (shown as percentage with 95% confidence interval) between laryngeal mask airways sizes predicted by the regression model and the manufacturers’ recommendation for each age group


Our main finding is that in adolescents and adults, anaesthetists do not adhere to manufacturers’ recommendations of LMA size. This could result from the influence of other patient variables (such as sex) or be based on indifference (e.g. the anaesthetist picks up whichever LMA is closest to hand). We suggest that the former is more likely. First, the model's development data set demonstrates that weight and sex are significant variables associated with the choice of LMA. Second, use of these variables in the model's validation set improves prediction of the LMA size chosen to a level significantly above manufacturers’ weight-based recommendations. Finally, anaesthetists closely follow manufacturers’ weight-based recommendation in children indicating that they do not choose LMA size randomly.

It seems that A.I.J. Brain, the inventor of the LMA, determined the weight-based size recommendations using a rule of thumb rather than on the basis of scientific study. Initially, he created four sizes (1, 2, 3 and 4), and only later were the half sizes added for greater flexibility in children and a size 5 for adults.19 More recently, an LMA size 6 has also been added by certain manufacturers for use in adults weighing more than 100 kg. Initially, the LMA size 3 was intended for adult females and the LMA size 4 for adult males.19 Today, the use of the LMA size 3 is recommended for children or adults weighing 30 to 50 kg, depending on the LMA brand [Teleflex Incorporated Instructions For Use – LMA Classic, LMA Flexible, LMA Flexible Single Use & LMA Unique; (Accessed 11 December 2014) and Ambu A/S Denmark Ambu AuraStraight Disposable Laryngeal Mask; (Accessed 11 December 2014)].

The higher adherence to manufacturers’ recommendations in children than in adults might also be explained by the greater number of LMA sizes available, each with a smaller weight-range. In children, there are five LMA sizes for a weight range up to 49 kg, while there are only three LMA sizes for all adults. This may suggest that the availability of LMA sizes such as 3.5 and 4.5 could be considered.

Although our data do not relate to the outcome of LMA insertion, they do reflect the practice of a large sample of anaesthetists from four hospitals and encompass a large number of patients. Several other studies have described weight, sex, height and other physical characteristics (such as combined widths of the patient's index, middle and ring fingers32 or size of external ear33) as the basis for the choice of LMA size in both adults and children. However, these were either based on studies with small numbers of patients20,22–25,29 or on personal opinions and letters without scientific references.11,17,19,34–36 Other proposed formulae to define the correct LMA size have focused on patient weight in isolation35,37. Berry et al.24 found no correlation between weight, sex or height and LMA size in adults, but their study was based on only 30 patients. Kim et al.38 showed that in overweight children between the age of 2 and 15 years, manufacturers’ weight-based recommendations are appropriate. However, in underweight children, LMA size selection based on estimated weight by age provided better ventilation conditions than actual patient weight. Asai and Brimacombe11 supported a strict sex-based selection of LMA size in adults (LMA size 5 for men and size 4 or 5 for women). However, their recommendations were not based on clinical studies. The shape of the larynx may also have an influence, as it seems to be different between tall and small adults39 and male and female patients.23 Racial differences in cephalometry40 may also influence the optimal LMA size in adults. According to Brimacombe et al.,41 smaller LMAs are easier to insert in Oriental patients. Rao et al.42 showed that in the female Malaysian population, the LMA size 4 fits best and that the LMA size 5 was often too big to be inserted or positioned. According to the study of Tan et al.43 on LMA size in Asian populations, the LMA size 5 in male and the LMA size 4 in female patients produced the most effective glottic seal. Most of the Israeli population is white.

Our study has a number of limitations. Other clinical factors that may have influenced choice of LMA size include height,24 habitus or BMI, mouth opening, Mallampati classification, no data of which were available to be included in the analyses. There were also no data available on the quality of LMA fit (leaks, volume and pressure of cuff inflation) and it is unknown in how many cases the LMA initially inserted had to be exchanged for a smaller or larger one. It is assumed that the AIMS record reflected the final LMA size chosen and that this represented the best LMA size for the given patient. No data were available on the experience level of the different anaesthetists involved, but as a large number of anaesthetists were included and that experience in LMA insertion is acquired within a short time period,44 its influence was considered to be negligible. Data on the specific brands of LMA used in each case were not available and, thus, could not be included as a variable. However, only LMAs with the standard design were used. All LMA manufacturers provide equal weight recommendations for choosing LMA size and the maximal cuff volumes (used as proxy for cuff dimensions) for equally sized LMAs are almost identical across different manufacturers. A brand effect is, therefore, considered unlikely. Although this is a retrospective study, the large number of patients, anaesthetists and clinical data included (so-called ‘Big Data’)45,46 from four hospitals add substantial validity to the findings.

Manufacturers’ weight-based recommendations for LMA size in adults and adolescents do not reflect clinical practice. LMA size in adults and adolescents can be effectively predicted using sex-specific patient weight categories, which are different to those recommended by manufacturers. The introduction of half sizes in adults may be helpful. This is in contrast to children, in whom the manufacturers’ weight categories are closely adhered to and are suitable for clinical practice.

Acknowledgements relating to this article

Assistance with the study: none.

Financial support and sponsorship: none.

Conflicts of interest: none.

Presentation: this study was presented in part at the 22nd International Congress of the Israel Society of Anesthesiologists, September 2011, Tel-Aviv, Israel.


1. Brimacombe JR, Brain AIJ, Berry AM. The laryngeal mask airway. 1997; London: W.B. Saunders, 6–7.
2. Verghese C, Brimacombe JR. Survey of laryngeal mask airway usage in 11 910 patients: safety and efficacy for conventional and nonconventional usage. Anesth Analg 1996; 82:129–133.
3. Rosenblatt W, Sukhupragarn W. Barash P, Cullen B, Stoelting R, Cahalan M, Stock M. Airway management. Clinical anesthesia. Philadelphia, PA: Lippincott Williams & Wilkins, A Wolters Kluwer Business; 2009; 758.
4. Wong JG, Heaney M, Chambers NA, et al. Impact of laryngeal mask airway cuff pressures on the incidence of sore throat in children. Paediatr Anaesth 2009; 19:464–469.
5. Nott MR, Noble PD, Parmar M. Reducing the incidence of sore throat with the laryngeal mask airway. Eur J Anaesthesiol 1998; 15:153–157.
6. Burgard G, Mollhoff T, Prien T. The effect of laryngeal mask cuff pressure on postoperative sore throat incidence. J Clin Anesth 1996; 8:198–201.
7. Brain AI. The laryngeal mask - a new concept in airway management. Br J Anaesth 1983; 55:801–805.
8. Brain AI, McGhee TD, McAteer EJ, et al. The laryngeal mask airway. Development and preliminary trials of a new type of airway. Anaesthesia 1985; 40:356–361.
9. Brain AI. Further developments of the laryngeal mask. Anaesthesia 1989; 44:530.
10. Mason DG, Bingham RM. The laryngeal mask airway in children. Anaesthesia 1990; 45:760–763.
11. Asai T, Brimacombe J. Review article: cuff volume and size selection with the laryngeal mask. Anaesthesia 2000; 55:1179–1184.
12. McEwan AI, Mason DG. The laryngeal mask airway. J Clin Anesth 1992; 4:252–257.
13. Pennant JH, White PF. The laryngeal mask airway. Its uses in anesthesiology. Anesthesiology 1993; 79:144–163.
14. Brimacombe JR, Berry A. Laryngeal mask size selection and the sixth LMA. J Clin Anesth 1995; 7:265–266.
15. Brimacombe JR, Brain AIJ, Berry AM. The laryngeal mask airway. 1997; London: W.B. Saunders, 62–63.
16. Brimacombe JR. Positive pressure ventilation with the size 5 laryngeal mask. J Clin Anesth 1997; 9:113–117.
17. Voyagis GS. Appropriate size of the laryngeal mask airway in adults. Br J Anaesth 1998; 81:656–657.
18. Brain AI. Size matters!. Can J Anaesth 2000; 47:1262.
19. Bricombe J, Brain A. The laryngeal mask airway. London: W.B. Saunders; 1997.
20. Asai T, Howell TK, Koga K, Morris S. Appropriate size and inflation of the laryngeal mask airway. Br J Anaesth 1998; 80:470–474.
21. Nott MR. A ‘his’ and ’hers’ laryngeal mask? Anaesthesia 1998; 53:714–715.
22. Asai T, Murao K, Yukawa H, Shingu K. Re-evaluation of appropriate size of the laryngeal mask airway. Br J Anaesth 1999; 83:478–479.
23. Brimacombe J, Keller C. Laryngeal mask airway size selection in males and females: ease of insertion, oropharyngeal leak pressure, pharyngeal mucosal pressures and anatomical position. Br J Anaesth 1999; 82:703–707.
24. Berry AM, Brimacombe JR, McManus KF, Goldblatt M. An evaluation of the factors influencing selection of the optimal size of laryngeal mask airway in normal adults. Anaesthesia 1998; 53:565–570.
25. Voyagis GS, Batzioulis PG, Secha-Doussaitou PN. Selection of the proper size of laryngeal mask airway in adults. Anesth Analg 1996; 83:663–664.
26. Kihara S, Brimacombe J. Sex-based ProSeal laryngeal mask airway size selection: a randomized crossover study of anesthetized, paralyzed male and female adult patients. Anesth Analg 2003; 97:280–284.
27. Kihara S, Brimacombe JR, Yaguchi Y, et al. A comparison of sex- and weight-based ProSeal laryngeal mask size selection criteria: a randomized study of healthy anesthetized, paralyzed adult patients. Anesthesiology 2004; 101:340–343.
28. Park C, Bahk JH, Ahn WS, et al. The laryngeal mask airway in infants and children. Can J Anaesth 2001; 48:413–417.
29. Loke GP, Tan SM, Ng AS. Appropriate size of laryngeal mask airway for children. Anaesth Intensive Care 2002; 30:771–774.
30. Bagshaw O. The size 1.5 laryngeal mask airway (LMA) in paediatric anaesthetic practice. Paediatr Anaesth 2002; 12:420–423.
31. Cromer B. Adolescent development. In: Kliegman RM, Stanton BF, Geme JWS, Schor NF, Behrman RE, editors. Nelson textbook of pediatrics. Philadelphia, PA; 2011. pp. 649–659.
32. Gallart L, Mases A, Martinez J, et al. Simple method to determine the size of the laryngeal mask airway in children. Eur J Anaesthesiol 2003; 20:570–574.
33. Zahoor A, Ahmad N, Sereche G, Riad W. A novel method for laryngeal mask airway size selection in paediatric patients. Eur J Anaesthesiol 2012; 29:386–390.
34. Hirai S, Nakamura R, Maekawa N. Alternative formula for Laryngeal Mask Airway trade mark size selection. Anesthesiology 2004; 100:460.
35. Kagawa T, Obara H. An easy formula to remember the laryngeal mask airway size-patient weight relationship. Anesthesiology 2000; 92:631–632.
36. Nott MR. Appropriate size of the laryngeal mask airway in adults. Br J Anaesth 1998; 81:657.
37. Ho AM, Karmakar MK, Dion PW. Choosing the correct laryngeal mask airway sizes and cuff inflation volumes in pediatric patients. J Emerg Med 2008; 35:299–300.
38. Kim HJ, Park MJ, Kim JT, et al. Appropriate laryngeal mask airway size for overweight and underweight children. Anaesthesia 2010; 65:50–53.
39. Brimacombe J, Keller C. Performance of the size 5 laryngeal mask airway in males and females. Anasthesiol Intensivmed Notfallmed Schmerzther 2000; 35:567–570.
40. Johnson PF. Racial norms: esthetic and prosthodontic implications. J Prosthet Dent 1992; 67:502–508.
41. Brimacombe J, Keller C, Sartain J. Appropriate size of laryngeal mask. Br J Anaesth 2000; 84:290–291.
42. Rao AS, Yew AE, Inbasegaran K. Optimal size selection of laryngeal mask airway in Malaysian female adult population. Med J Malaysia 2003; 58:717–722.
43. Tan SM, Sim YY, Koay CK. The ProSeal laryngeal mask airway size selection in male and female patients in an Asian population. Anaesth Intensive Care 2005; 33:239–242.
44. Mohr S, Weigand MA, Hofer S, et al. Developing the skill of laryngeal mask insertion: prospective single center study. Anaesthesist 2013; 62:447–452.
45. Sessler DI. Big data and its contributions to peri-operative medicine. Anaesthesia 2014; 69:100–105.
46. Litman RS. Complications of laryngeal masks in children: big data comes to pediatric anesthesia. Anesthesiology 2013; 119:1239–1240.
© 2015 European Society of Anaesthesiology